35395426

REVIEW ARTICLE

Loneliness and the health of older people

Conor O Luanaigh*y and Brian A. Lawlorz

Mercer’s Institute for Research in Ageing, St. James’s Hospital, Dublin 8, Ireland

SUMMARY

Background The intense focus on major psychiatric disorders in both contemporary psychiatric research and clinicalpractice has resulted in the relative neglect of less definable constructs such as loneliness and how such entities might impacton health outcomes. The purpose of this review is to raise awareness among physicians and psychiatrists of the medicalimpact and biological effects of loneliness as well as making the argument that loneliness should be a legitimate therepeutictarget.Methods Using Pubmed we searched the literature for research and review papers looking at loneliness as a construct, howit is measured and its health effects. We reviewed the relevant papers and have summarized their main findings.Results Loneliness has strong associations with depression and may in fact be an independent risk factor for depression.Furthermore loneliness appears to have a significant impact on physical health being linked detrimentally to higher bloodpressure, worse sleep, immune stress responses and worse cognition over time in the elderly. There is a relative deficiency inadequate evidence based treatments for loneliness.Conclusion Loneliness is common in older people an is associated with adverse health consequences both from a mentaland physical health point of view. There needs to be an increased focus on initiating intervention strategies targetingloneliness to determine if decreasing loneliness can improve quality of life and functioning in the elderly. Copyright # 2008John Wiley & Sons, Ltd.

key words— loneliness; depression; physical health; cognition; treatment

INTRODUCTION

The intense focus on major psychiatric disorders in bothcontemporary psychiatric research and clinical practicehas resulted in the relative neglect of less definable orclear-cut constructs such as loneliness and how suchentities might impact on health outcomes. Although thehealth consequences of loneliness equal or exceed thatof many major psychiatric disorders, there is littlemention of the construct of loneliness in the psychiatricliterature, with most publications appearing in thesocial psychology and social gerontology journals. The

purpose of this review is to raise awareness amongphysicians and psychiatrists of the medical impact andbiological effects of loneliness and make the argumentthat loneliness may represent a legitimate target forhealth care interventions.

WHAT IS LONELINESS?

Loneliness can be defined as a subjectively experi-enced aversive emotional state that is related to theperception of unfulfilled intimate and social needs(Peplau and Perlman, 1982). Weiss (1973), suggestedthat there are two types of loneliness that can co-existor occur independently; social and emotional lone-liness. Social loneliness occurs through isolation andis caused by a lack of social integration and em-beddedness. This type of loneliness may, for instance,

INTERNATIONAL JOURNAL OF GERIATRIC PSYCHIATRY

Int J Geriatr Psychiatry 2008; 23: 1213–1221.

Published online 9 June 2008 in Wiley InterScience

(www.interscience.wiley.com) DOI: 10.1002/gps.2054

*Correspondence to: C. O. Luanaigh, Mercer’s Institute forResearch in Ageing, St. James’s Hospital, Dublin 8, Ireland.E-mail: [email protected] Fellow in Old Age Psychiatry.zConsultant Old Age Psychiatrist.

Copyright # 2008 John Wiley & Sons, Ltd.Received 6 December 2007

Accepted 10 April 2008

be experienced following relocation and could be bestresolved by acquiring new contacts. Emotional lone-liness on the other hand develops because of anabsence of a reliable attachment figure, such as apartner. In line with attachment theory (Bowlby,1969), the absence or loss of an attachment figure canonly be substituted by another close and intimate bondand supportive friendships or social contact cannotcompensate for the loss. The distinction betweenemotional and social loneliness may be particularlyrelevant for studies among older people because, dueto the death of aging relatives and friends, theprobability of having or finding an intimate attach-ment figure decreases with age thus incidence andprevalence of emotional loneliness in particular maybe raised.

Some believe that loneliness can be conceptualizedas a unidimensional entity (Russell et al., 1980; Russell,1996) and assume that the experience of loneliness isthe same across all circumstances and causes. Forexample, the loneliness of a soldier for loved ones backhome is equivalent to the loneliness of the dejectedmember of a dissolved intimate relationship.

The above conceptualizations of loneliness do notinclude a temporal component or refer to thefrequency or severity of the experience in theirdefinitions. To differentiate normal from pathologicalloneliness, the duration, frequency and severity of theexperience are of key importance. Feelings ofloneliness may either be persistent or short-lived.Transient feelings of loneliness are often situationallydetermined and normative similar to the depressivesymptoms experienced in an adjustment reaction.However, chronic or more severe feelings of lone-liness may be more of a cause for concern.

