The Role of Health and Health Promotion in Labour Force - RePub

The Role of Health and Health Promotion

in Labour Force Participation

MerelSchuring

The printing of this thesis was financially supported by the Department of Public

Health,ErasmusMC,UniversityMedicalCentreRotterdam.

Schuring,M

TheRoleofHealthandHealthPromotioninLabourForceParticipation.ThesisEras-

musMC,UniversityMedicalCentreRotterdam

ISBN:978-90-8559-094-1

Lay-outandPrint:OptimaGrafischeCommunicatie

Cover-photograph:©GuidoBenschop

©2010,M.Schuring

All rights reserved. No part of this book may be reproduced or transmitted in any

formorbyanymeanswithoutwrittenpermissionofthecopyrightowner.Someofthe

chaptersarebasedonpublishedmanuscripts,whichwerereproducedwithpermis-

sionofthepublishersandco-authors.

The Role of Health and Health Promotion in Labour Force Participation

De rol van gezondheid en gezondheidsbevordering in arbeidsparticipatie

Proefschrift

terverkrijgingvandegraaddoctoraande

ErasmusUniversiteitRotterdam

opgezagvanderectormagnificus

Prof.dr.H.G.Schmidt

envolgensbesluitvanhetCollegevanPromoties.

Deopenbareverdedigingzalplaatsvindenop

donderdag14oktoberom11:30uur

door

Merel Schuring

GeborenteAlphenaandenRijn

PRomoTiecommissie

Promotoren: Prof.dr.J.P.Mackenbach

Prof.dr.ir.A.Burdorf

Overigeleden: Prof.dr.M.C.H.Donker

Prof.dr.W.R.F.Notten

Prof.dr.J.L.Severens

contents

1. GeneralIntroduction 7

2. Ethnicdifferencesinunemploymentandillhealth 19

International Archives of Occupational and Environmental Health

2009, 82:1023-30

3. TheimpactofillhealthonexitfrompaidemploymentinEurope

amongolderworkers

33

Occupational and Environmental Medicine, in press

4. Theeffectsofillhealthonenteringandmaintainingpaid

employment:evidenceinEuropeancountries

53

Journal of Epidemiology and Community Health 2007, 61:597-604

5. Theeffectofre-employmentonperceivedhealth 71

Journal of Epidemiology and Community Health, in press

6. Effectivenessofahealthpromotionprogrammeforlong-term

unemployedsubjectswithhealthproblems:arandomised

controlledtrial

87

Journal of Epidemiology and Community Health 2009, 63:893-899

7. Changesinphysicalhealthamongparticipantsina

multidisciplinaryhealthprogrammeforlong-termunemployed

persons

BMC Public Health 2009, 9:197

105

8. GeneralDiscussion 125

Summary 139

Samenvatting 143

Dankwoord 147

CurriculumVitae 148

Listofpublications 151

PhDPortfolio 153

1General introduction

Ch

apte

r1

Generalintroduction 9

1.1 inequaLiTies in HeaLTH

Duringthelastcentury,thecombinedeffectsofimprovementsinlivingandworking

conditionsandadvancesinmedicineandhealthcarehaveledtoaconsistentlyincreas-

inglifeexpectancyintheEuropeanUnion.[1]In2007intheEU,thelifeexpectancyofa

newbornboyatbirthwas76.1yearsandofanewborngirl82.2years.Lifeexpectancy

is,however,notequallydistributedinsociety.Personswithalowerlevelofeducation,

aloweroccupationalclass,oralowerlevelofincometendtodieatyoungerage,and

tohave,withintheirshorterlives,ahigherprevalenceofallkindsofhealthproblems.

Thisleadstotremendousdifferencesbetweensocioeconomicgroupsinthenumberof

yearsthatpersonscanexpecttoliveingoodhealth.InEurope,differencesinhealthy

lifeexpectancytypicallyamountto10yearsormore,countedfrombirth[2]According

tomany,suchdifferences inhealthareunacceptable,andrepresentoneofEurope’s

greatestchallengesforpublichealth.

Unemployed persons are a specific socioeconomically disadvantaged group. The

relationship between unemployment and poor health has been well established,

as demonstrated by a higher prevalence of illness and disability [3-4] and a higher

mortalityamongunemployedpersons.[5-6]Selectionandcausationmaycontributeto

theseinequalitiesinhealthamongemployedandunemployedpersons.Selectionmay

actthroughtwodifferentpathways:workerswithapoorhealthmaybemorelikelyto

leavethelabourforce,andunemployedpersonswithapoorhealthmaybelesslikely

to enter the workforce. Causation may also act in two different ways. Leaving the

workforcemayhaveanegativeinfluenceonhealthoftheex-workers.Theotherway

around,gainingpaidemploymentmayhaveapositiveinfluenceonhealth.

Paragraph1.2(HealthandWork)givesanoverviewofthecurrentstateofknowledge

concerningtheinfluenceofhealthonenteringorleavingtheworkforce.Paragraph1.3

(Work and Health) is focused on the effect of gaining paid employment on health.

Paragraph 1.4 (Health promotion among the unemployed) describes the current

evidenceontheeffectivenessofhealthpromotioninterventionsamongunemployed

personsforre-employment.

1.2 HeaLTH anD WoRk

Thereisampleevidencefortheselectionhypothesis.[7-11]Acommunitybasedsurvey

intheUnitedKingdomfoundevidenceforhealthrelatedjobloss,especiallyinrela-

tiontomusculoskeletaldisordersandmentalillness.[8]InaBritishhouseholdpanel

survey healthier persons were more likely to gain employment than persons with

minorpsychiatricmorbidity.[10]

10 Chapter1

In many industrialized countries the population is ageing, due to increasing life

expectancy and falling birth rates.[12] A rather paradoxical development is that,

despiteincreasesinlifeexpectancy,theaveragetimepersonsspendinpaidworkhas

decreased or been stable in most European countries. This paradox is explained by

prolongededucationamongyoungercohorts,and-evenamoreimportantcontribu-

tor-higherratesofexitfromthelabourmarketatolderages.[13]Asaconsequence,

manycountriesaredevelopingpoliciestoencourageolderworkerstoremainlonger

inthelabourmarketanddelayretirement.[14]In2009,theDutchgovernmentplanned

toincreasethestateretirementagefrom65to67.

Consequencesof illhealth for the likelihoodofbecomingor stayingunemployed

maydependonsocialandlabourmarketpolicieswhichvaryacrossEuropeancoun-

tries.Forexample,strikingdifferencesinaccesstobenefits,suchasearlyretirement

ordisabilitypension,fordisabledpersonshavebeendescribedwithintheEuropean

Union.IntheNetherlandslessthan10%ofdisabledpersonshavetheirmainsource

of incomethroughlabourwhereas inSwedenthisproportionamountstoover50%.

[15]Therearedifferentpathways thatmay leadworkersoutof theworkforce,such

as disability pension, retirement, unemployment or becoming a homemaker. There

maybedifferencesbetweencountrieswithrespecttothepathwaypersonswithhealth

problemstakeoutoftheworkforce.Healthproblemsaresurelyapredictorforexitout

ofthelabourforceintodisabilitypension.Evidencefortheinfluenceofhealthonexit

outofthelabourforceintoretirement,unemploymentortakingcareofthehousehold

islessclear.

Thereis littleevidenceaboutthetime-dynamicsofhealthrelatedselectionintoor

outofthelabourforce.Itmaytakesometimebeforehealthcomplaintsmayresultin

lossofpaidemployment.Similarly,improvementsinthehealthstatusofunemployed

personsmayneedsometimetoinfluencethelikelihoodofbecomingre-employed.

The associations between health and employment may not be similar across all

socio-economicgroups.Personalcharacteristics,suchasage,education,ethnicback-

groundand sex, andhousehold characteristics, suchasmarital statusorhousehold

income,willinfluencethesocialcontextofhealthandemploymentstatusand,thus,

influencetheassociationsbetweenhealthandemployment.[9]

1.3 WoRk anD HeaLTH

Thereisalsoevidenceforthecausationhypothesis.Joblossmayleadtoimpairedrole

andemotionalfunctioning,poorhealth,depression,andlowselfesteem.[16-17]Ina

recentstudyamongBritishhouseholdswith11yearfollow-upunemploymenthada

strongeffectontheincidenceofanylimitingillnessandre-employmentwasrelated

Ch

apte

r1


to recovery from these illnesses.[3] A review of 16 longitudinal studies concerning

mentalhealtheffectsofunemploymentconcluded that lossofemploymentaffected

mentalhealth,butalsothatgainingemploymentimprovedmentalhealth.[18]ABrit-

ishlongitudinalstudyfoundthattransitionsfrompaidemploymenttovariousforms

of non-employment (unemployment, long-term sick leave, maternity leave) had a

negative impact on mental health. Transitions from non-employment to formal em-

ploymentresultedinanimprovementofmentalhealth.[10]Thiswasalsofoundina

five-yearfollowupstudyamonglong-termunemployedNorwegianswhichreported

recoveryofmentalhealthwithinsixmonthsafterre-employment.[19]

In a review Waddel and Burton[20] concluded that re-employment leads to clear

benefits inpsychologicalhealthandsomemeasuresofwell-being,althoughthere is

adearthofinformationonphysicalhealth.Astudyonthepsychologicalandphysi-

cal well-being during unemployment also demonstrated that the bulk of research

is focused on mental health outcomes, suggesting that other aspects of health need

moreattention.[21]Researchisneededtoinvestigatewhetherre-employmentalsohas

beneficialeffectsonphysicalhealth,andifthisistrue,whetherre-employmenthasan

immediateeffectonphysicalhealth.

1.4 HeaLTH PRomoTion amonG THe unemPLoyeD

Apoorhealthwillactasabarriertoreturntopaidemployment.[16-17,22]Inorderto

increasethepossibilitiesforre-employment,improvementsinhealthofunemployed

personsmay,therefore,beanimportantstep.

Unemployed persons with chronic health complaints, such as musculoskeletal

disorders,maybecomeprogressivelylesshealthy,sincefearofpainandreinjurymay

leadtoreducedactivities[23],resultinginapassivelifestylewithlowlevelsofphysi-

calactivity.[7,24]Hence,exercisestoimprovephysicalactivitymaybebeneficial,not

only for those subjects with disorders of the locomotive system, but also for other

chronicdiseasesaswell,includingheartandpulmonarydiseasesanddepression.[25]

InareviewPedersenandSaltin[25]concludedthatexercisetherapyhaspositiveef-

fectsongeneralwell-being,physicalhealthanddepressivesymptomsofpersonswith

chronicdiseases.

Cognitive-behaviouraltherapymaybeneededtotargetspecificpain-relatedbeliefs

andcopingstrategiesformodification.[26]RoseandHarris[27]haveidentifiedcogni-

tive behavioural therapy as a promising intervention to improve the psychological

healthofpersonswhoareunemployed.Watsonandcolleagues [28]providedsome

indications that an occupationally oriented rehabilitation programme consisting

of physical exercise and cognitive behavioural training for long-term unemployed

12 Chapter1

persons with chronic low back pain improved physical fitness as well as increased

employmentamongunemployedparticipantswithhealthcomplaints.However,these

resultswerebasedonanobservationalstudy.

Thereisaneedforarandomisedcontrolledtrialtoinvestigatetheeffectsofacom-

bined physical exercise and cognitive behavioural programme among unemployed

persons with a poor health. In addition, factors that determine non-participation,

drop-out and non-compliance in a health promotion programme for unemployed

personsneedtobeaddressed.

1.5 objecTives

Theobjectivesofthisthesisarethreefold:

1. To study the influence of poor health on entering and maintaining paid employment

2. To study the influence of entering paid employment on different dimensions of perceived

health.

3. To evaluate the effectiveness of a health promotion programme among unemployed persons

with health complaints on physical and mental health and re-employment

1.6 DaTaseTs useD in THis THesis

The analyses of this thesis were based on four different datasets. Two large Euro-

peanlongitudinaldatasets(SHARE;ECHP)wereusedforsecondarydata-analyses.A

longitudinalstudywasconductedtoinvestigatetheassociationbetweenhealthand

re-employmentamongunemployedcitizensofthecityofRotterdam.Arandomised

controlled trial was conducted to evaluate the effectiveness of a health promotion

programmeforunemployedpersonswithhealthcomplaintsinthecityofRotterdam.

1.6.1SurveyonHealthandAgeinginEurope(SHAREstudy)

Inchapter3theeffectofillhealthonexitfromthelabourforceamongolderworkers

inEuropeisinvestigatedontwowaves(2004,2006)oftheSurveyonHealthandAge-

inginEurope(SHAREstudy).TheSHAREstudyisalongitudinalsurveythataimsto

collectmedical,social,andeconomicdataonthepopulationagedover50yearsin11

EuropeanUnioncountries(Sweden,Denmark,TheNetherlands,Belgium,Germany,

Austria,Switzerland,France,Italy,Spain,andGreece).[11,12]Thefirstwaveofdata

wascollectedbyinterviewsbetweenAprilandOctober2004.Theoverallhousehold

Ch

apte

r1


responseacrossthe11SHAREcountriesinwhichdatacollectiontookplacein2004was

approximately62%,althoughsubstantialdifferencesamongcountrieswereobserved.

[12]Theavailabledatasetfromthefirstwaveofdatacollection(SHARERelease2.0)

contains28,517participants,with12,965subjects(45%)agedbetween50and63years.

Aftertwoyears8,729subjectsparticipatedagaininthequestionnairesurvey(SHARE

Release 1.0), resulting in a response of 67%. Complete information on employment

status in 2006 was available for 8,568 subjects. For the longitudinal analysis of the

influence of ill health on exit of the labour market, a cohort was available of 4,611

subjectswithpaidemploymentin2004andcompleteinformationonindividualand

workrelatedcharacteristicsatbaselineandworkstatusatfollow-upin2006.

1.6.2EuropeanCommunityHouseholdPanel(ECHP)

Inchapter4theeffectofillhealthontransitionsbetweenpaidemploymentandvarious

formsofnon-employment,includingretirement,unemployment,andhomemaker,is

investigatedinfivewaves(1994-1998)oftheEuropeanCommunityHouseholdPanel

(ECHP). The ECHP study is a social survey, designed for the member states of the

EuropeanUnion,withauniformdesignthatallowsforadaptationtonationalrequire-

ments. Through its longitudinal design it aims to represent the social dynamics in

Europe,from1994,theyearofthefirstwave,throughouttheperiodthatiscoveredby

thesubsequentwaves.ThedatawerecollectedbyNationalInstitutesforStatisticsor

researchcenters,whiledatachecks,weightingsandimputationsweredonecentrally

by the Statistical Office of the European Communities (Eurostat).All surveys were

basedonanon-stratifiedrandomsamplingdesign.Thetargetpopulationwasmade

upofallnationalprivatehouseholds.Allpersonsinthepanelhouseholdwereindi-

viduallyinterviewed.Thedatacollectionwascarriedoutinmostcountriesbypaper-

and-pencil interviewing, but in four countries (United Kingdom, The Netherlands,

Portugal,andGreece)bycomputer-assistedpersonalinterviewing.Theoverallhouse-

hold response in the firstwavewas72%,butvariedconsiderablyamongcountries.

TheresponseinlaterwavesoftheECHPstudywashigher.Adetaileddescriptionof

samplingproceduresandresponserateshasbeenpublishedelsewhere.[16,17]

Forthelongitudinalanalysisoftheinfluenceofillhealthonenteringandmaintain-

ingpaidemployment,twodifferentpopulationsweredefined.Thefirststudypopula-

tionincluded4446personswhowereunemployedforatleasttwoconsecutiveyears

ofwhich1590(36%)personsenteredtheworkforceduringthelastyearoffollow-up.

Thesecondstudypopulationconsistedof57436workerswhowereemployedforat

leasttwoconsecutiveyearsofwhich6191(11%)personslefttheworkforceinthelast

yearoffollow-upduetounemployment(n=3000),retirement(n=2017),ortakingcare

ofthehousehold(n=1174).

14 Chapter1

1.6.3Re-employmentandhealthdataset

Inchapter5theassociationbetweenhealthandre-employmentwasstudiedinalongi-

tudinalstudywith6monthsfollow-upamongunemployedpersonsonsocialsecurity

benefits,whowerereferredtoare-employmenttrainingcentreforare-employment

trainingbytheEmploymentCentreoftheCityofRotterdam,TheNetherlands.Some

oftheparticipantsdidhavechronichealthproblems,butweredeclaredfitenoughto

becapableoffulltimeemploymentafterinvestigationbyaphysician,apsychologist,

andanemploymentspecialist.FromDecember2004untilDecember2007,everyweek

anaverageof19subjectswasenrolledinthestudy.Intotal,2754eligibleparticipants

were included in the study. Information about socio-demographic characteristics,

health related quality of life (measured by the SF-36), and psychological measures

ofparticipantswerecollectedwithpostalquestionnairesandface-to-faceinterviews

indifferent languages(Dutch,Turkish,Arabic(interview)).Atbaselinetheresponse

was 66% (1829/2754) and after a follow-up period of approximately 6 months the

response was 53% (965/1829). Information about start dates of a re-employment

training and start and end dates of social security benefits and reasons for ending

benefits were derived from registries of the Employment Centre of the City of Rot-

terdam. Re-employment was defined as leaving the social security benefit services

forat least threemonthsbecauseof startingwithpaidemployment,verifiedby the

nationalSocialSecurityAgency.

1.6.4.Randomisedcontrolledtrialdataset

Inchapter6theeffectivenessofahealthpromotionprogrammeamongunemployed

persons is investigated in a randomized controlled trial. The study population

consisted of persons on social security benefits in the City of Rotterdam, who did

havechronichealthproblems,butweredeclaredfitenoughtobecapableoffulltime

employmentafter investigationbyaphysician, apsychologist, andanemployment

specialist.FromDecember2004untilDecember2007,eligibleparticipantswereran-

domly assigned to the intervention (n= 465) or control group (n=456). Information

about socio-demographic characteristics, health related quality of life (measured by

theSF-36),psychologicalmeasures,workvalues,andjobsearchactivitiesofpartici-

pantswerecollectedwithpostalquestionnairesandface-to-faceinterviewsindiffer-

entlanguages(Dutch,Turkish,Arabic(interview)).Atbaseline,theresponsewas74%

(343/465) intheinterventiongroupand68%(310/456) inthecontrolgroup.Aftera

follow-upperiodofapproximately6months,theresponsewas51%(176/343)inthe

interventiongroupand48%(150/310) in thecontrolgroup. Informationaboutstart

datesofare-employmenttrainingandstartandenddatesofsocialsecuritybenefits

Ch

apte

r1


andreasonsforendingbenefitswerederivedfromregistriesoftheEmploymentCen-

treoftheCityofRotterdam.

1.7 ouTLine oF THis THesis

Followingthisgeneral introduction,chapter2willdescribe theassociationbetween

unemployment and health among different socioeconomic groups. Chapter 3 and 4

willaddressthefirstobjectiveofthisthesis,i.e.theinfluenceofpoorhealthonenter-

ingandmaintainingpaidemployment.Chapter5willfocusonthesecondobjective

ofthisthesis,i.e.theinfluenceofenteringpaidemploymentondifferentdimensions

ofperceivedhealth.Chapter6and7willfocusonthethirdobjectiveofthisthesis,i.e.

theeffectivenessofahealthpromotionprogrammeamongunemployedpersonswith

healthcomplaintsonphysicalandmentalhealthandre-employment.

Chapter 2 presents a cross-sectional study on the association between health and

employmentstatusamongdifferentsocio-economicgroupsintheCityofRotterdam,

theNetherlands.Chapter3presentsasecondarydata-analysiswithtwowaves(2004,

2006) of the SHARE study on the effect of ill health on exit from the labour force

amongolderworkers inEurope.Chapter4presentsasecondarydata-analysiswith

fivewaves(1994-1998)oftheEuropeanCommunityHouseholdPanel(ECHP)onthe

effectofillhealthontransitionsbetweenpaidemploymentandvariousformsofnon-

employment, includingretirement,unemploymentandhomemaker.Chapter5pres-

entsaprospectivestudyontheassociationbetweenhealthandre-employmentamong

965unemployedpersonsintheCityofRotterdam.Chapter6presentsarandomised

controlledtrialtoevaluatetheeffectivenessofahealthpromotionprogrammeamong

unemployedpersonswithhealthcomplaintsonphysicalandmentalhealthand re-

employment.Chapter7presentsanobservationalstudytoevaluatechangesinphysi-

calhealthamongparticipantsinahealthpromotionprogramme.Chapter8discusses

the main findings of the previous chapters within the light of the objectives of this

thesis.Methodologicalconsiderationsandrecommendationsforclinicalpracticeand

futureresearcharepresented.

16 Chapter1

ReFeRences

1. Mackenbach JP: Ziekte in Nederland. Gezondheid tussen politiek en biologie.Amsterdam:UitgeverijMouria2010:59-67.

2. MackenbachJP:HealthInequalities:EuropeinProfile.Anindependent,expertreportcom-missionedby,andpublishedundertheauspicesof,theUKPresidencyoftheEU2005.

3. BartleyM,SackerA,ClarkeP:Employmentstatus,employmentconditions,andlimitingillness:prospectiveevidencefromtheBritishhouseholdpanelsurvey1991-2001.Journal of epidemiology and community health 2004,58(6):501-506.

4. JanlertU:Unemploymentasadiseaseanddiseasesoftheunemployed.Scand J Work Environ Health 1997,23Suppl3:79-83.

5. MorrisJK,CookDG,ShaperAG:Lossofemploymentandmortality.BMJ (Clinical research ed 1994,308(6937):1135-1139.

6. Lundin A, Lundberg I, Hallsten L, Ottosson J, Hemmingsson T: Unemployment andmortality--alongitudinalprospectivestudyonselectionandcausationin49321Swedishmiddle-agedmen.Journal of epidemiology and community health 2010,64(1):22-28.

7. ClaussenB,BjorndalA,HjortPF:Healthandre-employmentinatwoyearfollowupoflongtermunemployed.Journal of epidemiology and community health 1993,47(1):14-18.

8. SolomonC,PooleJ,PalmerKT,CoggonD:Health-relatedjobloss:findingsfromacommu-nity-basedsurvey.Occup Environ Med 2007,64(3):144-149.

9. McDonoughP,AmickBC,3rd:Thesocialcontextofhealthselection:alongitudinalstudyofhealthandemployment.Soc Sci Med 2001,53(1):135-145.

10. ThomasC,BenzevalM,StansfeldSA:Employmenttransitionsandmentalhealth:ananalysisfromtheBritishhouseholdpanelsurvey.Journal of epidemiology and community health 2005,59(3):243-249.

11. JusotF,KhlatM,RochereauT,SermeC:Joblossfrompoorhealth,smokingandobesity:anationalprospectivesurvey inFrance. Journal of epidemiology and community health 2008,62(4):332-337.

12. IlmarinenJE:Agingworkers.Occup Environ Med 2001,58(8):546-552. 13. StattinM:Retirementongroundsofillhealth.Occup Environ Med 2005,62(2):135-140. 14. CaiL,KalbG:Healthstatusandlabourforceparticipation:evidencefromAustralia.Health

Econ 2006,15(3):241-261. 15. OECD:Transformingdisabilityintoability.Policiestopromoteworkandincomesecurityfor

disabledpeople.Paris, OECD,2003. 16. PriceRH,ChoiJN,VinokurAD:Linksinthechainofadversityfollowingjobloss:how

financialstrainandlossofpersonalcontrolleadtodepression,impairedfunctioning,andpoorhealth.Journal of occupational health psychology 2002,7(4):302-312.

17. EdenD,AviramA:Self-efficacytrainingtospeedreemployment:Helpingpeopletohelpthemselves.Journal of Applied Psychology 1993,78(3):352-360.

18. MurphyGC,Athanasou,J.A.:Theeffectofunemploymentonmentalhealth.Journal of Oc-cupational and Organizational Psychology 1999,72:83-99.

19. ClaussenB:Healthandre-employmentinafive-yearfollow-upoflong-termunemployed.Scand J Public Health 1999,27(2):94-100.

20. WaddellG,Burton,A.K.:Isworkgoodforyourhealthandwell-being?London: The Stationery Office 2006(www.tsoshop.co.uk):ISBN:9780117036949.

Ch

apte

r1


21. McKee-RyanF,SongZ,WanbergCR,KinickiAJ:Psychologicalandphysicalwell-beingdur-ingunemployment:ameta-analyticstudy.J Appl Psychol 2005,90(1):53-76.

22. CreedPA:Improvingthementalandphysicalhealthofunemployedpeople:whyandhow?Med J Aust 1998,168(4):177-178.

23. BendixAF,BendixT,OstenfeldS,BushE,Andersen:Activetreatmentprogramsforpatientswithchroniclowbackpain:aprospective,randomized,observer-blindedstudy.Eur Spine J 1995,4(3):148-152.

24. LakkaTA,KauhanenJ,SalonenJT:Conditioningleisuretimephysicalactivityandcardiore-spiratoryfitnessinsociodemographicgroupsofmiddle-agesmenineasternFinland.Int J Epidemiol 1996,25(1):86-93.

25. PedersenBK,SaltinB:Evidenceforprescribingexerciseastherapyinchronicdisease.Scand J Med Sci Sports 2006,16Suppl1:3-63.

26. TurnerJA,JensenMP,RomanoJM:Dobeliefs,coping,andcatastrophizingindependentlypredictfunctioninginpatientswithchronicpain?Pain 2000,85(1-2):115-125.

27. RoseV,HarrisE:Fromefficacytoeffectiveness:casestudiesinunemploymentresearch.J Public Health (Oxf) 2004,26(3):297-302.

28. WatsonPJ,BookerCK,MooresL,MainCJ:Returningthechronicallyunemployedwithlowbackpaintoemployment.Eur J Pain 2004,8(4):359-369.

2ethnic differences in unemployment

and ill health

SchuringM,BurdorfA,KunstAE,VoorhamAJJ,MackenbachJP.

international Archives of Occupational and Environmental Health 2009, 82:1023-30.

20 Chapter2

absTRacT

objective:Theaimofthestudyistoevaluatewhetherhealthinequalitiesassociated

withunemploymentarecomparableacrossdifferentethnicgroups.

method:Arandomsampleof inhabitantsofthecityofRotterdamfilledoutaques-

tionnaire on health and its determinants, with a response of 55.4% (n=2057). In a

cross-sectional design the associations of unemployment, ethnicity, and individual

characteristics with a perceived poor health were investigated with logistic regres-

sionanalysis.Theassociationsofthesedeterminantswithphysicalandmentalhealth,

measuredbytheShortForm36HealthSurvey,wereevaluatedwithlinearregression

analyses. Interactions between ethnicity and unemployment were investigated to

determine whether associations of unemployment and health differed across ethnic

groups.

Results: Ill health was more common among unemployed persons (OR 2.6; 95% CI

1.7-3.8) than workers in paid employment. Health inequalities between employed

and unemployed persons were largest among native Dutch persons (OR=3.2) and

Surinamese/Antilleanpersons(OR=2.6),andsmaller inTurkish/Moroccanpersons

(OR=1.6)andoverseasrefugees(OR=1.6).Theproportionsofpersonswithpoorhealth

thatcouldbeattributedtounemploymentwere14,26,14,and13%,respectively.

conclusions: Differences in ill health between employed and unemployed persons

were less profound in ethnic groups compared to the majority population, but the

prevalence of unemployment was much higher in ethnic groups. The population

attributable fractions varied between 14% and 28%, supporting the argument that

policiesforhealthequityshouldpaymoreattentiontomeasuresthatincludepersons

in the labourmarketandthatpreventworkerswith illhealth fromdroppingoutof

theworkforce.

Ch

apte

r2

Ethnicdifferencesinunemploymentandillhealth 21

inTRoDucTion

Thepresenceofsocioeconomicinequalitiesinhealthhasbeenwidelyacknowledged.

Lower education, unskilled labour, and a low income are associated with higher

mortalityandmorbidity. [1]Labour forceparticipation isan importantdeterminant

ofhealthinequalities,asdemonstratedbyahigherprevalenceof illness.[2]anddis-

ability [3] and a higher mortality among unemployed persons.[4] A poor health is

strongly associated with non-participation in the labour force, both unemployment

anddisability.[5-6]Theassociationbetweenhealthandemploymentisbi-directional:

unemploymentmaycausepoorhealth (causationhypothesis),andpoorhealthmay

increasetheprobabilityofbecomingunemployed(selectionhypothesis).[7-8]

Within many countries, substantial inequalities in health between ethnic groups

exist.[9-10]Theextent towhichsocioeconomic inequalitiesunderlieethnic inequali-

ties in health remains debated. Many researchers argue that ethnic inequalities in

health are predominantly determined by socioeconomic inequalities.[11-12] Others

argue that ethnicity is an independent risk factor for self-reported illness, with an

importanceequaltoriskfactorssuchassocialclass,age,havingapoorsocialnetwork,

nottakingregularexercise,andnotfeelingsecureindailylife.[13]Whenthestrength

oftherelationbetweensocioeconomicstatusandhealthvariesacrossethnicgroups,

this will have consequences for the extent to which socioeconomic inequalities can

underlie ethnic differences in health. A Canadian study found that socioeconomic

factorsweremoreimportanttoself-ratedhealthstatusandpresenceofchronicillness

among immigrants than in non-immigrants [14] There are some indications from a

Germanstudythatunemployedforeignworkerswerelesssatisfiedwiththeirhealth

than unemployed Germans.[15] Schuring et al.[8] observed that in countries with a

lownationalunemploymentrateapoorhealthwasstronglyassociatedwithentering

orretainingpaidemployment,whereasincountrieswithahighnationalunemploy-

ment rate the effect of a poor health on selection in and out of the workforce was

much smaller.A possible explanation is that with high unemployment various fac-

torsdeterminelabouropportunities,suchaseducation,training,andage,andthata

poorhealthonlyplaysaminor role relative to these socio-demographic factors.[16]

Withlowunemploymentpersonsofallageandeducationallevelareretainedinthe

workforce and, thus, the influence of a poor health becomes more prominent. This

reasoningwouldimplythat,withinagivencountry,amongthosegroupswithhigh

unemployment,suchasminoritygroups,socio-demographic factorswillexceedthe

importanceofhealth.Hence,ahighunemploymentrateinminoritygroupsmaymask

theassociationbetweenhealthandemploymentstatusinthesegroups.

Inorder tobetterunderstandtherelationbetweenethnicity,socioeconomicstatus

andhealth,itisimportanttoassesswhethersocioeconomicstatusisassociatedwith

22 Chapter2

healthinasimilarwayacrossethnicgroups.Inthispaperweexaminetheassociations

betweenunemploymentandhealthinthethreelargestethnicminoritygroupsofthe

Netherlandsandtheindigenouspopulation.Theaimsofthestudywere(1)toevaluate

whethertheassociationsbetweenpoorhealthandemploymentstatusarelessstrong

amongethnicgroupswithhighunemployment thanamongDutchpersons,and (2)

toassessthedifferencesinproportionsofunemploymentattributedtopoorhealth.

meTHoDs

Population

BetweenMarchandJune2003ahealthquestionnairesurveywasundertakenbythe

municipalhealthserviceofRotterdaminarandomsampleof6404inhabitantsofthe

cityofRotterdam,aged16to84.[17]Aquestionnairewassendtothehomeaddress,

followedbytworeminderstwoandfourweekslater,respectively.Atotalof3406sub-

jectsreturnedthequestionnaire(response55.4%).Thoserespondentswhowereaged

between16and65yearsoldandnotengagedas student ina secondaryor tertiary

educational programme were selected for the current study with a cross-sectional

design.Atotalof2057subjectsmettheseinclusioncriteria.

Socio-demographicvariables

Thesocio-demographicvariablesincludedinthisstudywerelabourforceparticipa-

tion, ethnic background, highest educational level, age, sex, and marital status. In

this study, labour status was based on self-reported current economic status with

five mutually exclusive categories: full-time employment (>32 h/week), part-time

employment(<32h/week),unemployment,disabilitypension,andhomemaker.

Theethnicbackgroundoftherespondentwasbasedonthecountryoforiginofthe

mother.IncasethemotherwasborninTheNetherlands,thecountryofbirthofthe

fatherwasleading.[18]Differentethnicgroupsweredefined,basedondifferencesin

experiencesofmigration(refugeesorlabourmigrants)anddifferencesingeographi-

cal and cultural distance from the Netherlands. Three ethnic minority groups were

defined: (1) Turks and Moroccans, (2)Antilleans and Surinamese, and (3) refugees.

TurksandMoroccans initiallycameas labourmigrants to theNetherlandsfromthe

early1960s,whilethemigrationofSurinameseandAntilleans/Arubansisrelatedto

thecolonialpast.Refugeesareanotherimportantgroupofmigrantsfromdesignated

countriesAfghanistan,Algeria,Angola,Bosnia,China,Chile,Croatia,DemocraticRe-

publicoftheCongo,Eritrea,HongKong,Iran,Iraq,Kosovo,Liberia,Nigeria,Sedan,

Ch

apte

r2


Serve, Sierra Leone, Somalia, South Korea, Syria, or former-Yugoslavia. Immigrants

fromothercountrieswerenotincludedintheanalysis(n=296).

Subjectsweredivided into threegroupsaccordingto theirhighest levelofeduca-

tionalattainment.Ahigheducationallevelwasdefinedashighervocationaltraining

oruniversity,intermediateeducationallevelwasdefinedashighersecondaryschool-

ing or intermediate vocational training, and low educational level was defined as

noeducation,primaryschool, lowerandintermediatesecondaryschoolingorlower

vocationaltraining.Maritalstatuswasusedtodistinguishthosesubjectsmarriedor

livingtogetherwithothers.

Healthmeasures

Self-reported health (SRH) was measured by asking subjects to rate their overall

healthonafive-pointscale,rangingfrom‘excellent’,verygood’,‘good’and‘fair’to

‘poor’.Thosereportinglessthan‘goodhealth’weredefinedashavingapoorhealth.

[16]Healthwasalsomeasuredwith theDutchversionof theShortForm36Health

Survey(SF-36).[19]TheSF-36consistsof36 itemsthatwereusedtocalculatescores

oneightdimensions:physicalfunctioning,generalhealth,mentalhealth,bodilypain,

socialfunctioning,vitality,rolelimitationduetoemotionalhealthproblems,androle

limitationduetophysicalhealthproblems.Scorescouldrangefrom0to100,witha

higherscoreindicatingabetterhealthrelatedqualityoflife.

Statisticalanalysis

Characteristicsofsubjectswereanalyzedusingdescriptivestatistics.Fortwosubscales

oftheSF-36,rolelimitationsduetophysicalhealthproblemsandrolelimitationsdue

toemotionalhealthproblems,astrongceilingeffectexistedandthevariableswerenot

normallydistributed.Therefore,thesesubscaleswereexcludedfromfurtheranalysis.

The statistical analysis was conducted on the study population with complete

informationonallvariablesincludedinthemultivariateanalyses.Sinceeducational

levelwasnotavailable for207subjects (10%)and forothervariablesa fewmissing

valuesoccurred,thenumberofsubjectsintheanalysesmayvaryslightly.Theassocia-

tionsbetweenunemployment,ethnicityandothersocio-demographiccharacteristics

and perceived poor health were investigated with logistic regression analysis, with

the odds ratio (OR) as measure of association. The analysis started with univariate

logisticregressionmodelstodeterminewhichindependentvariableswereofinterest

toconsiderinthefinalmodel.VariableswithaPvalueofatleast0.10wereselected

for further analysis. A multivariate logistic regression analysis was conducted to

determinetheassociationofemploymentstatus,ethnicbackground,sex,age,educa-

24 Chapter2

tionallevel,andmaritalstatuswiththedichotomousoutcomemeasureofpoorhealth.

Explanatoryvariableswere included into themainmodelonebyonebya forward

selection procedure, in order of magnitude of explained variance in the univariate

analyses,and independentvariableswithaPvalueofat least0.05wereretained in

the model. Interaction effects between ethnicity and unemployment were analyzed

inordertodeterminewhethertheeffectsofunemploymentonhealthdifferedacross

ethnicgroups.Theproportionofpersonswithpoorhealththattheoreticallycouldbe

attributedtounemploymentwascalculatedwiththepopulationattributablefraction,

expressed by the formula PAF% = 100 × [p × (OR-1)] / [1 + p × (OR-1)], whereby

p is the proportion of unemployed persons and the OR is the association between

unemploymentandpoorhealth.[20]

Table 2.1 Characteristicsandperceivedhealthofsubjectswithdifferentethnicbackgroundsinacommunity-basedhealthsurveyintheNetherlands(n=2057)

Dutchn=1448

T/mn=228

s/an=281

Refugeen=100

Women 808(55.9%) 119(52.2%) 170(60.5%) 50(50.0%)

Age*18-24yr25-44yr45-54yr55-65yr

96(6.6%)662(45.7%)347(24.0%)343(23.7%)

34(14.9%)137(60.1%)31(13.6%)26(11.4%)

39(13.9%)145(51.6%)68(24.2%)29(10.3%)

13(13.0%)54(54.0%)19(19.0%)14(14.0%)

Married* 882(61.8%) 168(74.3%) 113(40.8%) 56(57.1%)

Educationallevel*HighIntermediateLow

Missing

394(28.7%)350(25.5%)628(45.8%)76

10(6.3%)42(26.4%)107(67.3%)69

24(10.0%)59(24.7%)156(65.3%)42

18(22.5%)30(37.5%)32(40.0%)20

Employmentstatus*Employed>32h/wkEmployed<32h/wkUnemployedDisabilitypensionHomemaker

812(56.1%)289(20.0%)111(7.7%)111(7.7%)125(8.6%)

83(36.4%)28(12.3%)60(26.3%)14(6.1%)43(18.9%)

139(49.5%)56(19.9%)63(22.4%)13(4.6%)10(3.6%)

51(51.0%)13(13.0%)25(25.0%)3(3.0%)8(8.0%)

Poorhealth* 261(18.1%) 97(42.7%) 88(31.7%) 21(21.0%)

Generalhealth*Physicalfunctioning*Socialfunctioning*Bodilypain*Vitality*Mentalhealth*Rolelimitations,physical*Rolelimitations,emotional*

70.1(19.7)87.4(19.9)81.7(23.2)78.7(24.2)62.6(19.2)73.9(17.6)80.2(34.5)84.7(32.1)

55.7(22.8)69.1(27.0)69.4(24.7)65.1(28.3)50.6(18.0)61.8(18.8)66.3(36.9)69.8(39.6)

63.3(20.6)78.8(25.8)73.7(27.2)72.2(26.6)54.9(18.9)68.3(20.6)77.5(35.0)78.8(37.2)

65.5(19.5)79.2(26.3)75.9(24.6)73.5(24.7)55.0(18.9)66.4(18.0)80.6(31.6)81.4(33.8)

*chi-squaretestP<0.05,comparingminoritygroupstothenativeDutchpopulation

Ch

apte

r2


Theassociationsoflabourstatus,ethnicity,andothersocio-demographiccharacter-

isticswithphysicalandmentalhealthwereinvestigatedwithmultiplelinearregres-

sionanalysis,withasdependentvariablesthescoresonthesixsubscalesoftheSF-36;

general health, physical health, bodily pain, mental health, social functioning, and

vitality.

All statistical analyses were performed with the statistical package SPSS 11.0 for

Windows.

ResuLTs

CharacteristicsofsubjectsarepresentedinTable2.1,stratifiedbyethnicbackground.

Immigrantsubjectswereyoungerofage,moreoftenunemployedand,withtheexcep-

tionofrefugees,lowereducatedthannativeDutchsubjects.SubjectswithaTurkishor

Morrocanbackgroundweremoreoftenmarriedandhomemakercomparedwiththe

otherethnicgroups.HealthstatuswaslowerinmigrantsthannativeDutchsubjects

formostdimensionsofhealth.TurkishandMoroccansubjectshadtheworsthealth,

whereasthehealthstatusofrefugeesresembledthatofnativeDutchsubjectsforsome

dimensionsofhealth.

Figure 2.1 shows that within each ethnic group, with the exception of refugees,

unemployed subjects had a worse health than employed subjects. Subjects with a

0102030405060708090

100

native Dutch Turkish /Moroccan

Surinamsese /Antillean

Refugee

Employed >32 h/wk Employed <32 h/wkUnemployed Disability Homemaker

Figure 2.1 Perceivedhealthofsubjectswithdifferentethnicbackgroundsinacommunity-basedhealthsurveyintheNetherlands(n=2057)specifiedfordifferentcategoriesoflabourforceparticipationorbeingoutoftheworkforce

26 Chapter2

disabilitypensionhadtheworsthealth ineveryethnicgroup.Amongsubjectswith

aTurkishorMoroccanbackgroundthehealthstatusofhomemakerswasequaltothe

healthstatusofunemployedsubjects.

Table 2.2 shows that all socio-demographic variables in this study were included

inthemultivariatemodel.MigrantsmoreoftenhadapoorhealththannativeDutch

subjects,evenafteradjustingforage,sex,educationallevel,maritalstatus,andlabour

forceparticipation.ThehealthstatusofTurkishorMoroccansubjectswastheworst

(OR=3.9(2.6-6.0)),whereasthehealthstatusofrefugeeswasnotsignificantlydifferent

(OR=1.8(0.9-3.3))fromthatofnativeDutchsubjects.