There would appear to be both state and trait aspectsto loneliness. The state of loneliness that results fromrelocation may be remedied by reintegration with newcontacts; however, there may exist in any oneindividual temperament or trait characteristics thatmilitate against improvement in loneliness withrestructuring of contacts. (Cacioppo et al., 2006a),present an evolutionary theory for loneliness in whichthey conclude that loneliness is a human trait that hasevolved and is heritable, suggesting that certainindividuals are more vulnerable to developing feelingsof loneliness in response to environmental trig-gers than others. In this theory, they postulate thatthose of our early ancestors who were inclined to formsocial connections, communicate and work togetherand share food had a selective advantage to survive.

In conditions of hardship, early hunter-gathererswho had a genetic disposition to experience loneliness

may have been more likely to return to share theirfood, shelter, or defense with their family and alliesto diminish the pain of loneliness. Individuals withno such feelings of loneliness when separated fromothers may have roamed the earth better nourishedthan those who felt distressed by social separation, butthe abandoned offspring—and the genetic predis-position of the parents—would have been less likelyto survive. Thus loneliness as a trait ensured thesurvival of the ‘gene’ in accordance with RichardDawkins’ ‘selfish gene’ theory at the expense of theindividual’s own distress. This theory is backed upsomewhat by recent genetic studies that have showna significant degree of heritability for loneliness(Boomsma et al., 2005).

Perhaps a more comprehensive way to viewloneliness would be as a heterogeneous construct thatcan be defined in terms of the ‘bio-psycho-social’model. One can have a biological predisposition toexperiencing loneliness possibly related to inheritedpersonality traits, while others may experience lone-liness related to other psychological precipitants suchas grief, or depression. Then from a social perspectiveone can experience loneliness if socially isolated. Theadvantages of explaining loneliness in such a way isthat it highlights clearly the factors behind why theindividual is experiencing loneliness and thus aids indeciding the most appropriate therapeutic interven-tion.

It is important finally, to distinguish betweenloneliness, ‘aloneness’, isolation and lack of socialsupport. Being alone or ‘aloneness’ is not alwaysviewed negatively by the individual. Furthermore, aperson may experience loneliness while in thecompany of others. While loneliness by definition isan undesirable experience, aloneness or solitude maybe desirable fostering creativity, facilitating self-reflection, self-regulation, concentration and learning.Social isolation is an objective measure of poor socialintegration without subjective appraisal. While there isoverlap in these terms, they should not be usedinterchangeably as they reflect different constructs.

ASSESSMENT OF LONELINESS

Measures of loneliness, by definition, are generallybased on subjective report. Self report measures, bytheir very nature, capture only the publicly declaredexperience of loneliness which may differ from theprivately felt extent of the feeling.

The assessment of loneliness can vary from the useof a single question to the use of detailed self reportinstruments. The simplest way to assess loneliness is

Copyright # 2008 John Wiley & Sons, Ltd. Int J Geriatr Psychiatry 2008; 23: 1213–1221.

DOI: 10.1002/gps

1214 c. o luanaigh and b. a. lawlor

to use a single question—‘Do you feel lonely?’ toelicit a self rated scaled measure of loneliness withanswers from ‘never’ to ‘often/always’. The advan-tages of a such self report measure are that it is easyto use in clinical and research settings, is generallyacceptable to people and it asks directly about feelingsof loneliness. However, its simplicity is also aweakness. The question presumes an understandingof the concept of loneliness by patients and parti-cipants, when it is likely that the nature and meaningof the concept will vary among different groups ofpeople and over time. The use of a single item selfreport instrument also fails to differentiate betweenemotional or social loneliness, and the state and traitaspects of the experience. Although this may seemsomewhat academic, when considering interventionsor treatment for loneliness knowing the ‘type’ ofloneliness can be important in treatment success.

Alternative self-report approaches to measuringloneliness have been developed using derivativequestionnaire scales. Table 1 summarizes some ofthe main scales used along with their properties. The

University of California Los Angeles (UCLA) Lone-liness Scale (Russell, 1996) is the most commonlyused self-report loneliness instrument, preferred byboth researchers and clinical practitioners. It wasdeveloped to assess satisfaction with social relation-ships. It is conceptualized as one-dimensional in itsstructure and the scale items tap both the frequencyand intensity of salient aspects and events of the lonelyexperience (e.g. ‘How often do you feel alone?’). Itmeasures loneliness as a unitary phenomenon that isthe same regardless of what may have caused it. Inorder to reduce response bias, the word ‘lonely’ neverappears in the instrument. Of note also the scale doesnot specify a time frame for respondents so it isunclear whether a state or trait measure has beendesigned.

Unlike the UCLA Loneliness scale, a number of theother scales (de Jong-Gierveld Scale, WittenbergEmotional vs Social Loneliness Scale, the RussellEmotional and Social Loneliness Scale and the Socialand Emotional Loneliness Scale for Adults) canmeasure social and emotional loneliness. In the de

Table 1. Loneliness scales and their properties

Name of scale Author(s) Brief description Psychometric properties

de Jong-Gierveld Scale de Jong-Gierveld (1987) Eleven-item scale – six itemsassess for emotional lonelinesswhile the other five assesssocial loneliness

a coefficients between 0.7 and0.76 for six-item scale.Correlation between thesix-item scale and 11-itemscale very high – between 0.93and 0.95

UCLA Loneliness Scale Russell (1996) Twenty-item scale, each itemrated from 1 (never) to 4 (often)

High internal consistencya¼ 0.92. Good test reliability(r¼ 0.73 after 12 months)

Social and EmotionalLoneliness Scale for Adults

DiTommaso and Spinner(1993, 1997)

Thirty-seven-item seven-pointLikert scale. Divided intosubscales measuring romantic,family and social loneliness.

Internal consistency of 0.89.Moderate subscale intercorre-lations.

Loneliness Rating Scale Scalise et al. (1984) Forty-adjective Likert scale –assesses four ten-itemdimensions – agitation,dejection,depletion andisolation.

High internal consistencies forthe four subscales (0.82< a <0.89)

Wittenberg Emotional vsSocial Loneliness Scale

Wittenberg (1986; in Shaverand Brennan, 1991)

Five-item Likert scales toassess social and emotionalloneliness

Good internal consistencies forboth subscales (a¼ 0.78 and0.76 respectively). The twoScales were moderatelycorrelated (r¼ 0.44)

Russell emotional and SocialLoneliness Scale

Russell et al. (1984) Single items each consisting ofa two sentence description of atype of loneliness(emotionalor social) followed by anine-point scale.

No internal consistency datareported due to titemsingularity. Authors reportnon-significant relationbetween item scores.

Differential Loneliness Scale Schmidt and Sermat (1983) Sixty-item true-false scale –divided into four subscales –romantic, friendship, family,and large group loneliness

High internal consistencya> 0.89 with subscaleestimates above 0.70


DOI: 10.1002/gps

loneliness and the health of older people 1215

Jong-Gierveld Scale three dimensions of loneliness areidentified: intensity, time perspective, and emotionalcharacteristics. Of note, the ‘Wittenberg Emotional vsSocial Loneliness Scale’ and the ‘Social and EmotionalLoneliness Scale for Adults’ are both moderatelycorrelated with the UCLA scale with the subsectionsmeasuring social loneliness especially correlated.

The Differential Loneliness Scale (DLS) identifiesspecific areas and dimensions of relationships in whichloneliness might be experienced and measures theextent to which a person is satisfied or dissatisfied witheach of four specific social relationships; romantic,sexual, friendship, relationships with family andrelationships with larger groups or the community.

When it comes to deciding whether to use the‘single question self-report measure’ or one of theloneliness instruments it is important to consider whatit is you want to assess about the experience ofloneliness. If it is simply to gauge the prevalenceof loneliness then the self-report single questionshould be sufficient. However, if the type of lonelinessthe person is suffering from it important becausetherapeutic interventions are being considered, thena scale that assesses the types of loneliness is moreappropriate. In a primary care setting where time isoften limited, use of the ‘single question’ is timeefficient and will identify people who are experiencingloneliness and alert the doctor to a potential problemthat needs looking into. In psychiatry clinics orresearch settings, the loneliness scales can be usedto characterize the quality of loneliness in more detail.

EPIDEMIOLOGY OF LONELINESS

Prevalence

British community studies have reported rates ofloneliness among those aged 65 þ of between 5–16%(Sheldon, 1948; Townsend, 1957; Tunstall, 1966;Townsend, 1968; Hunt, 1978; Bond and Carstairs,1982; Qereshi and Walker, 1989). The medianreported rate is approximately 9–10% (Victor et al.,2000). A more recent study found that 7% were oftenor always lonely (Victor et al., 2005). In an Irishcontext, the ongoing Dublin Healthy Ageing Studyshowed 10% reported being often or always lonely atthe baseline assessments (O’Connell et al., unpub-lished data).