Table2.3describestheassociationswithhealth-relatedqualityoflife,whichresem-

bles thepatternobservedforaperceivedpoorhealth in table2.Allmigrantgroups

hadsignificantlylowerscoresforthephysicalaswellasthementalhealthdimensions

of the SF36 than the native Dutch group, indicating a poorer health. Unemployed

personsandpersonsreceivingadisabilitypensionhadsignificantlylowerscoresfor

alldimensionsofphysicalandmentalhealthcomparedwithemployedpersons.The

magnitudeofthesedifferenceswascomparabletotheobserveddifferencesinhealth

acrossageandeducation.

Table 2.2 Associationsbetweendemographicfactorsandemploymentstatuswithpoorhealthofsubjectswithdifferentethnicbackgroundsinacommunity-basedhealthsurveyintheNetherlands(n=1815)bymultivariatelogisticregressionanalysis

n oR (95% ci)

NativeDutchTurkish/MoroccanSurinamese/AntilleanRefugee

134615723379

1.03.9(2.6-6.0)2.5(1.7-3.6)1.8(0.9-3.3)

Age18-24yr25-44yr45-55yr55-64yr

143886417369

1.02.4(1.3-4.3)5.5(2.9-10.4)4.7(2.5-9.1)

Women 1016 1.6(1.2-2.1)

EducationallevelHighIntermediateLow

443473899

1.01.6(1.0-2.5)3.2(2.1-4.9)

Married 1106 1.4(1.0-1.8)

EmploymentstatusEmployed>32h/wkEmployed<32h/wkUnemployedDisabilitypensionHomemaker

996349194119157

1.01.0(0.7-1.5)2.6(1.7-3.8)14.4(8.8-23.6)1.1(0.7-1.8)

ORoddsratio,CIconfidenceinterval

Ch

apte

r2


Theinteractionbetweenemploymentstatusandethnicbackgroundhadasignificant

contributiontothelogisticregressionmodel(χ2=10.4;df=3;p=0.018),demonstrat-

ingthattheassociationsbetweenemploymentstatusandhealthvariedwithinethnic

groups (Table2.4).Health inequalitiesbetweenemployedandunemployedsubjects

werelargestamongtheDutchsubjects(OR=3.2(1.9-5.4),followedbySurinameseand

Antilleans(OR=2.6(1.3-5.2),andlesspronouncedamongTurkish/Moroccansubjects

(OR=1.6(0.7-3.7))andrefugees(OR=1.6(0.4-6.2)).ThePAFofunemploymentinpoor

healthwas14%amongDutch,26%inSurinameseandAntilleans,14%amongTurkish

andMoroccan,and13%amongrefugees.

Discussion

Illhealthwassubstantiallymorecommonamongunemployedpersonsthanworkers

inpaidemployment.Healthinequalitiesassociatedwithemploymentdifferedwithin

ethnic groups, with the strongest association between employment and health for

Table 2.3 Associationsbetweendemographicfactorsandemploymentstatuswithhealthrelatedqualityoflife(sixsubscalesoftheSF-36)ofsubjectswithdifferentethnicbackgroundsinacommunity-basedhealthsurveyintheNetherlands(n=1845)bymultivariatelinearregressionanalysis

General health

Physical functioning

bodily pain mental health

social functioning

vitality

Intercept 81.8(2.3) 102.4(2.4) 100.0(2.9) 82.9(2.2) 97.5(2.8)* 77.5(2.3)*


0-11.0(1.6)*-6.5(1.4)*-4.5(2.2)*

0-13.1(1.7)*-8.2(1.4)*-7.0(2.3)*

0-9.6(2.0)*-5.0(1.7)*-6.5(2.7)*

0-8.2(1.5)*-3.9(1.3)*-5.9(2.0)*

0-9.5(2.0)*-7.3(1.6)*-6.6(2.6)*

0-7.3(1.6)*-5.5(1.4)*-7.5(2.1)*

Age18-24yr25-44yr45-54yr55-64yr

0-1.7(1.7)-5.5(1.8)*-6.6(1.9)*

01.0 (1.7)-4.2(1.9)*-6.0(1.9*

0-2.2(2.1)-7.1(2.3)*-4.9(2.3)*

0-0.2(1.6)-0.3(1.7)-1.6(1.8)

0-2.4(2.0)-3.9(2.2)-2.7(2.3)

0-3.5(1.6)*-3.0(1.8)-2.5(1.8)

Women -1.4(1.0) -2.9(1.0)* -6.6(1.2)* -2.8(0.9)* -4.9(1.2)* -4.3(1.0)*


0-1.9(1.2)-6.1(1.2)*

0-1.9(1.3)-8.7(1.2)*

0-3.5(1.5)*-7.2(1.5)*

0-0.5(1.1)-4.4(1.1)*

0-1.0(1.5)-4.0(1.4)*

0-1.3(1.2)-6.1(1.1)*

EmploymentstatusEmployed>32h/wkEmployed<32h/wkUnemployedDisabilitypensionHomemaker

00.2(1.2)-7.7(1.5)*-28.3(1.9)*-1.0(1.8)

0-1.1(1.3)-4.7(1.6)*-33.2(2.0)*-3.1(1.9)

0-0.3(1.6)-8.8(1.9)*-29.3(2.3)*-0.09(2.2)

0-0.1(1.2)-10.5(1.4)*-16.1(1.8)*0.6(1.7)

0-1.3(1.5)-9.3(1.8)*-31.9(2.3)*1.0(2.1)

0-1.1(1.2)-7.3(1.5)*-19.3(1.9)*0.04(1.7)

*P<0.05

28 Chapter2

nativeDutchpersons,followedbySurinameseandAntilleansandalesspronounced

associationbetweenemploymentandhealthforTurkish/Moroccanpersonsandrefu-

gees. The population attributable fraction varied between 13% and 26%, indicating

thatemploymentstatusisanimportantfactorinsocioeconomichealthinequalities.

Thedesignof this studywas cross-sectional, and thereforenoassumptioncanbe

madeaboutthedirectionoftheassociationbetweenpoorhealthandunemployment

amongmigrantgroups.Unemploymentmaycausepoorhealthandpoorhealthmay

increase the probability of becoming unemployed.[7-8]A another limitation of this

studywasthelackofinformationonnon-respondents.Withrespecttounemployment

in the study population, the proportion of unemployed persons within each ethnic

groupresembledcloselytheregisteredunemploymentinthecityofRotterdamand,

thus,responsedoesnotseembiasedtowardsemployedorunemployedpersons.

In this study ethnic groups reported higher prevalences of poor health and also

lower scoresonhealth-relatedqualityof life.There is awidespreadagreement that

a single question asking subjects to rate their overall health on a scale from excel-

lenttopoorprovidesausefulsummaryofhowsubjectsperceivetheiroverallhealth

status.[16]However,thevalidityofthissingle-itemquestioninsubjectswithdifferent

cultural backgrounds has been questioned.[21] Differences in self-concepts between

ethnicgroupsmayinfluencetheresultsofthesingleitemgeneralhealthquestion.The

Table 2.4 Associationsbetweenunemploymentandpoorhealthwithindifferentethnicbackgroundsinacommunity-basedhealthsurveyintheNetherlands(n=1558)

oR (95% ci)

Age18-24yr25-44yr45-55yr55-64yr

11.9(1.1-3.6)4.2(2.3-8.0)4.1(2.2-7.9)

Women 1.6(1.2-2.2)


11.8(1.1-3.1)3.7(2.3-6.0)


14.3(2.4-7.4)2.8(1.8-4.3)2.0(0.9-4.1)

Effectofunemploymentwithinethnicgroup:NativeDutchTurkish/MoroccanAntillean/SurinameseRefugee

3.2(1.9-5.4)1.6(0.7-3.7)2.6(1.3-5.2)1.6(0.4-6.2)

Employed(full-timeandpart-time)andunemployedpersonswereincluded,whereashomemakersendisabledpersons(n=327)werenotincludedinthisanalysis.ORoddsratio,CIconfidenceinterval

Ch

apte

r2


observationthatafteradjustingforthewell-establishedsocio-demographicdetermi-

nantsofhealthinequalities,stillsystematicdifferencesinoccurrenceofpoorhealthin

ethnicgroupsrelativetotheDutchgroupwereobserved,mayindicateover-estimation

of poor health. In the current study similar conclusions on unemployed, ethnicity,

andhealthweredrawnwhenusingthesinglequestiononperceivedgeneralhealth

questionandtheother35questionsonphysicalandmentalhealthdimensionsofthe

SF36.Thiscorroboratestheopinionthatthegeneralhealthquestionprovidesagood

summery of the mental and physical health in migrant groups and the indigenous

population.Thisfindingis,ofcourse,alsosupportedbythehighcorrelationsbetween

perceivedgeneralhealthandallhealthdimensionsintheSF-36.

A high proportion of persons with a poor health among ethnic groups has been

observed invariousstudies indifferent countries.[9-13]Differentexplanationshave

beenputforward.ASwedishstudyamongimmigrantsfromPoland,Turkey,andIran

foundthatacculturation(definedbytheknowledgeoftheSwedishlanguage)wasan

important mediator in the pathway between ethnicity and poor health.[22] Indeed,

in our study population differences in mastering the Dutch language may have

influencedhealth.ForSurinameseandAntilleansDutch isusuallya firstor second

language,whereasforTurksandMoroccansknowledgeoftheDutchlanguageisoften

limitedorabsent,especiallyamongolderwomen.Languageproblemsmayhamper

effective communication with physicians and also inhibit access to information on

healthandhealthcare.[23]Inthecurrentstudy,masteryof theDutchlanguagewas

notincludedintheanalyses,buttheobservationthatthehealthstatusofhomemakers

withaTurkishorMoroccanbackgroundwasworse thanthehealthstatusofhome-

makerswithanotherethnicbackgroundmayreflectaloweracculturation.

Differencesinmigrationexperiencesmayalsocontributetothedifferencesinhealth

betweentheethnicminoritygroups.Refugeeshaveadifferentmigrationhistorythan

Turks,Moroccans,Surinamese,andAntilleans.Forrefugees,experiencesofviolence,

theflightforasylumandforcedbrokensocialnetworksmayhaveaffectedhealth.[13]

However,inthisstudywefoundthatrefugeeshadarelativelygoodhealthcompared

with the other ethnic minority groups. The relatively good health of refugees can

partlybeexplainedbytherelativelyhigheducationallevelofrefugeesrelativetothe

otherethnicminoritygroups.

Employedmigrantsarestillconcentratedinbluecollarjobsinindustry.[15]Especially

foremployedrefugees,whohavearelativelyhigheducationallevel(87%intermedi-

ate/higheducated)comparedtotheemployednativeDutch(77%intermediate/high

educated), adverse health effects of unsatisfactory jobs have been suggested (Smith

2000).AnAustralianstudyhas shown thatunsatisfactory jobscanbeasdepressing

asunemployment.[24]Unfortunately,informationaboutthepotentialmisfitbetween

personalcapabilitiesandjobrequirementswasnotcollectedinthecurrentstudy.

30 Chapter2

It was hypothesized that the associations of poor health and employment status

would be less profound in ethnic groups with a high prevalence of unemployment

comparedtotheDutchpopulation.Whentheunemploymentrateishigh,theeffectof

healthselectionoutoftheworkforceisrelativesmallcomparedtootherfactorsthat

determinelabouropportunitiesforpeople.[16]Ingeneral,ourresultsindeedshowed

that the association between unemployment and poor health was strongest in the

Dutchpopulation(OR=3.2)withthelowestunemployment,whereastheassociations

between unemployment and health were less profound in ethnic minority groups

(ORsbetween1.6and2.6),whichwerecharacterisedbyahigherunemploymentlevel.

In the current study the logistic regressionanalysis showed that theassociationbe-

tweenunemploymentandhealthwasnotstatisticallysignificantwithintheTurkish/

Moroccan group. However, when adjustment for sex and educational level did not

takeplace,asignificantassociationbetweenunemploymentandhealth(OR=2.5)was

found.Hence,theabsenceofhealthinequalitiesacrossemploymentstatuswithinthis

ethnicgroupmaybeexplainedby thestrongcorrelationsbetweensexandemploy-

mentstatusandbetweeneducationallevelandemploymentstatus.Theseadditional

analyses showed that especially female, low educated Turkish/Moroccan persons

wereoftenunemployedandalsoreportedthehighestoccurrenceofapoorhealth.

ThePAFofunemploymentinpoorhealthwithinthefourethnicgroupsvariedbetween

13% among refugees to 26% among Surinamese/Antillean subjects. The PAF values

among Dutch persons (14%) was strongly influenced by the high OR for unemploy-

ment,whereasthePAFvaluesamongtheethnicminoritygroupsweremoreinfluenced

bythehighprevalenceofunemployment.Althoughthiscross-sectionalstudydoesnot

permitconclusionsoncausality,thesefindingssuggestthat,undertheassumptionthat

unemploymentleadstoapoorhealth,healthinequalitiesrelatedtounemploymentarea

majorpublichealthprobleminallethnicgroups.Thus,policiesforhealthequityshould

paymoreattentiontomeasuresthatincludepersonswithapoorhealthinthelabour

marketandthatpreventworkerswithillhealthfromdroppingoutoftheworkforce.

Acknowledgements

FundingforthisstudywasprovidedbythePublicHealthFund(FondsOGZ).

ReFeRences

1. MarmotMG,SmithGD,StansfeldS,PatelC,NorthF,HeadJ,WhiteI,BrunnerE,FeeneyA: Health inequalities among British civil servants: the Whitehall II study. Lancet 1991,337:1387-1393.

Ch

apte

r2


2. ClaussenB:Healthandre-employmentinafive-yearfollow-upoflong-termunemployed.Scand J Public Health 1999,27:94-100.


4. MorrisJK,CookDG,ShaperAG:Lossofemploymentandmortality.BMJ (Clinical research ed 1994,308:1135-1139.

5. AlaviniaSM,BurdorfA:Unemploymentandretirementandill-health:across-sectionalanalysisacrossEuropeancountries.Int Arch Occup Environ Health 2008,82:39-45.

6. BootCR,HeijmansM,vanderGuldenJW,RijkenM:Theroleofillnessperceptionsinlaborparticipationofthechronicallyill.Int Arch Occup Environ Health 2008,82:13-20.

7. BartleyM,SackerA,ClarkeP:Employmentstatus,employmentconditions,andlimitingillness:prospectiveevidencefromtheBritishhouseholdpanelsurvey1991-2001.Journal of epidemiology and community health 2004,58:501-506.

8. SchuringM,BurdorfA,KunstAE,MackenbachJP:Theeffectofillhealthonenteringandmaintainingpaidemployment:evidenceinEuropeancountries.Journal of epidemiology and community health 2007,61:597-604.

9. SmithGD:Learningtolivewithcomplexity:ethnicity,socioeconomicposition,andhealthinBritainandtheUnitedStates.Am J Public Health 2000,90:1694-1698.

10. BosV,KunstAE,Keij-DeerenbergIM,GarssenJ,MackenbachJP:Ethnicinequalitiesinage-andcause-specificmortalityinTheNetherlands.Int J Epidemiol 2004,33:1112-1119.

11. NazrooJY:Thestructuringofethnicinequalitiesinhealth:economicposition,racialdiscrimi-nation,andracism.Am J Public Health 2003,93:277-284.

12. ChandolaT:EthnicandclassdifferencesinhealthinrelationtoBritishSouthAsians:usingthenewNationalStatisticsSocio-EconomicClassification.Soc Sci Med 2001,52:1285-1296.

13. SundquistJ:Ethnicity,socialclassandhealth.Apopulation-basedstudyontheinfluenceofsocialfactorsonself-reportedillnessin223LatinAmericanrefugees,333Finnishand126southEuropeanlabourmigrantsand841Swedishcontrols.Soc Sci Med 1995,40:777-787.

14. DunnJR,DyckI:SocialdeterminantsofhealthinCanada’simmigrantpopulation:resultsfromtheNationalPopulationHealthSurvey.Soc Sci Med 2000,51:1573-1593.

15. ElkelesT,SeifertW:Immigrantsandhealth:unemploymentandhealth-risksoflabourmi-grantsintheFederalRepublicofGermany,1984-1992.Soc Sci Med 1996,43:1035-1047.

16. FayersPM,SprangersMA:Understandingself-ratedhealth.Lancet 2002,359:187-188. 17. KuilmanM,vanDijkAP,vanderLeeG,SchrijversCTM:Resultatengezondheidsenquete

Rotterdam2003[ResultshealthquestionnaireRotterdam2003].internrapportGGDRot-terdameo2005.

18. CBS:Herkomstvanpersonen;allochtonenenmigratie[Countryoforiginofpersons;mi-grantsandmigration].CentraalBureauvoordeStatistiek,Voorburg/Heerlen2003.

19. WareJE,Jr.,SherbourneCD:TheMOS36-itemshort-formhealthsurvey(SF-36).I.Concep-tualframeworkanditemselection.Med Care 1992,30:473-483.

20. LastJM:Adictionaryofepidemiology,4thedn.OxfordUniversityPress,NewYork.2001. 21. AgyemangC,DenktasS,BruijnzeelsM,FoetsM:Validityofthesingle-itemquestionon

self-ratedhealthstatusinfirstgenerationTurkishandMoroccansversusnativeDutchintheNetherlands.Public health 2006,120:543-550.

22. WikingE,JohanssonSE,SundquistJ:Ethnicity,acculturation,andselfreportedhealth.Apopulation based study among immigrants from Poland, Turkey, and Iran in Sweden.Journal of epidemiology and community health 2004,58:574-582.

32 Chapter2

23. UnikenVenemaHP,GarretsenHF,vanderMaasPJ:Healthofmigrantsandmigranthealthpolicy,TheNetherlandsasanexample.Soc Sci Med 1995,41:809-818.

24. GraetzB:Healthconsequencesofemploymentandunemployment:longitudinalevidenceforyoungmenandwomen.Soc Sci Med 1993,36:715-724.

3The impact of ill health on exit from paid

employment in europe among older workers

vandenBergT.I.J.,SchuringM,AvendanoM,MackenbachJP,BurdorfA

Occupational and Environmental Medicine, in press

34 Chapter3

absTRacT

objective: To determine the impact of ill health on exit from paid employment in

Europeamongolderworkers.

methods: ParticipantsoftheSurveyonHealthandAgeinginEurope(SHARE)in11

Europeancountriesin2004and2006wereselectedwhenbetween50and63yearsold

andinpaidemploymentatbaseline(n=4611).Datawerecollectedonself-ratedhealth,

chronicdiseases,mobility limitations,behavioural factors (obesity,smoking,alcohol

use, andphysical activity), andworkcharacteristics.Participantswere followed for

twoyearsandclassifiedintoemployed,retired,unemployed,anddisabledatendof

follow-up.Multinomiallogisticregressionwasusedtoestimatetheeffectofdifferent

measuresofillhealthonexitfrompaidemployment.

Results: During the two-year follow-up period, 17% of employed workers quitted

paidemployment,primarilydue toearlyretirement.Controlling for individualand

workrelatedcharacteristics,poorself-perceivedhealthwasstronglyassociatedwith

exitfrompaidemploymentduetoretirement,unemployment,ordisability(ORsfrom

1.32to4.37).Adjustmentbyworkrelatedfactorsandlifestylereducedthesignificant

associations between ill health and exit from paid employment by 0 to 18.7%.Low

education,obesity,lowjobcontrol,andeffort-rewardimbalancewereassociatedwith

measuresofillhealth,butalsoriskfactorsforexitfrompaidemploymentafteradjust-

mentforillhealth.

conclusion:Poorself-perceivedhealthwasstrongestassociatedwithexitfrompaid

employment among European workers aged 50-63 years, compared to three other

measuresof illhealth.This studysuggests that theeffectsof illhealthonexit from

paid employment can be diminished by a variety of preventive measures towards

obesity,problematicalcoholuse,jobcontrol,andeffort-rewardbalance.

Ch

apte

r3

TheimpactofillhealthonexitfrompaidemploymentinEuropeamongolderworkers 35

inTRoDucTion

Inmanyindustrializedcountries thepopulation isageing,dueto increasing lifeex-

pectancyandfallingbirthrates.[1]Aratherparadoxicaldevelopmentisthat,despite

increasesinlifeexpectancy,theaveragetimepeoplespendinpaidworkhasdecreased

inmostEuropeancountries.Althoughpartofthisdecreaseisexplainedbyprolonged

educationamongyoungercohorts,amoreimportantcontributoristhehigherrateof

exit from the labourmarketatolderages.[2]Asa consequence,manycountriesare

developingpoliciestoencourageolderworkerstoremainlongerinthelabourmarket

anddelayretirement.[3]Clearly,thesuccessofthesepolicieswilldependonabetter

understandingofageingintheworkforceandtheparticularroleofhealthandwork

characteristics in continuing work or exit from the labour market. Recent evidence

suggeststhatworkcanbegoodforhealth,reversingtheharmfuleffectsoflong-term

unemploymentandprolongedsicknessabsence.[4]However,thecurrentassumption

seemsthatillnessisincompatiblewithbeinginwork.[4]

Thereisampleevidencethatillhealthmaycauseexitfromthelabourforce.[3,5-10]

Theearlieststudiesfocusedonpoorhealthasriskforunemployment,forexamplein

constructionworkers,wherebyseveralhealthproblemspredictedtheriskoflong-term

unemployment.[6]Morerecently, there isanincreasingawarenessthatamongolder

workers ill health does not only affect unemployment and disability, but may also

driveselectionoutoftheworkforceduetoearlyretirementandstayinghometotake

careofthefamily.[3,10]Forexample,inaFinnishcohortofmaleworkerswith11year

follow-up,self-assessedpoorhealthwasastrongpredictorofbothenrolmentintodis-

abilitybenefitaswellasnon-illnessbasedearlyretirement.[7]OntheEuropeanlevel

itwasshownthatillhealthwasariskfactorfortransitionsbetweenpaidemployment

and various forms of non-employment, including retirement, unemployment, and

takingcareofthehousehold.[5,9]

However,itremainsunclearwhichtypeofhealthmeasureismostpredictiveforfu-

tureexit,andhowlargethedirecteffectofpoorhealthisonfutureexit.Theaimofthe

studyistoexploretheroleofillhealthinexitfrompaidemployment.Thefollowing

researchquestionswereformulated.First,whichmeasuresofhealtharepredictivefor

exitfrompaidemployment?Second,howmuchoftheobservedassociationsbetween

illhealthandfutureexitcanbeexplainedbyworkrelatedfactorsandlifestyle?

36 Chapter3

meTHoDs

Studypopulation

ThestudypopulationconsistedofparticipantsoftheSurveyonHealthandAgeingin

Europe(SHAREstudy).SHAREisalongitudinalsurveythataimstocollectmedical,

social,andeconomicdataonthepopulationagedover50yearsin11EuropeanUnion

countries(Sweden,Denmark,TheNetherlands,Belgium,Germany,Austria,Switzer-

land, France, Italy, Spain, and Greece).[11-12] In the participating SHARE countries

theinstitutionalconditionswithrespecttosamplingweresodifferentthatauniform

samplingdesignfortheentireprojectwasnotfeasible.Differentregistriesofnational

orlocallevelwereusedthatpermittedstratificationbyage.Thesamplingdesignsvar-

iedfromsimplerandomselectionofhouseholdstocomplicatedmultistagedesigns.

ThefirstwaveofdatawascollectedbyinterviewsbetweenAprilandOctober2004.

Theoverallhouseholdresponseacrossthe11SHAREcountriesinwhichdatacollec-

tiontookplacein2004was57.4%,althoughsubstantialdifferencesamongcountries

wereobserved.[12]Theavailabledatasetfromthefirstwaveofdatacollection(SHARE

Release2.0)contains28,517participants,with12,965subjects(45%)agedbetween50

and 63 years. Individuals aged 63 years and older were excluded from the current

study,sinceitwasassumedthatworkersnormallyretiredwhentheybecame65years

oldattheendoffollow-up.Whilethisassumptioncertainlyhaslimitations,giventhe

complexity to define retirement at the individual level and the small proportion of

workersabovetheageof63yearsinthestudypopulation(about2%),itwasconsid-

eredtobe thedefinitionthatwasmostcomparableacrosscountries.For93persons

employmentstatuswasunknown,resultinginastudypopulationof12,872subjects,

ofwhich7119 (55%) subjectswithpaidemployment.After twoyears8,729 subjects

participated again in the questionnaire survey (SHARE Release 1.0), resulting in a

responseof67%.Completeinformationonemploymentstatusin2006wasavailable

for8,568 subjects.For the longitudinalanalysisof the influenceof illhealthonexit

ofthelabourmarket,acohortwasavailableof4,611subjectswithpaidemployment

in 2004 and complete information on individual and work related characteristics at

baselineandworkstatusatfollow-upin2006.

Labourforceparticipation

Theoutcomeof thisstudyisworkstatus,whichwasbasedonself-reportedcurrent

economic status that best described respondent’s situation based on four mutually

exclusivecategories:paidwork,retired,unemployed,disabled.Thedefinitionofbe-

ingemployedinSHAREencompassesallindividualswhodeclaredtohavedoneany

Ch

apte

r3


kind of formal paid work in the last four weeks, including self-employed work for

familybusiness.Unemployedwerethosewhowerelaidofffromtheirlastjobbefore

beingabletobenefitfromnormalpensionbenefits,andthereforewereforcedtospend

some time in unemployment before effectively being retired. Sickness or disability

insurance applied to people who exited the labour force for reasons of recognized

health problems.[11] The category of disabled participants predominantly includes

persons whose health problems at work were an eligibility criterion for receiving a

disabilitypension.Total exit from theworkforcewasdefinedasexit either through

earlyretirement,unemploymentordisability.

Healthmeasurements

TheEuropeanversionofself-perceivedhealth,a5-pointscalequestionrangingbetween

verygoodto(very)bad,wasusedtodefinepoorhealth(lessthangood).Thisfrequently

usedquestionhasbeenshowntobeagoodindicatorofgeneralphysicalandmental

health.[13-14]Asecondhealthmeasurewashavingatleastoneofthefollowingchronic

diseases diagnosed by a doctor during lifetime; heart disease, stroke, diabetes, lung

disease,asthma,arthritisorrheuma,andosteoporosis.Functionallimitations,reflect-

ingtheabilityofindividualstoperformnormallyinsociety,werecharacterizedwith

twodichotomousmeasuresofhealth.Thefirstmeasureofinterest,mobilityproblems,

reflectslimitationswithmobility,armorfinemotorfunctions.Mobilityproblemswere

definedasoneormoreaffirmativeanswersonalistof10mobilityproblems,suchas

walking100metersandreachingorextendingarmsaboveshoulderlevel.Thesecond

measure,instrumentallimitations,waspositiveforsubjectswithoneormoreofthe13

instrumentalactivitiesofdailylife,suchaspreparingmealsandmakingphonecalls.

Individualcharacteristics

The highest education successfully completed was coded according to the 1997

International Standard Classification of Education (ISCED-97) and categorized into

low (pre-primary, primary and lower secondary education), intermediate (upper

secondaryeducation)andhigh (post secondaryeducation).Bodymass index (BMI)

wascalculatedbydividingbodyweightinkilogrambythesquareofbodyheightin

meters. BMI was recoded into normal (<25 kg/m2), overweight (>=25 and <30 kg/

m2), or obese (>=30 kg/m2). Marital status was used to categorize individuals into

thosewhowere livingwithaspouseorapartner in thesamehousehold (reference

category)andthoselivingalone.Smokersweresubjectswhowerecurrentlysmoking;

allotherswerecategorizedasnon-smokers.Problematicalcoholusewasdefinedby

analcoholconsumptionoftwoormoreglassesofalcoholicbeverageatleast5daysa

38 Chapter3

weekinthelastsixmonths.Physicalactivitywasmeasuredwithsinglequestionson

regularparticipationinmoderateactivitiesandvigorousactivities,bothona4-point

scale ranging from ‘more than once a week’ to ‘hardly ever, or never’. Those who

reportedlessthanonceaweekmoderateorvigorousactivitywereconsideredtolack

inleisure-timephysicalactivity.[12]

Workrelatedcharacteristics

Workrelatedcharacteristicswereassessedbyashortbatteryofitemsderivedfrom(i)the

JobContentQuestionnairemeasuringthedemand-controlmodel[15]and(ii)theeffort-

rewardimbalancemodelquestionnaire[16].Allitemswereonafourpointscaleranging

from1‘stronglyagree’to4‘stronglydisagree’.Singleitemmeasuredhightimepressure

(“I’munderconstanttimepressureduetoaheavyworkload.”)Lackofjobcontrolwas

measuredby thesumscoreof two items (“Ihavevery little freedomtodecidehowI

domywork”,“Ihaveanopportunitytodevelopnewskills”).Country-specificmedian

valueswereusedtodefinethepresenceofhightimepressure,andlackofjobcontrol.

Effort-rewardimbalancewasmeasuredby2itemson‘effort’(‘physicallydemand-

ing’and‘timepressure’)and5itemson‘reward’(‘receiveadequatesupport’,‘receive

recognition’, ‘adequate salary’, ‘job promotion prospects’, ‘job security’). ‘Effort-re-

wardimbalance’wasdefinedbytheratioofthesumscoreofthe‘effort’itemsandthe

sumscoreofthe‘reward’items,adjustedforthenumberofitems.[17]Effort-reward

imbalancewasdefinedasascorewithintheuppertertileofthisratiopercountry.[17]

Ahighphysicalworkdemandwasmeasuredwithoneitem(“Myjobisphysically

demanding.”).Country-specificmedianvaluewasusedtodefinethepresenceofhigh

physicalworkdemand.

Statisticalanalysis

Logistic regression was used to evaluate cross-sectional associations at baseline be-

tween four measures of ill health as dependent variables and individual and work

characteristicsasindependentvariables,adjustingforcountry.

Risk factors for exit from paid employment during the two year follow-up were

evaluatedbymeansofamultinomiallogisticregressionanalysis.Thestudypopula-

tion consisted of subjects with paid employment at baseline and odds ratios were

calculatedforthelikelihoodoftransitiontoeverystateofnon-participation,i.e.early

retirement,unemployment,anddisabilityduringthe2yearfollow-up.Theresultsfor

homemakerswerenotshownasthisgroupwashighlydominatedbyfemalegender,

butsubjectswhoexitedpaidemploymentthroughbecomingahomeworkerremained

inthesample.Thefirststepintheanalysiswastoestablishunivariateassociationsbe-

Ch

apte

r3


tweenthedependentvariableworkstatus,andhealthmeasures,socio-demographic

factors, lifestyle factors, and work characteristics as independent factors, including

country as fixed effect. In the second step multivariate analyses were conducted to

modelemploymentstatusat theendof follow-upasa functionof fourmeasuresof

health. For the initial selection of potential covariates for the multivariate model,

univariate associations with a significant level of p<0.05 were considered. For each

independent variable measure of health, we calculated odds ratios for dependent

variable exit for work, adjusted for age, sex and education (reference model) and

furtheradjustedforlifestylefactorsandworkrelatedcharacteristicsseparately,andin

combination.Foreachregressionmodelthepercentagechangeinoddsratioofeach

Table 3.1 Individualcharacteristics,lifestylefactors,healthstatus,andworkcharacteristicsamong4611employedpersonsaged50-63yearsoldin11EuropeancountriesduringthefirstwaveoftheSurveyonHealthandAgeinginEurope(SHARE)

employed(n=4611)

IndividualcharacteristicsFemaleAge

50-54yr55-59yr60-63yr

EducationlevelLowIntermediateHigh

Livingwithoutpartner

45%(2088)

48%(2224)40%(1826)12%(561)

31%(1443)33%(1513)36%(1655)20%(937)

LifestylefactorsBMI

<25kg/m2

25-30kg/m2

≥30kg/m2

CurrentsmokerProblematicalcoholuseLackofleisure-timephysicalactivity

44%(2011)41%(1910)15%(690)27%(1252)14%(664)56%(2561)

Work-relatedfactorsHightimepressureatwork(1/0)Highphysicalworkdemands(1/0)Lackofjobcontrol(1/0)Effort-rewardimbalanceatwork(1/0)

56%(2567)46%(2138)57%(2622)33%(1531)

PerceivedhealthVerygoodGoodFair(Very)bad

32%(1475)51%(2343)15%(708)2%(85)

Chronicdisease(1/0) 25%(1130)

Mobilityproblems(1/0) 28%(1287)

Instrumentallimitationsindailyactivities(1/0) 4%(170)

40 Chapter3

pathwayofexitwascalculated(100x[ORreferencemodel–OR+explanatoryfactors]/[ORreferencemodel-1].

[18]Oneofthemainadvantagesofthismethodisthatitcanbeusedtoestimatedirect

and indirectcontributionsofexplanatory factors.One limitation is that thepercent-

age change can be similar for different absolute changes in odds ratios. However,

allcontributionswerecalculatedrelativelytothesameoddsratios,whichwerealso

presented.Therefore,webelievethatthislimitationhasalimitedeffectonourresults.

Population Attributable Fractions were calculated for significant determinants of

exit from paid employment, using the formula PAF= Pe (OR-1)/(1+PE(OR-1))[19],

wherebyPerepresentstheprevalenceofexposureinthestudypopulation.

All statistical models were based on the number of persons with complete data

available.ThestatisticalanalyseswerecarriedoutwithSPSSversion15.0.[20]

ResuLTs

About 17% of the employed workers reported less than good health (table 3.1). In-

terrelationsof the fourhealthmeasuresweremoderate,withSpearmancorrelations

varyingfrom0.06to0.33.Intotal,55%ofthesubjectswithapoorhealthhadachronic

disease, 57% mobility problems, and 9% instrumental limitations. Chronic diseases

withhighestprevalenceweredepression(17.7%n=814),arthritis/osteoporosis(12.3%

n=565),andrespiratorydiseases(5.7%n=265)(datanotshown).About61%ofsubjects

withachronicdiseaseperceivedtheirhealthasgood.

Table 3.2 Exitfrompaidemploymentamong4611participantsaged50-63yearsoldin11Europeancountriesduringtwoyearsfollow-upduringthefirsttwowavesoftheSurveyonHealthandAgeinginEurope(SHARE)

employed in 2004 Labour market position 2006

country n employed Retired unemployed Disabled Homemaker

Sweden 720 88.5%(637) 5.8%(42) 1.7%(12) 3.3%(24) 0.7%(5)

Denmark 409 81.2%(332) 13,0%(53) 3.9%(16) 1.7%(7) 0.2%(1)

TheNetherlands 484 79.8%(386) 12.6%(61) 2.1%(10) 2.3%(11) 3.3%(16)

Belgium 617 84.8%(523) 8.8%(54) 0.8%(5) 3.4%(21) 2.3%(14)

Germany 411 76.6%(315) 12.9%(53) 7.1%(29) 1.0%(4) 2.4%(10)

Austria 209 71.8%(150) 23.0%(48) 2.9%(6) 1.4%(3) 1.0%(2)

Switzerland 238 88.2%(210) 5.5%(13) 2.5%(6) 0.4%(1) 3.4%(8)

France 490 80.4%(394) 12.7%(62) 3.9%(19) 2.2%(11) 0.8%(4)

Italy 276 72.1%(199) 19.2%(53) 4.3%(12) 0.4%(1) 4.0%(11)

Spain 228 80.7%(184) 7.0%(16) 4.8%(11) 2.6%(6) 4.8%(11)

Greece 529 92.1%(487) 5.1%(27) 0.2%(1) 0.2%(1) 2.5%(13)

Total 4611 82.8%(3817) 10.5%(482) 2.8%(127) 2.0%(90) 2.1%(95)

Ch

apte

r3


Duringthetwoyearfollow-upperiod17%(n=794)ofemployedworkersexitedthe

workforce,primarilyduetoretirement(11%)(table3.2).Considerabledifferences in

prevalenceofexitfrompaidemploymentandpathwaysofexitwerefoundbetween

countries.At baseline all four measurements of ill health were associated with low

education (OR’s 1.27-2.00) (table 3.3). Obesity (ORs 1.53-2.74) and lack of physical

activity in leisure time (ORs 1.24-1.87) were associated with most health outcomes.

Effort-rewardimbalanceatworkwasincreasedforallfourhealthoutcomes(ORs1.25-

1.64).Hightimepressure(ORs0.98-1.09),andproblematicalcoholuse(ORs0.87-1.12)

werenotassociatedwithanyofthehealthoutcomes.