Gender

Most studies found that women report loneliness morecommonly than men, although some have found nogender differences (Andersson, 1982; Peplau and

Perlman, 1982; Qereshi and Walker, 1989). Severalinterpretations of these differences have beensuggested. Factors such as older age, greater like-lihood of being widowed and depression, are all morecommon in women as well as being individuallyassociated with loneliness and thus they can confoundthe findings that loneliness is more common inwomen. It has also been proposed that there is also agender bias in self-disclosure—i.e. men may be lesslikely to admit feelings of loneliness.

Marital status

The group that reports the highest frequency ofloneliness is non-married males followed by non-married females, and then married females and finallymarried males (Andersson, 1990). While this associ-ation with different demographic groups is not entirelyconsistent and exceptions do occur, these findingswould appear to indicate that marriage is somewhatprotective against loneliness, concurring with Weiss’sview of emotional loneliness occurring in the absenceof a reliable attachment figure such as a spouse.

Bereavement

Bereavement in the elderly appears to be a major riskfactor for the development of loneliness (Perlman andPeplau, 1984; Wenger et al., 1996; Victor et al., 2005).It is plausible that loneliness can result from the deathof a loved one through the loss of an attachment figureand then the state of being widowed in which theperson may live alone and have no replacementattachment figure.

Age

There is no clear relationship between loneliness andincreasing age. In fact, self reported feelings ofloneliness may be actually higher in adolescents.When the prevalence figures are plotted as a curve theproportions of different age groups reporting lone-liness assumes the character of a shallow ‘U’ (Peplauet al., 1982; Andersson, 1990, 1993). Thus, althoughthe data is less robust, it seems as if there is an increasein loneliness in the highest age-groups—fromapproximately age 75 (Perlman and Peplau, 1984).The increase in loneliness in the higher age groupsmight be explained by higher incidence of emotionaltype loneliness as a result of increased likelihood ofwidowhood (with resultant loss of attachment figure).It may also be possible that similar to late onsetdepression that brain abnormalities (deep white matter


DOI: 10.1002/gps


changes, atrophy) plays a role in the personexperiencing more loneliness the older they are.Research into such a possibility is lacking butcertainly may be worth pursuing.

Genetic/personality factors

There is some evidence that loneliness can bedetermined by genetic and temperament factors. Inone study, individual differences in lonelinessdemonstrated considerable temporal stability andheritability with an estimate of heritability being48% (Boomsma et al., 2005). Cacioppo et al. (2006a)looked at the possible association between lonelinessand personality type and found that higher levels ofloneliness were found in individuals with lower levelsof extraversion and higher levels of neuroticism.Preliminary data from the ongoing Dublin HealthAgeing study also indicates that loneliness(measuredusing the single question—‘Do you feel lonely?’) isassociated with higher scores of neuroticism evenwhen controlled for depressive symptoms (O’Connellet al., unpublished data).

Other risk groups for loneliness have beenidentified, examples of which are: single parents,family caregivers, unemployed, low level of edu-cation, poor income, people who have moved recently,people who live alone, poor vision and loss of hearing(Perlman and Peplau, 1984; Savikko et al., 2005).

While the majority of these risk factors are notmodifiable (gender, temperament, bereavement), as isthe case in depression, they can help to identify at riskgroups for early intervention.

LONELINESS AND DEPRESSION: CHICKENOR SCRAMBLED EGG?

Depression and loneliness in older people are stronglyassociated and often co-occur (Creecy et al., 1985;Hansson et al., 1986; Heikkinen and Kauppinen,2004). Furthermore, loneliness is a consistent andstrong risk factor for depression. However, in spite ofthe overlap, loneliness and depression are consideredto be distinct entities. From a clinical perspective, it iswell known that older people who are lonely will oftenpresent with depressive symptoms but depressedindividuals do not always report feeling lonely. Factoranalytic studies also provide some support the ideathat depression and loneliness are separate constructs(Cacioppo et al., 2006b).

Table 2 summarizes the main studies that haveexamined loneliness as a potential risk factor fordepression. Prince et al. (1997) explored the relation-

ship between social support deficits, loneliness andlife events as risk factors for depression in the elderlyand found that frequent loneliness was one of thestrongest cross-sectional associations wih depression.Further cross-sectional studies have been carried outthat have looked at specific groups in the elderlynamely people in independent living retirementcommunities and elderly males respectively (Alpassand Neville, 2003; Adams et al., 2004).