Table 3.3 Cross-sectionalmultivariateassociationsbetweenindividualcharacteristics,lifestyleandworkcharacteristics,anddifferenthealthoutcomesamong4611employedpersonsaged50-63yearsoldin11Europeancountries,duringthefirstwaveoftheSurveyonHealthandAgeinginEurope(SHARE)

Less than good perceived health (n=793)

chronic disease(n=1130)

mobility problems(n=1287)

instrumental limitations in daily activities(n=170)

oR 95%ci oR 95%ci oR 95% ci oR 95% ci

Age50-54yr55-59yr60-63yr

11.24*(1.04-1.47)1.35*(1.06-1.73)

11.55*(1.33-1.79)1.72*(1.39-2.13)

11.21*(1.05-1.40)1.43*(1.16-1.77)

11.33(0.95-1.85)1.40(0.86-2.29)


11.55*(1.26-1.90)2.00*(1.62-2.47)

11.21*(1.02-1.44)1.30*(1.08-1.58)

11.22*(1.03-1.44)1.27*(1.06-1.51)

11.34(0.90-2.03)1.61*(1.07-2.43)

Female 1.18(1.00-1.39) 1.34*(1.16-1.54) 2.05*(1.78-2.37) 2.44*(1.74-3.43)

Withoutpartner 0.99(0.81-1.21) 1.24*(1.05-1.46) 1.05(0.89-1.24) 1.17(0.81-1.68)

BMI<25kg/m2

25-30kg/m2

≥30kg/m2

11.03(0.86-1.24)2.01*(1.62-2.50)

11.18*(1.01-1.39)1.88*(1.55-2.29)

11.49*(1.28-1.74)2.74*(2.25-3.32)

11.13(0.79-1.61)1.53(0.99-2.35)

Currentsmoker 1.18(0.98-1.40) 0.93(0.80-1.09) 1.15(0.99-1.34) 1.55*(1.11-2.16)

Problematicalcoholuse

0.98(0.78-1.24) 1.12(0.92-1.37) 0.87(0.71-1.07) 0.98(0.60-1.60)

Lackofleisure-timephysicalactivity

1.46*(1.23-1.72) 1.24*(1.08-1.43) 1.57*(1.37-1.81) 1.87*(1.33-2.63)

Hightimepressureatwork

1.04(0.87-1.24) 0.98(0.84-1.15) 1.06(0.91-1.24) 1.09(0.76-1.56)

Highphysicalworkdemands

1.14(0.94-1.38) 1.09(0.92-1.28) 1.26*(1.08-1.48) 1.07(0.73-1.56)

Lackofjobcontrol 1.26*(1.06-1.49) 1.15*(1.00-1.33) 1.04(0.90-1.19) 1.23(0.88-1.71)

Effort-rewardimbalanceatwork

1.64*(1.33-2.01) 1.25*(1.04-1.50) 1.39*(1.17-1.66) 1.55*(1.03-2.33)

*P<0.05,ORoddsratio,CIconfidenceinterval

42 Chapter3

Table 3.4 Univariateassociationsbetweenhealth,individualcharacteristics,lifestyle,andworkcharacteristics,andtransitionsintounemployment,retirement,anddisabilityamong4611initiallyemployedsubjectsaged50-63yearsoldduringtwoyearsfollow-upintheSurveyonHealthandAgeinginEurope(SHARE)(stayinginpaidemploymentasreferencecategory)

unemployed (n=127)

Retired(n=482)

Disabled(n=90)

Total exit (n=699)


Lessthangoodperceivedhealth(1/0)

2.49*(1.70-3.66) 1.50*(1.19-1.90) 5.04*(3.28-7.74) 2.08*(1.72-2.51)

Chronicdisease(1/0)

1.62*(1.10-2.37) 1.74*(1.42-2.14) 3.00*(1.96-4.59) 2.00*(1.68-2.37)

Mobilityproblems(1/0)

1.29(0.88-1.90) 1.37*(1.11-1.68) 3.44*(2.26-5.26) 1.56*(1.31-1.85)

Instrumentallimitationsindailyactivities(1/0)

1.69(0.76-3.73) 1.25(0.78-2.02) 3.52*(1.82-6.83) 1.73*(1.19-2.50)

Age50-54yr55-59yr60-63yr

11.72*(1.17-2.53)2.51*(1.42-4.43)

18.08*(5.91-11.04)33.30*(23.36-47.46)

12.36*(1.50-3.72)1.31(0.56-3.07)

13.96*(3.21-4.88)8.75*(6.83-11.21)


11.87*(1.17-3.00)2.80*(1.69-4.64)

11.12(0.89-1.43)1.65*(1.29-2.12)

11.56(0.89-2.72)2.20*(1.28-3.77)

11.39*(1.14-1.71)1.68*(1.37-2.05)

Female 1.16(0.81-1.66) 0.83(0.68-1.01) 0.95(0.62-1.45) 0.92(0.78-1.08)

Withoutpartner 1.36(0.90-2.05) 0.83(0.64-1.06) 0.78(0.45-1.37) 0.92(0.75-1.13)

BMI<25kg/m2

25-30kg/m2

≥30kg/m2

11.00(0.66-1.51)1.92*(1.21-3.07)

11.23*(1.00-1.52)1.40*(1.05-1.84)

10.97(0.60-1.56)1.71(0.98-2.97)

11.11(0.93-1.33)1.51*(1.20-1.89)

Currentsmoker 1.21(0.82-1.80) 0.87(0.70-1.09) 1.46(0.93-2.30) 0.97(0.81-1.17)

Problematicalcoholuse

1.04(0.63-1.72) 1.37*(1.07-1.76) 1.65(0.95-2.87) 1.57*(1.27-1.94)

Lackofleisure-timephysicalactivity

1.29(0.89-1.86) 1.24*(1.01-1.51) 0.98(0.64-1.49) 1.19*(1.01-1.44)

Hightimepressureatwork

0.72(0.50-1.04) 1.02(0.83-1.25) 1.23(0.80-1.90) 1.00(0.85-1.18)

Highphysicalworkdemands

1.04(0.73-1.50) 1.17(0.96-1.42) 1.57*(1.03-2.40) 1.19*(1.01-1.40)

Lackofjobcontrol 1.59*(1.07-2.37) 1.23*(1.00-1.51) 2.68*(1.59-4.54) 1.62*(1.37-1.91)

Effort-rewardimbalanceatwork

1.51*(1.05-2.16) 1.09(0.89-1.34) 1.62*(1.06-2.48) 1.22*(1.03-1.44)

*P<0.05,ORoddsratio,CIconfidenceinterval

Ch

apte

r3


Table 3.5 Multivariateassociationsbetween4differenthealthmeasuresandtransitionsintounemployment,retirement,anddisabilityamong4611initiallyemployedsubjectsaged50-63yearsoldin11Europeancountriesduringtwoyearsoffollow-upintheSurveyonHealthandAgeinginEurope(SHARE)(stayinginpaidemploymentasreferencecategory)

unemployed (n=127)

Retired(n=482)

Disabled(n=90)

Total exit (n=699)


Lessthangoodperceivedhealth(1/0)1

2.16*(1.47-3.19) 1.38*(1.07-1.79) 4.59*(2.97-7.10) 1.95*(1.59-2.39)

Adjustedforlifestylefactors 2.04*(1.37-3.02) 1.35*(1.04-1.76) 4.52*(2.91-7.02) 1.88*(1.53-2.31)

change 10.3% 7.9% 2.0% 7.4%

Adjustedforworkcharacteristics

2.09*(1.41-3.10) 1.35*(1.04-1.76) 4.36*(2.80-6.77) 1.87*(1.52-2.29)

change 6.0% 7.9% 6.4% 8.4%

Adjustedforlifestyle+work 1.96*(1.32-2.92) 1.32*(1.01-1.72) 4.24*(2.71-6.62) 1.78*(1.45-2.20)

change 14.7% 15.8% 6.1% 17.9%

Chronicdisease(1/0)1 1.42(0.96-2.09) 1.38*(1.10-1.73) 2.74*(1.78-4.22) 1.74*(1.45-2.09)

Adjustedforlifestylefactors 1.33(0.90-1.97) 1.36*(1.08-1.71) 2.71*(1.76-4.19) 1.67*(1.39-2.02)

change 21.4% 5.3% 1.7% 9.5%


1.39(0.94-2.05) 1.36*(1.08-1.71) 2.66*(1.73-4.11) 1.69*(1.41-2.04)

change 7.1% 5.3% 4.6% 6.8%

Adjustedforlifestyle+work 1.30(0.88-1.93) 1.28*(1.01-1.62) 2.62*(1.69-4.07) 1.63*(1.35-1.96)

change 28.6% 13.2% 4.0% 14.9%

Mobilityproblems(1/0)1 1.15(0.78-1.71) 1.20(0.95-1.51) 3.32*(2.15-5.12) 1.46*(1.21-1.75)

Adjustedforlifestylefactors 1.07(0.71-1.59) 1.17(0.93-1.48) 3.35*(2.15-5.22) 1.40*(1.16-1.70)

change 53.3% 15.0% -1.3% 13.0%


1.12(0.75-1.66) 1.18(0.94-1.49) 3.22*(2.08-4.99) 1.43*(1.19-1.72)

change 20.0% 10.0% 4.3% 6.5%

Adjustedforlifestyle+work 1.03(0.69-1.54) 1.15(0.91-1.46) 3.22*(2.06-5.03) 1.37*(1.13-1.65)

change 80.0% 25.0% 4.3% 19.6%

Instrumentallimitationsindailyactivities(1/0)1

1.43(0.64-3.19) 1.06(0.63-1.78) 3.19*(1.62-6.25) 1.55*(1.04-2.30)

Adjustedforlifestylefactors 1.35(0.60-3.02) 1.03(0.61-1.74) 3.19*(1.62-6.29) 1.48(0.99-2.20)

change 18.6% 50.0% 0% 12.7%


1.39(0.62-3.11) 1.03(0.61-1.73) 3.00*(1.52-5.93) 1.47(0.99-2.20)

change 9.3% 50.0% 8.7% 14.6%

Adjustedforlifestyle+work 1.31(0.58-2.93) 0.99(0.59-1.69) 2.98*(1.50-5.91) 1.40(0.93-2.08)

change 18.6% 133.3% 18.7% 28.2%

1Adjustedforindividualcharacteristicsage,sexandeducationallevel

44 Chapter3

Table 3.4 shows that self-perceived poor health was a risk factor for transition to

unemployment(OR2.49),retirement(OR1.50),andworkdisability(OR5.04).Allfour

healthmeasureswereassociatedwithanyexitfromwork(ORs1.56-2.08).Alowedu-

cationallevelwasariskfactorforallthreepathwaysofnon-participation(ORs1.65-

2.80).Lifestylefactorsseemimportantforexittroughretirement.Exceptsmoking,all

lifestylefactorswereassociatedwithexitfrompaidemploymentthroughretirement

(ORs 1.23-1.40). Among work-related factors lack of job control showed increased

risksforallthreepathwaysofexit(ORs1.23-2.68).Highphysicalworkdemandswere

ariskfactoronlyforbecomingdisabled(OR1.57).Workrelatedfactorsseemedmost

importantforexitthroughworkdisabilityincomparisontootherpathways.

Table 3.5 shows that the observed associations between different measures of ill

healthandtransitionstonon-participation,afteradjustmentfor lifestyle factorsand

work characteristics. Significant odds ratios between ill health and exit from paid

employmentdecreasedby0%to10%afteradjustmentforlifestylefactors,4%to9%

afteradjustmentforworkfactors,and4%to19%afteradjustmentforlifestylefactors

andworkcharacteristicssimultaneously.Adjustmentwithlifestylefactorsandwork

related characteristics had a smaller influence on the association between ill health

andworkdisabilityleastcomparedtotheotherpathwaysofexitfrompaidemploy-

ment. Lifestyle factors and work related characteristics had a comparable influence

on the relation between ill health and exit from paid employment. In the fully ad-

justed models for each of the four health measures the lifestyle factors obesity and

problematicalcoholuseremainedsignificantinatleastoneofthemodels.Regarding

workrelatedcharacteristics,lackofjobcontrolandeffort-rewardimbalanceatwork

remainedsignificantafterfulladjustmentinatleastoneofthefourmodels.

Thepopulationattributablefractionsofaless-than-goodself-perceivedhealthfortran-

sitionintounemployment,retirement,anddisabledwere27%,9%,and61%,respectively.

Discussion

Duringatwoyearfollow-up,17%ofworkersemployedatbaselineleftpaidemploy-

ment,primarilyduetoearlyretirement.Controllingforindividualandworkrelated

characteristics,poorself-perceivedhealthwasstronglyassociatedwithexitfrompaid

employmentduetoretirement,unemployment,ordisability(OR’sfrom1.32to4.24).

In order of decreasing importance, chronic diseases, mobility problems and instru-

mentallimitationsalsoinfluencedexitfrompaidemployment,mostnotablythrough

disability.Significantassociationsbetweenillhealthandexitfrompaidemployment

changed0to19%afteradjustmentforlifestyleandworkcharacteristics.

Ch

apte

r3


Some limitations must be taken into account in this study. First, although the at-

trition rate between baseline and follow-up was high (68%)[21], attrition was not

influencedby sex,but lower inhighagegroups.Except inGreeceandSwitzerland

therewasnorelationshipbetweenworkstatusandattrition,andthosereportinggood

orbetterhealthatbaselinehadahigherpropensitytoparticipateatthefollow-up.

Secondly,therearelargevariationsbetweenEuropeancountriesintheassociation

betweenillhealthandvariousformsofexit frompaidemployment.[5]Thesevaria-

tions may reflect differences between countries in institutional arrangements (e.g.

availability of disability benefit schemes for those with health problems), or other

factors (e.g. more or less selectivity of unemployment dependent on over-all levels

of unemployment). All analyses were therefore adjusted for country. Due to small

numbers,country-specificorregion-specificanalyseswerenotfeasible.

Third, all variables were based on self-reported data, which could have caused

reportingbias.Theproblemwithusingself-reportedhealthinanempiricalanalysisof

labourforceparticipationisthatitmaybeanendogenousexplanatoryvariable.[13,22-

23]Accordingtothejustificationhypothesisindividualsjustifytheirnon-participation

byclaimingthattheyareinillhealth.Subjectswithintentionsatbaselinetoquitpaid

employmentinthenearfuturemayalsohavebeenmorepronetoreporthighwork

demandsoralessbeneficialeffort-rewardbalanceinordertojustifytheirfutureexit

frompaidemployment.[17]

Fourth,thecurrentstudyusedafollow-upperiodoftwoyearsand,therefore,had

limiteddiscriminatorypoweranddoesnotgive insight in long-termeffectsofpoor

healthonexitfrompaidemploymentortherelevanttimewindowsfortheseeffects.

AEuropeanstudyshowed thatpoorhealthhad thestrongesteffectson leaving the

workforce in theyearbefore the transition.[8]Thus, it isexpected that the reported

influenceofillhealthonexitfrompaidemploymentisafairreflectionoftheeffectsof

illhealthonworkparticipation.

Severalstudieshaveanalyzedtheeffectsofhealthonexitfrompaidemploymentof

olderworkers.[3,5-7,9,24-27]Theresultsofthisstudysupporttheselectionhypoth-

esis,wherebypeoplewithpoorhealtharemorelikelytoquitpaidemployment.[28]

Theinfluenceoftypeofhealthmeasurediffersbyrouteofexit,butanoveralleffecton

totalexitwasconsistentlypresentforallmeasuresofillhealth.

Therelationbetweenpoorhealthandexitfrompaidemploymentmaybeexplained

byamismatchbetweenanindividual’scapacitiesandtherequirementsofthejob.[2]

Functional limitations might therefore be more important than self-perceived poor

healthforfuturelossofpaidemployment.However,theanalysesshowedthatapoor

self-perceived health was a stronger predictor for pathways of exit than functional

limitations,expressedbyeithermobilityproblemsorinstrumentallimitationsindaily

activities.Anexplanationcouldbethatself-perceivedhealthincludesmentalhealth

46 Chapter3

as well, whereas functional limitations concern primarily physical health. The high

prevalence of depression in the cohort may have contributed to the association be-

tweenself-perceivedhealthandfutureexitfrompaidemployment.

Theanalysesalsoshowedthathavingeverbeingdiagnosedwithachronicdisease

played a less profound role in exit from paid employment. This may be explained

by the fact that people diagnosed with these chronic conditions who remained in

paidemploymentareaselectionofthefittestsurvivors[29],whilethosewhoalready

leftpaidemploymentduetothesediseaseshavenotbeenincludedinoursampleas

theyhadalreadyleftpaidemploymentbeforethebaselineinvestigation.Analyseson

theroleofonsetofdiseaseduring the follow-upperiodwasnot feasibleasonly12

subjects reported that theonsetof their chronicdiseasehadbeendiagnosedduring

thefollow-upperiod.

The direct influence of ill health on exit from paid employment had odds ratios

varyingbetween1.37and5.04.Thecorrespondingpopulationattributablefractionsof

a less-than-goodself-perceivedhealthfortransitionintounemployment,retirement,

and disabled were 27%, 9%, and 61%, respectively. Under the assumption that the

observed associations represent a causal process, these associations and population

attributablefractionsindicatethatagoodhealthisanimportantfactorinmaintaining

paid employment. Based on this finding interventions aimed at prevention of exit

from paid employment should prevent or minimize ill health. Given the strong as-

sociationsatbaselinebetweenobesityandlackofleisuretimephysicalactivitywith

severalmeasuresof illhealth,healthpromotion interventionsshouldbeconsidered

that increase physical activity and support a healthy diet.[30-31] The consistent as-

sociationsatbaselinebetweenlackof jobcontrol,highphysicalworkdemands,and

effort-rewardimbalancewithseveralmeasuresofillhealth,outlinetheimportanceof

improvementofworkingconditionsandworkorganizationaswell.

Weobservedthatadjustmentwithlifestylefactorsandworkrelatedcharacteristics

showedreasonablechangesinhealthrelatedexitfrompaidemployment.Thechange

wasonlyimportantforstatisticallysignificantassociationsbecauseasmalldifference

inoddsratiocouldotherwiseresultinahighproportionofchange.Theinfluenceof

lifestylefactorsandworkcharacteristicsontheimpactofillhealthonlabourforceexit

pointsattheimportanceofprovidingworkerswithhealthproblemswithpossibilities

thatwillenablethemtocontinueworking,forexamplebyempoweringworkerswith

chronicdiseases.[32]

In the fullyadjustedmodelobesity,problematicalcoholuse, low job control, and

effort-rewardimbalanceremainedstatisticallysignificantforatleastoneofthepath-

waysofexitinatleastoneofthemodels.

Differentstudiessupporttheassociationbetweenunhealthylifestyles,suchaslack

ofphysicalactivity,obesity,andproblematicalcoholuse,andexitfrompaidemploy-

Ch

apte

r3


ment.[5, 27, 33-35] In the fully adjusted multinominal models problematic alcohol

usewasconsistentlyassociatedwithenteringworkdisabilitywithORsvaryingfrom

1.84-1.88.Ithasbeensuggestedthatthismaybeexplainedbyproblemswithworking

times,workoutput,concentration,occupationalsafety,andcooperation,irrespective

ofhealthstatus.[8] In themultivariatemodelobesitywasassociatedwithbecoming

unemployed,(OR1.67).ThisisinagreementwithaFrenchstudythatreportedobesity

asariskfactorforunemploymentaftercontrollingforself-reportedhealth.[34]

Smokingwasnotassociatedwithearlyexit.Earlierstudieshaveshowncontradictory

results,withsignificantassociationsforsmoking[8,36-39]aswellasnon-significant

associationswithdifferentformsofexitfrompaidemployment.[34,40-41]

In the fully adjusted models lack of job control remained a significant predictor

forexitthroughretirementanddisability,whereaseffort-rewardimbalancepredicted

unemployment. Several studies have corroborated the observed direct influence of

strenuousworkingconditionsonexit frompaidemployment. [7, 27, 36, 42-43] Ina

cross-sectionalanalysisoftheSHAREsurveyatbaseline,ahighimbalancebetweenef-

fortsandrewardswasalsoassociatedwithintendedearlyretirementaftercontrolling

forpoorself-perceivedhealth[17].Hence,preventivemeasurestowardsproblematic

alcoholuse,obesity,jobcontrol,andeffort-rewardimbalancewillcontributetodimin-

ishtheoccurrenceofhealthrelatedearlyexitfrompaidemployment.

This study only focused on exit from paid employment, but poor health could

haveanadditionalimpactintermsofchangeofjobsandstalledcareers.TheHealth

andRetirementSurvey[25]showedthatworkersafter theonsetofhealthproblems

oftenchangedjobswithinseveralyears.Thismightalsobetruefortheonsetofpoor

health inearlierphasesof thecareer (youngerworkers).Poorhealthmayalsohave

adverseeffectsonperformanceatwork,asobservedintheinfluenceofpoorhealthon

sicknessabsence[44]andproductivitylossatwork.[45-46]Durationofemployment

contractcouldbeofinfluenceonsustainingpaidemployment.However,only7%of

thesubjectswithpaidemploymenthadatemporaryemploymentcontract,andthus

thisparametercouldnotbeevaluatedinthisstudy.

ThehealthstatusofolderEuropeanworkershasamajorinfluenceonthelikelihood

of sustaining paid employment. Self-perceived poor health and, to a lesser extent,

havingachronicdisease,perceivingmobilityproblemsandlimitationsseempredic-

tiveforfutureworkparticipation.Thereisconsistentevidencethatsocialinequalities

inhealthdependonwork related factorsaswellas lifestylebehaviours.[47-48]The

resultsofthisstudysuggestthatlabourmarketparticipationofolderworkerswithill

healthmaybesustainedbyinterventionsthatpromoteahealthierlifeandhealthier

workingconditions.Asexitfrompaidemploymentisoftenirreversibleatolderage,

prevention of work loss by improving worker’s health or improving ill workers’

work circumstances and lifestyles should be a key priority. Important entry-points

48 Chapter3

for policy could be lifestyle interventions, improvements of job control and effort-

reward balance, and social policies to encourage employment among older persons

withhealthproblems.

Acknowledgements

This paper uses data from release 2 of SHARE 2004 and release 1 of SHARE 2006.

The SHARE data collection has been primarily funded by the European Commis-

sion through the 5th framework programme (project QLK6-CT-2001-00360 in the

thematicprogrammeQualityofLife).AdditionalfundingcamefromtheUSNational

InstituteonAgeing(U01AG09740-13S2,P01AG005842,P01AG08291,P30AG12815,

Y1-AG-4553-01andOGHA04-064).DatacollectioninAustria(throughtheAustrian

ScienceFoundation,FWF),Belgium (through theBelgianSciencePolicyOffice)and

Switzerland (through BBW/OFES/UFES) was nationally funded. The SHARE data

collectioninIsraelwasfundedbytheUSNationalInstituteonAging(R21AG025169),

bytheGerman-IsraeliFoundationforScientificResearchandDevelopment(G.I.F.),and

bytheNationalInsuranceInstituteofIsrael.FurthersupportbytheEuropeanCom-

missionthroughthe6thframeworkprogram(projectsSHARE-I3,RII-CT-2006-062193,

andCOMPARE,CIT5-CT-2005-028857)isgratefullyacknowledged.Formethodologi-

cal details see Boersch-Supan and Juerges (2005). Mauricio Avendano is supported

by a grant from the Netherlands Organisation for Scientific Research (NWO, grant

no. 451-07-001) and a Bell Fellowship from the Harvard Center for Population and

Developmentstudies.

Ch

apte

r3


ReFeRences

1. IlmarinenJE:Agingworkers.Occupational and environmental medicine 2001,58:546-552. 2. StattinM:Retirementongroundsofillhealth.Occupational and environmental medicine 2005,

62:135-140. 3. CaiL,KalbG:Healthstatusandlabourforceparticipation:evidencefromAustralia.Health

economics 2006,15:241-261. 4. BlackC:Workingforahealthiertomorrow.2008. 5. AlaviniaSM,BurdorfA:Unemploymentandretirementandill-health:across-sectional

analysisacrossEuropeancountries.International archives of occupational and environmental health 2008,82:39-45.

6. BartleyM:Unemploymentandillhealth:understandingtherelationship.Journal of epidemiol-ogy and community health 1994,48:333-337.

7. KarpansaloM,ManninenP,KauhanenJ,LakkaTA,SalonenJT:Perceivedhealthasapredic-torofearlyretirement.Scandinavian journal of work, environment & health 2004,30:287-292.

8. Leino-ArjasP,LiiraJ,MutanenP,MalmivaaraA,MatikainenE:Predictorsandconsequencesof unemployment among construction workers: prospective cohort study. BMJ (Clinical research ed 1999,319:600-605.

9. SchuringM,BurdorfL,KunstA,MackenbachJ:Theeffectsofillhealthonenteringandmaintainingpaidemployment:evidenceinEuropeancountries.Journal of epidemiology and community health 2007,61:597-604.

10. ThomasC,BenzevalM,StansfeldSA:Employmenttransitionsandmentalhealth:ananalysisfromtheBritishhouseholdpanelsurvey.Journal of epidemiology and community health 2005,59:243-249.

11. Borsch-SupanA,JurgesH,MackenbachJ,SiegristJ,WeberG:Health,AgeingandRetirementin Europe. First results from the survey of Healht, Ageing and Retirement in Europe.Firstresults fromthesurveyofHealth,AgeingandRetirement inEurope. .Mannheim:MannheimResearchInstitutefortheEconomicofAgeing(MEA);2005.

12. Borsch-SupanA,KartsenH,JurgesH:Anewcomprehensiveandinternationalviewonage-ing:introducingthe‘SurveyofHealth,AgeingandRetirementinEurope’..Eur J Ageing 2005,2:245-253.

13. DwyerDS,MitchellOS:Healthproblemsasdeterminantsofretirement:areself-ratedmea-suresendogenous?Journal of health economics 1999,18:173-193.

14. IdlerEL,BenyaminiY:Self-ratedhealthandmortality:areviewoftwenty-sevencommunitystudies.Journal of health and social behavior 1997,38:21-37.

15. KarasekR,BrissonC,KawakamiN,HoutmanI,BongersP,AmickB:TheJobContentQues-tionnaire(JCQ):aninstrumentforinternationallycomparativeassessmentsofpsychoso-cialjobcharacteristics.Journal of occupational health psychology 1998,3:322-355.

16. SiegristJ:Adversehealtheffectsofhigh-effort/low-rewardconditions.Journal of occupational health psychology 1996,1:27-41.

17. SiegristJ,WahrendorfM,vondemKnesebeckO,JurgesH,Borsch-SupanA:Qualityofwork,well-being,and intendedearlyretirementofolderemployees:baselineresults fromtheSHAREStudy.European journal of public health 2007,17:62-68.

18. vanOortFV,vanLentheFJ,MackenbachJP:Material,psychosocial,andbehaviouralfactorsintheexplanationofeducationalinequalitiesinmortalityinTheNetherlands.Journal of epidemiology and community health 2005,59:214-220.

50 Chapter3

19. Hennekens C, Buring B, Mayrent S: Epidemiology in Medicine. Boston: LippincottWilliams&Wilkins;1987.

20. Statisticalpackageforsocialsciences(SPSS).15edition.Chicago,IL;2006. 21. Schröder M: Attrition. In Healt, ageing and retirement in Europe (2004-2007) Starting the

longitudinal dimension. Edited by Borsch-Supan A, Brugiavini A, Jürges H, KapteynA, Mackenbach J, Siegrist J, Weber G. Mannheim: Mannheim Research Institute for theEconomicsofAging(MEA);2008:325-330.

22. BoundJ:Self-reportedversusobjectivemeasuresofhealthinretirmentmodels.Journal of Human Resources 1991,26:106-138.

23. KalwijA,VermeulenF:HealthandlabourforceparticipationofolderpeopleinEurope:whatdoobjectivehealthindicatorsaddtotheanalysis?Health economics 2008,17:619-638.

24. BaandersAN,RijkenPM,PetersL:Labourparticipationof thechronically ill.Aprofilesketch.European journal of public health 2002,12:124-130.

25. BoundJ,SchoenbaumM,StinebricknerTR,WaidmannT:Thedynamiceffectsofhealthonthelaborforcetransitionsofolderworkers.Labour economic 1999,6:179-202.

26. McDonoughP,AmickBC,3rd:Thesocialcontextofhealthselection:alongitudinalstudyofhealthandemployment.Social science & medicine (1982) 2001,53:135-145.

27. FriisK,EkholmO,HundrupYA,ObelEB,GronbiekM:Influenceofhealth,lifestyle,workingconditions,andsociodemographyonearlyretirementamongnurses:TheDanishNurseCohortStudy.Scandinavian Journal of Public Health 2007,35:23-30.

28. ArrowJO:Estimatingtheinfluenceofhealthasariskfactoronunemployment:asurvivalanalysisofemploymentdurationsforworkerssurveyedintheGermanSocio-EconomicPanel(1984-1990).Social science & medicine (1982) 1996,42:1651-1659.

29. DetailleSI,HeerkensYF,EngelsJA,vanderGuldenJW,vanDijkFJ:Commonprognosticfactorsofworkdisabilityamongemployeeswithachronicsomaticdisease:asystematicreviewofcohortstudies.Scandinavian journal of work, environment & health 2009,35:261-281.

30. DauchetL,AmouyelP,HercbergS,DallongevilleJ:Fruitandvegetableconsumptionandriskofcoronaryheartdisease:ameta-analysisofcohortstudies.J Nutr 2006,136(10):2588-2593.

31. Matson-Koffman DM, Brownstein JN, Neiner JA, Greaney ML:A site-specific literaturereview of policy and environmental interventions that promote physical activity andnutritionforcardiovascularhealth:whatworks?Am J Health Promot 2005,19:167-193.

32. VarekampI,HeutinkA,LandmanS,KoningCE,deVriesG,vanDijkFJ:FacilitatingEm-powerment in Employees with Chronic Disease: QualitativeAnalysis of the Process ofChange.Journal of occupational rehabilitation 2009.

33. Biering-SorensenF,LundJ,HoydalsmoOJ,DarreEM,DeisA,KrygerP,MullerCF:Riskindicatorsofdisabilitypension.A15yearfollow-upstudy.Dan Med Bull 1999,46:258-262.

34. JusotF,KhlatM,RochereauT,SermeC:Joblossfrompoorhealth,smokingandobesity:anationalprospectivesurvey inFrance. Journal of epidemiology and community health 2008,62:332-337.

35. RissanenA,HeliovaaraM,KnektP,ReunanenA,AromaaA,MaatelaJ:Riskofdisabilityandmortalityduetooverweight inaFinnishpopulation.BMJ (Clinical research ed 1990,301:835-837.

36. EriksenW,NatvigB,RutleO,BruusgaardD:Smokingasapredictoroflong-termworkdisabilityinphysicallyactiveandinactivepeople.Occup Med (Lond) 1998,48:315-320.

Ch

apte

r3


37. LiiraJ,Leino-ArjasP:Predictorsandconsequencesofunemploymentinconstructionandforestworkduringa5-yearfollow-up.Scandinavian journal of work, environment & health 1999,25:42-49.

38. LundT,CsonkaA:Riskfactorsinhealth,workenvironment,smokingstatus,andorganiza-tionalcontextforworkdisability.Am J Ind Med 2003,44:492-501.

39. RothenbacherD,ArndtV,FraisseE,ZschenderleinB,FliednerTM,BrennerH:Earlyretire-mentdue topermanentdisability inrelation tosmoking inworkersof theconstructionindustry.J Occup Environ Med 1998,40:63-68.

40. HaahrJP,FrostP,AndersenJH:Predictorsofhealthrelatedjobloss:atwo-yearfollow-upstudy inageneralworkingpopulation. Journal of occupational rehabilitation 2007,17:581-592.

41. LundT,IversenL,PoulsenKB:Workenvironmentfactors,health,lifestyleandmaritalstatusaspredictorsofjobchangeandearlyretirementinphysicallyheavyoccupations.Am J Ind Med 2001,40:161-169.

42. BlekesauneM,SolemPE:Workingconditionsandearlyretirement.Aprospectivestudyofretirementbehavior.Research on aging 2005,27(1):3-30.

43. SolomonC,PooleJ,PalmerKT,CoggonD:Health-relatedjobloss:findingsfromacommu-nity-basedsurvey.Occupational and environmental medicine 2007,64:144-149.

44. DuijtsSF,KantI,SwaenGM,vandenBrandtPA,ZeegersMP:Ameta-analysisofobserva-tionalstudiesidentifiespredictorsofsicknessabsence.Journal of clinical epidemiology 2007,60:1105-1115.

45. MeerdingWJ,WIJ,KoopmanschapMA,SeverensJL,BurdorfA:Healthproblemsleadtoconsiderable productivity loss at work among workers with high physical load jobs.Journal of clinical epidemiology 2005,58:517-523.

46. SchultzAB,EdingtonDW:Employeehealthandpresenteeism:asystematicreview.Journal of occupational rehabilitation 2007,17:547-579.

47. VahteraJ,VirtanenP,KivimakiM,PenttiJ:Workplaceasanoriginofhealthinequalities.Journal of epidemiology and community health 1999,53:399-407.

48. KuperH,MarmotM:Jobstrain,jobdemands,decisionlatitude,andriskofcoronaryheartdisease within the Whitehall II study. Journal of epidemiology and community health 2003,57:147-153.

4The effects of ill health on entering

and maintaining paid employment:

evidence in european countries

SchuringM,BurdorfA,KunstAE,MackenbachJP.

Journal of Epidemiology and Community Health2007,61:597-604.

54 Chapter4

absTRacT

objectives:Toexaminetheeffectsofillhealthonselectionintopaidemploymentin

Europeancountries.

methods:Fiveannualwaves(1994-8)oftheEuropeanCommunityHouseholdPanel

wereusedtoselecttwopopulations:(1)4446subjectsunemployedforatleast2years,

ofwhich1590(36%)subjectsfoundemploymentinthenextyear,and(2)57436sub-

jectsemployedforatleast2years,ofwhich6191(11%)subjectslefttheworkforcein

thenextyearbecauseofunemployment,(early)retirement,orhavingtotakecareof

household.Theinfluenceofaperceivedpoorhealthandachronichealthproblemon

employmenttransitionswasstudiedusinglogisticregressionanalysis.

Results:An interactionbetweenhealthandsexwasobserved,withwomen inpoor

health(oddsratio(OR)0.4),meninapoorhealth(OR0.6),andwomen(OR0.6)hav-

ing lesschance toenterpaidemployment thanmen ingoodhealth.Subjectswitha

poorhealthandlow/intermediateeducationhadthehighestriskofunemploymentor

(early)retirement.Takingcareofthehouseholdwasonlyinfluencedbyhealthamong

unmarried women. In most European countries, a poor health or a chronic health

problem predicted staying or becoming unemployed and the effects of health were

strongerwithalowernationalunemploymentlevel.

conclusion: In most European countries, socio-economic inequalities in ill health

were an important determinant for entering and maintaining paid employment. In

public health measures for health equity, it is of paramount importance to include

peoplewithpoorhealthinthelabourmarket.

Ch

apte

r4

Theeffectsofillhealthonenteringandmaintainingpaidemployment:evidenceinEuropeancountries 55

inTRoDucTion

Thepresenceofsocioeconomicinequalitiesinhealthhasbeenwidelyacknowledged.

[1] Lower education, unskilled labour and low income are associated with higher

mortalityandmorbidity.[2]Labourforceparticipationisanimportantdeterminantof

health inequalities,asdemonstratedbyahigherprevalenceof illnessanddisability

[3,4]andahighermortalityamongunemployedpeople.[5]Theassociationbetween

health and employment is bi-directional: unemployment may cause poor health

(causation hypothesis) and poor health may increase the probability of becoming

unemployed(selectionhypothesis)[6,7]

ThecausationmechanismisillustratedamongBritishhouseholds,whereunemploy-

menthadastrongeffectontheincidenceofanylimitingillnessandemploymentwas

relatedtorecoveryfromtheseillnesses.[6]Areviewoflongitudinalstudiesconcluded

that loss of employment affected mental health, but also that gaining employment

improvedmentalhealth.[8]

Thereisalsoevidencefortheselectionhypothesis.Amonglong-termunemployed

Norwegians,mentalmorbiditywasstronglyassociatedwithareductioninobtaininga

paidjob.[9]InaBritishstudy,healthierpeoplewerefoundmorelikelytogainemploy-

ment than people with minor psychiatric morbidity.[7] Ill health may also increase

otherformsofnon-employment,suchasearlyretirementandhavingtostayinghome

totakecareofthefamily.[7,10]InaFinnishcohortofmen,self-assessedpoorhealth

wasastrongpredictorofbothdisabilityaswellasnon-illnessbasedearlyretirement.

[11]

The associations between health and employment will not be similar across all

socio-economic groups, as age, education, sex, marital status or household income

may influence the social context of health and employment status.[12][13] The con-

sequencesofillhealthforthelikelihoodofbecomingorstayingunemploymentmay

dependonsocialandlabourmarketpolicies,whichvaryacrossEuropeancountries.

Forexample,strikingdifferencesinaccesstobenefits,suchasearlyretirementordis-

abilitypension,fordisabledpeoplehavebeendescribedwithintheEuropeanUnion.

In the Netherlands, < 10% of disabled people have labou as their main source of

incomewhereasinSwedenthisproportionamountstoover50%.[14]

Against thisbackground, theaimof thisstudywasto investigate theeffectsof ill

healthontheselectionprocessintopaidemploymentandwhethertheseeffectsinter-

actedwithsocioeconomicgroups.Thesecondaryaimofthestudywastoinvestigate

whetherobservedassociationsbetweenillhealthandemploymenttransitionsdiffered

amongEuropeancountries.

56 Chapter4

meTHoDs

Studypopulation

Thedatawerederivedfromthefirstfivewaves(1994-8)oftheEuropeanCommunity

Household (ECHP).[15] The ECHP is a social survey among member states of the

EuropeanUnionwithalongitudinaldesigntodescribethesocialdynamicsinEurope.

DatawerecollectedbyNationalInstitutesforStatisticsorresearchcenters,whiledata

checks, weightings and imputations were done centrally by the Statistical Office of

the European Communities (Eurostat). All surveys were based on a non-stratified

randomsamplingdesignamongallnationalprivatehouseholds.Allmembersofthe

householdwereindividuallyinterviewed.Thedatacollectionwascarriedoutinmost

countriesbypaper-and-pencil interviewing,butinfourcountries(UnitedKingdom,

TheNetherlands,Portugal,andGreece)bycomputer-assistedpersonalinterviewing.

Theoverallhouseholdresponserateinthefirstwavewas72%,butvariedconsiderably

amongcountries.Theresponserates in laterwavesoftheECHPstudywerehigher.

Adetaileddescriptionofsamplingproceduresandresponserateshasbeenpublished

elsewhere.[16,17]

For the purpose of this study, subjects aged between 16-65 years were selected,

with available information on employment status and health status during at least

three consecutive annual measurements, whereby the employment status remained

unchangedinthefirstandsecondmeasurementandapossibleemploymenttransition

hadoccurredinthethirdmeasurement.Althoughemploymentstatuswasascertained

annually,weconsideredthisstatusasrepresentativeforthewholeyearprecedingthe

administrationofthequestionnaire.Thisprocedureresultedinacohortwith3years

offollow-up,withtwoconsecutiveyearsbeforeapossibleemploymenttransition.For

asmallpartofthecohort,informationwasalsoavailableinthethirdandfourthan-

nualmeasurementbeforetheyearofemploymenttransition.Thus,foreverysubject,

anemploymenttransitionwaspossibleinagivenyearandlabourstatuswasregarded

asconstantinthe2-4yearsbeforepossibleemploymenttransition.

Self-defined employment status was classified into four mutually exclusive cat-

egories: employed (paid employment, paid apprenticeship or self-employment),

unemployment,retiredortakingcareofhousehold.[15]Employmentwasdefinedas

workinginajobforatleast15h/week.Subjectsworking<15h/weekwereautomati-

callyclassified intothecategoriesunemployed,retired, takingcareofhousehold, in

militaryservice,followingeducationoreconomicallyinactive.Unemployedsubjects

were defined as those people who worked <15 h/week and who considered them-

selves as being unemployed. Subjects who classified themselves into the last three

categorieswerenotselectedforanalysisinthisstudy,henceexcludingeconomically

Ch

apte

r4


inactivepeoplewhoarewithoutworkandwhodonotwishtoconsiderthemselvesas

unemployed.Subjectswithadisabilitypensionasanotherformofnon-employment

were also excluded from the analysis, as poor health is an essential requirement to

qualifyforadisabilitypension.However,disabledpersonswithapaidjobfor>15h/

weekwereincludedinthecategory“employed”.

Two different populations were defined to study employment transitions. The

first studypopulation included4446peoplewhowereunemployed forat least two

consecutiveyears,ofwhich1590(36%)peopleenteredtheworkforceduringthelast

yearoffollow-up.Thesecondstudypopulationconsistedof57436workerswhowere

employedforatleasttwoconsecutiveyears,ofwhich6191(11%)peopleleftthework-

forceinthelastyearoffollow-upduetounemployment(n=3000),retirement(n=2017)

orhavingtotakecareofthehousehold(n=1174).

Questionnaire

Thequestionnairecomprisedquestionsonindividualandhouseholdcharacteristics,

perceived general health, chronic health problem, and employment status. Indi-

vidualcharacteristics includedage,sex,educationandpersonal income.Household

characteristics included marital status, the presence of children aged ≤12 years and

household income. Subjects were divided into three groups according to their level

ofeducationalattainmentonthebasisoftheInternationalStandardClassificationof

Education (ISCED).[18] Informationonmarital statuswasused tocomparesubjects

marriedorlivingwithothers.Householdandpersonalincomewerecategorizedinto

three percentile groups (<25th centile, 25-75th centile and >75th centile), relative to

theincomedistributionofthestudysampleineachcountry.Ifthepersonalincomeof

asubjectwas≥75%ofthehouseholdincome,thesubjectwasdefinedasbreadwinner

withinthehousehold.

Twomeasuresofself-reportedhealthwereused.First,subjectswereaskedtorate

their own general health on a five-point scale, ranging from “very good”, “good”,

“fair”and“bad”to“verybad”.Thosereportinglessthan“goodhealth”weredefined

ashavingapoorhealth.[19]Second,subjectswereaskedwhethertheyhadanychronic

physicalormentalhealthproblem, illnessordisability (yes/no).Anaffirmativean-

swerwasclassifiedaschronichealthproblem[15]

Statisticalmethods

Logisticregressionanalysiswasperformedtostudytheimpactofpoorhealthanda

chronichealthproblemonemploymentstatus,adjustedforcountryandpersonaland

householdcharacteristics.Thefirststepintheanalysiswastoestablishtheassociations

58 Chapter4

betweenpoorhealthandemploymentstatus,includingcountryasacategoricalvariable

(fixedfactor)toadjustfortheeffectofcountryontheassociationbetweenpoorhealth

and employment status. In the second step, other independent socio-demographic

variableswereincludedinthemodelbyastep-forwardprocedure.Thevariablewith

thestrongestassociationwithemploymentstatuswasputinthemodelfirst,followed

bythenextstrongest,andsoon.Variableswithasignificantreductionintheoverall

scaleddevianceofthemodelwereretainedinthemultivariatemodel.Inthethirdstep,

interactions between poor health and significant socio-demographic variables were

investigated for their influence on the overall fit of the model. The interaction with

thestrongesteffectonthescaleddevianceofthemodelwasincludedinthefinalmul-

tivariatemodel.Thus, interactiontermswithouthealthwerenotconsidered,thereby

excludingpossibleinteractionsamongsocio-demographicvariables.Forallanalyses,

fourdifferenttimewindowswereapplied,investigatingtheeffectsoftheindependent

variablesmeasured1,2,3,and4yearsbeforetheemploymenttransition.Allstatistical

modelswerebasedonthe(varying)numberofpeopleavailableforthethreedifferent

measuresofemploymentstatus,withoutweighingtheregressioncoefficientaccording

toattritionrateinthecountrysampleorpopulationsizeineachcountry.

Fortheanalysisontransitionfromemploymentto(early)retirement,onlysubjects

in the age group 55-65 years were used, as occurrence of retirement among those

aged≤54wasverylow.Owingtothisselection,agewasnotincludedinthisanalysis.

Likewise,theanalysisontransitionfromemploymenttohouseholdwasrestrictedto

women,asonlyveryfewmenclassifiedthemselvesastakingcareof thehousehold

withouthavinganypaidjob.

Thevariationamongcountriesintheassociationbetweenpoorhealthandemploy-

ment statuswas investigatedby introductionof the interaction termofpoorhealth

andcountry ineachmultivariatemodel.This interactiontermwasusedtoestimate

thecountry-specificassociationsforpoorhealthandemploymentstatus(expressedby

oddsratios (ORs)),adjustedforrelevantsocio-demographiccovariates.Allanalyses

wererepeatedwiththepresenceofachronichealthproblemasoutcomemeasure.The

analyseswerecarriedoutwiththestatisticalpackageSASV.8.2.

ResuLTs

Thebackgroundcharacteristicsofbothpopulationsarepresented inTable4.1.Both

health measures were strongly associated with almost 50% of all persons reporting

aperceivedpoorhealth,alsoindicatingthepresenceofachronicphysicalormental

healthproblem,illnessordisability.Illhealth,sex,education,havingyoungchildren

inhouseholdandbeingabreadwinnerwereassociatedwithemploymenttransition.