Loneliness was again significantly associated withdepression and both of these studies indicated thatloneliness was a potential risk factor for depression.All of these studies point to a definite associationbetween loneliness and depression but because of theirdesign are unable to indicate whether lonelinesscauses depression or depression causes loneliness, orindeed whether loneliness is simply part of thedepressive syndrome.

Longitudinal studies are more helpful in establish-ing whether loneliness is firstly separate fromdepression and secondly whether it is an actual riskfactor for depression. Green et al. (1992) carried out alongitudinal study over 3 years and found that feelingsof loneliness were significantly associated withdevelopment of depression 3 years later. Heikkinenand Kauppinen (2004) in a 10-year longitudinal studyon very old Finnish adults found that lonelinesspredicted long term trajectories in depressive symp-toms. In a more recently published study, the abovefinding was extended by looking at loneliness as aspecific risk factor for depression using a crosssectional and longitudinal approaches in a sample ofadults (Cacioppo et al., 2006b). In the cross-sectionalarm of the study, higher levels of loneliness wereassociated with higher levels of depressive symptoms,independent of the effects of age, gender, ethnicity,education, income, marital status, social support, andperceived stress. In the longitudinal arm of the studywhich was carried out over 3 years loneliness wasagain associated with more depressive symptoms,independent of demographic covariates, maritalstatus, social support, hostility and perceived stress.The analysis showed reciprocal influences over timebetween loneliness and depressive symptomatology,suggesting that loneliness and depressive symptoma-tology can act in a synergistic way to diminish wellbeing in older adults.

LONELINESS AND PHYSICAL HEALTHSTATUS

As well as the more apparent connection betweenloneliness and psychological and psychiatric morbid-


DOI: 10.1002/gps


ity, loneliness appears to be a risk factor for poorphysical health. (Stek et al., 2004) showed that whileloneliness on its own does not affect mortality rates, itsignificantly increases mortality rates in the elderlywhen it co-exists with depression. The mechanism forthis effect on mortality is unclear but may be mediatedthrough altered immune or physiological responses.

Table 3 summarizes the studies that have beencarried out examining the association of lonelinesswith various physiological and immune measures.

Loneliness is associated with hypertension, poorsleep and abnormal stress responses (Cacioppo et al.,2002a, 2002b; Steptoe et al., 2004). These findingssuggest that loneliness is associated detrimentally withphysiological and immune stress responses that mayaccount for the excess cardiovascular morbidity seenin people who are lonely. Although all of these studies

controlled for depression as well as other lifestylevariables, none of them controlled for anxiety, whichcould be a confounding factor.

LONELINESS AND COGNITIVE DECLINE

There have been a number of studies that examinedwhether loneliness is a risk factor for cognitivedecline. As part of the Helsinki Ageing study Tilviset al. (2004) found that the subjective experience ofloneliness could over time result in greater cognitivedecline in the elderly than people who are not lonely.More recently Wilson et al. (2007) analysed theassociation between loneliness and Alzheimer’sdisease in 823 older adults over a 4-year period andfound that the risk for developing Alzheimer’sdementia was substantially increased in those who

Table 2. Loneliness as a risk factor for depression

Authors Type of study Measure ofloneliness

Size of study Main findings

KB Adams, S Saundersand EA Auth

Cross-sectional UCLA Scale 234 participants(53% response rate)

Loneliness was a risk factor fordepression using multipleregression compared with othervariables.

J T Cacioppo, M EHughes, L J Waite, LCHawkley, R A Thisted

Cross-sectionaland longitudinal

UCLA Scale Cross sectional study—1,945 participants—54þLongitudinalstudy—over 3 years212 cases

Study 1 and 2—Higher levels ofreported loneliness associatedwith higher levels of depressivesymptoms.

F M Alpass, S Neville Cross-sectional UCLA Scale 217 men (72.3%response)

Strongest relationship todepression using regressionanalysis was loneliness Loneliness(but not social supports) relatedto depression

M L Stek, D J Vinkers,J Gussekloo, A T FBeekman, R C van derMast, r G J westendorp

Cohort regular1 year folow-up

De Jong GierveldScale

599 (87% response) Mortality risk of depression withperceived loneliness was twotimes higher than normal, with nosignificant mortality risk fordepression in absence of loneliness

P Tikkainen andR-L Heikkinen(34)

5 year cohort Single itemquestion—‘Do youfeel lonely?’

262 (92%response),with 5years later148 patricipants.

Lonely people had significantlymore depressive symptoms atbaseline and at follow up. Socialintegration explains lonelinessbut not depressive symptoms.

Green BH, CopelandJR, Dewey ME,Sharma V etal

3 year cohort Single item question—‘Do you feel lonely?’

1070 participants Low life satisfaction, lonelinessand smoking were associatedwith development of depressionafter 3 years. Multivariateanalysis confirmed theirindependent effects.

Prince MJ, HarwoodRH, Blizard RA,Thomas A andMann AH

Cross-sectional Single item question—‘Do you feel lonely?’—extracted from AgeCatprogram.

654 participantsinterviewed

Loneliness strongly associatedwith depression – Odds Ratio12.4 (7.6–20.0)

Heikkinen R andKauppinen M

10 year cohort Single itemquesetion—‘Doyou feel lonely?’

337 at baseline, 131at 10 year follow-up

Loneliness predicted depressivesymptomatology


DOI: 10.1002/gps


were lonely as compared to those who were not lonely,even when controlling for the level of social activity.At autopsy, loneliness during life was not actuallyrelated to the severity of Alzheimer’s disease patho-logy. It would seem that loneliness may increase therisk of clinical expression of dementia for the samedegree of Alzheimer’s neuropathology.

LONELINESS AS A TARGET FORTHERAPEUTIC INTERVENTIONS

Loneliness is associated with adverse health out-comes, including mortality, depression, poor sleep,systolic hypertension, heart disease and cognitivedecline. The direction of causality is unclear as themajority of the evidence comes from cross sectionalstudies and therefore more longitudinal studies areneeded to further elucidate potential mechanisms ofcausation. However, we would feel also that furtherresearch should be focused on possible interventionsagainst loneliness. A number of different interventionapproaches for loneliness have been suggested in thepast ranging from increasing social supports topsychotherapy approaches, however the vast majorityof these interventions have not been adequately trialedfor effectiveness. Research into interventions forloneliness should be based upon a clear model ofloneliness. Most interventions that have been pro-posed in the past seem to be focused more towardstreating the social type of loneliness or even moreconfusingly social isolation. Future research should,therefore, seek to differentiate clearly between lone-liness and isolation as well as differentiating betweenthe two types of loneliness along the bi-dimensionalmodel of loneliness as suggested by Weiss. Thebi-dimensional approach to loneliness is advantageousin that it differentiates between the potential causes of

loneliness and thus can help direct to more appropriateinterventions depending on the type of loneliness.Further questions that research should seek to addressinclude whether the different types of lonelinesspredict different outcomes and different responses tointerventions. For example, it is easy to envisagepeople with emotional loneliness not responding aswell to increased social supports compared to peoplewho have social loneliness. Novel interventionsshould also be trialed such as making use of moderntechnology such as the Internet and programs such asSkype#whereby the person can have visual as well asverbal contact. Research could also assess whetheruntreated co-morbid loneliness plays a key role intreatment-resistant depression and if so what inter-ventions are required to address the loneliness. From aclinical perspective based on our current knowledgewe would recommend that loneliness should be soughtfor in any medical/psychiatric assessment, particularlyif the patient is also depressed. Secondly, seekingpotential reasons for this loneliness during the courseof the clinical history can lead to a better overallunderstanding of the patients’ problems. This can alsohelp to clarify the type of loneliness which can focusthe doctor on possible intervention strategies.

CONCLUSIONS

Loneliness is common in older people and isassociated with adverse health consequences, bothfrom a mental and physical health point of view. It isan area that has been relatively neglected in themedical literature. The current evidence as outlined inthis review would support the initiation of appropriateintervention strategies that target loneliness, based ona clear understanding of the construct of loneliness asoutlined in this review.

Table 3. Relationship between loneliness and physiological measures

Authors Physical healthmeasure

Lonelinessmeasure

Findings associated with loneliness

Cacioppo JT, Hawkley LC, Crawford E,Ernst JM, Burleson MH, et al.

Blood Pressure UCLA Scale Increased systolic Blood pressure in theelderly.

Hawkley LC, Masi CM, Berry JD,Cacioppo JTCacioppo JT, Hawkley LC, Bernston GG,Ernst JM, et al.

Sleep UCLA Scale Reduced time in bed spent asleep. Increasedwake time after sleep onset.

Cacioppo JT, Hawkley LC, Crawford E,Ernst JM, Burleson MH et al.

Reduced sleep quality, increased daytimesleepiness

Steptoe A, Owen N,Kunz-Ebrecht SR, Brydon L

Neuroendocrine/immune system

UCLA Scale Increased fibrinogen and natural killer cellresponse to stress

Sorkin D, Rook KS, Lu JL Heart condition status UCLA Scale Loneliness associated with 3-fold increasein odds of having a heart condition.