Ch

apte

r4


Table 4.1 BackgroundcharacteristicsofsubjectsfromtheEuropeanHouseholdPanelSurveyintheyearofanemploymenttransitionafteratleasttwoconsecutiveyearsofunemploymentoremployment

characteristic

Subjectswithatleast2yearsofunemploymentwhoremainedunemployed(n=2856)orenteredemployment(n=1590)inthe3rdyear

Subjectswithatleast2yearsofemploymentwhoremainedemployed(n=51245)orlefttheworkforce(n=6191)inthe3rdyear

unemployed became employed employed Left workforce

n (%) n (%) n (%) n (%)

Age(years)16–2425–4445–5455–65

758(27)1241(45)463(17)291(11)

586(38)*785(49)170(11)38(2)*

5610(11)29344(60)10764(22)3240(7)

736(13)2200(38)*1102(19)1805(31)*

Sex(men) 1315(46) 886(56)* 31642(62) 3216(52)*

Married(yes) 1333(47) 700(44) 36347(71) 4399(71)


263(10)956(25)1507(55)

192(13)595(39)732(48)*

13297(27)18838(38)17594(35)

896(15)*2029(34)2982(51)*

Children<12years 800(28) 498(31)* 17797(35) 1674(27)*

Breadwinner 481(17) 251(16) 16341(32) 1585(26)*

Perceivedpoorhealth 968(34) 327(21)* 12437(24) 2311(37)*

Chronichealthproblem 662(23) 200(13)* 8100(16) 1544(25)*

*Waldχ2–test,P<0.05

40

45

50

55

60

65

70

75

80

85

4 3 2 1 *Years before employment transition

Perc

enta

ge o

f per

sons

with

go

od p

erce

ived

hea

lth (%

)

(Re)employed (n=183)Left workforce to take care of household (n=232)Left workforce due to unemployment (n=463)Continuously employed (n=30763)Continuously unemployed (n=850)Left workforce due to retirement (n=538)

Figure 4.1 ProportionofsubjectswithaperceivedgoodhealthamongrespondentswhocompletedallfivewavesoftheEuropeanHouseholdPanelSurveyandanemploymenttransitioninthelastyearofthesurvey

40

45

50

55

60

65

70

75

80

85

4 3 2 1 *Years before employment transition

Perc

enta

ge o

f per

sons

with

go

od p

erce

ived

hea

lth (%

)

(Re)employed (n=183)Left workforce to take care of household (n=232)Left workforce due to unemployment (n=463)Continuously employed (n=30763)Continuously unemployed (n=850)Left workforce due to retirement (n=538)

60 Chapter4

Figure4.1showsthatinallgroupstheproportionofpeopleingoodhealthdecreased

intheperiod1994-8.Theproportionofpeopleingoodhealthwasconsistentlyhigher

amongthosewhowereemployedorbecameemployedthanamongpeoplewhowere

notemployedorlefttheworkforce.

Table 4.2 shows that an interaction was observed between health status and sex.

Womenwithpoorhealth,menwithpoorhealthandwomenwithgoodhealthwere

significantlymorelikelytoremainunemployedinthenextyearthanmenwithgood

health.Thisinteractionwaspresentupto4yearsbeforeenteringpaidemployment.

Apositiveimpacton(re)employmentwasobservedforpeoplebeingmarried,being

ofyoungerage,havinghighereducationandhavinghigherpersonalincome.Beinga

breadwinner(OR0.9,95%CI0.8-1.2)andhavingchildren<12yearsinthehousehold

(OR1.0,95%CI0.8-1.2)werenotassociatedwithenteringpaidemployment.Chronic

healthproblemsshowedverysimilarresultsforalltimewindows(resultsnotshown).

Table4.3showsthatpoorhealthwasmoreimportantamongthosewithhigheredu-

cationthanamongthosewith lowereducationbecomingunemployed.Theeffectof

poorhealthwithinthestrataofintermediateandhighereducationbecamelessstrong

Table 4.2 Multivariatemodelswithdifferenttimewindowsofsignificantdeterminantsof(re)employmentamongsubjectsfromtheEuropeanHouseholdPanelSurveywithatleast2consecutivepreviousyearsofunemploymentbeforeenteringpaidemployment

Determinant

years before entering paid employment

1 yearoR (95% ci)

2 yearsoR (95% ci)

3 yearsoR (95% ci)

4 yearsoR (95% ci)

NstayedunemployedNenteredemployment

26341481

26461483

1444609

792174

Poorhealth-Women†

Goodhealth-WomenPoorhealth-MenGoodhealth-Men

0.4(0.3-0.5)*0.6(0.5-0.7)*0.6(0.4-0.7)*1

0.4(0.3-0.5)*0.6(0.5-0.7)*0.6(0.5-0.7)*1

0.6(0.4-0.8)*0.5(0.4-0.7)*0.6(0.4-0.8)*1

0.5(0.3-0.9)*0.7(0.5-1.0)*0.6(0.4-1.1)1

Age(years)16-2425-4445-5455-64

8.1(5.4-12.2)*5.7(3.9-8.3)*3.3(2.2-5.0)*1

9.0(6.0-13.5)*6.0(4.2-8.8)*3.4(2.3-5.0)*1

11.0(5.2-23.4)*8.3(4.1-17.0)*4.7(2.2-9.9)*1

4.7(1.4-14.9)*3.6(1.2-10.8)*1.7(0.5-5.7)1

Married 1.4(1.2-1.7)* 1.4(1.2-1.6)* 1.4(1.1-1.8)* 1.6(1.0-2.6)*


1.7(1.3-2.1)*1.6(1.3-1.8)*1

1.6(1.3-2.0)*1.5(1.3-1.8)*1

2.1(1.5-2.9)*1.6(1.3-2.0)*1

2.6(1.5-4.7)*1.9(1.2-2.8)*1

Personalincome>75thcentile25-75thcentile<25thcentile

1.6(1.3-1.9)*1.5(1.2-1.7)*1

1.9(1.6-2.3)*1.4(1.2-1.6)*1

1.8(1.4-2.5)*1.4(1.1-1.8)*1

1.4(0.9-2.3)1.0(0.6-1.6)1

*Waldχ2-test,P<0.05.†significantinteraction.

Ch

apte

r4


inmoredistantyears. Job losswasalso influencedbyolderage,notbeingmarried,

havingchildren<12years (asper thepreviousyear)andhaving lowerpersonal in-

come.Beingbreadwinner(OR1.0,95%CI0.9-1.1)ormen(OR1.1,95%CI1.0-1.2)did

notinfluenceunemployment.Theimpactofchronichealthproblemswasremarkably

similartoaperceivedpoorhealth(resultsnotshown).

Among subjects aged ≥55 years, those with a poor health had a higher chance of

retiring the next year than subjects with a good health (table 4.4). Among highly

educatedsubjects,thisassociationwasnotobservedforthehealthstatus2,3,and4

yearsbeforetheyearofretirement.Beingmarriedseemedtoreducethelikelihoodof

retirement.Breadwinner,sex,andpersonalincomeallhadoddsratiosclosetounity.

Theinfluenceofachronichealthproblemon(early)retirementcloselymirroredthe

resultsofaperceivedpoorhealth.

Table4.5demonstratesthattakingcareofthehouseholdwasinfluencedbyhealth

amongunmarriedwomen(OR=1.6)butnotamongmarriedwomen.Thisassociation

wasonlyobservedinearlieryears.Leavingtheworkforcetotakecareofthehouse-

holdwasconsistentlyassociatedwitholderage,havingyoungchildren,lowpersonal

income,andloweducationallevel.

Table 4.3 MultivariatemodelswithdifferenttimewindowsofsignificantdeterminantsofbecomingunemployedamongsubjectsfromtheEuropeanHouseholdPanelSurveywithatleast2consecutivepreviousyearsofemploymentbeforebecomingunemployed

Determinant

years before becoming unemployed

1 yearoR (95% ci)

2 yearsoR (95% ci)

3 yearsoR (95% ci)

4 yearsoR (95% ci)

N(stayedemployed)N(becameunemployed)

473782700

366101629

30706515

28780427

Poorhealth-loweducation†

Goodhealth-loweducationPoorhealth-intermediateeducationGoodhealth-intermediateeducationPoorhealth-higheducationGoodhealth-higheducation

2.2(1.8-2.6)*1.9(1.7-2.2)*2.4(2.0-2.9)*1.4(1.2-1.6)*2.1(1.6-2.6)*1

2.5(2.0-3.1)*1.9(1.6-2.3)*2.1(1.7-2.6)*1.5(1.2-1.8)*1.6(1.2-2.2)*1

2.3(1.6-3.4)*2.2(1.6-3.1)*2.5(1.7-3.7)*1.6(1.2-2.3)*1.4(0.8-2.4)*1

2.5(1.6-3.9)*2.2(1.5-3.1)*2.3(1.5-3.5)*1.8(1.3-2.6)*1.2(0.7-2.3)1

Age(years)16-2425-4445-5455-64

1.0(0.9-1.3)0.8(0.7-0.9)*0.8(0.7-0.9)*1

1.0(0.8-1.3)0.8(0.6-1.0)*0.8(0.6-1.0)1

0.9(0.6-1.4)0.8(0.5-1.1)0.9(0.6-1.3)1

0.9(0.6-1.4)0.7(0.5-1.0)0.8(0.5-1.2)1

Married 0.7(0.6-0.7)* 0.7(0.6-0.8)* 0.7(0.6-0.9)* 0.8(0.6-0.9)*

Children<12years 1.2(1.1-1.3)* 1.1(1.0-1.3) 1.1(0.9-1.4) n/a


0.2(0.2-0.3)*0.5(0.4-0.5)*1

0.2(0.2-0.3)*0.4(0.4-0.5)*1

0.3(0.3-0.5)*0.5(0.4-0.6)*1

0.4(0.3-0.6)*0.5(0.4-0.7)*1

n/a=notavailableindatacollection.*Waldχ2-test,P<0.05.†significantinteraction.

62 Chapter4

In table 4.6, the country-specific associations are presented for 11 countries (Lux-

embourgnotincludedasassociationswerenotestimabletherebecauseofverysmall

numbers).In9of11Europeancountries,aperceivedpoorhealthwasariskfactorfor

staying unemployed, varying from OR=0.2 in Denmark to OR=0.7 in Germany and

France,althoughthisassociationreachedtheconventionallevelofstatisticalsignifi-

canceinonlyfourcountries.In7of11countriesaperceivedpoorhealthincreasedthe

riskofbecomingunemployed,varyingfromOR=1.2inFrancetoOR=2.7intheNeth-

erlands.Amongolderworkers,aperceivedpoorhealthraisedthelikelihoodofretire-

mentinninecountries,withsmalleffectsobservedinFranceandtheUK(OR=1.1)and

significantlargeeffectsinBelgium(OR=2.6)andGermany(OR=2.6).Forpoorhealth

andachronichealthproblem,ingeneral,comparableassociationswerefound,with

theexceptionoftwocountries.InItaly,achronichealthproblempredictedremaining

unemployed (OR=0.4), but poor health had no discernable impact (OR=1.0). In the

Netherlands,achronichealthproblemseemedtoincreasethelikelihoodofretirement

amongolderworkers(OR=1.7),whereasapoorhealthdecreasedretirement(OR=0.6),

althoughbothassociationswerenotstatisticallysignificant.Figure4.2illustratesthat

a lower unemployment rate at national level in 1998 was associated with larger ef-

fectsofpoorhealthonnotenteringemployment(regressioncoefficient0.06,R2=0.46,

p=0.02)andlargereffectsofpoorhealthonbecomingunemployed(regressioncoeffi-

cient-0.10,R2=0.27,p=0.10).Thistrendovercountrieswasnotobservedfortheeffects

ofpoorhealthonretirementortakingcareofhousehold.

Table 4.4 Multivariatemodelswithdifferenttimewindowsofsignificantdeterminantsofretirementamongworkersaged55-65yearsintheEuropeanHouseholdPanelSurveywithatleast2consecutivepreviousyearsofemploymentbeforeretirement

Determinant

years before retirement

1 yearoR (95% ci)

2 yearsoR (95% ci)

3 yearsoR (95% ci)

4 yearsoR (95% ci)

N(stayedemployed)N(becameretired)

31401374

31671384

2097806

1778345

Poorhealth-loweducation†

Goodhealth-loweducationPoorhealth-intermediateeducationGoodhealth-intermediateeducationPoorhealth-higheducationGoodhealth-higheducation

1.7(1.4-2.2)*1.3(1.0-1.6)*2.0(1.6-2.7)*1.4(1.1-1.6)*1.5(1.0-2.2)*1

1.7(1.3-2.1)*1.2(1.0-1.5)1.8(1.3-2.4)*1.3(1.1-1.7)*1.1(0.8-1.6)1

1.4(1.0-1.9)*1.2(0.9-1.5)1.7(1.2-2.6)*1.2(0.8-1.7)0.8(0.5-1.3)1

1.6(1.0-2.4)*1.2(0.8-1.8)1.7(1.0-2.9)*1.1(0.7-1.7)*1.1(0.6-2.1)1

Married 0.8(0.7-1.0)* 0.8(0.7-1.0)* 0.8(0.7-1.0) 0.9(0.6-1.2)

*Waldχ2-test,P<0.05.†significantinteraction.

Ch

apte

r4


Discussion

Thisstudyshowsthattheeffectsofhealthontransitionsbetweenpaidemploymentand

various formsofnon-employmentwere influencedbysocio-demographicvariables.

Apoorhealthwasariskfactorforremainingunemployedamongmen,buthadless

effectamongwomen.Apoorhealthwasalsoariskfactorforbecomingunemployed

or retiring, especially among highly educated workers. Finally, among women the

transitionfromemploymentintotakingcareofahouseholdwasinfluencedbyapoor

healthonlyamongwomennotmarried.InmostEuropeancountries,poorhealthand

achronichealthproblemwereriskfactorsfornotenteringtheworkforceorbecoming

unemployed,butlargedifferencesamongcountrieswereobservedthatcouldpartly

beexplainedbytheunemploymentrateatnationallevel.

TheseconclusionsarebasedontheEuropeanHouseholdPanelSurvey.Anobvious

disadvantageofthecurrentanalysisistherestrictiontotwomeasuresofself-reported

health.Bothmeasuresofhealthwerestronglycorrelated,asonewouldexpect.Unfor-

Table 4.5 MultivariatemodelswithdifferenttimewindowsofsignificantdeterminantsofleavingemploymentforhouseholdcareamongwomenintheEuropeanHouseholdPanelSurveywithatleast2consecutivepreviousyearsofemploymentbeforeleavingtheworkforcetotakecareofhousehold

years before leaving the workforce to take care of household

1 yearoR (95% ci)

2 yearsoR (95% ci)

3 yearsoR (95% ci)

4 yearsoR (95% ci)

N(stayedemployed)N(household)

181001000

13818608

11351229

10419193

Poorhealth-marriedGoodhealth-marriedPoorhealth-notmarriedGoodHealth-notmarried

2.6(2.0-3.3)*2.4(2.0-3.0)*1.6(1.1-2.3)*1

2.1(1.5-2.8)*1.9(1.4-2.4)*1.3(0.9-2.0)1

1.7(1.1-2.7)*1.7(1.1-2.6)*0.7(0.3-1.4)1

1.3(0.8-2.2)1.7(1.1-2.5)*0.9(0.4-1.7)1

Age16-2425-4445-54

0.5(0.4-0.7)*0.4(0.3-0.5)*0.4(0.3-0.5)*

0.4(0.3-0.7)*0.4(0.3-0.5)*0.4(0.3-0.6)*

0.6(0.3-1.2)0.4(0.3-0.7)*0.5(0.3-0.9)*

0.6(0.3-1.2)0.4(0.2-0.6)*0.4(0.2-0.7)*


0.6(0.5-0.8)*0.8(0.7-1.0)*1

0.7(0.5-0.9)*0.9(0.7-1.1)1

0.8(0.5-1.2)0.9(0.7-1.3)1

0.7(0.5-1.2)0.9(0.6-1.3)1

Children<12years 1.9(1.6-2.2)* 1.6(1.3-1.9)* 1.1(0.8-1.5) n/a


0.3(0.2-0.4)*0.4(0.3-0.4)*1

0.2(0.2-0.4)*0.4(0.3-0.5)*1

0.4(0.2-0.6)*0.4(0.3-0.6)*1

0.3(0.2-0.6)*0.6(0.4-0.8)*1

n/a=notavailableindatacollection.*Waldχ2-test,P<0.05.†significantinteraction

64 Chapter4

tunately,noinformationwasavailableonself-reportedordiagnosedspecificdiseases

and,hence,itwasnotpossibletoevaluatewhichparticulardiseaseslargelyexplained

theroleofhealthinenteringorleavingtheworkforce.Theordinalmeasurementand

subsequentdichotomisingofself-perceivedhealthmayhampercomparabilityacross

countriesand,indeedinourstudypopulations,theprevalenceofpoorhealthvaried

fromabout8%inGreeceandIrelandtoover40%inPortugalandGermany.However,

themagnitudeoftheprevalencewasnotassociatedwiththeobservedORsforpoor

health.Inaddition,whenusingamorestringentcut-offpointforapoorhealth(only

Table 4.6 InfluenceofperceivedpoorhealthandchronichealthproblemsinthepreviousyearonemploymenttransitionsinthenextyearamongsubjectsfromtheEuropeanHouseholdPanelSurveyin11Europeancountries

country

From unemployed to employed

From employed to unemployed

From employed to retired (> 55 year)

n oR n oR n oR

GermanyPerceivedpoorhealthChronichealthproblem

4830.7(0.5-1.1)0.9(0.5-1.3)

95302.6(2.1-3.8)*2.2(1.8-2.7)*

7962.6(1.8-3.8)*2.4(1.7-3.4)*

DenmarkPerceivedpoorhealthChronichealthproblem

1300.2(0.1-0.7)*0.5(0.2-1.2)

25642.0(1.3-3.2)*1.2(0.8-1.8)

2231.6(0.8-3.0)1.1(0.7-2.0)

NetherlandsPerceivedpoorhealthChronichealthproblem

3440.3(0.2-0.6)*0.4(0.3-0.8)*

38042.7(1.8-4.0)*2.8(198-4.3)*

1300.6(0.1-5.8)1.7(0.3-10.8)

BelgiumPerceivedpoorhealthChronichealthproblem

2800.6(0.2-1.3)0.3(0.1-1.4)

26991.6(1.0-2.6)1.3(0.6-2.5)

1342.6(1.1-6.3)*1.5(0.6-4.4)

FrancePerceivedpoorhealthChronichealthproblem

4850.7(0.4-1.1)0.8(0.5-1.4)

56091.2(1.0-1.6)1.2(0.9-1.7)

3701.1(0.7-1.7)1.5(0.9-2.6)

UnitedKingdomPerceivedpoorhealthChronichealthproblem

2290.4(0.2-0.8)*0.4(0.2-0.8)*

71091.8(1.3-2.6)*1.2(0.9-1.8)

6051.1(0.7-1.8)1.5(1.0-2.3)

IrelandPerceivedpoorhealthChronichealthproblem

2600.6(0.3-1.3)0.7(0.3-1.6)

28600.6(0.2-1.5)0.8(0.4-1.7)

2991.4(0.7-2.7)1.4(0.7-2.7)

ItalyPerceivedpoorhealthChronichealthproblem

8911.0(0.7-1.5)0.4(0.2-0.9)*

62381.0(0.7-1.3)0.9(0.5-1.6)

5540.8(0.6-1.1)0.9(0.5-1.5)

GreecePerceivedpoorhealthChronichealthproblem

3851.2(0.5-2.5)0.7(0.2-2.1)

37401.3(0.8-2.0)1.0(0.6-1.8)

4891.3(0.9-2.0)0.9(0.6-1.6)

SpainPerceivedpoorhealthChronichealthproblem

7840.6(0.4-0.9)*0.4(0.3-0.7)*

48970.9(0.7-1.2)1.1(0.8-1.4)

4101.5(1.0-2.4)*1.3(0.8-2.0)

PortugalPerceivedpoorhealthChronichealthproblem

1910.5(0.3-1.0)*0.9(0.4-2.1)

48511.0(0.7-1.2)1.5(1.1-2.0)*

6691.5(1.0-2.4)1.8(1.2-2.7)*

*Waldχ2-test,P<0.05,adjustedforage,sex,maritalstatus,education,andpersonalincome.

Ch

apte

r4


including subjects with bad or very bad health), results remained largely similar,

although the Ors for the effects of poor health and intermediate or high education

onbecomingunemployedorretiringweresomewhathigherthanthosepresentedin

tables4.3and4.4.This suggests that thedistributionaldifferencesamongcountries

havenotinvalidatedourfindings.

Aseconddrawbackwasthelargevariationinhouseholdresponses,whichinsome

countrieswereverylow,mostnotablyinGermany(48%)andIreland(56%).Inaddi-

tion,amongsubjectslosttofollow-upbetweenannualquestionnaires(attritionrate)

those with younger age, lower education, unmarried, and male gender were over-

represented.[16]Thelowhouseholdresponseinsomecountriescouldbeinfluenced

byhealth,butwhetherdifferentialbiasispresentremainsunknown.Asthenational

attritionratewasnotassociatedwiththeprevalenceofperceivedpoorhealthatbase-

line,wedonotexpectasubstantialbiasinthenon-responseduringthefollow-up.

Athirddisadvantagewastheuseofself-reportedlabourstatus,whichmaydiffer

fromofficialdefinitions.Forexample,apersonmayconsiderhimselfasunemployed

onlywhenheisactivelylookingforwork,whereasothersonunemploymentbenefit

may have categorized themselves as economically inactive.Another example is the

inability to distinguish illness-based unemployment or retirement from non-illness

based unemployment or retirement. Subjects becoming or remaining unemployed

0

0.5

1.0

1.5

2.0

2.5

3.0

0 5 10 15 20

OR

- va

lue

OR Poor health - unemploymentOR Poor health - employmentLinear (OR Poor health - unemployment)Linear (OR Poor health - employment)

Unemployment rate at national level

Figure 4.2 Themagnitudeoftheassociationbetweenaperceivedpoorhealthandbecomingunemployedorenteringemployment(OR)againstthenationalunemploymentratein1998among11countriesintheEuropeanHouseholdPanelSurvey

66 Chapter4

mayincludeasubstantialproportionofsubjectswith(workrelated)disability.Inthe

baselinesurvey,approximately2%oftherespondentsclassifiedasunemployedstated

thatthemainreasonfornotseekingworkwastheirownillness,injuryorincapacita-

tion.Hence,thiswillhaveonlyasmalleffectontheobservedinfluenceofpoorhealth

onthelikelihoodofenteringpaidemployment.

Theresultsofthisstudysupporttheexistenceofahealthselectionintheworkforce

ashealthierpeoplearemorelikelytobecomeorremainemployedthanlesshealthy

people.Severalstudieshavereportedasimilarimpactofhealthonenteringpaidem-

ploymentorleavingtheworkforce.IversenandSabroe[20]showedthatpsychological

healthremainedstableforthoseemployedorunemployedduringa2yearfollow-up,

whereas for those losingorgainingemployment,psychologicalhealthdecreasedor

increased,respectively.Thelikelihoodoffindingpaidworkwasnegativelyinfluenced

by mental health problems in British workers [7] and by health-related difficulties

in Finland.[21] A Swedish longitudinal study showed that subjects with limiting

longstandingillnesshadanincreasedriskofbecomingunemployed[10]andamong

Finnishmiddle-agedmen,apoorhealthpredictednon-illnessbasedearlyretirement.

[11]Theinfluenceofage,education,andmaritalstatusonemploymentstatusreflects

well-knownlabourmarketconditions-forexample,inallmosteveryEuropeancountry,

unemploymentishighestamongyoungerpeopleandamongthosewithalowlevel

of educational attainment.[22] The specific contribution of the current study is that

itshowsthatpoorhealthhadalong-termeffectonenteringpaidemployment,buta

muchmoretransienteffectonleavingtheworkforce.Poorhealthwasariskfactorfor

notenteringpaidemploymentalreadyfouryearsbeforethepotentialtransitioninto

theworkforce.Itisofinteresttonotethatinthecurrentstudytheeffectofpoorhealth

on entering paid employment was consistently present up to four years before the

actualtransitiontookplace.Thissuggeststhatthehealthstatusamongunemployed

subjectswithpoorhealthremainsstableovertimeand,hence,hasalong-termeffect

ontheprobabilityofenteringpaidemployment.Incontrast,theeffectsofpoorhealth

on unemployment, retirement or taking care of household were less pronounced in

moredistantyears,suggestingamoretransientpattern,withachangeinhealthbeing

theriskfactorratherthanpoorhealth.

Fewstudieshavereportedoninteractionsbetweenhealthstatusandsocioeconomic

position. McDonough and Amick[13] demonstrated that, in the USA, the effect of

perceivedillhealthonlabourmarketexitdependedonsex,race,andeducation.Poor

healthwasamoreimportantfactorinleavingtheworkforcethanamongwomen,and

theeffectofpoorhealthonlabourforceexitincreasedwitheducation.Bothfindings

werecorroboratedinourstudy(table4.2and4.3,respectively).Theeffectofhealthon

employmentstatuswilldifferacrossdifferentsocioeconomicpositionsastheconse-

quencesofpoorhealthalsodependonsocialandlabourmarketcircumstances.These

Ch

apte

r4


lattercircumstancesvaryacrossthe11Europeancountriesinourstudy-forexample,by

thelevelofprotectionforworkerswithchronicdiseasesagainstworkforceexclusion

andrehabilitationpolicies,toincludepeoplewithpoorhealthinregularorsheltered

employment.[23]Itmaybehypothesizedthatcountrieswhereinapoorhealthstatus

wasastrongpredictorforlabourforceexitwerelesssuccessfulinretainingthosewith

apoorhealthstatusinthelabourforce.[14]Analternativeexplanationmaybethatin

countrieswherethisassociationwasnotobserved,healthislessimportantthanother

factorsindeterminingemploymentstatus.Figure2suggeststhat,incountrieswitha

lownationalunemploymentrate,healthwillcompetewithotherlabourmarketfactors

intheprocessofenteringorretainingpaidemployment,whereasincountrieswitha

highnationalunemploymentrate,theeffectofhealthselectionoutoftheworkforce

isrelativesmallcomparedwithotherfactorsthatdeterminelabouropportunitiesfor

people. However, this finding does not suggest that is these latter countries, health

selection per sé is not important, but rather that it cannot be demonstrated in the

currentanalysis.

An importantquestion iswhat society shoulddowithunhealthy job seekersand

unhealthy workers. Preventive public health interventions should not only address

thehealthselectionoutoftheworkforce,butalsotakeintoconsiderationthatbeing

unemployedcan lead to furtherdeteriorationof thehealth status, long-termunem-

ployment,andsocialexclusion.[23]Illhealthisanimportantdeterminantforentering

andmaintainingpaidemploymentinmanyEuropeancountries.Theseconsequences

of ill health will increase socioeconomic inequalities in health. It has been reported

before that unemployment, retirement, and work disability explain a great deal of

health inequalities across European countries.[24]. In policies for health equity, it is

ofparamountimportancetodeveloppublichealthmeasures,tailoredtosocio-demo-

graphiccharacteristics, that includepeoplewithapoorhealth in the labourmarket

andpreventworkerswithillhealthfromdroppingoutoftheworkforce.

68 Chapter4

ReFeRences

1. AchesonD.Independentinquiryintoinequalitiesinhealth.London:theStationaryOffice,1998.http://www.doh.gov.uk/ih/ih.htm(accessedOctober172005).

2. MarmotMG,SmithG,StansfieldS,etal.HealthinequalitiesamongBritishcivilservants:theWhitehallIIstudy.Lancet1991;337:1387-93.

3. ClaussenB.Healthandre-employmentinafive-yearfollowupoflong-termunemployed.Scand J Public Health1999;27:94-100.

4. JanlertU.Unemploymentasadiseaseanddiseasesoftheunemployed.Scand J Work Environ Health1997;23suppl3:79-83

5. MorrisJK,CookDG,ShaperAG.Lossofemploymentandmortality.BMJ1994;308:1135-9. 6. BartleyM,SackerA,ClarkeP.Employmentstatus,employmentconditions,andlimitingill-

ness:prospectiveevidencefromtheBritishhouseholdpanelsurvey1991-2001.J Epidemiol Community Health2004;58:501-506

7. ThomasC,BenzevalM,StansfeldSAEmploymenttransitionsandmentalhealth:ananalysisfromtheBritishhouseholdpanelsurvey.J Epidemiol Community Health2005;59:243-9.

8. MurphyGC,AthanasouJA.Theeffectofunemploymentonmentalhealth.J Occup Organ Psychol1999;72:83-99.

9. ClaussenB,BjorndalA,HjortPF.Healthandre-employmentinatwo-yearfollowupoflong-termunemployed.J Epidemiol Community Health1993;47:14-8.

10. LindholmC,Burström,DiderichsenF.Classdifferencesinthesocialconsequencesofillness?J Epidemiol Community Health2002;56:188-92.

11. KarpansaloM,ManninenP,KauhanenJ,LakkaTA,SalonenJ.Perceivedhealthasapredictorofearlyretirement.Scand J Work Environ Health2004;30:287-292

12. LoprestP,DavidoffA.Howchildrenwithspecialhealthcareneedsaffecttheemploymentdecisionsoflow-incomeparents.Matern Child HealthJ2004;8:171-82.

13. McDonoughP,AmickIIIBC.Thesocialcontextofhealthselection:alongitudinalstudyofhealthandemployment.Soc Science Med2001;53:135-45.

14. OECD.Transformingdisabilityintoability.Policiestopromoteworkandincomesecurityfordisabledpeople.Paris,OECD,2003.

15. Eurostat.(1999).EuropeanCommunityHouseholdPanel,Users’DatabaseManual(VersionofNovember1999),Luxembourg

16. PeracchiF.TheEuropeanCommunityHouseholdPanel:Areview.Emp Econ2002;27:63-90. 17. HuismanM,KunstAE,MackenbachJP.Inequalitiesintheprevalenceofsmokinginthe

EuropeanUnion:comparingeducationandincome.Prev Med2005;40:756-64. 18. UNESCO,(1997).InternationalStandardClassificationofEducation1997. 19. FayersPM,SprangersMAG.Understandingself-ratedhealth.Lancet2002;359:187-8. 20. IversenL,SabroeS.Psychologicalwell-beingamongunemployedanemployedpeopleafter

acompanyclosedown:alongitudinalstudy.J Soc Issues1988;44:141-52 21. PohjolaA.Healthproblemsandlong-termunemployment.Soc Work Health Care2001;34:101-12. 22. Eurostat.EuropeanUnionLabourForceStudy.http://forum.europa.eu.int/irc/dsis/em-

ployment/info/data/eu_lfs(accessedOctober172005) 23. GrammenosS.Illness,disabilityandsocialinclusion2003.EuropeanFoundationforthe

ImprovementofLivingandWorkingConditions.ISBN92-897-022104. 24. VanDoorslaerE,KoolmanX.Explainingthedifferencesinincome-relatedhealthinequalities

acrossEuropeancountries.Health Econom2004;13:609-28.

5The effect of re-employment on perceived health

SchuringM,MackenbachJP,VoorhamAJJ,BurdorfA.

Journal of Epidemiology and Community Health, in press.

70 Chapter5

absTRacT

background: Therelationshipbetweenunemploymentandpoorhealthhasbeenwell

established. Unemployment causes poor health and poor health increases the prob-

abilityofunemployment.

methods:A prospective study with 6 months follow-up was conducted among un-

employed subjects receiving social security benefits, who were capable of full time

employmentandwerereferred toa re-employment trainingcentre.Re-employment

was defined as ending social security benefits for at least three months because of

starting with paid employment. Health related quality of life was measured by the

ShortForm36HealthSurvey(SF-36).ACoxProportionalHazardsanalysiswasused

todetermine the factors thatpredicted re-employmentduring follow-up.The influ-

enceofre-employmentonchanges inperceivedhealthwas investigatedwith linear

regressionanalysis.

Results: Unemployedsubjectswithapoorhealthatbaselinewerelesslikelytoreturn

topaidemploymentduringfollow-up.Almostalldimensionsofhealthatbaselinehad

aninfluenceonthelikelihoodofbecomingemployed.Amongthere-employedsub-

jects,generalhealth,physicalfunctioning,socialfunctioning,vitality,mentalhealth,

bodily pain, and role-limitations due to emotional or physical problems improved,

withaneffectsizevaryingfrom0.11to0.66.

conclusion: Thisstudyprovidesevidencethatre-employmentleadstoimprovement

ofself-perceivedhealthwithinashorttimewindow.Thissuggeststhat labourforce

participationshouldbeconsideredastherapeuticinterventionwithinhealthpromo-

tionprogrammesamongunemployedpersons.

Ch

apte

r5

Theeffectofre-employmentonperceivedhealth 71

inTRoDucTion

Therelationshipbetweenunemploymentandpoorhealthhasbeenwellestablished,

as demonstrated by a higher prevalence of illness and disability[1-2] and a higher

mortalityamongunemployedpeople.[3]Theassociationbetweenhealthandemploy-

mentisbi-directional:unemploymentmaycausepoorhealth(causationhypothesis),

andpoorhealthmayincreasetheprobabilityofunemployment(selectionhypothesis).

There is evidence for the selectionhypothesis.A longitudinal studyamongEuro-

peancountries showed that in themajorityofEuropeancountriesaperceivedpoor

healthorachronichealthproblempredictedbecomingorstayingunemployed.[4]A

community based survey in the United Kingdom found that health related job loss

hadbecomeincreasinglycommon,especiallyinrelationtomusculoskeletaldisorders

andmentalillness.[5]Atwo-yearfollowupstudyoflongtermunemployedNorwe-

giansreportedthathealthrelatedselectionto longtermunemploymentexplaineda

substantialpartoftheexcessmentalmorbidityamongunemployedpeople.[6]

However, there is also evidence for the causation hypothesis. Various studies

haveshownthatunemploymentgaverise tohealtheffects,especiallypsychological

distress,depression,andreducedmentalhealth.Areviewof16longitudinalstudies

concerningmentalhealtheffectofunemploymentconcludedthatlossofemployment

affected mental health, but also that gaining employment improved mental health.

[7] A British longitudinal study found that transitions from paid employment to

various forms of non-employment (unemployment, long-term sick leave, maternity

leave)hadanegativeimpactonmentalhealth.Transitionsfromnon-employmentto

formalemploymentresultedinanimprovementofmentalhealth.Theeffectswerefelt

most stronglywithinsixmonthsafter the transition.[8]A five-year followupstudy

amonglong-termunemployedNorwegiansreportedrecoveryofmentalhealthafter

re-employment.[9]

“Work, matched to one’s knowledge and skills and undertaken in a safe, healthy

environment, can reverse theharmfuleffectsofprolongedsicknessabsenceor long

term unemployment, and promote health, well-being and prosperity”. This is the

main message of the so-called Black report, which reviewed the health of Britain’s

working-age population[10-11] In a review Waddel and Burton[12] concluded that

re-employment leads to clear benefits in psychological health and some measures

ofwell-being,althoughthere isadearthof informationonphysicalhealth.Ameta-

analytic study of the psychological and physical well-being during unemployment

also demonstrated that the bulk of research is focused on mental health outcomes,

suggesting that other aspects of health need more attention.[13] One study on re-

employmentindicatedthatphysicalaswellasmentalhealthimprovedamongthose

workingatoneyearfollow-upcomparedtothosenotworking.[14]Anotherstudyalso

72 Chapter5

showedapositiveassociationbetweengainingemploymentandphysicalfunctioning

among older workers who were displaced.[15] However, both studies were among

olderworkerswhowererecentlydisplacedorearlyretired.Itisnotknownwhether

theseresultscanbegeneralizedtothewholeworkingagepopulationwithadifferent

unemploymenthistory.

The impact of re-employment on mental health is reported by two meta-analytic

studies.[7,13]Thereislimitedinsightintotheeffectsizeofre-employmentonother

dimensions of health.[13] The aim of this study was to investigate the effect of re-

employment on different dimensions of health within a short period after entering

paidemployment.

meTHoDs

Studypopulation

A prospective study with 6 months follow-up was conducted among unemployed

subjects. The study population consisted of persons on social security benefits who

were capable of full time employment and who were referred by the Employment

CentreoftheCityofRotterdam,TheNetherlands,tooneofthefourre-employment

trainingcentresintheareaforare-employmenttraining.Someoftheparticipantsdid

havechronichealthproblems,butweredeclaredfitenoughtobecapableoffulltime

employmentafter investigationbyaphysician, apsychologist, andanemployment

specialist. From December 2004 until December 2007, every week an average of 19

subjectswasenrolledinthestudyafterreferraltoare-employmenttrainingcentre.In

total,2754eligibleparticipantswereincludedinthestudy.Participationinthisstudy

wasvoluntary.TheMedicalEthicsCommitteeofErasmusMCprovidedadeclaration

ofnoobjection.

Datacollection

Thefirstquestionnairewassenttoprospectiveparticipantsimmediatelyafterthere-

ferraltothere-employmenttrainingcentre.Thefollow-upquestionnairewassentsix

monthslater.Theprocedureofdatacollectionwassimilaratbaselineandfollow-up.

Thefirstquestionnairewassenttothehomeaddressoftheparticipants,followedby

tworeminderstworespectivelyfourweekslater.Additionalactionswereundertaken

to include more subjects.As a large part of the study population had a non-Dutch

background,thequestionnaireandcoveringletterweretranslatedinTurkishandsend

inadditiontotheDutchquestionnairetosubjectswithaTurkishsurname.Ifsubjects

Ch

apte

r5


ofthestudypopulationneededhelpwithfillinginthequestionnaire,theycouldgetin

touchwithaninterviewer.Subjectswhodidnotreplytothepostalquestionnairewere

visitedbyan interviewerat theirhomeaddresswith fourattemptsatdifferentday

times during a two week period. The interviewers were matched with the subjects,

basedonethnicity,age,andsex,andcouldofferan interview in themother tongue

(Dutch,Arabic,orTurkish).


Socio-demographic variables, such as ethnic background, highest educational level,

age,sex,andmaritalstatuswereincludedinthequestionnaire.Ethnicbackgroundof

therespondentwasbasedonthecountryofbirthof themother. Incase themother

wasborninTheNetherlands,thecountryofbirthofthefatherwasleading.[16]Dif-

ferentethnicgroupsweredefined,basedondifferencesingeographicalandcultural

distancefromtheNetherlands.Threeethnicminoritygroupsweredefined:1)Turks

and Moroccans, 2) Antilleans and Surinamese, and 3) a miscellaneous group with

all other countries of origin. Subjects were divided into three groups according to

theirhighestlevelofeducationalattainment.Ahigheducationallevelwasdefinedas

highervocationaltrainingoruniversity,intermediateeducationallevelwasdefinedas

highersecondaryschoolingorintermediatevocationaltraining,andloweducational

levelwasdefinedasnoeducation,primaryschool,lowerandintermediatesecondary

schoolingor lowervocational training.Maritalstatuswasusedtodistinguish those

subjectsmarriedorlivingtogetherfromothers.

Psychologicalmeasures

Mastery was measured by the Personal Mastery Scale[17], which consists of seven

items (eg “I have little control over the things that happen to me”, “There is little

I can do to change many of the important things in my life”), answered on a four

point Likert scale (strongly agree to strongly disagree).Average scores across items

werecalculated,rangingfrom1to4,withahigherscoreindicatingahigherlevelof

mastery.Incasethreeormoreitemswereunanswered,noscorewascomputed.

Self-esteemwasmeasuredwiththeRosenbergSelf-EsteemScale[18],with10items

(e.g.,“Onthewhole,Iamsatisfiedwithmyself”,“Allinall,Iaminclinedtofeelthat

I am a failure”), answered on a four point Likert scale (strongly agree to strongly

disagree).Averagescoresacross itemswerecalculatedagain, ranging from1 to4;a

higherscoreindicatedahigherlevelofself-esteem.Incasethreeormoreitemswere

unanswered,noscorewascomputed.

74 Chapter5

Re-employmenttraining

Are-employmenttrainingcentreprovidedastandardizedapproachofare-employ-

ment training, characterizedbyabroadre-orientationonemploymentandemploy-

ability,enhancementofjobsearchskills,andintensificationofjobsearchefforts.

Re-employment

StartandenddatesofthesocialsecuritybenefitswereregisteredattheEmployment

Centre of the City of Rotterdam, The Netherlands. In these registers, additional in-

formationaboutreasonsforendingbenefitswasalsoadministered.Re-employment

was defined as leaving the social security benefit services for at least three months

because of starting with paid employment, verified by the national Social Security

Agency.Subjectswhoquittedtheirregistrationforasocialsecuritybenefitforother

reasons,forexamplemovinginwithapartnerormovingoutofthecityofRotterdam,

werecensoredfromthemomentthattheirbenefitpaymentwasquittedbythesocial

securityservices.

Healthmeasures

HealthrelatedqualityoflifewasmeasuredwiththeDutchversionoftheShortForm

36 Health Survey (SF-36).[19-20] Self-reported health (SRH) was measured with the

first itemof theSF-36byasking subjects to rate theiroverallhealthona five-point

scale,rangingfrom‘excellent’,verygood’,‘good’and‘fair’to‘poor’.Thosereporting

lessthan‘goodhealth’weredefinedashavingapoorhealth.[21]

Theother35itemsoftheSF-36wereusedtocalculatescoresoneightdimensions:

physical functioning, general health, mental health, bodily pain, social functioning,

vitality, role limitation due to emotional health problems, and role limitation due

to physical health problems. Scores could range from 0 to 100, with a higher score

indicatingabetterhealthrelatedqualityoflife.

Statisticalanalysis

AllstatisticalanalyseswereconductedbymeansofthestatisticalpackageSPSS(ver-

sion15)forWindowsandthelevelofsignificancewassetat0.05.

In the analyses two groups of respondents were compared, those who were still

unemployedatfollow-upandthosewhowerere-employedatfollow-up.Thebaseline

characteristics of both groups were compared with the chi-square test for dichoto-

mousdataandthet-testforcontinuousdata.Non-responsetothefirstquestionnaire

Ch

apte

r5


and loss-to-follow up during the six months until the second questionnaire were

investigated by logistic regression analysis with potential determinants; individual

characteristics, perceived health, participation in re-employment training, and re-

employment.