DOI: 10.1002/gps


An increased understanding into the complexities ofloneliness, its causes, effects as well as possibleinterventions may lead to improved patient care aswell as better health outcomes, quality of life andfunctioning in older people.

CONFLICT OF INTEREST

None known.

REFERENCES

Adams KB, Sanders S, Auth EA. 2004. Loneliness and depression inindependent living retirement communities: risk and resiliencefactors. Aging Mental Health 8(6): 475–485.

Alpass FM, Neville S. 2003. Loneliness, health and depression inolder males. Aging Mental Health 7(3): 212–216.

Andersson L. 1982. Interdisciplinary study of loneliness with evalu-ation of social contacts as a means towards improving compe-tence in old/age. Acta Sociologica 25: 75–80.

Andersson L. 1990. Om ensamhetens innehall och aldres upplevel-ser. Tidskriften Aldrecentrum 1–2: 9–13.

Andersson L. 1993. Loneliness. In The Encyclopedia of AdultDevelopment, Kastenbaum R (ed.). The Oryx Press: Phoenix,AZ; 282–285.

Boomsma DI, Willemsem G, Dolan CV, et al. 2005. Genetic andenvironmental contributions to loneliness in adults: The Nether-lands Twin Register Study. Behav Genet 35(6): 745–752.

Bond J, Carstairs V. 1982. The Elderly in Clackmannan. ScottishHealth Services Studies 42, Scottish Home and Health Depart-ment: Edinburgh.

Bowlby J. 1969. Attachment and Loss, Vol. 1: Attachment. Hogarth:London.

Cacioppo JT, Hawkley LC, Bernston GG, et al. 2002a. Do lonelydays invade the nights? Potential social modulation of sleepefficiency. Psychol Sci 13(4): 384–387.

Cacioppo JT, Hawkley LC, Crawford E, et al. 2002b. Loneliness andhealth: potential mechanisms. Psychosom Med 64: 207–417.

Cacioppo JT, Hawkley LC, Ernst JM, et al. 2006a. Loneliness withina nomological net: an evolutionary perspective. J Res Personality40: 1054–1085.

Cacioppo JT, Hughes ME, Waite LJ, et al. 2006b. Loneliness as aspecific risk factor for depressive symptoms: Cross-sectional andlongitudinal analyses. Psychol Aging 21(1): 140–151.

Creecy R, Berg W, Wright R. 1985. Loneliness among the elderly: acausal approach. J Gerontol 40: 487–493.

de Jong Gierveld J. 1987. Developing and testing a model ofloneliness. J Personal Soc Psychol 53: 119–128.

DiTommaso E, Spinner B. 1993. The development and initialvalidation of the Social and Emotional Loneliness Scale forAdults (SELSA). Personality Individual Differ 14: 127–134.

DiTomasso, Spinner. 1997. Social and Emotional Loneliness: areexamination of Weiss’s typology of loneliness. Pers Indiv differ22: 417–422.

Green BH, Copeland JR, Dewey ME, et al. 1992. Risk factors fordepression in elderly people: a prospective study. Acta PsychiatrScand 86(3): 213–217.

Hansson R, Jones W, Carpenter B, Remonder I. 1986. Lonelinessand adjustment to old age. Int J Aging Hum Develop 24: 41–53.

Hawkley LC, Masi CM, Berry JD, Cacioppo JT. 2006. Loneliness isa unique predictor of age-related differences in systolic bloodpressure. Psychol Aging 21(1): 52–164.

Heikkinen RL, Kauppinen M. 2004. Depressive symptoms in latelife: a 10-year follow-up. Arch Geront Geriatr 38: 239–250.

Hunt A. 1978. The Elderly at Home. Her Majesty’s StationeryOffice: London.

O’Connell H, Chin AV, Cunningham C, Lawlor BA. 2003. TheDublin Healthy Ageing Study. Mercer’s Institute for Research inAgeing, St James’s Hospital, Dublin 8, unpublished data.

Peplau LA, Bikson I, Rook K, Goodshilds J. 1982. Being old andliving alone. In Loneliness: A Sourcebook of Current Theory,Research and Therapy, Peplau LA, Perlman D (eds). Wiley-Interscience: New York.

Peplau L, Perlman D. 1982. Perspectives on loneliness. In Lone-liness: A Sourcebook of Current Theory, Research and Therapy,Peplau LA, Perlman D (eds). Wiley-Interscience: New York.