A Cox Proportional Hazards analysis was used to determine the factors that pre-

dicted re-employment during the follow-up period. The follow-up period was cen-

soredatdateofre-employment.Independentfactorswereindividualcharacteristics,

selfesteem,mastery,andtheeightdimensionsof theSF36.Variableswerecoded in

such a manner that a Hazard Ratio above 1 indicates an increased likelihood of re-

entering paid employment. For each dimension of the SF36 a standardized Hazard

Ratiowascalculatedbasedonanaveragescoreacrossitems,representingtheeffectof

anincreaseofonestandarddeviationintheaveragescoreatbaselineonthelikelihood

ofenteringpaidemployment.Individualcharacteristicswithasignificantunivariate

effectonre-employment,wereenteredfirstinthemultivariatemodel.Subsequently,

the health measure with the largest goodness-of-fit in the univariate analysis was

enteredinthemultivariatemodel.Afterthat,theeffectoftheothermeasuresofhealth

wereinvestigatedfortheiradditionaleffectongoodness-of-fit.

The association of re-employment with changes in perceived health was investi-

gated with linear regression analysis. The association of re-employment with each

dimensionofhealthwasadjustedforage,sex,ethnicbackground,education,duration

onbenefit,participationinare-employmenttraining,andalsoforthebaselinevalue

of the health dimension under study. For each dimension of health Cohen’s d was

calculated as measure of effect size by dividing the difference in health before and

afterre-employmentbytheirpooledstandarddeviation.[22]

ResuLTs

Figure5.1shows that2754subjectswereenrolled in thestudyafter referral toa re-

employmenttrainingcentre.Onethirdofthesubjects(n=932)hadchronichealthcom-

plaints,butweredeclaredtobefitenoughtobecapableoffulltimeemploymentafter

investigationbyaphysician,apsychologistandanemploymentspecialist.Fromthe

2754subjectswhoreceivedthefirstquestionnaire,1829subjects(66%)filledoutand

returnedthequestionnaire.Morethantwothirdoftherespondents(70%)returnedthe

baselinequestionnairebypost,whereasalmostonethirdoftherespondents(30%)had

a face-to-face interview. Non-response was statistically significantly higher among

youngersubjectsandmen.

Theresponseatfollow-upwas53%(965/1829).Loss-to-followupwasstatistically

significantlyhigheramongyoungersubjects,men,andsubjectsofnon-Dutchorigin,

76 Chapter5

butnotrelatedtomaritalstatus,employmenthistory,durationonbenefit,orhealthat

baseline.Subjectswhohadstartedwithare-employmenttraining(n=461)were less

oftenlosttofollowup(OR=0.8,95%CI0.7-1.0),whereassubjectswhohadreturned

topaidemployment(n=123)weremoreoftenlosttofollowup(OR=1.9,95%CI1.3-

2.7).However,amongthosewhoreturnedtopaidemployment,healthatbaselinedid

not differ significantly between non-respondents (n=76) and respondents (n=47) at

follow-up.

Table5.1showsthecharacteristicsoftherespondentstothefirstquestionnaire.The

meanageofrespondentswas39.7(sd9.5)years,49%wasman,57%hadalowlevelof

education,75%belongedtoanethnicminoritygroup,and43%receivedsocialsecurity

benefitsformorethanfiveyears.Duringthefollow-upperiod30%oftherespondents

hadstartedwithare-employmenttraining.

Figure5.2showsthathealthatbaselinewasbetteramongsubjectswhoreturnedto

paid employment during the follow-up period.All dimensions of health improved

Persons on social security benefits.Chronic health problems?

Referral to a re-employment training centerN=2754

Exemption from the obligation to work

Investigation by a physician, psychologist and an employment specialist.Capable of full time employment?

Enrollment in the studyN=2754

Response at baselineN=1829- 70% returned by post- 30% face-to-face interview

Non-response at baselineN=925

Response at follow-upN=965- 79% returned by post- 21% face-to-face interview

Non-response at baselineN=864

Yes

YesN=932

NoNoN=1822

Figure 5.1 Flowofparticipantsandresponseatbaselineandfollow-up

Ch

apte

r5


among re-employed subjects, whereas the health status of unemployed subjects re-

mainedunchanged.

Table5.2showsthatunemployedsubjectswithapoorhealthatbaselinewereless

likely to return to paid employment during follow-up (OR=0.39, 95% CI 0.2-0.7).

Alldimensionsofperceivedhealthatbaselinehadaninfluenceonthelikelihoodof

becomingemployed,exceptformentalhealth.Physicalfunctioninghadthestrongest

influence on the likelihood of becoming employed (OR=2.76, 95% 1.8-4.3). Ethnic

background,maritalstatus,selfesteem,mastery,andparticipatinginare-employment

trainingwerenotsignificantintheunivariatemodel.

Inthemultivariateanalysis,physical functioningatbaselinehadthestrongestas-

sociationwith re-employment.Due to the correlationsof theotherhealthmeasures

withphysicalhealth(generalhealthr=0.56,bodilypainr=0.62,mentalhealthr=

0.26, social functioning r=0.43, vitality r=0.41, role functioning, emotional r = 0.27,

rolefunctioning,physicalr=0.55),thesemeasuresofhealthdidnothaveasignificant

contributiontothemultivariatemodelincludingphysicalfunctioning.

Table5.3showsthatamongthere-employedsubjectsgeneralhealth,physicalfunc-

tioning, social functioning, vitality, mental health, bodily pain, and role-limitations

duetoemotionalorphysicalproblemsimprovedduringthefollow-upperiod,withan

effectsizevaryingfrom0.11to0.66.Thelargestrelativeimprovementswereobserved

Table 5.1 Characteristicsofrespondentstothefirstquestionnaire(n=1829)

variableRespondents n=1829

Age(n-%)18-44yr45-64yr

1230(67)599(33)

Men(n-%) 894(49)

Maritalstatuslivingwithpartner(n-%) 601(33)

Educationallevel(n-%)HigherandintermediatelevelLowerlevelMissing

634(35)1049(57)146(8)

Ethnicbackground(n-%)NativeDutchTurkish/MoroccanAntillean/SurinameseRefugee/Otherimmigrants

450(25)448(25)488(27)404(23)

Durationonsocialbenefit(n-%)LessthanoneyearBetween1and5years5yearsandmore

478(27)512(29)759(43)

Startedwithre-employmenttraining(n-%) 541(30)

Selfesteem(scale1-4:mean-sd) 2.48(0.6)

Mastery(scale1-4:mean-sd) 2.89(0.5)

78 Chapter5

formentalhealth,socialfunctioning,androle-limitationsduetoemotionalorphysical

problems, whereas physical functioning showed the smallest relative improvement.

For those subjects who remained unemployed the effect sizes varied from -0.04 to

0.06,indicatingthattheirhealthstatusremainedvirtuallyunchangedduringthesix

monthsfollow-upperiod.Participation inare-employmenttrainingwasnotassoci-

atedwithchangeofhealthstatus(datanotshown).

Discussion

Unemployedsubjectswithapoorhealthatbaselinewerelesslikelytoreturntopaid

employment during follow-up.Almost all dimensions of health at baseline had an

influence on the likelihood of becoming employed. The strongest association was

foundbetweenphysical functioningatbaselineand re-employment.Among the re-

employed subjects, general health, physical functioning, social functioning, vitality,

mentalhealth,bodilypain,androle-limitationsduetoemotionalorphysicalproblems

improved,withaneffectsizevaryingfrom0.11to0.66.Hence,re-employmentposi-

tivelyinfluencedmentalhealthaswellasphysicalhealth.

0

20

40

60

80

100

120Se

lf-r

ated

hea

lth o

n 0-

100

scal

e

gene

ral

heal

th

role

lim

itatio

n-ph

ysic

alpr

oble

m

bodi

lypa

in

men

tal

heal

th

soci

alfu

nctio

ning

vita

lity

phys

ical

func

tioni

ng

baseline follow-up Re-entered paid employment (n=47)

baseline follow-up Continued to be unemployed (n=918)

role

lim

itatio

n-em

otio

nal

prob

lem

Figure 5.2 Healthatbaselineandfollow-upforsubjectswhocontinuedtobeunemployed(n=918)andsubjectswhore-enteredpaidemployment(n=47)duringthefollow-upperiodofsixmonths

Ch

apte

r5


Table 5.2 Theinfluenceofindividualcharacteristicsanddimensionsofself-perceivedhealthamonglong-termunemployedpersons(n=965)onthelikelihoodofre-employment(n=47)duringafollow-upperiodofsixmonths.(CoxProportionalHazardsanalysis)

Re-employment (univariate model)

Re-employment (multivariate model)

Hazard Ratio (95%ci) Hazard Ratio (95%ci)

Higherage(>45years) 0.40(0.20-0.83)** 0.45(0.19-1.10)*

Men 2.19(1.21-3.97)** 1.74(0.91-3.31)

Loweducationallevel 0.52(0.30-0.93)** 0.60(0.32-1.12)

Poorhealth 0.39(0.22-0.69)** -

Generalhealth† 1.57(1.20-2.07)** -

Physicalfunctioning† 2.76(1.80-4.24)** 2.18(1.40-3.38)**

Bodilypain† 1.92(1.42-2.60)** -

Mentalhealth† 1.19(0.89-1.59) -

Socialfunctioning† 1.54(1.13-2.10)** -

Vitality† 1.48(1.12-1.95)** -

Rolefunctioning,emotional† 1.44(1.06-1.96)** -

Rolefunctioning,physical† 1.72(1.25-2.37)** -

**P<0.05,*P<0.10†StandardizedHazardRatio,representingtheeffectofanincreaseofonestandarddeviationintheaveragescore.

Table 5.3 Theinfluenceofre-employmentonchangesineightdimensionsofhealthmeasuredbytheSF-36healthquestionnaireamonglong-termunemployedpersonsduringafollow-upperiodofsixmonths.

effect of re-employment change† (se)(n=965)

effect size‡(cohen’s d)(n=47)

Generalhealth +7.0(2.7)** +0.18

Physicalfunctioning +9.2(3.4)** +0.11

Bodilypain +11.3(3.6)** +0.20

Mentalhealth +11.0(2.7)** +0.66

Socialfunctioning +14.2(3.8)** +0.32

Vitality +7.8(2.5)** +0.26

Rolefunctioning,emotional +22.7(6.8)** +0.46

Rolefunctioning,physical +20.0(6.0)** +0.33

†.Eachlinearregressionmodelwasadjustedforage,gender,ethnicbackground,education,durationonbenefit,andhealthatbaseline.‡:Effectsizeswerebasedonthemeanvaluesofhealthatbaselineandfollow-upofthere-employedsubjects

80 Chapter5

Losstofollowupwassignificantlyhigheramongyoungersubjects,men,subjects

ofnon-Dutchoriginandsubjectswhohadreturned topaidemployment.However,

among those who found employment (n=123), health at baseline did not differ be-

tween non-respondents (n=76) and respondents (n=47) at follow up. Therefore, we

assumethattheeffectsofre-employmentonhealthwerenotinfluencedbyselective

losstofollowup.

Subjects who were referred to a re-employment training centre were included in

thestudy.ThepolicyoftheEmploymentCentreoftheCityofRotterdamwastorefer

everybodywhowasabletoworkandcouldusesomehelpwithsearchingforajob.

Hence,subjectswerenotincludedinthestudywhentheywerenotable(duetohealth

problems)ornotobliged(duetofamilyobligationsoreducation)towork.Inaddition,

recentlyunemployedpersonswhoweresupposedtobeabletofindpaidemployment

bythemselves,werenotincluded.Thismayhaveaffectedtheresultswithrespectto

healthstatusandre-employmentprobabilities.

There-employedworkersgottheirjobsatdifferenttimesduringthefollow-upperi-

odofthestudy.Therefore,somepersonsmayhavebeenbackatworkformuchlonger

thanothersandthus,thelengthofexposuretothehealthbenefitsofre-employment

variedamongstthere-employedsample.Unfortunately,duetothesmallnumberof

personswhogainedemploymentitisnotfeasibletoanalysetherelationbetweentime

atworkandhealthbenefitsduetore-employment.

Thestudyisbasedonself-reportedhealthstatus.Therefore, justificationbiasmay

haveinfluencedthepresentresults.Respondentswhowerestillunemployedatfollow

upmayhavereportedmoreeasilythattheyhadapoorhealththanthosewhofound

paidemploymentinordertojustifythefactthattheydidnotfindajob.[23]

Duetotheobservationaldesignofthestudytheresultsmayhavebeeninfluenced

by selection bias. Subjects who found employment differed from those who stayed

unemployed with respect to health at baseline and personal characteristics. These

variableswereadjustedfor intheanalysis,buttheremayhavebeenothervariables

thatinfluencedbothhealthchangeandre-employment,e.g.motivationtowork,that

were not included in the analysis. However, two psychological measures that were

includedinthestudy,selfesteemandmastery,didnothaveasignificantinfluenceon

the likelihoodof re-employment. Inaddition, thesepsychologicalmeasuresdidnot

influencetheestimatedeffectsofre-employmentonhealth.

To control for the systematic differences in health at baseline between the unem-

ployed and re-employed subjects, stratified analyses were conducted showing that

re-employmentresultedinanincreaseofhealthforbothsubjectswithapoorhealth

atbaselineaswellassubjectswithagoodhealthatbaseline.Ingeneral,subjectswith

apoorhealthatbaselineshowedalargerincreaseinhealthafterre-employmentthan

subjectswithagoodhealthatbaseline(datanotshown).However,afterfindingpaid

Ch

apte

r5


employmentthehealthstatusofrecentlyre-employedsubjectsstilllaggedthehealth

statusoflong-termemployedpersons.[24]

Unemployedsubjectswithagoodhealthweremorelikelytoreturntopaidemploy-

mentduringfollow-up.Therefore,theresultsofthisstudysupporttheselectionhy-

pothesis.ThesefindingsareinaccordancewithastudyoftheEuropeanCommunity

HouseholdPanel,whichfoundthatapoorhealthorchronichealthproblempredicted

staying or becoming unemployed in European countries.[4] A two year follow up

study of long term unemployed in Norway showed that health related selection to

longtermunemploymentexplainedasubstantialpartoftheexcessmentalmorbidity

amongunemployedpeople.[6]

This study also showed that the health status of re-employed subjects improved,

whereasthehealthstatusofunemployedsubjectsremainedunchanged.Re-employ-

ment did not only have a positive influence on mental health, but also on physical

functioning, which was in accordance with findings from two other studies[14-15]

Basedonthesefinding,itseemsmostlikelythatstartingwithpaidemploymentresults

inhealthimprovement.However,duetothedesignofthestudyitcannotberuledout

thatitistheotherwayaround;achangeinhealthstatusmayincreasethelikelihoodof

findingpaidemployment.However,areviewshowedthatanincreaseinhealthstatus

oflongtermunemployedisratherunlikely.[2]Therefore,itisassumedthatthehealth

statusofparticipantsincreasedshortlyafterre-employment,supportingthecausation

hypothesis.This is incongruencewith findings fromotherstudies,which foundan

increaseinmentalhealthstatusafterre-employment.[8-9]

Inthepresentstudy,aneffectsizeof0.66wasfoundfortheimprovementofmen-

tal health after re-employment. This result is in accordance with findings from two

meta-analytic studies [7,13], reporting thatgainingemployment impactsonmental

health with an effect size of 0.54 respectively 0.89. The effects on other dimensions

wereslightlylower,butsincedifferentdimensionsofhealthwereassociatedwitheach

other it is difficult to infer that mental health seems more important than physical

health.ThisisinaccordancewiththeresultsoftheWhitehallIIstudyofcivilservants,

whichshowedthatphysicalandmentalhealthwerecorrelatedforparticipantswitha

lowsocioeconomicposition.[25]

Inthecurrentstudyonly5%(47/965)oftheparticipantsfoundpaidemployment

withinsixmonths.Thislowre-employmentrateisinaccordancewithastudyamong

personsreceivingsocialsecuritybenefitsinthecityofAmsterdam,theNetherlands,

whichshowedthatonly8%ofthepersonsfoundpaidemploymentinatwo-yearpe-

riodfrom2004-2006.[26]Thecurrentstudypopulationischaracterisedbyalong-term

unemployment history. Long-term unemployed subjects often have much difficulty

infindingpaidemployment,especiallyforthosewithapoorhealth.Allparticipants

in the study were referred to a re-employment training centre for a re-employment

82 Chapter5

training.However,thisre-employmenttrainingdidnothaveapositiveeffectonthe

probabilityofre-employmentoronthehealthstatusofparticipants.Therefore,other

measuresseemtoberequiredtoincreasethechanceofre-employmentforlongterm

unemployedwithapoorhealth.

IntherecentBlackreportonthehealthoftheBritishworkingagepopulation,Black

appealedforchangingperceptionsoffitnessforwork.Insteadofstickingtotheidea

thatonecannotworkunless100%fit, itisrecommendedthatacampaignshouldbe

launchedtomakeemployers,healthcareprofessionalsandthegeneralpublicaware

thatworkis ingeneralgoodforhealth.[10]Thisstudyprovidesevidencethatwork

isindeedgoodforyourhealthand,thus,workshouldbeconsideredasanimportant

part of health promotion programmes among unemployed persons. In policies for

health equity public health measures are required to include persons with a poor

healthinthelabourmarketandtopreventworkerswithillhealthfromdroppingout

oftheworkforce.

Ch

apte

r5


ReFeRences





5. SolomonC,PooleJ,PalmerKT,CoggonD:Health-relatedjobloss:findingsfromacommu-nity-basedsurvey.Occup Environ Med 2007,64:144-149.

6. ClaussenB,BjorndalA,HjortPF:Healthandre-employmentinatwoyearfollowupoflongtermunemployed.Journal of epidemiology and community health 1993,47:14-18.

7. MurphyGC,Athanasou,J.A.:Theeffectofunemploymentonmentalhealth.Journal of Oc-cupational and Organizational Psychology 1999,72:83-99.



10. BlackC:Workingforahealthiertomorrow.London: The Stationery Office 2008(http://www.workingforhealth.gov.uk/Carol-Blacks-Review/):ISBN:9780117025134

11. HulshofCT:Workingforahealthiertomorrow.Occup Environ Med 2009,66:1-2. 12. WaddellG,Burton,A.K.:Isworkgoodforyourhealthandwell-being?London: The Stationery

Office 2006(www.tsoshop.co.uk):ISBN:9780117036949. 13. McKee-RyanF,SongZ,WanbergCR,KinickiAJ:Psychologicalandphysicalwell-beingdur-

ingunemployment:ameta-analyticstudy.J Appl Psychol 2005,90:53-76. 14. PattaniS,ConstantinoviciN,WilliamsS:Predictorsofre-employmentandqualityoflife

inNHSstaffoneyearafterearlyretirementbecauseofillhealth;anationalprospectivestudy.Occup Environ Med 2004,61:572-576.

15. GalloWT,BradleyEH,SiegelM,KaslSV:Healtheffectsofinvoluntaryjoblossamongolderworkers: findingsfromthehealthandretirementsurvey.J Gerontol B Psychol Sci Soc Sci 2000,55:S131-140.

16. CBS:Herkomstvanpersonen;allochtonenenmigratie[Countryoforiginofpersons;mi-grantsandmigration].Centraal Bureau voor de Statistiek, Voorburg / Heerlen 2003.

17. PearlinLI,SchoolerC:Thestructureofcoping.J Health Soc Behav 1978,19:2-21. 18. RosenbergM:Societyandtheadolescentself-image.Princeton:PrincetonUniversityPress

1965. 19. WareJE,Jr.,SherbourneCD:TheMOS36-itemshort-formhealthsurvey(SF-36).I.Concep-

tualframeworkanditemselection.Med Care 1992,30:473-483. 20. VanderZeeKI,SandermanR:Hetmetenvandealgemenegezondheidstoestandmetde

RAND-36: een handleiding [Measuring health status with the RAND-36: a manual]. In

84 Chapter5

Groningen:NoordelijkCentrumvoorGezondheidsvraagstukken,RijksuniversiteitGron-ingen.1993.

21. FayersPM,SprangersMA:Understandingself-ratedhealth.Lancet 2002,359:187-188. 22. CohenJ:Statisticalpoweranalysisforthebehavioralsciences(2nded.).Hillsdale,NJ:Law-

renceEarlbaumAssociates1988. 23. KapteynA,SmithJP,vanSoestA:Dynamicsofworkdisabilityandpain.J Health Econ 2008,

27:496-509. 24. SchuringM,BurdorfA,KunstA,VoorhamT,MackenbachJ:Ethnicdifferencesinunemploy-

mentandillhealth.Int Arch Occup Environ Health 2009,82:1023-1030. 25. SackerA,HeadJ,GimenoD,BartleyM:SocialInequalityinPhysicalandMentalHealth

ComorbidityDynamics.Psychosom Med 2009. 26. Re-integratie-vanbijstandnaarwerk.Rekenkamer Amsterdam 2007.

6effectiveness of a health promotion programme

for long-term unemployed subjects with health

problems: a randomized controlled trial

SchuringM,BurdorfA,VoorhamAJJ,derWeduweK,MackenbachJP

Journal of Epidemiology and Community Health 2009, 63:893-899.

86 Chapter6

absTRacT

background:Employmentstatusisanimportantdeterminantofhealthinequalities.

Amongunemployedpersonsapoorhealthdecreasesthelikelihoodofre-employment.

methods:Arandomizedcontrolledtrialwith6monthsfollowupamongunemployed

personswithhealthcomplaintsreceivingsocialsecuritybenefitsfromthecityofRot-

terdam,theNetherlands.Intotal,456personswereassignedtothecontrolgroupand

465 persons to the intervention group. The intervention consisted of three sessions

weeklyover12weeks.Onesessionaweekwasfocusedoneducationtoenhancethe

abilitytocopewith(health)problemsandtwoweeklysessionsconsistedofphysical

activities. The primary outcome measures were perceived health, measured by the

ShortForm36HealthSurvey,andpsychologicalmeasuresmastery,selfesteem,and

pain-relatedfearofmovement.Secondaryoutcomemeasureswereworkvalues, job

searchactivities,andre-employment.

Results:Enrollmentintheinterventionprogrammewas65%and72%completedthe

programme with over 70% attendance to all sessions. The intervention had a good

reachamongsubjectswithlowereducation,buthadnoeffectonmentalandphysical

health,mastery,selfesteem,andpain-relatedfearofmovement.Participationinthe

programmehadnoinfluenceonworkvalues,jobsearchactivities,orre-employment.

conclusion:Theinterventionprogrammeaimedatpromotionofphysicalandmental

healthofunemployedpeoplewithhealthcomplaintsdidnotshowbeneficialeffects.

The lackof integration into regularvocational rehabilitationactivitiesmayhave in-

terferedwiththesefindings.Itcannotberecommendedtoimplementthisparticular

healthprogramme.

Ch

apte

r6

Effectivenessofahealthpromotionprogrammeforlong-termunemployedsubjectswithhealthproblems 87

inTRoDucTion


as demonstrated by a higher prevalence of illness and disability [1-2] and a higher

mortalityamongunemployedpeople.[3]Theassociationbetweenhealthandemploy-

mentisbi-directional:unemploymentmaycausepoorhealth(causationhypothesis),

andpoorhealthmayincreasetheprobabilityofunemployment(selectionhypothesis).

[4-6]

Workprovidesavarietyoffeatures,includingtheuseofskills,interpersonalcontact,

andprovisionofeconomicresources,thatareresponsibleforpsychologicalwell-being

andareadverselyinfluencedbyjoblossandunemployment.[7]Joblossmayleadto

impairedroleandemotionalfunctioning,poorhealth,anddepression.[8]Self-esteem

declineswithjobloss[9]andalowself-esteemisadeterminantofself-reportedpoor

health.[10]Apoorpsychologicalhealthwillactasabarriertoreturntopaidemploy-

mentthroughadecreasedmotivation,loweredexpectationsinfindingemployment,

andineffectivejobseeking.[8-9,11]Thus,unemploymentmayleadtoapoorerhealth,

whichinturnwillreducethelikelihoodofre-employment.

Inorderto increasethepossibilitiesforre-employment, improvementinhealthof

unemployedpersonsmay,therefore,beanimportantstep.Unemployedpersonswith

chronichealthcomplaints, suchasmusculoskeletaldisorders,maybecomeprogres-

sivelylesshealthy,sincefearofpainandreinjurymayleadtoreducedactivities[12],

resulting in a passive life style with low levels of physical activity.[13-14] Hence,

exercises to improvephysicalactivitymaybebeneficial,notonly for those subjects

withdisordersofthelocomotivesystem,butalsoforotherchronicdiseasesaswell,

including heart and pulmonary diseases and depression.[15] Cognitive-behavioural

therapymaybeneededtotargetspecificpain-relatedbeliefsandcopingstrategiesfor

modification.[16]RoseandHarris[17]haveidentifiedcognitivebehaviouraltherapy

asapromisinginterventiontoimprovethepsychologicalhealthofpersonswhoare

unemployed. Recently, Watson and colleagues [18] have provided some indications

that a combined physical exercise and cognitive behavioural programme improved

physical fitness as well as increased employment rates among unemployed partici-

pantswithhealthcomplaints.Theseresultsshouldbeinterpretedwithcaution,how-

ever,sincethevoluntaryparticipationintheprogrammemighthavebiasedtowards

participantswithahighmotivationandapositiveattitudetowards(returnto)work.

Hence,thereisaneedforrandomizedtrialsonmultidisciplinaryhealthintervention

onunemployedpersons.Theaimofthecurrentstudy(“Workonyourhealth”)was

toevaluatetheeffectivenessofahealthpromotionprogramme,consistingofphysical

exerciseandcognitivetraining,onphysicalandmentalhealthofunemployedsubjects

withhealthcomplaints.

88 Chapter6

meTHoDs

Designandstudypopulation

Thestudywasdesignedasarandomizedcontrolledtrial(RCT)andapprovedbythe

MedicalEthicsCommitteeoftheUniversityMedicalCenterRotterdamErasmusMC.

Personsonsocialsecuritybenefitswho(partly)attributedtheirinabilitytofindapaid

jobtotheirchronichealthproblems,werereferredbytheEmploymentCentreofthe

CityofRotterdam,TheNetherlands,forafit-to-worktest,conductedbyaphysician,

psychologist, and an employment specialist. All participants with health problems

anddeclaredtobecapableoffulltimeemploymentwereselectedastargetpopulation

forthisstudy.Formorethantwothirdofthepopulation(68%)thepresenceofchronic

painwasascertainedbyaphysician.Inaddition,subjectswererequiredtounderstand

andspeakDutchatbasiclevel.

Randomization

Randomization was performed by a researcher with a computer-generated list of

randomnumbers (SASSoftware,version8.12,Cary,NC).At the company thatper-

formed the fit-to-work tests, every week the researcher allocated all participants to

theinterventiongrouporthereferencegroup.FromDecember2004untilDecember

2007,everyparticipantwhomettheinclusioncriteriawasrandomized.Afterrandom-

izationeligibleparticipantswereapproachedby the research team forparticipation

in the study and asked to provide their written informed consent. This procedure

of informedconsentafterrandomizationwasnecessary,sincetheCityofRotterdam

requiredanimmediatereferraltoanemploymentserviceforvocationalrehabilitation

withoutfurtheradministrativedelay,andalsorequiredthatwillingnesstoparticipate

inthestudywouldnotinterferewithexistingrequirementsforjobsearch.

Theallocationsequencewasconcealeduntiltheparticipanthadbeenassigned.The

participants and the professionals providing the intervention could not be blinded.

However, the persons who were involved in data collection and data entry were

blinded,sincedatacollectionwasconductedcompletelyindependentfromtheinter-

ventionandregularvocationalrehabilitation.

Figure 6.1 shows the diagram of the flow of participants through the phases of

thetrial.Intotal,465subjectswereassignedtotheinterventionprogrammeand456

subjectstotheusualcaregroup.

Ch

apte

r6


Intervention

Theinterventionwasaimedatchangingthewayunemployedpersonsperceiveand

cope with their health complaints. The rationale was based on the biopsychosocial

modelofchronicpainandsubsequentinterdisciplinarypainmanagementapproach.

Patients with chronic pain are at increased risk for emotional disorders (such as

anxiety,depressions,andanger),maladaptivecognitions(suchascatastrophizingand

poorcopingskills),functionaldeficitsandphysicaldeconditioning(duetodecreased

physicalactivityandfearofinjury).Theseeffectsareofteninterdependent,sothatone

cannotsimplytreatonetotheexclusionofothers.Interdisciplinarypainmanagement

Non-respons Q1N=122

Lost to follow-upN=111

Lost to follow-upN=160

No participation N=104

no longer benefit n=8reintegration programme n=22unknown n=74

Non-respons Q1N=146

Enrollment of eligible subjectsN=921

Randomization

Allocated to interventionN=465

Allocated to care as usualN=456

Filled out questionnaire Q1N=343




Analysis per protocolN=128

AnalysisN=150

Analysis intention to treatN=176

Initial participationN=239

Figure 6.1 Flowofparticipantsthroughthephasesofthetrial

90 Chapter6

embracesthefactthatthecomprehensivetreatmentofallthesedimensionsisneeded

inordertobeeffective.[19]

The intervention consisted of three sessions each week during a 12-week period.

Every week one session of three hours was focused on behavioural education and

twosessionswerefocusedonphysicalactivity.Thebehaviouraleducationalcompo-

nent was designed to enhance a participant’s insight in his/her health complaints,

to increasepositivecopingwithhealthproblemsbyreducingfearandavoidanceof

movement,toenhanceself-esteemandfeelingsofmastery,andtoimprovefunction-

ingbylearningtothinkpositivelyandenhancesocialskills.Thispartwasconducted

bytwopreventionworkers.

Thephysicalactivitycomponentconsistedoftwosessionsofthreehours.Thefirst

sessioncomprisedof1.5hours fitness training (cardioandweight training)and1.5

hoursofindoorsports(e.g.basketball,swimming).Thesecondsessioncomprisedof

1.5 hours fitness training and 1.5 hours of outdoor activities. The intervention was

designedtomaketheparticipantsawarethat itwassafe tomoveandhealthytobe

physically active, to extend the social network, to improve daily structure, and to

improvegeneralwellbeing.Theexerciseprogrammewasdevelopedaccordingtothe

graded-activity principle.[20-22] The exercises started below the average functional

capacity assessed during the first session and were increased gradually during the

courseoftheintervention,accordingtothetime-contingencyprinciple.Thesesessions

wereconductedbyphysicaleducationteachers.

Usualcare

Subjectsinthereferencegroupwerereferredtooneofthethreevocationalrehabilita-

tionservicesinthearea.Theyprovidedastandardizedapproachofvocationalreha-

bilitation,characterizedbyabroadre-orientationonemploymentandemployability,

enhancementof jobsearchskills,and intensificationof jobsearchefforts.Theusual

careapproachdidnotincludeanyactivityrelatedtohealth,suchashealthpromotion

toincreasephysicalactivityinleisuretime.

Datacollection

Aquestionnaireandan informedconsentwassendto thehomeaddressof thepar-

ticipants, followed by two reminders two respectively four weeks later.Additional

actions were undertaken to include more subjects. The questionnaire and covering

letterweretranslatedinTurkishandsendinadditiontotheDutchquestionnaireto

subjectswithaTurkishsurname.Ifsubjectsofthestudypopulationneededhelpwith

fillinginthequestionnaire,theycouldgetintouchwithaninterviewer.Subjectswho

Ch

apte

r6


didnotreplytothepostalquestionnairewerevisitedbyaninterviewerattheirhome

address with four attempts at different day times during a two week period. The

interviewerswerematchedwithsubjects,basedonethnicity,age,andsex,andcould

offeraninterviewinthemothertongue(Dutch,Arabic,orTurkish).


Socio-demographic variables, such as ethnic background, education, age, sex, and

maritalstatuswereincludedinthestudy.Ethnicbackgroundoftherespondentwas

basedonthecountryofbirthofthemother.IncasethemotherwasborninTheNeth-

erlands, the country of birth of the father was leading.[23] Different ethnic groups

were defined, based on differences in experiences of migration (refugees or labour

migrants)anddifferencesingeographicalandculturaldistancefromtheNetherlands.

Three ethnic minority groups were defined: 1) Turks and Moroccans, 2) Antilleans

and Surinamese, and 3) a miscellaneous group with all other countries of origin.

Subjectsweredividedintothreegroupsaccordingtothehighestlevelofeducational

attainment. A high educational level was defined as higher vocational training or

university, intermediate educational level was defined as higher secondary school-

ing or intermediate vocational training, and low educational level was defined as

noeducation,primaryschool, lowerandintermediatesecondaryschoolingorlower

vocationaltraining.Maritalstatuswasusedtodistinguishthosesubjectsmarriedor

livingtogetherfromothers.

Primaryoutcomemeasures

Health measuresHealthrelatedqualityoflifewasmeasuredwiththeDutchversionoftheShortForm

36HealthSurvey(SF-36).[24-25].Self-reportedhealth(SRH)wasmeasuredwiththe

first itemof theSF-36byasking subjects to rate theiroverallhealthona five-point

scale,rangingfrom‘excellent’,verygood’,‘good’and‘fair’to‘poor’.Thosereporting

lessthan‘goodhealth’weredefinedashavingapoorhealth.[26]

Theother35itemsoftheSF-36wereusedtocalculatescoresoneightdimensions:


vitality, role limitation due to emotional health problems, and role limitation due

to physical health problems. Scores could range from 0 to 100, with a higher score

indicatingabetterhealthrelatedqualityoflife.

92 Chapter6

Psychological measuresMastery was measured by the Personal Mastery Scale [27], which consists of seven




were calculated, ranging from 1 to 4, with a higher score indicating a higher level

of mastery. In case three or more items were unanswered, no score was computed

(Cronbach’salpha=0.69).

Self-esteemwasmeasuredwiththeRosenbergSelf-EsteemScale[28]],with10items





unanswered,noscorewascomputed(Cronbach’salpha=0.84).

Kinesiophobia was measured with the Tampa Scale of Kinesiophobia [29], which

consists of 17 items on fear of movement and injury (e.g. “It’s really not safe for a

personwithaconditionlikeminetobephysicallyactive”,“PainalwaysmeansIhave

injuredmybody”)onafourpointLikertscale(stronglyagreetostronglydisagree).

Averagescoresacrossitemswereagaincalculated,rangingfrom1to4,withahigher

scoreindicatingahigherlevelofkinesiophobia.Incasefiveormoreitemswereunan-

swered,noscorewascomputed(Cronbach’salpha=0.80).

Secondaryoutcomemeasures

Attitudesandvalues towardspaidemploymentweremeasuredwith fivequestions

(e.g.,“Iwoulddoanythingtogetajob”,“Ifyoudon’twanttowork,youtakeadvan-

tageofothers”)onafivepointscale(stronglyagreetostronglydisagree).Asumscore

wasalsocalculated(Cronbach’salpha=0.30).Jobsearchactivitiesweremeasuredwith

fivequestionsconcerningdifferenttypesofjobsearchactivities(e.g.,“writinganap-

plicationletter”,“searchingforvacanciesinthenewspaperorontheinternet”).Asum

scorewascalculated,rangingfrom0(no jobsearchactivities) to5(many jobsearch

activities) (Cronbach’s alpha=0.68). Finally, re-employment was measured with one

question(“Doyoucurrentlyhaveapaidjob?”)onadichotomousscale(yes/no).

Processevaluation

At the end of the intervention programme semi-structured interviews were under-

taken with ten participants and ten trainers to obtain more qualitative insight into

differentaspectsoftheinterventionthatcouldbeimprovedinthefuture.

Ch

apte

r6


Statisticalanalysis

Basedonthesamplesizecalculationwithaninitialparticipationof70%anda loss-

to-follow-upof30%(powerof80%,one-sidedsignificancelevel0.05),adifferenceof

5.0points(10%)inthescoresonhealthmeasuresbetweentheinterventiongroupand

referencegroupcouldbedetectedwith400personsassignedtoboth trialarmsand

196personswith completedata collection ineachgroup.With theseassumptionsa

differenceof10%inproportionofenteringpaidemploymentbetweenthe interven-

tion(15%)andcontrolgroup(5%)couldbedetected.

The effects of individual characteristics and physical and mental health on par-

ticipation in the intervention were investigated by logistic regression analysis. The

dependentvariableswerefailuretostartwiththeprogramme(yes/no)anddropping

outoftheprogramme(yes/no).Independentvariableswithap-valueof0.10orless

wereretainedinthemultivariatemodelsaswellasageandsexbydefault.Inorder

to compare both analyses, a variable retained in one multivariate model was also

includedintheothermultivariatemodel.

Thebaselinecharacteristicsofbothgroupswerecomparedwiththechi-squaretest

fordichotomousdataandthet-testforcontinuousdata.Theeffectsoftheinterven-

tion on outcome measures at 6 months follow-up were analyzed according to the

intention-to-treat principle, including all subjects regardless of whether or not they

actually received the complete intervention. The analysis was conducted with all

availablerespondentsatthetimeoffollow-upandanon-responseanalysiswascon-

ductedtoevaluatewhetherdrop-outduringthefollow-upperiodwasassociatedwith

healthstatusorinterventionstatus.Animputationtechniqueformissingresponseson

healthoutcomesduringfollow-upmeasurementswasnotused,sincethechoicefora

particularimputationmethodmayinfluencetheestimationoftheinterventioneffect.

[30]Theeffectsoftheinterventiononthecontinuousoutcomemeasureswereevalu-

atedwithamixedeffectmodelforrepeatedmeasurements,withtheinterventionas

fixedeffectandacompoundsymmetrycovariancestructurefortherandomvariation

betweenpersonsandacrosspersons.Thisapproachestimates thechange inanout-

comemeasurebetweenbaselineandfollow-up,takingintoaccountthebaselinevalue

oftheoutcomemeasureofinterestandthepotentialconfoundersage,sex,ethnicity,

education,andtimeonbenefits(SASversion8.12-procedureMixed).

Theeffectsoftheinterventiononthedichotomousoutcomemeasurere-employment

wasanalyzedbyachi-squaremethod (SASversion8.12 -procedureSurveymeans),

adjusted for sex and age.All analyses were carried out with the statistical package

SASversion8.12.

94 Chapter6

ResuLTs

Figure 6.1 shows that at baseline 921 subjects were enrolled in the study. In the

intervention group 343 (74%) persons returned the first questionnaire, which was

statisticallysignificantlyhigherthanthe310(68%)respondentsinthereferencegroup.

Non-responsewasnotinfluencedbyageorsex.

The response at follow-up did not differ between the intervention group (n=176;

51%) and reference group (n=150; 48%). Loss-to-follow up was statistically sig-

nificantlyhigheramongmen,butnotrelatedtoage,maritalstatus,education,ethnic

Table 6.1 Characteristicsofthestudypopulationandbaselinevaluesofoutcomemeasures

variableintervention group(n=343)

usual care group(n=310)

Age(n-%)18-44yr45-64yr

194(52.8)149(43.4)

175(52.9)135(43.6)

Men(n-%) 172(50.2) 160(51.6)

Maritalstatuslivingwithpartner(n-%) 127(37.0) 119(38.4)

Educationallevel(n-%)HigherandintermediatelevelLowerlevel

113(32.9)230(67.1)

123(39.7)187(60.3)

Ethnicbackground(n-%)NativeDutchTurkish/MoroccanAntillean/SurinameseRefugee/Otherimmigrants

88(25.7)90(26.2)89(26.0)76(22.2)

65(21.0)87(28.1)90(29.0)68(21.9)

Employmentexperience(n-%)NeverworkedLessthan5years5yearsandmore

67(19.9)117(34.7)153(45.4)

60(19.8)107(35.3)136(44.9)

Durationonsocialbenefit(n-%)LessthanoneyearBetween1and5years5yearsandmore

58(16.9)85(24.8)200(58.3)

53(17.1)73(23.6)184(59.4)

Healthoutcomemeasures(mean-sd)Generalhealth(0-100)Physicalfunctioning(0-100)Bodilypain(0-100)Mentalhealth(0-100)Socialfunctioning(0-100)Vitality(0-100)

37.4(18.5)52.8(23.7)41.0(23.1)52.5(19.1)52.2(26.1)43.4(16.6)

37.4(19.8)53.5(24.8)43.0(24.5)54.0(20.0)52.6(27.7)45.7(18.1)

Psychologicaloutcomemeasures(mean-sd)Self-esteem(1-4)Mastery(1-4)Kinesophobia(1-4)

2.9(0.6)2.4(0.6)2.7(0.5)

2.9(0.6)2.4(0.6)2.7(0.5)

Poorhealth(n-%) 281(81.9) 252(81.3)

Ch

apte

r6


background,workexperience,durationonbenefit,orhealthatbaseline.Within the

interventiongroup,participationintheinterventionwasnotassociatedwithresponse

onthefollow-upquestionnaire.

Table6.1showsthattherandomizationwassuccessfulincreatingstudygroupswith

similardemographiccharacteristicsandphysicalandpsychologicalhealthatbaseline.

Amongthe465subjectswhowereallocatedtotheinterventiongroup,300subjects

(65%)initiallystartedwiththehealthprogramme.Duringthehealthprogramme85

subjects(28%)droppedoutoftheinterventionduetoanattendancelevelbelow70%

ofallsessionoffered(table6.2). Initialparticipationwasnot influencedbyage,eth-

nicbackground,educational level,marital status, employmenthistory,orperceived

health. Men were more likely to not enter the intervention programme (OR=1.5).

Subjectsdroppingoutoftheprogrammereportedpoorerphysicalandmentalhealth

atbaseline(table6.3).Highfearofmovementatbaselinedidhaveasignificanteffect

ondroppingoutoftheprogrammeintheunivariatemodel(OR=2.095%CI1.0-3.9).

However,duetotheassociationsoffearofmovementwithmentalhealth(r=0.3)and

Table 6.2 Numberofsubjectswhostartedwiththehealthprogramme,whocontinuedtheirparticipationthroughouttheprogramme,andsubjectswhodroppedoutduringtheprogramme.

number of subjectsn (%)

number of respondents to the first questionnaire n

Allocatedtointervention 465 343

Initialparticipation 300(65) 239

Continuedparticipation(>70%attendancetoallsessions)

215(72) 172

Dropout:Between50-70%participationLessthan50%participation

41(14)44(15)

3037

Table 6.3 Logisticregressionanalysisontheeffectsofindividualcharacteristicsandmentalandphysicalhealthonfailuretostartwiththeprogrammeordroppingoutduringtheprogramme

variable not starting with programme(n=104/343)oR (95% ci)

Dropping out(n=67/239)

oR (95% ci)

Age18-44yr45-64yr

1.00.7(0.4-1.1)

1.00.9(0.5-1.7)

Men 1.5(0.9-2.5)* 0.9(0.5-1.7)

Healthoutcomemeasures(standardisedvalues)PhysicalfunctioningMentalhealth

1.0(0.8-1.3)1.0(0.8-1.3)

0.8(0.5-1.1)*0.7(0.5-1.0)*

*0.05<P<0.10ORoddsratio,CIconfidenceinterval.

96 Chapter6

physical functioning (r=0.4), the effect of fear of movement on dropping out of the

programmewasnotsignificantinthemultivariatemodel.

Effectsoftheintervention

Table6.4showsthattheintention-to-treatanalysisdemonstratednobeneficialeffects

oftheinterventiononhealthandpsychologicaloutcomemeasures.Therewasnoef-

fectonemployment statusat 6months followup (re-employment: 2.0% in refrence

groupversus2.2%ininterventiongroup,estimateddifference0.3%(-1.9%-2.3%)).In

addition, attitudes and values towards paid employment, and job search activities

werenotalteredbytheintervention.Theeffectsizesoftheindividualparametersas

wellasthesumscoresacrossscaleswereallclosetounity(datanotshown)

Subgroupanalyses

Aperprotocolanalysisofsubjectswhoinitiallyparticipatedintheinterventionanda

subgroupanalysisofsubjectswhoparticipatedatleast70%oftheinterventiondidnot

showpositiveeffectsoftheintervention.(datanotshown)

Subgroupanalysesbasedonsubjectswithmusculoskeletal complaintsor subjects

withpsychologicalcomplaintsshowedthattheinterventionwasnoteffectiveineither

subgroup.Theeffectsizeswereallclosetounityanddidnotdifferfromtheoverall

effectsize(datanotshown)

Table 6.4 Outcomemeasuresatfollowupintheinterventionandusualcaregroupandtheestimatedeffectoftheintervention

no. of participants intervention/usual care

intervention groupmean (sd)

usual care groupmean (sd)

estimated effect (difference)

HealthoutcomemeasuresGeneralhealth(0-100)Physicalfunctioning(0-100)Bodilypain(0-100)Mentalhealth(0-100)Socialfunctioning(0-100)Vitality(0-100)

172/146174/147173/147173/148173/149173/148

40.7(18.6)54.9(24.2)45.6(24.5)54.8(17.8)55.1(23.3)45.1(16.7)

36.9(22.3)53.8(25.2)44.5(23.9)53.4(21.7)53.7(28.2)43.8(18.7)

1.1(-0.9-3.0)0.1(-2.2-2.5)0.7(-1.7-3.2)0.4(-1.5-2.3)0.5(-2.3-3.3)0.6(-1.2-2.4)

PsychologicaloutcomemeasuresSelf-esteem(1-4)Mastery(1-4)Kinesophobia(1-4)

172/146161/135172/146

2.8(0.6)2.5(0.6)2.6(0.5)

2.9(0.6)2.5(0.7)2.7(0.5)

-0.06(-0.1-0.0)-0.05(-0.1-0.0)-0.01(-0.1-0.1)

Differencewasadjustedforage,sex,ethnicbackground,education,durationonbenefitandalsoforbaselinevaluesofhealthandpsychologicalmeasures.

Ch

apte

r6


Effectofaco-intervention

During the follow-up of 6 months, 80 subjects in the intervention group (23%) and

56 subjects in the reference group (19%) started with a job search training. This co-

interventiondidnotinfluencetheobservedlackofanyeffectoftheintervention.

Discussion

Thehealthpromotionprogrammedidnotshowpositiveeffectsonperceivedmental

andphysicalhealth,selfesteem,mastery,andfearofmovement.Inaddition,values

andattitudestowardspaidemployment,jobsearchactivities,andemploymentstatus

atsixmonthsfollowupwerenotaffectedbythehealthprogrammeeither.

Therearethreepossiblereasonswhytheinterventionwasnoteffective:1)thestudy

couldnotdemonstrateaneffectduetomethodologicallimitations;2)theintervention

wasnotsuccessfullyimplemented;or3)theinterventionwasindeednoteffectivein

thisform.

Methodologicallimitations

Eligibleparticipantswererandomizedbefore theywereapproachedbytheresearch

team for participation in the study. This procedure was necessary since the City of

Rotterdamrequiredanimmediatereferraltoanemploymentservicewithoutfurther

administrative delay.As a consequence of this procedure, the non response on the

firstquestionnaireafterrandomizationwasrelativelyhigh.Theresponsetothefirst

questionnairewasslightlyhigherintheinterventiongroup(74%)comparedwiththe

referencegroup(68%).Receivinganinvitationtoparticipate inahealthprogramme

mayhaveinfluencedthedecisiontofilloutthequestionnaire,butthishadlittleinflu-

enceonthecomparabilityofinterventionandreferencegroup.

Datacollectionwasconductedcompletelyindependentfromtheinterventionpro-

gramme and regular vocational rehabilitation, since participation in these activities

waspartlymandatory,whereasparticipationinthisstudywascompletelyvoluntary.

Asa consequence, somesubjectswho filledout thequestionnairedidnot takepart

in the intervention, whereas other subjects took part in the intervention but did

not respond on the questionnaire. For most subjects the reason for not taking part

in the intervention or the vocational rehabilitation was unknown, although a small

proportion was due to termination of the benefits for different reasons (moving in

withapartnerormovingoutofthecityofRotterdam).However,participationinthe

interventionwasnot influencedbypersonal characteristics, employmenthistory,or

98 Chapter6

perceived health. Therefore, it is assumed that there was no selection bias in initial

participationintheprogramme.

Thepowercalculationwasbasedon196subjectspergroupwithcompletedatacol-

lection.Infact,theachievedsamplewasconsiderablyless.Sincetheestimatedeffects

of the interventionwereclose tounity forallparameters,a largerstudypopulation

wouldnothaveresultedinstatisticallysignificanteffectsoftheintervention.There-

fore,itisassumedthatthefailuretodetectbetween-groupdifferencesisnotduetoa

lackofpowerofthestudy.

In this study perceived health was an important outcome measure, whereas the

focusoftheinterventionprogrammewasalsoonimprovingobjectivephysicalhealth.

Intheinterventiongroupitwasshownthatabettercardiorespiratoryfitnessdidnot

resultinabetterperceivedhealth.[31]

Anothermethodologicalreasonfornotfindinganyeffectofthehealthprogramme

mayhavebeentheabsenceofanassessmentofchangeimmediatelyaftertheendof

the intervention.Theoretically, there couldhavebeenapositiveeffectonhealthdi-

rectlyaftertheprogrammewasfinished,whichhadalreadyfadedawayatthetimeof

followup,onaverageaboutthreemonthsaftertheprogrammetermination.However,

thiswouldimplyalackofsustainabilityoftheintervention.Inthedesignofthestudy

avocational trainingdirectlyaftertheendof thehealthprogramwasthoughttobe

abletosustainpositiveeffectsoftheintervention.However,inpracticethevocational

trainingwasoftendelayedordidnotstartatall.Duetoorganisationalproblemsofthe

socialsecurityserviceandvocationalrehabilitationcentresonly23%oftheinterven-

tiongroupand19%ofthecontrolgroupstartedwithavocationaltrainingwithinthe

followupperiodofsixmonths.Thislackoffollow-upactivitiesconcerningvocational

rehabilitationcanbeconsideredasanimplementationfailureandmayhaveinterfered

withourresults.

Theinterventionwasnotsuccessfullyimplemented

The intervention was offered during two years in eight periods with three groups

perperiod.Intotal,22groupsstartedwiththeprogrammewith11to22participants

pergroup.Theoretically,eachprogrammeconsistedof12weeksofmultidisciplinary

rehabilitation.Inpractice,however,participantsreceivedonlynineweeksofeffective

training,sincethreeweekswerelostduetothetimeittookfortheintroductionand

intakeandouttakeactivities.

Participationinthehealthprogrammewasmandatory,butthesocialsecurityservice

ofthecityofRotterdamdidnotstrictlyenforceactualparticipation.Themandatory

naturehasundoubtedlyresultedinan increasedparticipation.Participantswhofell

obligedto join,maynothaveexperiencedaneedto improvetheirhealth.Feedback

Ch

apte

r6


provided by the physical education teachers indicated that getting participants in-

volvedwasamajorchallengeinitselfandthatanincreaseintrainingeffortinatime-

contingentmannerwillcertainlynothavebeenachievedbyallparticipants.Hence,

thegradedactivityprinciplewasnotadheredtoforallparticipants.

Interviewswithtenparticipantsandthetrainersrevealedthatthecognitivetraining

wasnotwelladaptedtothisstudypopulationofpersonswithloweducationandlow

socio-economicstatus.Bothtrainersandparticipantsacknowledgedthatthecognitive

training should be improved with less focus on theory and being more adapted to

experiencesindailylifeoftheparticipants.

Among the persons who initially started with the intervention, 72% completed

the programme with more than 70% attendance to all sessions offered. Hence, the

interventionhadagoodreachamongsubjectswhoareusuallydifficulttoengagein

healthpromotionactivities.Only28%didnotsucceedtoparticipateinatleast70%

oftheprogrammeactivities.Participantswithapoorphysicalandmentalhealthand

high fear of movement at baseline had a higher chance of dropping out during the

programmecomparedtopersonswithagoodhealthandnofearofmovement.This

mayhaveaffectedtheeffectivenessofthehealthprogramme,becausethosepersons

whocouldpotentiallybenefit themostdroppedoutof theprogramme.However,a

subgroupanalysisof thesubjectswhoparticipated inat least70%ofall sessions in

theprogrammeversus thereferencegroup,didnotshowanypositiveeffectsof the

intervention.

Theinterventionwasindeednoteffectiveinthisform.

A systematic review concluded that only intensive (>100 hours) multidisciplinary

rehabilitationwillreducepainandimprovefunctioninpatientswithchroniclowback

pain.[32]Basedon this criterion, it is assumed that the intensityofourprogramme

was high enough (around 108 hours), but the subgroup analysis showed that par-

ticipantswithlowbackpaindidnotimproveintheirpain,physicalfunction,orany

of theotheroutcomemeasures in thecurrentstudy.Bendixetal.[12]showedthata

multidisciplinaryintensiveprogrammethatranfor3successiveweeksof39hoursper

weekwaseffective,whereasthehealthprogrammeunderstudyranfor12successive

weekswith9-12hoursperweek.Amoreintensiveprogrammewithashorterduration

maybemoreeffectivethanalessintensiveprogrammewithalongerduration.

Ananticipatedresultofthehealthprogrammewasahigherre-employmentinthe

interventiongroup.Amultidisciplinaryrehabilitationprogramme,describedbyWat-

sonetal.[18],withastrongfocusonimprovementintheabilitytoworkandonactual

return towork, showed thatat6months follow-up38%of subjectswereemployed

and another 23% were in voluntary work of education/training. In contrast, in the

100 Chapter6

healthprogrammeofferedinRotterdamreturntoworkwasnotpartoftheindividual

goalsetting.Thelackofastrongintegrationofthehealth-orientedinterventioninto

the regular vocational rehabilitation activities may have impeded beneficial effects.

Pateletal. [33]hassuggested thatmultidisciplinaryapproachesshouldnotonlybe

concerned with medical and psychological issues, but should also address the ob-

staclestoreturntoworkasseenbythepatient.

Theprocessevaluationshowedthatafter theendof theprogramme,mostpeople

fellbackintotheiroldlifestylewithlowlevelsofphysicalactivity. Inordertohave

sustainableeffectsofahealthpromotionprogramme,itisimportantthatparticipants

continuetobephysicallyactive.Theabsenceofasustainedeffortbyhealthcounsel-

lorsafterterminationoftheinterventiontoencourageparticipantstostayphysically

activemayhavecontributedtothefactthatnohealtheffectswerefound.

In conclusion, the intervention aimed at the promotion of physical and mental

healthofunemployedpersonswithhealthcomplaintsdidnotshowanybeneficialef-

fects.Thus,itcannotberecommendedtoimplementthisparticularhealthpromotion

programmetocounteractatindividuallevelthenegativeeffectsofunemploymenton

health.Measuresonsocietalleveltoreducethenegativeeffectsofunemploymenton

populationhealtharerequired [11]. Inpolicies forhealthequity, it remainsofpara-

mount importance todevelopmeasures to includepeoplewithapoorhealth in the

labourmarket.

Acknowledgements

FundingforthisstudywasprovidedbythePublicHealthFund(FondsOGZ)ofthe

Netherlands(reg.nr.P161).Wegratefullyacknowledgethededicationoftheprofes-

sionalsprovidingthehealthpromotionprogramme(Ergocontrol)andtheircontinu-

oussupporttothisstudy.

Ch

apte

r6


ReFeRences







7. WarrP,JacksonP:Adaptingtotheunemployedrole:alongitudinalinvestigation.Soc Sci Med 1987,25:1219-1224.

8. PriceRH,ChoiJN,VinokurAD:Linksinthechainofadversityfollowingjobloss:howfinancialstrainandlossofpersonalcontrolleadtodepression,impairedfunctioning,andpoorhealth.Journal of occupational health psychology 2002,7:302-312.

9. EdenD,AviramA:Self-efficacytrainingtospeedreemployment:Helpingpeopletohelpthemselves.Journal of Applied Psychology 1993,78:352-360.

10. CottCA,GignacMA,BadleyEM:DeterminantsofselfratedhealthforCanadianswithchronic disease and disability. Journal of epidemiology and community health 1999, 53:731-736.

11. CreedPA:Improvingthementalandphysicalhealthofunemployedpeople:whyandhow?Med J Aust 1998,168:177-178.

12. BendixAF,BendixT,OstenfeldS,BushE,Andersen:Activetreatmentprogramsforpatientswithchroniclowbackpain:aprospective,randomized,observer-blindedstudy.Eur Spine J 1995,4:148-152.


14. LakkaTA,KauhanenJ,SalonenJT:Conditioningleisuretimephysicalactivityandcardiore-spiratoryfitnessinsociodemographicgroupsofmiddle-agesmenineasternFinland.Int J Epidemiol 1996,25:86-93.


16. TurnerJA,JensenMP,RomanoJM:Dobeliefs,coping,andcatastrophizingindependentlypredictfunctioninginpatientswithchronicpain?Pain 2000,85:115-125.

17. RoseV,HarrisE:Fromefficacytoeffectiveness:casestudiesinunemploymentresearch.J Public Health (Oxf) 2004,26:297-302.

18. WatsonPJ,BookerCK,MooresL,MainCJ:Returningthechronicallyunemployedwithlowbackpaintoemployment.Eur J Pain 2004,8:359-369.

102 Chapter6

19. Gatchel RJ, PengYB, Peters ML, Fuchs PN, Turk DC: The biopsychosocial approach tochronicpain:scientificadvancesandfuturedirections.Psychol Bull 2007,133:581-624.

20. LindstromI,OhlundC,EekC,WallinL,PetersonLE,FordyceWE,NachemsonAL:Theef-fectofgradedactivityonpatientswithsubacutelowbackpain:arandomizedprospectiveclinicalstudywithanoperant-conditioningbehavioralapproach.Phys Ther 1992,72:279-290;discussion291-273.

21. FordyceW:Behavioralmethodsforchonicpainandillness.StLouis:Mosby.1976. 22. OsteloRW,KokeAJ,BeurskensAJ,deVetHC,KerckhoffsMR,VlaeyenJW,WoltersPM,

Berfelo MW, van den Brandt PA: Behavioral-graded activity compared with usual careafterfirst-timedisksurgery:considerationsofthedesignofarandomizedclinicaltrial.J Manipulative Physiol Ther 2000,23:312-319.

23. CBS:Herkomstvanpersonen;allochtonenenmigratie[Countryoforiginofpersons;mi-grantsandmigration].CentraalBureauvoordeStatistiek,Voorburg/Heerlen2003.

24. WareJE,Jr.,SherbourneCD:TheMOS36-itemshort-formhealthsurvey(SF-36).I.Concep-tualframeworkanditemselection.Med Care 1992,30:473-483.

25. VanderZeeKI,SandermanR:HetmetenvandealgemenegezondheidstoestandmetdeRAND-36: een handleiding [Measuring health status with the RAND-36: a manual]. InGroningen:NoordelijkCentrumvoorGezondheidsvraagstukken,RijksuniversiteitGron-ingen.1993.

26. FayersPM,SprangersMA:Understandingself-ratedhealth.Lancet 2002,359:187-188. 27. PearlinLI,SchoolerC:Thestructureofcoping.J Health Soc Behav 1978,19:2-21. 28. RosenbergM:Societyandtheadolescentself-image.Princeton: Princeton University Press

1965. 29. VlaeyenJWS,KoleSnijdersAMJ,BoerenRGB,vanEekH:Fearofmovement/(re)injuryin

chroniclowbackpainanditsrelationtobehavioralperformance.Pain 1995,62:363-372. 30. HollisS,CampbellF:Whatismeantbyintentiontotreatanalysis?Surveyofpublished

randomisedcontrolledtrials.BMJ (Clinical research ed 1999,319:670-674. 31. SchutgensCAE,SchuringM,VoorhamAJ,BurdorfA:Changesinphysicalhealthamong

participantsinamultidisciplinaryhealthprogrammeforlong-termunemployedpersons.BMC Public Health2009,9:197.

32. GuzmanJ,EsmailR,KarjalainenK,MalmivaaraA,IrvinE,BombardierC:Multidisciplinaryrehabilitationforchroniclowbackpain:systematicreview.BMJ (Clinical research ed 2001,322:1511-1516.

33. PatelS,GreasleyK,WatsonPJ:Barrierstorehabilitationandreturntoworkforunemployedchronicpainpatients:aqualitativestudy.Eur J Pain 2007,11:831-840.

7changes in physical health among participants

in a multidisciplinary health programme

for long-term unemployed persons

SchutgensCAE,SchuringM,VoorhamAJJ,BurdorfA.

BMC Public Health 2009, 9:197

104 Chapter7

absTRacT

background: Therelationshipbetweenpoorhealthandunemploymentiswellestab-

lished.Healthpromotionamongunemployedpersonsmayimprovetheirhealth.The

aims of this study were to investigate characteristics of non-participants and drop-

outs inamultidisciplinaryhealthpromotionprogrammefor long-termunemployed

personswithhealthcomplaints,toevaluatechangesinphysicalhealthamongpartici-

pants,andtoinvestigatedeterminantsofimprovementinphysicalhealth.

methods:Alongitudinal,non-controlleddesignwasused.Theprogrammeconsisted

of twoweeklyexercisesessionsandoneweeklycognitivesessionduring12weeks.

Themainoutcomemeasureswerebodymassindex,bloodpressure,cardiorespiratory

fitness,abdominalmusclestrength,andlowbackandhamstringflexibility.Potential

determinantsofchangeinphysicalhealthweredemographicvariables,psychological

variables(self-esteem,mastery,andkinesiophobia),andself-perceivedhealth.

Results:Theinitialresponsewas73%and252personshadcompletedatacollection

at baseline. In total, 36 subjects were lost during follow-up. Participants were pre-

dominantly loweducated, long-termunemployed,and inpoorhealth.Participation

in theprogrammewasnot influencedbydemographicandpsychological factorsor

by self-reportedhealth.Drop-outswereyoungerandhada lowerbodymass index

at baseline than subjects who completed the programme.At post-test, participants’

cardio respiratory fitness, abdominal muscle strength, and flexibility had increased

by6.8%-51.0%,whereasdiastolicandsystolicbloodpressureshaddecreasedby2.2%-

2.5%.Theeffectsizesrangesfrom0.17-0.68.

conclusions: Participants with the poorest physical health benefited most from the

programmeandsexdifferencesinimprovementwereobserved.Physicalhealthofun-

employedpersonswithhealthcomplaintsimprovedafterparticipationinthishealth

promotionprogramme,butnotsufficiently,consideringtheirpoorphysicalhealthat

baseline.

Ch

apte

r7

Changesinphysicalhealthamongparticipantsinamultidisciplinaryhealthprogramme 105

inTRoDucTion

Therelationshipbetweenunemploymentandpoorerhealthhasbeenwellestablished.

[1-3] This relationship is bi-directional with both a selection mechanism with poor

health reducing the likelihood on paid employment, and a causation mechanism

wherebyunemploymentwillresultsinapoorerhealth.[1,4]Theseassociationsmay

bemediatedbyothervariables,suchashealthbehaviourandpsychosocialvariables.

Alowself-esteem,forinstance,isadeterminantofself-reportedpoorhealth[5]and

alsodecreasesthelikelihoodofemployment.[6-8]Thus,unemploymentmayleadto

poorerhealth,whichinturnreducesthechancesofreemployment.

In order to improve the possibility for reemployment, improvement in health of

unemployed persons may, therefore, be an important step. Pedersen and Saltin [9]

haveconcludedintheirextensivereviewthatexercisetherapyhaspositiveeffectson

maximumoxygenuptake(VO2max),musclestrength,generalwell-being,bloodpres-

sure,weight,bodyfatpercentage,anddepressivesymptomsofpersonswithchronic

diseases.Thereissomeevidencethatphysicalandmentalhealthareinterrelatedand

thatdeterminantsofphysicalhealthmayalsopositivelyaffectmentalhealthandvice

versa.Masteryorthesenseofcontroloverone’slife,andself-esteemhavebeenassoci-

atedwithabetterself-reportedphysicalhealth.[5]

There is, however, limited research into the effects of exercise-based programmes

among groups with a poor health in a low socio-economic position. A low socio-

economic position and a poor health have consistently been associated with non-

participation and drop-out in health programmes. [10-14] In addition to this, it is

importanttoidentifydeterminantsofnon-compliancewhichmayinfluencetheeffects

ofexerciseprogrammes.[15]

Watson and colleagues [16] provided some indications that a combined physical

exerciseandcognitivebehaviouralprogrammeimprovedphysical fitnessaswellas

increasedemploymentratesamongunemployedparticipants.Theseresultsshouldbe

interpretedwithcaution,however,sincethevoluntaryparticipationintheprogramme

mighthavebiasedtowardsparticipantswithahighmotivationandapositiveattitude

towards(returnto)work.

Thereisaclearneedformoreinsightintowaystoimprovehealthofpersonsina

lowsocio-economicposition.Anintensive,multidisciplinaryhealthprogrammewas

developedforunemployedpersonswithhealthcomplaints(“Workonyourhealth”),

consistingofphysicalexerciseandcognitivetraining,withthegoaltoimprovephysi-

calandmentalhealthasacontributiontoincreasetheopportunitiesonpaidemploy-

ment.Theaimsofthepresentstudywere(1)toidentifythefactorsthatdetermined

non-participation,drop-out, andnon-compliance inahealthpromotionprogramme

106 Chapter7

forunemployedpersons, (2) to evaluate the changes inphysicalhealthamongpar-

ticipants,and(3)toinvestigatethedeterminantsofimprovementinphysicalhealth.

meTHoDs

Studydesignandpopulation

A longitudinal, non-controlled design was used among participants in a health

promotion programme. Unemployed persons with chronic health complaints were

referred by the Employment Centre of the City of Rotterdam, The Netherlands, for

a fit-to-work test, conductedbyaphysician,psychologist,andanemploymentspe-

cialist.Participantswereselectedonthebasisofthefollowingcriteria:unemployed,

diagnosed with chronic health complaints by the physician or a psychologist, but

consideredtobecapableoffulltimeemployment,andbeingatleastmoderatelyable

to understand and speak Dutch. From December 2004 until December 2007 partici-

pantswereincludedinthestudy.Theinvitationtoparticipateinthehealthpromotion

programmewassendoutbytheproviderofthisprogramme,withasupportingletter

fromthecityofRotterdamstipulatingthatattendingtheprogrammeforatleast70%

wasmoreorlessmandatory,andthatrefusalmightresultinacutinthesocialbenefits

received.Theresearchgroupcarriedoutthecurrentevaluationstudyandparticipa-

tionwasstrictlyvoluntary.Beforethestartoftheprogramme,participantsweresenta

questionnaireandprepaidreturnenvelope.ForthosewithaTurkishlastname(alarge

ethnic minority), a Turkish version of the questionnaire was sent as well. Another

largeethnicminoritygrouparetheMoroccanpeople.However,itwasnotpossibleto

makeaMoroccan-ArabicquestionnairebecausethemajorityoftheMoroccansinthe

NetherlandsspeakBerber,whichisnotawrittenlanguage.Aftertwoandfourweeks,

reminderlettersandquestionnairesweresenttotheparticipants.If,afterfourweeks,

still no questionnaire had been sent back, an interviewer visited the home address.

Whenfourvisitsduringdifferenthoursinatwo-weekperiodwerenotsuccessful,a

participant was considered a non-respondent. The interviewers were matched with

subjects,basedonethnicity,age,andsex,andcouldofferaninterviewinthemother

tongue (Dutch, Arabic, or Turkish). The Medical Ethics Committee of the Erasmus

MC,Rotterdamapprovedthestudy.

Of the 465 subjects who were invited to take part in the health promotion pro-

gramme, 338 participated in an assessment to evaluate medical and psychological

eligibility tostart theprogramme(response73%).Thereasons fornon-participation

(n=127)werenotreceivingasocialsecuritybenefitanymore(n=14),beingallocatedto

areintegrationoreducationalprogramme(n=24),andunknown(n=89).Inaddition,

Ch

apte

r7


22individualsweredeclaredmedicallyunfittosuccessfullyparticipateinthephysical

trainingandanother5individualswereexcludedformajorpsychologicalproblems.

Intotal,311personsstartedtheprogramme,ofwhich252alsofilledoutthequestion-

nairesendoutbytheresearchteam.Intotal,216outof252individualscompletedthe

health promotion programme (86%) and 36 subjects were lost to follow-up (Figure

7.1).

Theprogramme

Theinterventionwasaimedatchangingthewayunemployedpersonsperceiveand

cope with their health complaints. The rationale was based on the biopsychosocial

modelofchronicpainandsubsequentinterdisciplinarypainmanagementapproach.

Patients with chronic pain are at increased risk for emotional disorders (such as

Invited to participate in programmeN= 465

Participated in intake (medical & psychological assessment)N = 338

No longer benefit (n=14)Reintegration programme (n=24)Unknown (n=89)

Started programmeN = 311

Declared medically unfit (n=22)Declared psychologically unfit (n=5)

Filled out questionnaire N = 252

No questionnaire (n=59)

Available post-testN = 216

Lost to follow-up (n=36)

Figure 7.1 Flowdiagramofparticipationinthehealthpromotionprogramme

108 Chapter7

anxiety,depressions,andanger),maladaptivecognitions(suchascatastrophizingand

poorcopingskills),functionaldeficitsandphysicaldeconditioning(duetodecreased

physicalactivityandfearofinjury).Theseeffectsareofteninterdependent,sothatone

cannotsimplytreatonetotheexclusionofothers.Interdisciplinarypainmanagement

embracesthefactthatthecomprehensivetreatmentofallthesedimensionsisneeded

inordertobeeffective.[17]

Thehealthpromotionprogrammeconsistedofthreesessionsofthreehoursevery

week during a twelve-week period. One session a week was focused on cognitions

andtwoweeklysessionswerefocusedonphysicalactivity.Thecognitivecomponent

wasdesignedtoenhanceparticipants’ insightintheirhealthcomplaints(e.g.move-

mentmaybepainful,thoughharmless)andhowtocopewiththesecomplaints,toen-

hanceself-esteemandfeelingsofmastery,toreducefearandavoidanceofmovement,

andtoimprovesocialfunctioningbylearningtothinkpositivelyandincreasesocial

skills.Thecognitivecomponent,conductedbytwopreventionworkers,wasprimar-

ilyfacilitatingthephysicalactivitypartoftheintervention.Theexerciseprogramme

consistedof1.5hoursfitnesstrainingtwiceaweek(cardioandweighttraining),1.5

hoursof indoorsportsweekly,and1.5hoursofoutdooractivitiesweekly.Thispart

of theprogrammewasprimarilydesigned to improvephysical fitness.Theexercise

programmewasdevelopedaccordingtothegraded-activityprinciple.Theexercises

started below the average functional capacity assessed during the first session and

wereincreasedgraduallyduringthecourseoftheintervention,accordingtothetime-

contingencyprinciple.Thesesessionswereconductedbyphysicaleducationteachers.

Theinterventioncostswereapproximately€2300perparticipantwhoenrolledinthe

programme.

Outcomemeasures

Sevenphysicalhealthindicatorsweremeasuredatstartandendoftheprogrammeby

theprovideroftheprogramme:BodyMassIndex(BMI)inkg/m²,bodyfatpercent-

age,systolicanddiastolicbloodpressure,cardiorespiratoryfitness,abdominalmuscle

strength,andlowbackandhamstringflexibility.Bodyfatpercentagewasdetermined

bymeansofabioelectricalimpedanceanalysiswithabodyfatmeter.[18]Bloodpres-

sure(mmHG)wasmeasuredwithanautomaticsphygmomanometerattheleftwrist.

Cardio respiratory fitness was measured by the Åstrand Ergometer Bicycle test

of maximum oxygen absorption (VO2max in ml/kg/minute) (Åstrand and Kodahl,

1986).Participantscycledonabicycleergometerwithaconstantpeddlingrateof70-

75rotations/minute.Theworkloadwasadjustedtotheparticipant’sheartrate,which

had tobeapproximately120beatsperminuteafter twominutes.Subsequently, the

participantcycledforsixminutes.Iftheheartratefluctuatedmorethanfivebeatsin

Ch

apte

r7


thelastminute,thetestwasprolongeduntilasteadypulsewasobtainedforatleast

oneminute.VO2maxwasestimatedonbasisoftheaverageheartrateinthelastmin-

ute,workload,sex,andage.Thetestwascarriedoutunderstandardizedconditions

withthetemperaturebetween18and20°Candatmospherichumiditybetween40and

60%.Beforethetest,participantsrestedfiveminutes,andtheyhadtoabstainfromeat-

ing,drinkingcoffee,andsmokingfortwohours,fromalcoholfortwelvehours,and

fromvigorousphysicalactivityandsunbathingforsixhours. If theparticipantwas

notabletocyclewithaheartrateof120beatsperminuteoriftheheartrateexceeded

170beatsperminute,thetestwasterminated.

Abdominalmusclestrengthwasdeterminedasthenumberofsit-upsperminute,

withkneesbent(90°)andfootsandhandsonthefloor.Shouldershadtostayabove

thefloorduringthetest.Handshadtoreachalineat7.5cmfromthestartingposi-

tion (Fitness Canada). Low back and hamstring flexibility (cm) was measured with

a sit-and-reach test, selecting the best of three trials. Participants placed their foot

soles, without shoes, against the end of a box.Arms were stretched forward as far

aspossiblewithunbentkneesand the reachwasdetermined (cm)on thebox scale

(FitnessCanada).

Determinants

Determinants of (change in) health were demographic characteristics (sex, age,

educationallevel,ethnicity,andmaritalstatus),durationofunemployment,mastery,

self-esteem,kinesiophobia,andself-perceivedhealth.Educationallevelwasmeasured

asthehighestlevelofeducationalattainmentinthreecategories.Ahigheducational

level was defined as higher vocational training or university, intermediate educa-

tionallevelashighersecondarytrainingorintermediatevocationaltraining,andlow

educationallevelasnoeducation,primaryschool,lowerandintermediatesecondary

trainingorlowervocationaltraining.Ethnicitywasbasedonthemother’scountryof

origin;incasethemotherwasnativeDutch,thefather’scountryoforiginwaslead-

ing.Fourethnicgroupsweredefined:NativeDutch,TurkishandMoroccan,Antillean

andSurinamese,andother.TurkishandMoroccanpeoplehaveasimilarimmigration

history with limited acculturation.Antillean and Surinamese people originate from

former Dutch colonies and are reasonably integrated in Dutch society by virtue of

speakingDutch.Theotherethnicgroupisaheterogeneousmixtureofalargenumber

ofnationalities.Maritalstatusdistinguishedbetweensubjectslivingwithoutapartner

andsubjectsbeingmarriedorlivingwithapartner.

MasterywasmeasuredbythePersonalMasteryScale[19],whichconsistsofseven



110 Chapter7


were calculated, ranging from 1 to 4, with a higher score indicating a higher level

of mastery. In case three or more items were unanswered, no score was computed

(Cronbach’salpha=0.69).

Self-esteemwasmeasuredwiththeRosenbergSelf-EsteemScale[20],with10items





unanswered,noscorewascomputed(Cronbach’salpha=0.84).

Kinesiophobia was measured with the Tampa Scale of Kinesiophobia [21], which

consists of 17 items on fear of movement and injury (e.g. “It’s really not safe for a

personwithaconditionlikeminetobephysicallyactive”,“PainalwaysmeansIhave

injuredmybody”)onafourpointLikertscale(stronglyagreetostronglydisagree).

Averagescoresacrossitemswereagaincalculated,rangingfrom1to4,withahigher

scoreindicatingahigherlevelofkinesiophobia.Incasefiveormoreitemswereunan-

swered,noscorewascomputed(Cronbach’salpha=0.80).

Self-perceivedhealthwasmeasuredwiththeShortForm36HealthSurvey(SF-36)

[22]. The SF-36 consists of 36 questions about health, covering eight dimensions:


vitality,role limitationduetoemotionalhealthproblems,androle limitationdueto

physicalhealthproblems.Scoresmayrangefrom0to100withahigherscoreindicat-

ingabetter self-perceivedhealth.

Tomeasureprogrammecompliance,participants’trainingattendancerecordswere

kept.Subjectswereconsideredcompliantwhentheyattendedatleast70%ofallses-

sions.Thecut-offpointof70%attendanceofallsessionswasinlinewiththepolicy

of the social security services, which demanded an attendance of at least 70% from

participants.

Statisticalanalyses

AllstatisticalanalyseswereconductedbymeansofthestatisticalpackageSPSS(ver-

sion13)forWindowsandthelevelofsignificancewassetat0.05.

Differences between participants and non-participants, between drop-outs and

completers,andbetweencompliersandnon-complierswereevaluatedbychisquare

testsandone-wayanalysesofvariance.

The dependent measures of physical health in the statistical analyses were BMI,

bodyfatpercentage(%),systolicanddiastolicbloodpressures(mmHg),VO2max(ml/

kg/minute),abdominalmusclestrength(sit-ups/minute),andflexibility(cm).Toin-

Ch

apte

r7


vestigatewhichdeterminantswereassociatedwiththesemeasuresofphysicalhealth

atbaseline,univariate linearregressionanalyseswereconductedwithdemographic

characteristics,psychological factors,andself-perceivedhealthas independentvari-

ables.Subsequently,independentvariablesofinterest(p-value<0.10)wereincluded

inthemultipleregressionanalysisforeachphysicaloutcomemeasureandvariables

were retained in the final multivariate linear regression model when statistically

significant (p < 0.05) or statistically significant in a multivariate model on another

outcome measure. For each independent variable based on an average score across

items also the standardized regression coefficient was calculated, representing the

effectofanincreaseofonestandarddeviationintheaveragescoreonthemagnitude

oftheoutcomemeasure.Allanalyseswerecorrectedforthedurationbetweenthedate

offillingoutthequestionnaireandthedateofcollectingthephysicalhealthmeasures

inthetest.BodyfatpercentagecorrelatedhighlywithBMI(r=0.72)and,hence,only

BMI was further analysed. Mastery correlated with self-esteem (r = 0.46) and the

inclusionofbothvariablesinthesamemodelcreatedproblemswithmulticollinear-

ity, resulting in substantially higher confidence intervals and, thus, non-significant

results. Self-esteem was strongest associated with the outcome measures of interest

and selected for further presentation. The SF-subscales were interrelated (r varying

from0.25to0.63).Basedontheunivariateanalyses,physicalfunctioningwaschosen,

sinceithadthestrongestassociationswithseveraloutcomemeasures.

The changes in physical health during the health promotion programme were

evaluatedbysixpaired-samples t-testsandCohen’sdwascalculatedasmeasureof

effectsizebydividingthedifferencesbetweenpre-testandpost-testbytheirpooled

standarddeviation.[23]Duetoregressiontothemean,thephenomenonthatextreme

scoresfallbacktowardstheaveragewhenmeasuredagain,theinitialvalueatbaseline

willbeassociatedwiththeobservedchangeovertime.[24-25]Inordertoinvestigate

whetherchanges inphysicalhealthwereduetoregressionto themeanor todiffer-

entialresponsetothehealthpromotionprogramme,themeasuresofphysicalhealth

at baseline were classified into three categories: below 25% percentile, interquartile

range (p25-p75), and above 75% percentile. Regression to the mean will be present

when subjects with a poor physical health (lower quartile) improve and subjects

withagoodphysicalhealth(upperquartile)deterioratelikewise.Inlinearregression

analyseswithrepeatedmeasurements,thedeterminantsofimprovementinphysical

healthwereevaluatedbyintroducinginteractiontermsoftheinitialphysicalhealth

values,expressedascategoricalvariables,withtimeofmeasurementasfixedeffects

in the analysis. Similarly, interaction terms of significant determinants of physical

health at baseline with time of measurement were investigated, adjusted for initial

values of physical health. In these analyses the random variance components were

pooledacrossalldeterminantsandassumedtobeequalacrosstime.Thisassumption

112 Chapter7

ofacompoundsymmetrycovariancestructureresulted in themost restrictiveerror

structure possible, necessary because of the small number of subjects available for

somephysicalhealthmeasures.

ResuLTs

Baselinecharacteristicsoftheparticipants

Ofallparticipantswithcompletebaselineinformation(n=252),46%wasmale,75%

belongedtoanethnicminority,68%hadalowlevelofeducation,and72%reported

beingunemployedforat least5yearsorhadneverworked(Table7.1).Onaverage,

participants had a low self-perceived health, a low VO2max, and the prevalence of

overweight and obesity was high. No correlations were found between SF-36 sub-

scalesandphysicalhealthoutcomemeasures,exceptforabdominalmusclestrength

(physicalfunctioning:r =0.24,mentalhealth:r =0.20,generalhealth:r =0.16).

Characteristicsofnon-participants,drop-outs,andnon-compliers

Based on the information obtained by the questionnaires, subjects who started the

healthprogrammedidnotdifferstatisticallysignificantlyfromnon-participantswith

respect to demographic and psychological variables and self-perceived health. Sub-

jectswhocompletedtheprogramme(ie,attendedbothpre-testandpost-test)hada

higherBMIatbaseline(2.33,95%CI0.26-4.41)andwereolder(3.98,95%CI0.78-7.18)

thandrop-outs.Ofallsubjectswhocompletedtheprogramme,82%attendedatleast

70%ofthesessions.Compliantpersonshadastatisticallysignificantlyhigherphysical

functioning(7.97,95%CI0.15-15.78)andlesskinesiophobia(0.23,95%CI0.07-0.39)at

baselinethannon-compliantpersons.

Determinantsofphysicalhealthatbaseline

Sex, age, marital status, self-esteem, and self-perceived physical functioning were

determinants forphysicalhealthatbaseline,althoughnot foralloutcomemeasures

(Table7.2).Ethnicity,levelofeducation,unemploymentduration,andkinesiophobia

didnothaveasignificantcontribution.Theexplainedvariancewas lowest for flex-

ibility(R²=7.4%)andhighestforVO2max(R²=31.8%).Halfofthesubjectswerenot

abletofinishtheÅstrandErgometerBicycletestandfailurewasassociatedwitholder

age,lowerselfesteemandlowerphysicalfunctioning.

Ch

apte

r7


Table 7.1 Characteristicsofunemployedpersonswithhealthcomplaints(n=252)whoenrolledinahealthpromotionprogramme

variable

Demographiccharacteristics

Men 46.4%

Age(yr) 42.11(9.12)

Marriedorlivingwithpartner 35.2%

EthnicbackgroundNativeDutchTurkish/MoroccanSurinamese/AntilleanOther

24.9%25.7%29.0%20.4%

LevelofeducationLowIntermediateHigh

67.9%28.8%3.3%

Unemploymentduration(n=245)<5years>5yearsNeverworked

27.4%53.5%19.1%

Psychologicalmeasures

Self-esteem(1-4) 2.85(0.56)

Mastery(1-4) 2.45(0.55)

Kinesiophobia(1-4)(n=239) 2.68(0.47)

Self-perceivedhealth(SF-36)

Physicalfunctioning 52.69(23.11)

Rolefunctioning(physical) 32.18(39.66)

Bodilypain 41.39(23.42)

Vitality 43.74(16.03)

Socialfunctioning 53.64(25.88)

Rolefunctioning(emotional) 48.43(44.58)

Mentalhealth 53.00(18.85)

Generalhealth 37.62(18.23)

Physicalmeasures

BMI(kg/m2)overweight(25<=BMI<30)obese(BMI>=30)

27.61(5.68)32.9%29.4%

VO2max(ml/kg/minute)(n=130) 24.60(7.85)

Abdominalmusclestrength(sit-ups/minute)(n=216) 21.13(13.69)

Flexibility(cm)(n=223) 23.70(10.94)

Systolicbloodpressure(mmHg) 130.33(17.42)

Diastolicbloodpressure(mmHg) 82.96(10.89)

114 Chapter7

Changesinphysicalhealth

Participantsintheprogrammeshowedsignificantdecreasesindiastolicandsystolic

blood pressure by 2.2%-2.5% and significant increases in cardio respiratory fitness,

flexibility,andabdominalmusclestrengthby6.8%-51.0%(Table7.3). Effectsizeswere

smalltomedium(Cohen’sdrangedfrom0.17to0.68).Inaddition,theproportionof

participantsthatwasabletocompletethebicycletestincreasedfrom52%atbaseline

to71%atfollow-up.