Perlman D, Peplau LA. 1984. Loneliness research: a survey ofempirical findings. In Preventing the Harmful Consequences ofSevere and Persistent Loneliness, Peplau LA, Goldston SE (eds).National Institute of Mental Health (NIMH): WashingtonGovernment Printing Office: Washington, DC; 13–46.

Prince MJ, Harwood RH, Blizard RA, et al. 1997. Social supportdeficits, loneliness and life events as risk factors for depression inold age. The Gospel Oak Project VI. Psychologic Med 27:323–332.

Qereshi H, Walker A. 1989. The Caring Relationship. ElderlyPeople and Their Families. Macmillan: Basingstoke.

Russell D, Cutrona CE, Rose J, Yurko K. 1984. Social and emotionalloneliness: an examination of Weiss’s typology of loneliness.J Personality Soc Psychol 46: 1315–1321.

Russell D, Peplau LA, Cutrona CE. 1980. The Revised UCLALoneliness Scale: concurrent and discriminant validity evidence.J Personality Soc Psychol 39: 472–480.

Russell DW. 1996. UCLA Loneliness Scale (version 3): reliability,validity and factor structure. J Personality Assess 66: 20–40.

Savikko N, Routasalo P, Tilvis RS, et al. 2005. Predictors andsubjective causes of loneliness in an aged population. ArchGerontol Geriatr 41: 223–233.

Scalise JJ, Ginter EJ, Gerstein LH. 1984. A multidimensional lone-liness measure] the Loneliness Rating Scale (LRS). J PersonalityAssess 48: 525–530.

Schmidt N, Sermat V. 1983. Measuring loneliness in differentrelationships. J Personality Soc Psychol 44: 1038–1047.

Shaver PR, Brennan KA. 1991. Measures of depression and lone-liness. In Measures of Personality and Social PsychologicalAttitudes, Vol. 1, Robinson JP, Shaver PR, Wrightsman LS(eds). Academic Press: San Diego, CA; 195–200.

KEY POINTS

� Loneliness has been relatively neglected as aconstruct in old age psychiatry.

� Loneliness is strongly associated with depres-sion and may be an independent risk factor fordepession.

� Loneliness has been linked detrimentally with anumber of physical health measures as wellworse cognition.

� There is a need for evidence based interventions.


DOI: 10.1002/gps


Sheldon JH. 1948. The Social Medicine of Old Age. Oxford Uni-versity Press: Oxford.

Sorkin D, Rook KS, Lu JL. 2002. Loneliness, lack of emotional support,lack of companionship and the likelihood of having a heart conditionin an elderly sample. Ann Behavioural Med 24(4): 290–298.

Stek ML, VInkers DJ, Gussekloo J, et al. 2005. Is depression in oldage fatal only when people feel lonely? Am J Psychiatry 162: 1.

Steptoe A, Owen N, Kunz-Ebrecht SR, Brydon L. 2004. Lonelinessand neuroendocrine, cardiovascular,and inflammatory stressresponses in middle-aged men and women. Psychoneuroendo-crinol 29: 593–611.

Tilvis RS, Kahonen-Vare MH, Jolkkonen J, et al. 2004. Predictors ofcognitive decline and mortality of aged people over a 10-yearperiod. J Gerontol A Biol Sci Med Sci 59(3): 268–274.

Townsend P. 1957. The Family Life of Old People. Routledge andKegan Paul: London.

Townsend P. 1968. Isolation and loneliness. In Shanas E, TownsendP, Wedderton D, Friis H, Milhoj P and Stehouwer J. (eds), Old

people in Three Industrial Societies. Routledge and Kegan Paul:London; 258–287.

Tunstall J. 1966. Old and Alone. Routledge and Kegan Paul:London.

Victor CR, Scambler S, Bond J, Bowling A. 2000. Being alone inlater life: loneliness, isolation and living alone in later life. RevClin Gerontol 10: 407–417.

Victor CD, Scambler SJ, Bowling A, Bondt J. 2005. The prevalenceof and risk factors for loneliness in later life: a survey of olderpeople in Great Britain. Ageing Soc 25: 357–375.

Wenger GC, Davies R, Shahtahmesebi S, Scott A. 1996. Socialisolation and loneliness in old age: review and model refinement.Ageing Soc 16: 333–358.

Weiss RS. 1973. Loneliness: The Experience of Emotional andSocial Isolation. MIT Press: Cambridge, MA.

Wilson RS, Krueger KR, Arnold SE, et al. 2007. Lonelinessand risk of Alzheimer Disease. Arch Gen Psychiatry 64: 234–240.


DOI: 10.1002/gps


35395426

Documents

Transcript of 35395426