Table 7.2 Determinantsofphysicalhealth†atbaselineamongunemployedpersonswhoenrolledinahealthpromotionprogramme(n=252)basedonmultivariatelinearregressionanalyses

bmi vo2max(n = 130)

abdominal muscle strength

Flexibility systolic blood pressure

Diastolic blood pressure

Constant 26.26 41.08 14.13 17.97 97.66 67.04

Female 2.51** -3.91** -4.50** 4.04** -4.03 -0.51

Age(yr) 0.05 -0.36** -0.22** -0.09 0.52** 0.19**

Marriedorlivingwithpartner 2.08** -1.04 -6.67** -1.59 1.66 -0.28

Self-esteem(1-4) 0.71 -1.49 0.20 0.64 3.85* 3.79**

Physicalfunctioning(0-100) -0.01 0.05* 0.12** 0.08* -0.02 -0.05

Explainedvariance(R²) 8.9% 31.8% 15.6% 7.4% 9.8% 8.9%

*0.05<=P <=0.10,**P <0.05.†BMI(kg/m2),VO2max(ml/kg/minute),abdominalmusclestrength(sit-ups/minute),flexibility(cm),diastolicandsystolicbloodpressures(mmHg).

Table 7.3 Changesinphysicalhealthamongunemployedpersonswhoparticipatedinahealthpromotionprogramme

outcome measure Pre-test mean (sD)

Post-test mean (sD)

change (95% ci)

effect size (cohen’s d)

change (%)

BMI(n=216) 27.93(5.76) 27.86(5.70) -0.03(-0.12-0.06) 0.01 -0.1%

VO2max(n=97) 24.27(7.77) 25.60(8.08) 1.64**(0.53-2.76) 0.21 6.8%

Abdominalmusclestrength(n=196)

21.56(13.97) 31.24(17.94) 10.99**(9.02-12.96) 0.68 51.0%

Flexibility(n=191) 23.83(11.00) 25.10(10.68) 1.99**(1.17-2.81) 0.18 8.4%

Systolicbloodpressure(n=216)

130.63(17.57) 127.23(16.20) -3.28**(-5.48--1.08) 0.19 -2.5%

Diastolicbloodpressure(n=216)

83.39(10.98) 81.57(10.79) -1.83**(-3.40--0.26) 0.17 -2.2%

*0.05<=P <=0.10,**P <0.05,pairedt-test.

Ch

apte

r7


Determinantsofimprovementinphysicalhealth

Significant interactiontermsofsexwithtimeandbaselinevaluesofphysicalhealth

and time were consistently observed. Table 7.4 shows that men improved more in

VO2max, flexibility,andsystolicbloodpressure,whereaswomen improvedmore in

abdominal muscle strength. The effect of the category 25%-75% percentile of initial

physicalhealthdescribestheaverageimprovementinthestudypopulation,adjusted

forage,andwasclosetotheobserveddifferencesintable3.ForVO2max,abdominal

muscle strength, and flexibility, a statistically significant trend was observed with

subjectswithaninitiallylowerscoreonphysicalhealthimprovingmorethansubjects

with a better physical health. For systolic and diastolic blood pressure a strong re-

gressiontothemeanwasobservedwiththelowestgroupimprovingandthehighest

groupdeteriorating.

Discussion

Atthestartoftheprogramme,participantswereinpoorphysicalhealth,considering

theirlowVO2maxandthehighprevalenceofoverweightandobesity.Physicalhealth

oftheparticipantsimprovedsignificantly,exceptforBMI.Participants’cardiorespira-

toryfitness,abdominalmusclestrength,andflexibilityhadincreasedby6.8%-51.0%,

whereasdiastolicandsystolicbloodpressureshaddecreasedby2.2%-2.5%.Theeffect

sizesrangesfrom0.17-0.68,indicatingsmalltomoderateeffects.Participantswiththe

poorest physical health benefited most from the programme and sex differences in

improvementwereobserved.

Table 7.4 Determinantsofchangesinphysicalhealth†amongunemployedpersonswhoparticipatedinahealthpromotionprogramme(n=216)estimatedbymultivariatelinearregressionanalyseswithrepeatedmeasurements

change in vo2max

change in abdominal muscle strength

Flexibility systolic blood pressure

Diastolic blood pressure

Baselinevalue<25%percentile25%-75%percentile>75%percentile

3.37**2.06**-0.85

13.44**10.59**9.31**

3.61**2.00**0.25

8.33**-3.10-13.80**

5.22**-1.84-8.06**

SexMenWomen

2.19**1.23

10.30**11.59**

2.57**1.53**

-4.23**-2.46

-1.31-2.27**

*0.05<=P <=0.10,**P <0.05,pairedt-test.†VO2max(ml/kg/minute),abdominalmusclestrength(sit-ups/minute),flexibility(cm),diastolicandsystolicbloodpressures(mmHg).

116 Chapter7

Theparticipationinthishealthprogrammewas73%(n=338),whichwashigherthan

reportedinotherstudiesamonglowsocio-economicgroups[10,14]orinthegeneral

population.[13]Thehighparticipationwaspartlyduetothemoreorlesscompulsory

nature,whichmayalsoexplainthelackofanydifferencesbetweenparticipantsand

non-participants on demographic, psychological, or self-perceived health measures.

ParticipantswhocompletedtheprogrammehadahigherinitialBMIthandrop-outs,

indicatingthatthesubjectswhoneededtheprogrammethemostwerealsomostlikely

tofinishit.

Participantsintheprogrammewereaparticularlyunhealthygroup.Theprevalence

ofoverweightandobesitywas33%(n=83)and29%(n=74)respectively,ascompared

to40%and10%inthegeneralDutchpopulation.[26]Cardiorespiratoryfitnesswason

average30%lowerthaninhealthy,untrainedreferencegroups.[27]Inaddition,self-

perceivedhealthwasapproximately30%lowerthanarandomsampleofinhabitants

ofthesamecity(datanotshown).Althoughtheparticipants’physicalimprovements

werepromising,thechangesweregenerallymodest,andconsideringthepoorhealth

at baseline, the programme did not succeed to improve the participants’ physical

health to theaveragevalue in thegeneralDutchpopulation.Approximately25%of

therequiredimprovementwasreached.

Previous studies have shown similar improvements in physical health with flex-

ibilityincreasingwith9%afteraworksitehealthpromotionprogramme[28]anda3%

decreaseinsystolicanddiastolicbloodpressureafteranexerciseprogrammeamong

adults. [29] Slightly higher increases in maximum oxygen absorption have been re-

portedafterexerciseprogrammesamongdiabeticpatients(11.8%)[30]inobesewomen

(15%)[31],andstrokepatients (10%). [32]Bodymass indexremainedunchangedin

ourstudy,whichmaybeexplainedbythefactthatfoodintakewasnotaddressedin

theprogramme[28].Animportantconsiderationiswhetheralongerdurationofthe

programmeormoresessionsaweekwouldhaveresultedinlargergainsinphysical

health.Arecentreviewonseveralmodalitiesofphysicaltrainingprogrammesamong

diabeticpatientsshowedthattheinfluenceofprogrammedurationwaslimited,but

ahigherexerciseintensitywasassociatedwithgreaterincreaseinVO2max.[30]This

maybeconsideredasaguidelineforfutureexerciseprogrammes.

The improvement in physical health was predominantly associated with sex and

initialvalueofphysicalhealth,whereastrainingattendancenoranyofthedetermi-

nantsofphysicalhealthatbaselinewerenotassociatedwithimprovementsinphysi-

calhealth.Thefindingthatindividualsinpoorestphysicalconditionbenefitedmost

fromtheinterventionisinaccordancewithpreviousresearchintodecreasesinblood

pressure. [29,33]Forbloodpressureastrongregression to themeanwasobserved,

whereby subjects with high blood pressure decreased and subjects with low blood

pressureincreased.[24]Despitethisregressiontothemean,theimprovementamong

Ch

apte

r7


subjectswith intermediatebloodpressure indicates theoverall improvementdueto

participationinthehealthpromotionprogramme.

Theresultsarepromising,however, since theyare inaccordancewith findingsof

similar exercise interventions, even though subjects were not voluntarily enrolled

in the programme, and participants’ health improved consistently on all outcome

measures.Anearlier studyonunemployedpeoplewith lowbackpain [16] showed

improvements inphysicalhealthaswell,butmayhavebeenbiasedduetotheself-

selectionofmotivatedsubjects.Anotherlimitationisthelackofappropriateprocess

information on the implementation of the graded-activity principle in the physical

exercises.Feedbackprovidedbythephysicaleducationteachersindicatesthatgetting

theparticipantsinvolvedwasamajorchallengeinitselfandthatanincreaseintrain-

ingeffortwill certainlynothavebeenachievedbyallparticipants.Thismaypartly

explaintherathermoderategainsincardiorespiratoryfitness.Athirdweaknessofthis

studywasthechoiceforthebicycletestasmeasureofcardiorespiratoryfitness.At

baseline,halfoftheparticipantswerenotabletocarryoutthebicycletest.Thelackof

bikingskillsamongnon-Dutchpeoplemayhaveplayedarole,butalsoamongnative

Dutchpersonsatoolowcardiorespiratoryfitnesswasobservedtocarryoutthetest.

Personswhowerenotabletocarryoutthebicycleergometerpre-testdidnotdropout

oftheintervention,butwereallowedtoparticipateintheintervention.However,for

estimatingtheeffectoftheinterventiononcardiorespiratoryfitnesstheseparticipants

werenotincluded.Atpost-testsomeoftheparticipantswhowerenotabletocareout

the pre-test, where indeed able to carry out the post-test. The higher proportion of

participantcompleting this testat follow-up indicates that the improvement inVO-

2maxwillhavebeenunderestimated.Furthermore,itcannotbeexcludedthatpractice

effectsunderlietheimprovementsonabdominalmusclestrengthandflexibility.

Datacollectionwasconductedcompletelyindependentfromtheinterventionpro-

gramme,sinceattendingtheprogrammewasmoreorlessmandatorywhereaspartici-

pationinthisstudywascompletelyvoluntary.Asaconsequence,somesubjectswho

filledoutthequestionnairedidnottakepartintheprogramme,whereasothersubjects

tookpartintheprogrammebutdidnotrespondonthequestionnaire.Therefore,the

evaluationisbasedonlesspersons(46%,n=216)thanwhatwouldbeexpectedbased

ontheparticipation(73%)anddrop-outs(14%).

Persons with a low educational level and/or a non-Dutch origin may have had

difficultieswithfillingoutthequestionnaireduetoilliteracyordifficultieswiththe

Dutch language. To overcome these problems interviewers were used in this study.

The interviewers could offer an interview in the mother tongue (Dutch,Arabic, or

Turkish). However, the validity of the questionnaires may be less good for persons

withanon-Dutchoriginorlowliteracy,duetodifferencesininterpretationofques-

tionscausedbyculturaldifferences.

118 Chapter7

No associations between objective physical health and self-perceived health (SF-

36)werefoundinthisstudy.Itwasassumedthatbyimprovingphysicalhealthand

fitness, self-perceived physical and mental health would also increase. The lack of

anyassociationbetweenperceivedhealthandobjectivephysicalhealth,asmeasured

by cardio respiratory fitness, has been observed in several other studies. [34-35]

Perceivedhealthmaybeinfluencedbycognitions,forexamplethewaypeoplecope

with theirhealthproblems.Althoughphysicalhealthand self-reportedhealthwere

notmeasuredatexactly thesameday, timebetweenbothmeasurementsatbaseline

hadnoinfluenceonthelackofassociation.Duetothislackofassociation,itmightbe

questionedwhetherfocusingonphysicalhealthisthebestwaytoachievethemuch

needed improvement in self-reported health in this study population with health

complaintsthatwereoftenmentionedasabarriertostriveat(re)employment.Nev-

ertheless,improvementofphysicalhealthbyobjectivemeasurementsisbeneficialfor

thehealthstatus.

Toinvestigatetheeffectofthisinterventiononre-employmentasubstantiallylonger

follow-upperiodisneeded.However,weexpectthattheeffectoftheprogrammeon

re-employmentwillbemodestduetothefactthatphysicalfunctioningattheendof

theprogrammeisstillbelowtheaveragevalueinthegeneralDutchpopulation.

Thisstudyaddressesphysicalhealthwithinparticipantsofanexerciseprogramme.

TheoverallresultsoftheRCTongeneralhealthandsocialfunctioningispublished

elsewhere.[36]Schuring[36]foundthatthecurrenthealthpromotionprogrammedid

notshowbeneficialeffectsonperceivedhealth,psychologicalmeasures,workvalues,

job search activities or re-employment. This lack of positive effects of the interven-

tion, despite of the increased physical health, may be due to the fact that physical

functioningattheendoftheprogrammeisstillbelowtheaveragevalueinthegeneral

Dutchpopulationandthedurationoftheinterventionwasquitelimitedwithrespect

tosecondaryoutcomemeasuressuchasre-employment.Inadditiontothis,theseout-

comemeasureswereinvestigatedatleastthreemonthsaftertheendoftheprogram,

thebeneficialeffectsofthehealthprogrammemaybefadedawaybythattimedueto

alackoffollowupactivitiestosustainpossiblehealthbenefits.

Attheendoftheinterventionprogrammesemi-structuredinterviewswereunder-

taken with ten participants and ten trainers to obtain more qualitative insight into

differentaspectsoftheinterventionthatcouldbeimprovedinthefuture.Theprocess

evaluationshowedthataftertheendoftheprogramme,mostsubjectsfellbackinto

their old lifestyle with low levels of physical activity. In order to have sustainable

effects of a health promotion programme, it seems important for these participants

tohavecontinuedsupervisionandsupporttobeabletomaintainamorephysically

activelifestyle.

Ch

apte

r7


Thehealthsituationofunemployedpeoplemaydependonsocialandlabourmarket

policieswhichvaryacrossEuropeancountries.InthepastdecadesintheNetherlands,

healthproblemswereoftenalegitimatereasonforreceivingunemploymentbenefits.

Thesebenefitsareregardedsufficientlyhightocoverbasiccostsforalllivingexpenses.

Therefore,lessthan10%ofdisabledpersonshavetheirmainsourceofincomethrough

labour,whereasinSwedenthisproportionamountstoover50%.[37]Hence,results

fromstudiesconcerninghealthandemploymentstatusintheNetherlandsmaynotbe

easilybegeneralizedtoanotherEuropeancountry.

Conclusions

This study showed that (1) participation in a health promotion programme among

unemployed persons was not influenced by individual characteristics, but younger

personsweremorelikelytodropout,(2)physicalhealthmeasuredbycardiorespira-

toryfitnessincreasedonaverageby7%,and(3)participantswiththepooresthealth

at baseline benefited most from the programme. Although the health programme

consisting of an exercise and a cognitive component improved the health of unem-

ployedpersons,thisimprovementwasnotsufficientlyenoughtoraisephysicalhealth

tolevelsobservedamongindividualsinthegeneralpopulation.

Acknowledgements

FundingforthisstudywasprovidedbythePublicHealthFund(FondsOGZ)ofthe

Netherlands.

120 Chapter7

ReFeRences



3. WilsonSH,WalkerGM:Unemploymentandhealth:areview.Public health 1993,107:153-162. 4. SchuringM,BurdorfA,KunstAE,MackenbachJP:Theeffectofillhealthonenteringand

maintainingpaidemployment:evidenceinEuropeancountries.Journal of epidemiology and community health 2007,61:597-604.

5. CottCA,GignacMA,BadleyEM:DeterminantsofselfratedhealthforCanadianswithchronic disease and disability. Journal of epidemiology and community health 1999, 53:731-736.

6. VinokurAD,SchulY:MasteryandinoculationagainstsetbacksasactiveingredientsintheJOBSinterventionfortheunemployed.Journal of Consulting and Clinical Psychology 1997,65:867-877.

7. CreedPA:Improvingthementalandphysicalhealthofunemployedpeople:whyandhow?Med J Aust 1998,168:177-178.

8. WatersLE,MooreKA:Self-esteem,appraisalandcoping:Acomparisonofunemployedandre-employedpeople.Journal of Organizational Behavior 2002,23:593-604.


10. ChinnDJ,WhiteM,HowelD,HarlandJO,DrinkwaterCK:Factorsassociatedwithnon-participationinaphysicalactivitypromotiontrial.Public health 2006,120:309-319.

11. OleskeDM,KwasnyMM,LavenderSA,AnderssonGB:Participationinoccupationalhealthlongitudinalstudies:predictorsofmissedvisitsanddropouts.Annals of epidemiology 2007,17:9-18.

12. ProctorTJ,MayerTG,TheodoreB,GatchelRJ:Failuretocompleteafunctionalrestorationprogram for chronic musculoskeletal disorders: a prospective 1-year outcome study.Archives of physical medicine and rehabilitation 2005,86:1509-1515.

13. ToftUN,KristoffersenLH,AadahlM,vonHuthSmithL,PisingerC,JorgensenT:Dietandexerciseinterventioninageneralpopulation--mediatorsofparticipationandadherence:theInter99study.European journal of public health 2007,17:455-463.

14. WilburJ,McDevittJ,WangE,DancyB,BrillerJ,IngramD,NicolaT,LeeH,ZenkSN:Re-cruitmentofAfricanAmericanwomentoawalkingprogram:eligibility,ineligibility,andattritionduringscreening.Research in nursing & health 2006,29:176-189.

15. WinterEM,FowlerN:ExercisedefinedandquantifiedaccordingtotheSystemeInternationald’Unites.Journal of sports sciences 2009,27:447-460.

16. WatsonPJ,BookerCK,MooresL,MainCJ:Returningthechronicallyunemployedwithlowbackpaintoemployment.Eur J Pain 2004,8:359-369.

17. Gatchel RJ, PengYB, Peters ML, Fuchs PN, Turk DC: The biopsychosocial approach tochronicpain:scientificadvancesandfuturedirections.Psychol Bull 2007,133:581-624.

Ch

apte

r7


18. LukaskiHC,JohnsonPE,BolonchukWW,LykkenGI:Assessmentoffat-freemassusingbioelectricalimpedancemeasurementsofthehumanbody.The American journal of clinical nutrition 1985,41:810-817.

19. PearlinLI,SchoolerC:Thestructureofcoping.J Health Soc Behav 1978,19:2-21. 20. RosenbergM:Societyandtheadolescentself-image.Princeton: Princeton University Press

1965. 21. VlaeyenJWS,KoleSnijdersAMJ,BoerenRGB,vanEekH:Fearofmovement/(re)injuryin

chroniclowbackpainanditsrelationtobehavioralperformance.Pain 1995,62:363-372. 22. WareJE,Jr.,SherbourneCD:TheMOS36-itemshort-formhealthsurvey(SF-36).I.Concep-

tualframeworkanditemselection.Med Care 1992,30:473-483. 23. CohenJ:Statisticalpoweranalysisforthebehavioralsciences(2nded.).Hillsdale, NJ: Law-

rence Earlbaum Associates 1988. 24. BlandJM,AltmanDG:Someexamplesofregressiontowardsthemean.BMJ (Clinical research

ed 1994,309:780. 25. ChesterMR,BlandM,ChenL,KaskiJC:Therelationshipbetweenchangeandinitialvalue:

thecontinuingproblemofregressiontothemean.European heart journal 1995,16:289-290. 26. BlokstraA,SchuitAJ:Factsheetovergewicht.Prevalantieentrend[Factsheetoverweight.

Prevalenceandtrend].Bilthoven: Rijksinstituut voor Volksgezondheid en Milieu 2003. 27. ShvartzE,ReiboldRC:Aerobicfitnessnormsformalesandfemalesaged6to75years:a

review.Aviation, space, and environmental medicine 1990,61:3-11. 28. LiCL,TsengHM,TsengRF,LeeSJ:Theeffectivenessofanaerobicexerciseintervention

on worksite health-related physical fitness--a case in a high-tech company. Chang Gung medical journal 2006,29:100-106.

29. WheltonSP,ChinA,XinX,HeJ:Effectofaerobicexerciseonbloodpressure:ameta-analysisofrandomized,controlledtrials.Annals of internal medicine 2002,136:493-503.

30. BouleNG,KennyGP,HaddadE,WellsGA,SigalRJ:Meta-analysisoftheeffectofstructuredexercisetrainingoncardiorespiratoryfitnessinType2diabetesmellitus.Diabetologia 2003,46:1071-1081.

31. KraemerWJ,VolekJS,ClarkKL,GordonSE,IncledonT,PuhlSM,Triplett-McBrideNT,Mc-BrideJM,PutukianM,SebastianelliWJ:Physiologicaladaptationstoaweight-lossdietaryregimenandexerciseprogramsinwomen.J Appl Physiol 1997,83:270-279.

32. PangMY,EngJJ,DawsonAS,McKayHA,HarrisJE:Acommunity-basedfitnessandmobil-ityexerciseprogramforolderadultswithchronicstroke:arandomized,controlledtrial.Journal of the American Geriatrics Society 2005,53:1667-1674.

33. PescatelloLS,FranklinBA,FagardR,FarquharWB,KelleyGA,RayCA:AmericanCollegeofSportsMedicinepositionstand.Exerciseandhypertension.Medicine and science in sports and exercise 2004,36:533-553.

34. SorensenLE,PekkonenMM,MannikkoKH,LouhevaaraVA,SmolanderJ,AlenMJ:Asso-ciationsbetweenworkability,health-relatedqualityof life,physicalactivityandfitnessamongmiddle-agedmen.Applied ergonomics 2008,39:786-791.

35. SmeetsRJ,WittinkH,HiddingA,KnottnerusJA:Dopatientswithchroniclowbackpainhave a lower level of aerobic fitness than healthy controls?: are pain, disability, fear ofinjury,workingstatus,or levelof leisure timeactivityassociatedwith thedifference inaerobicfitnesslevel?Spine 2006,31:90-97.

122 Chapter7

36. SchuringM,BurdorfA,VoorhamAJ,derWeduweK,MackenbachJP:Effectivenessofahealthpromotionprogrammeforlong-termunemployedsubjectswithhealthproblems:arandomizedcontrolledtrial.Journal of epidemiology and community health 2009,63:893-899.

37. LakkaTA,KauhanenJ,SalonenJT:Conditioningleisuretimephysicalactivityandcardiore-spiratoryfitnessinsociodemographicgroupsofmiddle-agesmenineasternFinland.Int J Epidemiol 1996,25(1):86-93.

8General discussion

Ch

apte

r8

Generaldiscussion 125

8. GeneRaL Discussion


as demonstrated by a higher prevalence of illness and disability[1-2] and a higher

mortalityamongunemployedpersons.[3]Selectionandcausationmaycontributeto

theseinequalitiesinhealthamongemployedandunemployedpersons.Accordingto

theselectionhypothesisworkerswithapoorhealthmaybemorelikelytoleavethe

labourforce.Inaddition,unemployedpersonswithapoorhealthmaybelesslikely

toentertheworkforce.Thecausationhypothesissuggeststhatleavingtheworkforce

mayhaveanegativeinfluenceonhealthandthatgainingpaidemploymentmayhave

apositiveinfluenceonhealth.

Theobjectivesofthisthesisarethreefold:

1. To study the influence of poor health on entering (b) and maintaining (a) paid employment.

2. To study the influence of entering paid employment on different dimensions of perceived

health.

3. To evaluate the effectiveness of a health promotion programme among unemployed subjects

with health complaints on physical and mental health and re-employment.

8.1 main FinDinGs

In a descriptive study among unemployed inhabitants of the city of Rotterdam the

largesthealthinequalitiesbetweenemployedandunemployedpersonswereobserved

for native Dutch persons, whereas among persons with a non-Dutch origin, health

inequalities between employed and unemployed persons were less pronounced.

(Chapter2)

8.1.1HealthandWork

Objective 1a: To describe the influence of poor health on maintaining paid employment.

Inthisthesisapoorhealthwasdefinedaslessthangoodhealth.Alongitudinalstudy

with 2-year follow-up in the population aged over 50 years in 11 European Union

countries (SHARE) showed that the health status of older European workers had a

majorinfluenceonthelikelihoodofleavingtheworkforce.Apoorhealthwasarisk

factorforthetransitiontounemployment,retirementordisabilitypensionwithodd

ratiosvaryingfrom1.5to5.0.(Chapter3)

126 Chapter8

Secondary data-analysis on five waves (1994-1998) of the European Community

HouseholdPanel(ECHP)surveyontheworking-agepopulationin11Europeancoun-

tries also showed that transitions from paid employment to various forms of non-

employmentwereinfluencedbyhealth.Theeffectofpoorhealthontheprobabilityof

leavingtheworkforcewaslesspronouncedinmoredistantyearsandwasinfluenced

bysocio-economicvariables.Apoorhealthwasariskfactorforbecomingunemployed

orretiring,especiallyamonghighlyeducatedworkers.(Chapter4)

Objective 1b: To describe the influence of poor health on entering paid employment.

AprospectivestudyamongunemployedcitizensofthecityofRotterdamshowedthat

apoorperceivedhealthreducedtheprobabilityofenteringpaidemployment.Almost

alldimensionsofhealthatbaselinehadaninfluenceonthelikelihoodofenteringpaid

employment. The strongest association was found between physical functioning at

baselineandre-employment.(Chapter5)

TheaforementionedanalysisoffivewavesoftheECHPstudysupportedthefind-

ing that a poor health had a negative influence on the likelihood of entering paid

employment.Theeffectofpoorhealthonthelikelihoodofbecomingemployedwas

influencedbysocio-economiccharacteristics.Apoorhealthwasariskfactorfornot

enteringpaidemploymentamongmen,buthadlesseffectamongwomen.Theeffect

ofpoorhealthonenteringpaidemploymentwasconsistentlypresentuptofouryears

beforetheactualtransitiontookplace.Thehealthstatusamongunemployedsubjects

withpoorhealthremainedratherstableovertimeand,hence,hadalong-termeffect

ontheprobabilityofenteringpaidemployment.(Chapter4)

InmostEuropeancountries,apoorhealthwasariskfactorforleavingtheworkforce

andstayingunemployed,but largedifferencesamongcountrieswereobserved that

could partly be explained by the unemployment rate at national level. In countries

with a low national unemployment rate health competed with other labour market

factorsintheprocessofenteringorretainingpaidemployment,whereasincountries

withahighnationalunemploymentratetheeffectofhealthselectionoutofthework-

forcewasrelativesmallcomparedtootherfactorsthatdeterminelabouropportunities

forpersons.(Chapter4)

Inconclusion,apoorhealthofEuropeanworkersincreasedthelikelihoodofleaving

theworkforceduetounemployment,retirement,disabilitypension,ortakingcareof

ahousehold.Theeffectofpoorhealthontheprobabilityofleavingpaidemployment

waslesspronouncedinmoredistantyears.Unemployedsubjectswithapoorhealth

atbaselinewerealsolesslikelytoreturntopaidemployment.Almostalldimensions

ofhealth,butespeciallyphysical functioning,hadan influenceon the likelihoodof

entering paid employment. The effect of poor health on entering paid employment

Ch

apte

r8


wasconsistentlypresentup to fouryearsbefore theactual transition tookplace. In

mostEuropeancountries,poorhealthwasariskfactorforleavingpaidemployment

andstayingunemployed,butlargedifferencesamongcountrieswereobserved.

8.1.2WorkandHealth

Objective 2: To determine the influence of entering paid employment on different dimensions

of perceived health.

A prospective study among unemployed subjects receiving social security benefits,

who were capable of full time employment and were referred to a re-employment

centre in theCityofRotterdam, showed that5%of theparticipants (47/965) found

paidemploymentduringthefollow-upperiodofsixmonths.Amongthere-employed

participants all self-reported dimensions of health increased, with estimated effect

sizes varying from 0.1 to 0.7. The largest relative improvements were observed for

mental health, social functioning, and role-limitations due to emotional or physical

problems, whereas physical functioning showed the smallest relative improvement.

Allparticipantswerereferredtoatrainingcentreforare-employmenttraining.How-

ever,only30%oftheparticipantsactuallystartedwithare-employmenttraining.This

re-employmenttrainingdidnothaveapositiveeffectonthechanceofre-employment

oronthehealthstatusofparticipants.(Chapter5)

Inconclusion,amongre-employedpersonsalldimensionsofhealthimproved,with

thelargestimprovementformentalhealth,withinashortperiodofsixmonths.

8.1.3Healthpromotionoftheunemployed

Objective 3: To evaluate the effectiveness of a health promotion programme among unemployed

subjects with health complaints on physical and mental health and re-employment.

In a randomised controlled trial (RCT) the effectiveness of a health promotion pro-

grammeamongunemployedsubjects(n=921)withhealthcomplaintsreceivingsocial

security benefits from the city of Rotterdam was evaluated. The RCT showed that

thehealthpromotionprogrammehadnobeneficialeffectonself-reportedmentaland

physical health, self esteem, mastery, and pain-related fear of movement of unem-

ployedpersonswithhealthcomplaints.Inaddition,valuesandattitudestowardspaid

employment, job search activities, and employment status at six months follow up

werenotaffectedbythehealthprogrammeeither.(Chapter6)

The study among participants who received the intervention showed that cardio

respiratoryfitnessincreasedonaverageby7%.However,thisimprovementwasnot

128 Chapter8

sufficient to raise cardio respiratory fitnessofparticipants to levelsobserved in the

generalpopulation.(Chapter7)

Inconclusion,a randomisedcontrolled trial showed thatahealthpromotionpro-

grammeamongunemployedpersonswithhealthproblemsdidnot showbeneficial

effectsonself-reportedphysicalandmentalhealthandre-employment.

8.2 meTHoDoLoGicaL LimiTaTions

8.2.1HealthandWork

The SHARE study showed that the health status of older workers at baseline was

associatedwithemploymentstatustwoyearslater.(Chapter3)However,thisstudy

givesnoinsightintothetimewindowoftheinfluenceofapoorhealthontheprob-

ability to leave the workforce. The study includes only two measurements and the

exact moment of the employment transition within these two measurements is

unknown.Therefore,apoorhealthmayhaveanimmediateordelayedinfluenceon

theemploymenttransitionfromemploymenttonon-employment.Interruptedtime-

series with repeated measurements over longer periods can give more insight into

the time-dynamics of the influence of health on the transition from employment to

non-employment.TheECHPstudywasbasedonfivemeasurementsandshowedthat

the proportion of persons with a good health decreased among those who left the

workforce in theyearsbefore theemployment transition. (Chapter4)Therefore, the

ECHP study corroborates the findings of the SHARE study that a poor health is a

determinantoflabourforceexit.

TheECHPstudyshowedthattherewerelargevariationsbetweenEuropeancoun-

triesintheassociationbetweenillhealthandvariousformsofexitfrompaidemploy-

ment.[4]Duetosmallnumbers,thecountry-specificanalysesmaybelessreliablethan

theanalysisbasedon11Europeancountries.Thesevariationsmayreflectdifferences

between countries in institutional arrangements, e.g. the availability of disability

benefitschemesforthosewithhealthproblems.TheECHPstudyshowedthatalower

unemploymentrateatnationallevelwasassociatedwithlargereffectsofpoorhealth

onnotenteringpaidemploymentorbecomingunemployed.Thesestructuralfactors

atnationallevelmayalsoinfluenceresultswithinonecountry,suchasthestudiesthat

wereconductedintheCityofRotterdam,witharelativelyhighunemploymentrate

comparedtootherregionsoftheNetherlands.(Chapter2,5)

Ch

apte

r8


8.2.2WorkandHealth

AprospectivestudyamongunemployedcitizensoftheCityofRotterdamshowedan

increaseofhealthamongthosewhostartedwithpaidemploymentwithintheshort

follow-upperiod.(Chapter5)Duetotheobservationaldesignofthestudytheresults

mayhavebeeninfluencedbyselectionbias.Subjectswhofoundemploymentdiffered

fromthosewhostayedunemployedwithrespect tohealthatbaselineandpersonal

characteristics.Thesevariableswereadjustedforintheanalysis,buttheremayhave

beenothervariablesthatinfluencedbothhealthchangeandre-employmentthatwere

notincludedintheanalysis.

Theprospectivestudyconsistedoftwomeasurementsinbetweenwhichtheevent

ofenteringpaidemploymenttookplace.Basedonthesefinding,itseemsmostlikely

thatstartingwithpaidemploymentresultsinhealthimprovement.However,dueto

thedesignofthestudyitcannotberuledoutthatitistheotherwayaround;achange

inhealth statusmay increase the likelihoodof findingpaidemployment.However,

areviewshowedthatanincreaseinhealthstatusoflong-termunemployedisrather

unlikely.[2]Interruptedtime-serieswithrepeatedmeasurementscangivemoreinsight

intothetime-dependenttrendofhealthstatusbeforeenteringpaidemploymentand

theinfluenceofenteringpaidemploymentonthistrendofhealthstatus.


Arandomisedcontrolledtrialshowedthatahealthpromotionprogrammeaimedat

promotion of self-reported physical and mental health and re-employment among

unemployedpersonswithhealthproblemsdidnotshowbeneficialeffects.(Chapter6)

Thehealthprogrammemayhavehadnoeffectonself-reportedphysicalandmental

healthbecauseofalackofsustainabilityoftheintervention.Inthedesignofthestudy

avocationaltrainingdirectlyaftertheendofthehealthprogrammewasthoughttobe

abletosustainpositiveeffectsoftheintervention.However,inpracticethevocational

training was often delayed or did not start at all. This lack of follow-up activities

canbe consideredasan implementation failure.Theactivitiesof thehealthpromo-

tionprogrammewerenotproperlyintegratedwiththeactivitiesoftheemployment

services.

The study did not show any effect of the health promotion programme on re-

employment.Oneexplanationisthatthestudywasunderpowered.Theprospective

studyofunemployedpersonsreceivingsocialsecuritybenefitsinthecityofRotterdam

showedthatlessthan5%foundpaidemploymentwithinaperiodofsixmonths.The

probabilityof findingpaidemploymentwasevenworsefor theRCTstudypopula-

tion,becauseallparticipantswerefacedwithmentalorphysicalhealthproblems.The

130 Chapter8

follow-up period of six months (including three months participating in the health

promotionprogramme)mayhavebeentooshort tobeable to findanyeffectof the

interventionontheprobabilityofre-employmentof long-termunemployedpersons

withhealthproblems.

8.3 neW insiGHTs

8.3.1HealthandWork

1. Health selection out of the workforce is influenced by socioeconomic position of workers.

Healthselectionoutofthelabourforceiswelldocumented.Considerablylessatten-

tionhasbeendirectedtothewaysinwhichhealthselectionoutofthelabourforceis

moderatedbythesocioeconomicpositionofworkers.[5]

In this thesis it was found that the effect of poor health on labour force exit was

morepronouncedamonghigh-andintermediate-educatedworkersthanamonglow-

educatedworkers.Theeffectofpoorhealthamonghigh-andintermediate-educated

workersbecamelessstronginmoredistantyearsbeforeleavingtheworkforce.Low-

educated workers had a higher probability of becoming unemployed compared to

high-educated workers. Among low-educated workers, the influence of health on

theprobabilityofbecomingunemployedwasrelativelysmall,buttheprevalenceof

becoming unemployed was much higher than among high-educated workers. The

populationattributablefractionsofhealthinexitfrompaidemploymentwere4.3%,

4.0%,and3.1%for low, intermediate,andhigheducatedworkers.Therefore,apoor

healthhasan important influenceon labour forceexit amonghigh-aswell as low-

educatedworkers.

Itwasalsoshowninthisthesisthatamongethnicminoritygroupshealthinequali-

tiesbetweenemployedandunemployedpersonswerelessprofoundcomparedtothe

native-Dutch population. However, unemployment rates were higher among ethnic

minoritygroupscomparedtotheDutchpopulation.Whentheunemploymentrateis

high,theeffectofhealthselectionoutoftheworkforceisrelativelysmallcomparedto

otherfactorsthatdeterminelabouropportunities.[6]Theproportionofpersonswith

poorhealth that theoretically couldbeattributed tounemploymentvariedbetween

14%and26%amongdifferentethnicminoritygroups,comparedto13%fortheDutch

population.Thisindicatesthatemploymentstatusisanimportantfactorinsocioeco-

nomichealthinequalitiesinallethnicgroups.

Ch

apte

r8


2. The effect of ill health on labour force exit is influenced by labour market circumstances at

national level.

A comparative study of the relationship between unemployment and self-reported

healthin23Europeancountriesshowedthatthenegativerelationshipbetweenunem-

ploymentandhealth isconsistentacrossEuropebutvariesbywelfarestateregime,

suggestingthatlevelsofsocialprotectionmayhaveamoderatinginfluence.[7]Inthis

thesisitwasfoundthattheeffectofhealthonemploymentstatuswasinfluencedby

labourmarketcircumstances. Incountrieswithahighnationalunemploymentrate,

theeffectofhealthon labour forceexitwas lesspronouncedcompared tocountries

withalowunemploymentrate.Incountrieswithahighnationalunemploymentrate,

theeffectofhealth selectionoutof theworkforcewas relatively small compared to

otherfactorsthatdeterminedlabouropportunitiesforworkers.However,duetothe

highprevalenceoflabourforceexit,theamountofworkerswholefttheworkforcedue

tohealthproblemswasofcomparable importance incountrieswithahighnational

unemploymentrateaswellasincountrieswithalownationalunemploymentrate.It

wouldbeinterestingto investigatewhetherthelevelofprotectionforworkerswith

health problems against workforce exclusion and rehabilitation policies of different

Europeancountriescontributetotheinclusionofpersonswithhealthproblemsinthe

workforce.

3. Almost all dimensions of health, but especially physical functioning, had an influence on

the likelihood of becoming employed.

There is ample evidence to suggest that mental health influences the probability of

re-employment.[8] However, there is limited evidence about the influence of other

dimensionsofhealthontheprobabilityofre-employment.

Inthisthesisitwasfoundthatgeneralhealth,physicalfunctioning,bodilypain,so-

cialfunctioning,vitalityandrole-functioningduetoemotionalorphysicalproblems

allinfluencedthelikelihoodofre-employment.Thestrongestassociationwasfound

betweenphysicalfunctioningandre-employment.(Chapter5)

Health complaints may originate from a health condition, but the development

of chronic health problems often also depends on psychosocial factors. Perceptions

about the health problem and work may form a barrier to enter paid employment,

e.g.beliefsthatworkisharmful,andthatreturntoworkwilldofurtherdamageorbe

unsafe.[9]Inaddition,unemployedpersonswithhealthproblemsperformedlessjob

searchactivitiesandreceivedlessguidancetopaidemploymentfromtheemployment

services.[10-11](thisthesis,unpublishedresults)

132 Chapter8

To increase theprobabilityof re-employmentofunemployedpersonswithhealth

problems all barriers for entering paid employment should be addressed.[9] More

research is needed to develop evidence based practices to get unemployed persons

withhealthcomplaintsbackintotheworkforce.[12]

8.3.2Workandhealth

4. Among re-employed subjects mental as well as physical health improved within a short

time-period of six months.

In a review Waddel and Burton[13] concluded that re-employment leads to clear

benefitsinpsychologicalhealthandsomemeasuresofwell-being,althoughthereisa

dearthofinformationonphysicalhealth.Ameta-analyticstudyofthepsychological

and physical well-being during unemployment also demonstrated that the bulk of

researchisfocusedonmentalhealthoutcomes,suggestingthatotheraspectsofhealth

needmoreattention.[14]

Inthis thesis itwasfoundthatamongre-employedpersons,self-reportedgeneral

health, physical functioning, social functioning, vitality, mental health, bodily pain,

and role-limitations due to emotional or physical problems improved within six

monthsafterenteringpaidemployment,whereasthehealthstatusofthosewhocon-

tinuedtobeunemployedremainedunchanged.(Chapter5)

Workprovidesavarietyof features, including theopportunity for control,useof

skills, interpersonal contact, and provision of economic resources, that may be re-

sponsible for the increase of well-being among the re-employed persons.[15] In the

so-calledBlackreport(2008)itisstatedthatwork,matchedtoone’sknowledgeand

skillsandundertakeninasafe,healthyenvironment,canreversetheharmfuleffects

ofprolongedsicknessabsenceorlongtermunemployment,andpromotehealth,well-

beingandprosperity”.[16]Thisthesisprovidesevidencethatworkisindeedgoodfor

healthand,thus,workshouldbeconsideredasanimportantpartofhealthpromotion

programmesamongunemployedpersons.


5. It is of crucial importance among unemployed persons with health problems to integrate

activities to promote health with vocational training that promote re-employment.

Effortsoftheemploymentservicestoenhancere-employmentappearedtobenotef-

fectiveforunemployedpersonswithhealthproblems.(Chapter5)Thehealthproblems

haveanegativeinfluenceontheprobabilityofbecomingunemployed.Inaddition,the

Ch

apte

r8


unemployment situation has a negative influence on the health problems, through

inactivity,socialisolationandalackofdailyroutine.Tobreakthroughthisnegative

spiral, effortsneed tobe focusedonhealthpromotionaswellas re-employment.[9]

This thesis showed that it isof crucial importance to integrateactivities topromote

healthwithvocationaltrainingtopromotehealthandre-employmentofunemployed

personswithhealthproblems.

Arandomizedtrialontheeffectivenessandimplementationofsupportedemploy-

ment showed that high levels of integration of psychiatric and vocational services

increased theprobabilityof successfulworkoutcomesofunemployedpersonswith

severementalillness.[17]Supportedemploymentemphasisesrapidjobsearchonthe

basisofaperson’spreferenceandcontinuingsupporttoemployeeandemployerfrom

anemploymentspecialistworkingasanintegralmemberofthementalhealthservice.

Different studies have shown that supported employment is effective in helping

unemployed persons with severe mental illness to obtain competitive employment.

[18-19]

8.4 RecommenDaTions

8.4.1Recommendationsforpractice

1. Measures to keep workers with health complaints in the workforce should focus on high as

well as low educated workers.

A poor health status of European workers had a major influence on the likelihood

ofleavingtheworkforce.Theinfluenceofhealthontheprobabilityofbecomingun-

employed was more pronounced among high- and intermediate-educated workers,

buttheprevalenceofbecomingunemployedwasmuchhigheramonglow-educated

workers. (Chapter3)Therefore,apoorhealthhasan important influenceon labour

forceexitamonghigh-aswellas low-educatedworkers.Measures tokeepworkers

withhealthcomplaintsintheworkforceshouldfocusonhigh-aswellasloweducated

workers.

2. Measures to include unemployed persons in the workforce are recommended to reduce

socioeconomic inequalities in health.

In Europe, differences in healthy life expectancy between socioeconomic groups

typically amount to 10 years or more, counted from birth [20] According to many,

such differences in health are unacceptable, and represent one of Europe’s greatest

134 Chapter8

challenges for public health. Unemployed persons are a specific socioeconomically

disadvantagedgroup.Among re-employedpersonshealth improvedwithina short

period of six months.(Chapter 5) Therefore, in the ambition to narrow the existing

healthgapmeasurestoincludepersonsintheworkforcearerecommended.

3. It is recommended to integrate health promotion activities with labour market interven-

tions.

Apoorhealthreducestheprobabilityofenteringpaidemployment.Inarandomized

controlledtrialitwasshownthatahealthpromotionprogrammedidnothaveaneffect

onself-reportedphysicalandmentalhealthandre-employment(Chapter6).Thislack

ofeffectofthehealthpromotionprogrammemaybecausedbyalackofintegration

ofthehealthpromotionactivitieswiththeactivitiesoftheemploymentservices.Itis

recommendedtointegratedactivitiestopromotehealthwithactivitiesthatpromote

re-employment.ArecentpilotinRotterdamshowedthatsuchintegratedapproachis

possible and most likely also effective.[22-23] Work as therapeutic intervention has

beenshowntobeeffective insupportedemploymentprojects, that integratemental

healthcarewithlabourmarketsupport.[18-19]

8.4.2Recommendationsforresearch

1. It is of interest to study the influence of socio-demographic and psychosocial characteristics

on the effect of poor health on labour force exit.

Thelongitudinalstudywith5-yearfollow-up(ECHP)showedthattheeffectofpoor

healthontheprobabilityofleavingtheworkforcewasinfluencedbysex,educational

levelandmaritalstatusofworkers.(Chapter4)Itwouldbeinterestingtoinvestigate

whetherothersocio-demographic(e.g.age,ethnicbackground)orpsychosocialchar-

acteristics(e.g.selfesteem,socialsupport)influencetheassociationbetweenillhealth

andlabourforceexit.Itwouldbeinterestingtostudywhysomeworkerswithahealth

conditionleavetheworkforceandothersdonot.

2. It is of interest to investigate whether national regulations concerning workforce exclusion

and rehabilitation of persons with health problems contribute to the inclusion of persons

with health problems in the workforce.

The aforementioned longitudinal study with 5-year follow-up on the working-age

populationin11Europeancountries(ECHP)showedthattherewerelargevariations

betweenEuropeancountriesintheassociationbetweenillhealthandlabourforceexit,

Ch

apte

r8


thatcouldpartlybeexplainedby theunemployment rateatnational level.(Chapter

4)Thesevariationsmayreflectdifferencesbetweencountriesininstitutionalarrange-

ments,e.g.theavailabilityofdisabilitybenefitschemesforthosewithhealthproblems.

Itwouldbeinterestingtoinvestigatewhetherthelevelofprotectionforworkerswith

health problems against workforce exclusion and rehabilitation policies of different

Europeancountriescontributetotheinclusionofpersonswithhealthproblemsinthe

workforce.

3. The influence of entering paid employment on different dimensions of perceived health

should be corroborated

TheprospectivestudyamongunemployedcitizensoftheCityofRotterdamwithtwo

measurements inbetweenwhich theeventofenteringpaidemployment tookplace

showedanincreaseofhealthamongthosewhostartedwithpaidemploymentwithin

ashortfollow-upperiodofsixmonths.(Chapter5)Itisrecommendedtocorroborate

these findings with an interrupted time-series design with repeated measurements

overlongerperiods.Thisstudydesignwithmultiplemeasurementsofhealthbefore

andafterenteringpaidemploymentcangivemore insight into the time-dependent

trendofhealthstatusbeforeenteringpaidemploymentandtheinfluenceofentering

paidemploymentonthistrendofhealthstatus.Alongerfollow-upperiodisrecom-

mendedtoinvestigatewhetherthehealthimprovementissustainedintheyearsafter

re-employment.

4. It is recommended to evaluate the effectiveness of a highly integrated programme on health

and re-employment among unemployed persons with health problems in a randomised

controlled trial with a follow-up period of at least two years.

Arandomisedcontrolledtrialshowedthatahealthpromotionprogrammeaimedat

promotion of self-reported physical and mental health and re-employment among

unemployedpersonswithhealthproblemsdidnotshowbeneficialeffects.(Chapter6)

Itisadvisedtointegrateactivitiestopromotehealthwithactivitiesthatpromotere-

employment.Healthpromotionwithastrongfocusonincreasingsocial-andlabour

force participation may be a promising integrated approach to increase health and

re-employment of unemployed persons with health problems. Work as therapeutic

interventiontopromotehealthmayalsobeapromisingintegratedapproachtopro-

motehealthandre-employmentofunemployedpersonswithhealthproblems. It is

recommendedtoevaluatetheeffectivenessoftheseintegratedapproachesonhealth

andre-employmentamongunemployedpersonsinarandomisedcontrolledtrialwith

afollow-upofatleasttwoyears.

136 Chapter8

ReFeRences




4. AlaviniaSM,BurdorfA:Unemploymentandretirementandill-health:across-sectionalanalysisacrossEuropeancountries.International archives of occupational and environmental health 2008,82:39-45.

5. McDonoughP,AmickBC,3rd:Thesocialcontextofhealthselection:alongitudinalstudyofhealthandemployment.Soc Sci Med 2001,53:135-145.

6. FayersPM,SprangersMA:Understandingself-ratedhealth.Lancet 2002,359:187-188. 7. BambraC,EikemoTA:Welfarestateregimes,unemploymentandhealth:acomparative

studyoftherelationshipbetweenunemploymentandself-reportedhealthin23Europeancountries.Journal of epidemiology and community health 2009,63:92-98.


9. WaddellG,Burton,A.K.:Conceptsofrehabilitationforthemanagementofcommonhealthproblems.London: The Stationery Office 2004(www.tsoshop.co.uk):ISBN0117033944.

10. IWI:Kansenenbelemmeringen.Deondersteuningvanwerklozenmeteenuitkering.Inspec-tieWerkenInkomenMinisterievanSocialeZakenenWerkgelegenheid2008:oktober2008.

11. UWV:Kwartaalverkenning2009-I.2009. 12. BambraC,WhiteheadM,HamiltonV:Does‘welfare-to-work’work?Asystematicreviewof

theeffectivenessoftheUK’swelfare-to-workprogrammesforpeoplewithadisabilityorchronicillness.Soc Sci Med 2005,60:1905-1918.

13. WaddellG,Burton,A.K.:Isworkgoodforyourhealthandwell-being?London:TheStatio-neryOffice2006(www.tsoshop.co.uk):ISBN:9780117036949.

14. McKee-RyanF,SongZ,WanbergCR,KinickiAJ:Psychologicalandphysicalwell-beingdur-ingunemployment:ameta-analyticstudy.J Appl Psychol 2005,90:53-76.

15. WarrP,JacksonP:Adaptingtotheunemployedrole:alongitudinalinvestigation.Soc Sci Med 1987,25:1219-1224.

16. BlackC:Workingforahealthiertomorrow.London: The Stationery Office 2008(http://www.workingforhealth.gov.uk/Carol-Blacks-Review/):ISBN:9780117025134

17. CookJA,LehmanAF,DrakeR,McFarlaneWR,GoldPB,LeffHS,BlylerC,TopracMG,Razzano LA, Burke-Miller JK et al: Integration of psychiatric and vocational services: amultisite randomized, controlled trial of supported employment. Am J Psychiatry 2005,162:1948-1956.

18. CrowtherRE,MarshallM,BondGR,HuxleyP:Helpingpeoplewithseverementalillnesstoobtainwork:systematicreview.BMJ (Clinical research ed 2001,322:204-208.

19. BurnsT,CattyJ,BeckerT,DrakeRE,FiorittiA,KnappM,LauberC,RosslerW,TomovT,vanBusschbachJ et al:Theeffectivenessofsupportedemploymentforpeoplewithseverementalillness:arandomisedcontrolledtrial.Lancet 2007,370:1146-1152.

Ch

apte

r8


20. MackenbachJP:HealthInequalities:EuropeinProfile.Anindependent,expertreportcom-missionedby,andpublishedundertheauspicesof,theUKPresidencyoftheEU2005.

21. Lundin A, Lundberg I, Hallsten L, Ottosson J, Hemmingsson T: Unemployment andmortality--alongitudinalprospectivestudyonselectionandcausationin49321Swedishmiddle-agedmen.Journal of epidemiology and community health 2010,64:22-28.

22. JagmohansinghS:EvaluatieExITFeijenoord.RapportvandeSociaalWetenschappelijkeafdelingvandedienstSocialeZakenenWerkgelegenheid,Rotterdam2008.

23. OldenhofH,ColijnD:ExIT-ExtraIntensieveTrajectbegeleiding-Hetmultidisciplinairere-integratie instrumentvooreennaarvermogenparticiperendeengezondestad.RapportvandeDienstSocialeZakenenWerkgelenheidvandeGemeenteRotterdam2009.

Sum

mar

y

139Summary

summary


asdemonstratedbyahigherprevalenceofillnessanddisabilityandahighermortality

amongunemployedpersons.Selectionandcausationmaycontributetotheinequali-

tiesinhealthamongemployedandunemployedpersons.Selectionmayactthrough

twodifferentpathways.Workerswithapoorhealthmaybemorelikelytoleavethe

labourforce.Unemployedpersonswithapoorhealthmaybelesslikelytoenterthe

workforce.Causationmayalsoactintwodifferentways.Leavingtheworkforcemay

haveanegativeinfluenceonhealthoftheex-workers.Theotherwayaround,gaining

paidemploymentmayhaveapositiveinfluenceonhealth.

Theprimaryobjectivesofthisthesiswere(1)tostudytheinfluenceofpoorhealthon

enteringandmaintainingpaidemployment,(2)toinvestigatetheinfluenceofenter-

ingpaidemploymentondifferentdimensionsofperceivedhealth,and(3)toevaluate

theeffectivenessofahealthpromotionprogrammeamongunemployedpersonswith

healthcomplaintsonphysicalandmentalhealthandre-employment.Thefirstobjec-

tivewasaddressedinChapter3and4,thesecondobjectivewasaddressedinChapter

5,andthethirdobjectivewasaddressedinChapter6and7.

The associations between health and employment may not be similar across all

socio-economicgroups.chapter 2presentsacross-sectionalstudyamonginhabitants

oftheCityofRotterdamintheNetherlands,whichaimedtoevaluatewhetherhealth

inequalitiesassociatedwithunemploymentwerecomparableacrossdifferentethnic

groups. Differences in ill health between employed and unemployed persons were

lessprofoundinethnicminoritygroupscomparedwiththenativeDutchpopulation,

buttheprevalenceofunemploymentwasmuchhigherinethnicminoritygroups.The

proportions of persons with poor health that could be attributed to unemployment

variedbetween13%and28%acrossethnicgroups.

HeaLTH anD WoRk

Inmanyindustrializedcountries thepopulation isageing,dueto increasing lifeex-

pectancyandfallingbirthrates.Aratherparadoxicaldevelopmentisthat,despitein-

creasesinlifeexpectancy,theaveragetimepersonsspendinpaidworkhasdecreased

orbeenstable inmostEuropeancountries.Thisparadox isexplainedbyprolonged

educationamongyoungercohortsandamoreimportantcontributoristhehighrate

ofexitfromthelabourmarketatolderages.

In chapter 3, a longitudinal study with 2-year follow-up in the working popula-

tion aged between 50 and 63 years in 11 European Union countries (SHARE) was

Summary140

presented.Thislongitudinalstudyaimedtodeterminetheimpactofillhealthonexit

frompaidemploymentinEuropeamongolderworkers.Duringthetwo-yearfollow-

upperiod,17%ofemployedworkersquittedpaidemployment,primarilyduetoearly

retirement. Poor self-perceived health was strongly associated with exit from paid

employmentduetoretirement,unemployment,ordisability(ORsfrom1.32to4.37).

Consequencesofillhealthforthelikelihoodofbecomingorstayingunemployment

may also depend on socioeconomic position of workers and labour market circum-

stancesatnationallevel.Inchapter 4,alongitudinalstudywithfiveannualwavesof

theEuropeanCommunityHouseholdPanel(ECHP)surveyispresented,whichaimed

toexaminetheeffectofillhealthonenteringandmaintainingpaidemployment.The

effectofillhealthonlabourforceexitstronglydependedoneducation,sex,andfam-

ilysituation.InmostEuropeancountries,apoorhealthorachronichealthproblem

predicted staying or becoming unemployed and the effects of health were stronger

withalowernationalunemploymentlevel.

WoRk anD HeaLTH

Thereisampleevidencetosuggestthatre-employmentleadstoclearbenefitsinpsy-

chologicalhealth.However,itislessclearwhetherre-employmentalsohasbeneficial

effects on other measures of health, e.g. physical health. chapter 5 presents a pro-

spectivestudywith6monthsfollow-upamongunemployedsubjectsreceivingsocial

securitybenefitsfromtheCityofRotterdam,theNetherlands.Theaimofthisprospec-

tivestudywastoinvestigatetheeffectofre-employmentondifferentdimensionsof

healthwithinashortperiodafterenteringpaidemployment.Amongthere-employed

participantsallself-reporteddimensionsofhealthincreasedwithinsixmonths,with

estimatedeffectsizesvaryingfrom0.1to0.7.Thelargestrelativeimprovementswere

observedformentalhealth,socialfunctioning,androle-limitationsduetoemotional

or physical problems, whereas physical functioning showed the smallest relative

improvement.

HeaLTH PRomoTion amonG THe unemPLoyeD

Apoorhealthwillactasabarriertoreturntopaidemployment.Inordertoincrease

the possibilities for re-employment, improvement in health of unemployed persons

may,therefore,beanimportantstep.Inchapter 6,arandomisedcontrolledtrial(RCT)

with six months follow-up among unemployed persons with health complaints is

presented.TheaimofthisRCTwastoevaluatetheeffectivenessofahealthpromotion

Sum

mar

y

141Summary

programmeamongunemployedpersonswithphysicalandmentalhealthcomplaints

andre-employment.TheRCTshowedthatthehealthpromotionprogrammehadno

beneficial effect on self-reported mental and physical health, self esteem, mastery,

and pain-related fear of movement. In addition, values and attitudes towards paid

employment, job search activities, and employment status at six months follow up

were not influenced by the health programme either. The lack of integration of the

intervention intoregularvocationalrehabilitationactivitiesmayhavenegativelyaf-

fectedtheresultsofthehealthprogrammeamongunemployedpeople.

chapter 7 presentsa longitudinal studyamongparticipantswho received the in-

tervention.Theaimofthisstudywastoevaluatechangesinphysicalhealthamong

participants.Thestudyshowedthatcardiorespiratoryfitnessincreasedonaverageby

7%.However,thisimprovementwasnotsufficienttoraisecardiorepiratoryfitnessof

participantstolevelsobservedinthegeneralpopulation.

chapter 8,theGeneralDiscussion,startswithasummaryofthemainresultsofthis

thesis,followedbystudylimitationsthatshouldbeacknowledgedwheninterpreting

the results. Five specific new insights from the studies described above were dis-

cussed.Basedontheresearchpresentedinthisthesisthefollowingrecommendations

arepresented:

RecommenDaTions FoR PRacTice

1) Measurestokeepworkerswithhealthcomplaints intheworkforceshouldfocus

onhighaswellasloweducatedworkers.

2) Measures to includeunemployedpersons in theworkforceare recommended to

reducesocioeconomicinequalitiesinhealth.

3) It is recommended to integrate health promotion activities with labour market

interventions.

RecommenDaTions FoR ReseaRcH

1) Itisofinteresttostudytheinfluenceofsocio-demographicandpsychosocialchar-

acteristicsontheeffectofpoorhealthonlabourforceexit.

2) It isof interest to investigatewhethernationalregulationsconcerningworkforce

exclusion and rehabilitation of persons with health problems contribute to the

inclusionofpersonswithhealthproblemsintheworkforce.

3) Theinfluenceofenteringpaidemploymentondifferentdimensionsofperceived

healthshouldbecorroborated.

Summary142

4) Itisrecommendedtoevaluatetheeffectivenessofahighlyintegratedprogramme

onhealthandre-employmentamongunemployedpersonswithhealthproblems

inarandomisedcontrolledtrialwithafollow-upperiodofatleasttwoyears.

Sam

env

atti

ng

143Samenvatting

samenvatting

Derelatietussenwerkloosheideneenslechtegezondheidisveelvuldigaangetoond,

zoalsblijktuiteenhogereprevalentievanziekteenbeperkingeneneenhogeresterfte

onder werklozen. Twee verschillende mechanismen, selectie en causatie, dragen bij

aandegezondheidsverschillentussenwerkendeenwerklozemensen.Selectiewerkt

optweeverschillendemanieren.Teneerstehebbenwerknemersmeteenslechtege-

zondheideengroterekansomdearbeidsmarktteverlaten.Tentweedehebbenwerk-

lozen met een slechte gezondheid minder kansen om een betaalde baan te vinden.

Causatie werkt mogelijk ook op twee verschillende manieren. Stoppen met werken

heefteennegatieveinvloedopdegezondheidvanex-werknemers.Daarnaastkanhet

verkrijgenvaneenbetaaldebaandegezondheidpositiefbeïnvloeden.

Debelangrijkstedoelstellingenvanditproefschriftzijn(1)hetonderzoekenvande

invloedvaneenslechtegezondheidophetverkrijgenenbehoudenvaneenbetaalde

baan, (2)hetonderzoekenvande invloedvanhetverkrijgenvaneenbetaaldebaan

op verschillende aspecten van de ervaren gezondheid, en (3) het evalueren van de

effectiviteit van een gezondheidsbevorderingsprogramma op de fysieke en mentale

gezondheid en werkhervatting van werklozen met gezondheidsklachten. De eerste

doelstelling wordt behandeld in Hoofdstuk 3 en 4, de tweede doelstelling wordt

behandeldinHoofdstuk5endederdedoelstellingwordtbehandeldinHoofdstuk6

en7.

De samenhang tussen gezondheid en werkstatus is mogelijk niet gelijk voor ver-

schillende sociaaleconomische groepen. Hoofdstuk 2 beschrijft een cross-sectionele

studienaardesamenhangtussengezondheidenwerkstatusbijRotterdammersmet

verschillende etnische achtergronden. Gezondheidsverschillen tussen werkenden

enwerklozenwarenmindergrootbij etnischeminderheden invergelijkingmetau-

tochtoneNederlanders,maarhetwerkloosheidspercentagewashogeronderetnische

minderheden.Deproportiepersonenmeteenslechtegezondheiddoorwerkloosheid

varieerdeperetnischegroepvan13%tot23%.

GezonDHeiD en WeRk

Inveelgeïndustrialiseerdelandenveroudertdepopulatie,dooreentoenamevande

levensverwachtingeneenafnamevanhetgeboortecijfer.Eenparadoxaleontwikke-

ling is dat, ondanks de toenemende levensverwachting, het gemiddeld aantal jaren

datiemandbesteedtaanbetaaldwerkindeafgelopendecenniaisgedaaldofgelijkis

gebleven indemeesteEuropese landen.Dezeparadoxwordtgedeeltelijkverklaard

Samenvatting144

doordatjongeremensenlangeronderwijsvolgen,maareenbelangrijkereverklaringis

dehogeuitvaluitdearbeidsmarktonderouderemedewerkers.

In Hoofdstuk 3 wordt een longitudinale studie onder oudere werknemers in 11

Europeselanden.Doelvandezelongitudinalestudiewashetbepalenvandeinvloed

vaneenslechtegezondheidopuitstroomuitdearbeidsmarktinEuropaonderoudere

werknemers(leeftijd50-63jaar).Gedurendedefollow-upperiodevantweejaarverliet

17%vandeouderewerknemersdearbeidsmarkt,voornamelijkvanwegevervroegde

pensionering.Erwaseensterkesamenhangtusseneenslechteervarengezondheiden

uitvaluitdearbeidsmarktdoorpensionering,werkloosheidofarbeidsongeschiktheid

(oddsratio’svariërendvan1.32tot4.37).

Consequentiesvaneenslechtegezondheidopdekansomwerkloostewordenente

blijvenismogelijkookafhankelijkvandesociaaleconomischepositievanwerknemers

endearbeidsmarktomstandighedenopnationaalniveau.InHoofdstuk 4wordteen

longitudinalestudiemetvijf jaarlijksemeetmomentenvaneenvragenlijstonderzoek

onder Europese huishoudens (ECHP) beschreven, naar het effect van een slechte

gezondheidoneenbetaaldebaanteverkrijgenentebehouden.De invloedvaneen

slechtegezondheidopuitstroomuitdearbeidsmarktisafhankelijkvanopleidingsni-

veau,geslachtengezinssituatie.IndemeesteEuropeselandeniseenslechtegezond-

heidvaninvloedopdekansomwerkloostewordenenteblijven,waarbijdeinvloed

vangezondheidgroterisinlandenmeteenlaagwerkloosheidspercentage.

WeRk en GezonDHeiD

Er is veelvuldig aangetoond dat werkhervatting een positieve invloed heeft op de

mentalegezondheid.Echter,het isminderduidelijkofwerkhervattingook leidt tot

verbetering van de fysieke gezondheid. In Hoofdstuk 5 wordt een prospectief on-

derzoekonderwerklozeRotterdammersmeteenbijstandsuitkeringbeschrevenmet

eenfollow-upperiodevanzesmaanden.Hetdoelvandezeprospectievestudiewas

hetonderzoekenvande invloedvanhetstartenmeteenbetaaldebaanopverschil-

lendeaspectenvandegezondheidbinnenkortetijdnahetstartenmeteenbetaalde

baan.Alleaspectenvandeervarengezondheidvanmensendiegestartwarenmeteen

betaalde baan verbeterden binnen zes maanden. Het starten met een betaalde baan

hadrelatiefdegrootste invloedopdementalegezondheid,hetsociale functioneren

enrolbeperkingendooreenemotioneeloffysiekprobleem,terwijlfysiekfunctioneren

deminsteverbeteringlietzien.

Sam

env

atti

ng

145Samenvatting

GezonDHeiDsbevoRDeRinG bij WeRkLozen

Een slechte gezondheid vormt een belemmering voor het starten met een betaalde

baan. Het verbeteren van de gezondheid van werklozen is mogelijk een belangrijk

stapomdekansophetvindenvaneenbetaaldebaantevergroten. InHoofdstuk 6

wordteengerandomiseerdexperimentbeschrevenmeteenfollow-upperiodevanzes

maanden onder werklozen met gezondheidsklachten. Het doel van dit experiment

was het evalueren van de invloed van een gezondheidsbevoderingsprogramma op

defysiekeenmentalegezondheidenkansophetstartenmeteenbetaaldebaanvan

werklozenmetgezondheidsproblemen.Uithetonderzoekbleekdathetgezondheids-

bevorderingsprogrammageeninvloedhadopervarenmentaleenfysiekegezondheid,

zelfwaardering,hetomgaanmetproblemenenpijngerelateerdeangstomtebewegen.

Daarnaasthadhetgezondheidsprogrammaookgeeninvloedoppersoonlijkewaarden

enattitudestenaanzienvanbetaaldwerk,hetzoekennaareenbetaaldebaan,enhet

daadwerkelijkstartenmeteenbetaaldebaan.Hetontbrekenvaneengoedeintegratie

vanhetgezondheidsprogrammainderegulierere-integratieactiviteitenheeftmoge-

lijkeennegatiefeffectgehadopdeeffectiviteitvanhetgezondheidsprogrammavoor

werklozen.

Hoofdstuk 7 beschrijft een longitudinaal onderzoek naar verandering van de fy-

sieke gezondheid van deelnemers van het gezondheidsprogramma. Het onderzoek

laatziendathetfysiekeuithoudingsvermogenvandeelnemersvanhetgezondheids-

programmagemiddeld7%verbeterde.Echter,dezeverbeteringwasnietvoldoende

omdefysiekeconditieophetniveauvandealgemenebevolkingtekrijgen.

Hoofdstuk 8,dealgemenediscussie,begintmeteensamenvattingvandebelangrijk-

steresultatenvanditproefschrift,vervolgenswordenbeperkingenvanhetonderzoek

beschrevendieinhetachterhoofdgehoudenmoetenwordenbijhetinterpreterenvan

deresultaten.Vijfnieuweinzichtenvoortkomenduitbovenstaandeonderzoekenwor-

denbeschreven.Opbasisvanhetonderzoekinditproefschriftwordendevolgende

aanbevelingengegeven:

aanbeveLinGen vooR De PRakTijk

1. Maatregelen om werknemers met gezondheidsklachten aan het werk te houden

moetenzichrichtenopzowelhogeralsoplageropgeleidewerknemers.

2. Maatregelengerichtophetaanhetwerkgaanvanwerklozenmetgezondheids-

klachten kunnen bijdragen aan het verkleinen van sociaaleconomische gezond-

heidsverschillen.

Samenvatting146

3. Activiteitengerichtophetverbeterenvandegezondheidvanwerklozenzouden

geïntegreerdmoetenwordenmetwerkgerelateerdeinterventies.

aanbeveLinGen vooR onDeRzoek

1. Hetwordtaanbevolenomdeinvloedvansociaaldemografischeenpsychosociale

kenmerken op de relatie tussen een slechte gezondheid en uitstroom uit de ar-

beidsmarktteonderzoeken.

2. Het wordt aanbevolen om de invloed van nationale regelingen op het op de

arbeidsmarkt krijgen en behouden van mensen met gezondheidsklachten te on-

derzoeken.

3. Deinvloedvanhetstartenmeteenbetaaldebaanopverschillendeaspectenvande

ervarengezondheiddientdoorandereonderzoekenbevestigdteworden.

4. Hetwordtaanbevolenomdeeffectiviteitvaneen integraleaanpakvangezond-

heidsproblemenenre-integratienaarwerkvoorwerklozenmetgezondheidspro-

blemen te onderzoeken in een gerandomiseerd experiment met een follow-up

periodevanminstenstweejaar.

Dan

kw

oo

rd

147Dankwoord

Dankwoord

Alleenmijnnaamprijktopdevoorkant,maarvelenhebbenbijgedragenaandetot-

standkomingvanditproefschrift.Graagwilikeenaantalmensenbedankenvoorhun

bijdrage.

Allereerst wil ik mijn promotoren, Lex Burdorf en Johan Mackenbach, bedanken.

Lex,jouwambitieomhoogleraartewordenisrecentverwezenlijkt.Mijnpromoveren

markeertdaarmeetevensvoorjoueenspeciaalmoment,waarinjevoorheteerstals

promotorzaloptreden.Dankzijjouwsteunenvertrouwenishetvoormijmogelijkom

mijn ambities te verwezenlijken, als promovendus en inmiddels als postdoc onder-

zoeker.Ikwiljebedankenvoordeprettigeenleerzamesamenwerking.

Ikbeschouwhetalseenvoorrechtdatnietéén,maartweehooglerarenvandeafdel-

ingMaatschappelijkeGezondheidszorgalspromotoroptreden.Johan,het ismijeen

eer en genoegen dat jij van meet af aan betrokken bent geweest bij het onderzoek.

Bedanktvoorjeinhoudelijkebijdrageaanditproefschrift.GraagwilikookInezJoung

bedankenvoorhaarinzetalsbegeleiderbijdestartvanhetproject“Gezondaande

Slag”.

HetprojectGezondaandeSlag isontstaanuiteenunieksamenwerkingsverband

tussen de GGD Rotterdam-Rijnmond, de dienst Sociale Zaken en Werkgelegenheid

(SoZaWe) van de gemeente Rotterdam en het Erasmus MC. Dit project is de basis

geweest voor een belangrijk deel van dit proefschrift. Vanuit de GGD waren Toon

Voorham, Sven Graumans en Kees der Weduwe intensief betrokken bij het project.

VanuitdeSoZaWespeeldenAatBrandenPaulvandenBergeenbelangrijkerol.Ikwil

alle ledenvanhetprojectteamhartelijkbedankenvoorhunbijdrageaanhetproject

GezondaandeSlag.

VanuitErgocontrolhebbenmetnameJanPlat,AnnekeZwartenallebetrokkendo-

centeneenbelangrijkerolgespeeldbijhetprojectGezondaandeSlag.Hetuitvoeren

vandeinterventiehebbenjulliemetheelveeltoewijdinggedaan.Dankdaarvoor.

Graagwil ikookHalukArslanenalle interviewersbedankenvoordeuitdagende

taakominterviewsaftenemenbijeenbijzondermoeilijktebereikendoelgroep.Jul-

lie zijn in staat geweest om gewapend met de vragenlijst kwalitatief goede data te

verzamelen.Veeldankdaarvoor.

Ondermijncollega’svanhetMGZzijnernogeenaantalmensendieikgraagwil

bedanken voor hun bijdrage aan dit proefschrift. Heren van de helpdesk en dames

van het secretariaat, bedankt voor de goede ondersteuning. Marieke Hartman en

Christine Schutgens, graag wil ik jullie bedanken voor jullie inzet bij het uitvoeren

van het onderzoek voor jullie Master thesis. Halime, als onderzoeksassistent heb jij

veel werk voor het project Gezond aan de Slag verricht. Ik kon daarbij blindelings

Dankwoord148

opjevertrouwen.Daarnaastkondenwehetookpersoonlijkgoedmetelkaarvinden.

Dankvoordefijnesamenwerking.

Het zijn mijn directe collega’s die ervoor zorgen dat ik met plezier op het MGZ

werk.Graagwil ikmijnkamergenoten,Frank,Freek,Karien,Mohammed,Goedele,

Carlijn,EefjeenHester,bedankenvooralleleukeeninteressantegesprekken.Delunch-

wandelingiseengoedegewoontevanveleMGZ-ers.GraagwilikmijnMGZcollega’s

bedanken voor die gezellige wandelingen en de vele gesprekjes ‘onderweg naar de

koffieautomaat’.Fijnookdatikaltijdeveneenkamerkanbinnenlopenmeteenvraag

ofzomaar.Elin,Tilja,Suzan,Karien,Bart,Hilde,Klazine,Birgitte,Katrina,Willemieke,

Marianne en Loes, bedankt voor de gezelligheid door de jaren heen. Karien, jij zal

ééndagvóórmij jeproefschriftverdedigenen...westaanerallebeiopdiedagniet

alleenvoor!Carlijn,hetisfijnomjoualsvriendinéncollegatehebben.Bedanktvoor

jebetrokkenheidensteun.Hetiseengeruststellendideedatjijalsparanimfaanmijn

zijdezalstaanbijdeverdedigingvanmijnproefschrift.

Alhoewelfamilieenvriendengeendirectebijdragehebbengeleverdaanhetproef-

schrift,zijnzeindirectwelvanbelanggeweest.Andrea,Marie-JoséeenAnnemieke,

bedanktvoorjullievriendschapengezelligheidtijdenshetdrinkenvaneenkopjethee,

etentjeofwandeling.Annemieke,alsparanimfhebjeeenbelangrijkesymbolischerol

opdepromotiedag.Ikvindhetfijndatjijopmijnpromotiedagerbijzalzijn.

Lieve familieleden, bedankt voor de oprechte interesse door de jaren heen. Oma

Annie,nuisereindelijk,naastaldieingenieurs,ookeendoctorindefamilie!Mamma

enDick,bedanktvoorjulliebetrokkenheidensteun.Hetisfijnomtewetendatjullie

trotszijnopmij. Julliewetenalsgeenanderdathetsomsnietmeevaltomwerken

gezin tecombineren. Ikkomnogaltijdgraagbij jullie langsomeven ‘uit tepuffen’

vanonsdrukkebestaan.Roos,alsgrotezusheb jealtijdeenbelangrijke rol inmijn

leven gespeeld. Fijn dat we met onze gezinnen zo’n fijne familie vormen. Thea, ik

hebhet erggetroffenmetzo’n lieve schoonmoeder.Bedanktvoor jegastvrijheiden

behulpzaamheid.

Lieve Bastiaan, bedankt voor het leven dat wij samen leven. Inmiddels met drie

kleineboevenerbij endevierdeopkomst.Hetafrondenvanditproefschrift is een

mijlpaal, maar geen eindpunt. Er liggen nog vele uitdagingen op onze weg. Samen

metjoudurfikdezeaantegaan.

Cu

rric

ulu

mV

itae

149CurriculumVitae

curriculum vitae

Merel Schuring werd geboren op 24 maart 1977 teAlphen aan den Rijn. Zij genoot

haarbasisonderwijsopdeAlphenseMontessorischoolinAlphenaandenRijn.In1995

behaaldezijhaarVWO-diplomaaanhetScalaCollegeinAlphenaandenRijn.Daarna

startte zij met de studie Biologie aan de Universiteit Utrecht. In 2001 behaalde zij

haardoctoraalBiologie.Vanafapril2001werktezijalswetenschappelijkonderzoeker

bij het Coronel Instituut voorArbeid en Gezondheid van hetAMC inAmsterdam.

Vanaf november 2003 tot augustus 2008 was zij als wetenschappelijk onderzoeker

verbondenaandeafdelingMaatschappelijkeGezondheidszorgvanhetErasmusMC

inRotterdamenvoerdehetpromotieonderzoekuitdatresulteerdeinditproefschrift.

GelijktijdigvolgdezedeMasteropleidingPublicHealthaanhetNIHES(Netherlands

Institute for Health Sciences) en behaalde in augustus 2008 haar Master of Public

Health.Vanaugustus2008totmei2009waszijwerkzaamalsstatistischonderzoeker

bijhetCentraalBureauvoordeStatistiekinVoorburg.Sinds1mei2009iszeopnieuw

verbondenaandeafdelingMaatschappelijkeGezondheidszorgvanhetErasmusMC

inRotterdamals(postdoctoraal)onderzoeker.

Merel isgetrouwdmetBastiaandeBruinensamenhebbenzijdriezonen:Tobias

(2005),Storm(2007)enQuinten(2009).

Merel Schuring was born on March 24, 1977, in Alphen aan den Rijn, the Nether-

lands.Shereceivedherprimaryschooleducationat theAlphenseMontessorischool

in Alphen aan den Rijn. She obtained her secondary school education at the Scala

CollegeinAlphenaandenRijnin1995.Thereafter,shestartedstudyingattheFaculty

ofBiologyattheUniversityofUtrecht.In2001shegraduatedherMasterofScience

inBiology.Shewasthenemployedasascientificresearcherat theCoronelInstitute

ofOccupationalHealthattheAcademicMedicalCentreinAmsterdam.From2003to

2008,sheworkedasascientificresearcherattheDepartmentofPublicHealthatthe

ErasmusMedicalCentreinRotterdam,whereshecarriedouttheresearchpresented

inthisthesis.Duringthistime,shealsoenrolledintheMasterofScienceprogramme

at the Netherlands Institute for Health Sciences, and obtained her Master of Public

Health inAugust2008.FromAugust2008toMay2009,shewasemployedasasta-

tisticalresearcheratStatisticsNetherlandsinVoorburg.Sheiscurrentlyemployedas

postdoctoral researcherat theDepartmentofPublicHealthat theErasmusMedical

CentreinRotterdam.

MerelismarriedtoBastiaandeBruinandtheyhavethreesons:Tobias(2005),Storm

(2007)andQuinten(2009).

151Publicationsinthisthesis

Pu

bli

cati

on

s

Publications in this thesis

SchuringM,BurdorfA,KunstA,VoorhamT,MackenbachJ.(2009).Ethnicdifferences

inunemploymentandillhealth.International Archives of Occupational and Environmen-

tal Health,82:1023-30.

Schuring M, BurdorfA, VoorhamAJ, der Weduwe K, Mackenbach JP (2009). Effec-

tivenessofahealthpromotionprogrammefor long-termunemployedsubjectswith

healthproblems:arandomizedcontrolledtrial.Journal of Epidemiology and Community

Health,63:893-899.

Schutgens CAE, Schuring M, Voorham AJ, Burdorf A (2009). Changes in physical

health among participants in a multidisciplinary health programme for long-term

unemployedpersons.BMC Public Health,9:197.

SchuringM,BurdorfA,KunstAE,Mackenbach JP (2007).Theeffectof illhealthon

enteringandmaintainingpaidemployment:evidenceinEuropeancountries.Journal

of Epidemiology and Community Health, 61:597-604.

SchuringM,MackenbachJ,VoorhamAJ,BurdorfA.Theeffectofre-employmenton

perceivedhealth.Journal of Epidemiology and Community Health,inpress.

vandenBergT,SchuringM,AvendanoM,MackenbachJ,BurdorfA.Theimpactof

ill health on exit from the labour market in Europe. Occupational and Environmental

Medicine,inpress.

oTHeR PubLicaTions

derWeduweK,SchuringM,BurdorfL,VoorhamT(2006).GezondaandeSlag-ge-

zondheidbevorderenvanlangduriguitkeringsgerechtigden.TijdschriftvoorGezond-

heidswetenschappen,6:354-357

HaafkensJ,MoermanC,SchuringM,vanDijkF(2006)Searchingbibliographicdata-

basesfor literatureonchronicdiseaseandworkparticipation.Occupational Medicine

(London),56:39-45.

Publicationsinthisthesis152

Schuring M, Sluiter JK, Frings-Dresen MHW (2004) Evaluation of top-down imple-

mentationofhealthregulationsinthetransportsectorina5-yearperiod.International

Archives of Occupational and Environmental Health,77:53-9.

Schuring M, Sluiter JK, Frings-Dresen MHW (2003) Onregelmatige werktijden: het

ontstaanenherstelvanvermoeidheid.Tijdschrift voor Ergonomie,28:4-10.

Ph

DP

ort

foli

o

153PhDPortfolio

PhD Portfolio

summary of PhD training and teaching activities

NamePhDstudent:MerelSchuring PhDperiod:November2003–August2008ErasmusMCDepartment:Publichealth Promotors:J.P.Mackenbach,MD,PhD

A.Burdorf,PhDResearchSchool:Netherlandsinstituteforhealthsciences(Nihes)

year Workload(ecTs)

1. PhD training

General academic skills & Research skillsMasterofScienceinPublicHealth,NetherlandsInstituteforHealthSciences(Nihes),Rotterdam,TheNetherlands

2004-2008 70

Presentations SchuringM,MackenbachJP,VoorhamAJ,BurdorfA.“Theeffectofre-employmentonperceivedhealth”.PaperpresentedattheDepartmentofPublicHealth,ErasmusMC,Rotterdam,TheNetherlands

SchuringM,BurdorfA.“EvaluatieGezondaandeSlag”.ErgoControlCongres,Woudenberg,TheNetherlands

SchuringM,BurdorfA.“EvaluatieGezondaandeSlag-deinterventie”.ExpertmeetingGezondaandeSlag,Rotterdam,TheNetherlands

SchuringM,GraumansS.“GezondaandeSlag!Gezondheidbevorderenvanlangduriguitkeringsgerechtigden”.KenniscentrumGroteSteden;expertatelierSEGV,TheNetherlands.

VoorhamAJ,SchuringM,BoschR.Workshop:“GezondaandeSlag!Gezondheidbevorderenvanuitkeringsgerechtigden”.NederlandsCongresVolksgezondheid,Rotterdam,TheNetherlands.

2010

2008

2007

2006

2006

1

1

1

1

1

2. Teaching activities

SupervisingMaster’sthesis

Lecture(oninvitation)mastercourseWorkandOrganisationalPsychology,RadboudUniversityNijmegen,TheNetherlands

2006-2007

2010

6

1

The Role of Health and Health Promotion in Labour Force - RePub

Documents

Transcript of The Role of Health and Health Promotion in Labour Force - RePub