Vertical anthropometric measures and low back pain in adolescents

Vertical anthropometric measures and lowback pain in adolescents

SAMUEL STEELE, KAREN GRIMMER, MARIE WILLIAMS and TIFFANY GILLCentre for Allied Health Research, University of South Australia, Adelaide, Australia

ABSTRACT Background and Purpose: The present paper explores aspects of a largedataset of adolescent potentiates of low back pain. Standing, trunk and shoulder height,subischial height, weight and mean leg length are examined in relation to reports of lowback pain in the last two weeks, in strata of age and gender. Method: The present studyexamined 1254 adolescents, aged 11–18 years, drawn from 12 Adelaide public highschools during 1998. Results: At each year of age, girls reported significantly more lowback pain than boys. Girls and boys at each year of age provided different relationshipsbetween low back pain and anthropometric features, highlighting the unique nature ofadolescent growth. The most consistent findings were anthropometric differences between16-year-old boys, and 13- and 15-year-old girls, with and without low back pain, who maybe at critical stages of the peak growth curve. Conclusion: We suggest a high prevalenceof low back pain in adolescents, and that young sufferers with low back pain may bereflecting attributes of their stage of anthropometric growth.

Key words: adolescents, anthropometric, low back pain, physiotherapy

INTRODUCTION

Despite the regular media focus on thespinal health of school age children (usuallyrelated to load carrying), scientific literatureprovides conflicting information on theprevalence and precipitating factors of lowback pain in adolescents. The present paperreports on a subset of variables from a largestudy that assessed the effect of posteriorloads on adolescent posture and spinal pain.

Specif ically, it reports on differencesbetween vertical anthropometric parameters(standing height, trunk and shoulder heightin sitting, subischial height, weight and leglength) in adolescent boys and girls with andwithout recent low back pain. The predictivenature of anthropometric features, age andgender on low back pain were also tested.The anthropometric measures in the presentstudy have been reported previously, and

94 Physiotherapy Research International, 6(2) 94–105, 2001 © Whurr Publishers Ltd

PRI 6(2) crc 18/5/01 4:10 pm Page 94

there is a plausible biological argumentbetween them and spinal symptoms (Tanneret al., 1976; Pheasant, 1995). A causalmechanism underpinning the involvement ofanthropometrical features in low back painhas been suggested as incongruities in boneand muscle development during bonygrowth periods, resulting in diminishedmuscular support around vulnerable spinaljoints (Johnson, 1990; Bonjour et al., 1991).

Reports of prevalence of adolescent lowback pain range from 28% to 50%, with anaverage cumulative prevalence of approxi-mately 30% (Balague et al., 1988; Kujala etal., 1992; Olsen et al., 1992; Salminen et al.,1995; Burton et al., 1996). The effect ofgender on low back pain is unclear. Balagueet al. (1988) report that significantly moregirls than boys report low back pain; con-versely, Burton et al. (1996) report that moreboys than girls report low back pain. On theother hand, Olsen et al. (1992) found no dif-ference in reports of low back pain betweenboys and girls at any age. The lack of con-clusive findings could be attributed to a lackof standardized measurements, description,time frame and measurement of low backpain, as well as few concerted investigationsof other putative causes of adolescent lowback pain, such as female hormonal influ-ences, lifestyle, use of standard design furni-ture and posture.

Limitations of studies into adolescentlow back pain and anthropometry focus onselection and sampling biases which haveresulted in constrained age ranges, gendermisrepresentation and neglect of age-relateddifferences (Fairbank et al., 1984; Salminenet al., 1992; Ebrall, 1994). For instance,Ebrall (1994) uses a limited sample of maleadolescents aged 12–18 years (n = 125) andBurton et al. (1996) include both girls andboys in a limited sample (n = 147). Balagueet al. (1988) include both girls and boys in alarger sample size (n = 1715); however, no

anthropometric measures were taken. Notsurprisingly, there are considerable differ-ences in findings. For instance, Salminen etal. (1992) and Nissinen et al. (1994) pro-posed that boys with low back pain weretaller than their asymptomatic peers, andFairbank et al. (1984) and Nissinen et al.(1994) have related greater sitting height tolow back pain in adolescent males. Con-versely, Kujala et al. (1992) and Balague etal. (1993) report no correlation betweenheight and low back pain. Although evi-dence of leg length discrepancy is oftensought in a clinical setting when examiningyoung patients with low back pain (Corriganand Maitland, 1983) there is inconclusiveevidence of a link between low back painand leg length (Fairbank et al., 1984; Mil-grom et al., 1993).

Physiotherapists should be concerned byadolescent low back pain because of itspotential to become adult low back pain(which has been estimated to be as high as60–80% prevalence in a lifetime) (Dugglebyand Kumar, 1997). Risks associated withadult low back pain have been proposed assmoking, emotional state, job satisfaction,cardiovascular f itness, body type andweight, and socio-economic status (Biering-Sorensen, 1984; Deyo and Diehl, 1988;Klenerman et al., 1995). Consistent reportsof low back pain in both adolescent andadult years suggest that causal agencies ofadult low back pain may well be present inadolescence. If so, they need to be detectedearly in order to minimize low back painevents in later life. This may indicate theneed for school-based screening for lowback pain and its risk factors, in order forhealth education and promotion to com-mence appropriately early. However, on thebasis of the current literature, there is noclear evidence upon which physiotherapistscan base health promotion activities foranthropometric attributes of adolescent low

Anthropometry, low back pain and adolescents 95

PRI 6(2) crc 18/5/01 4:10 pm Page 95

96

back pain. The present paper attempts toclarify the picture of adolescent low backpain and vertical anthropometry.

METHOD

Data from a large cross-sectional observa-tional study were used (Grimmer et al.,1999) which investigated a range of featuresof students in 12 metropolitan state highschools in Adelaide. The relevant universityand government department ethics commit-tees provided ethical approval for the study.Reports of low back pain experienced in thepreceding two weeks were recorded inbinary form, and measurements of weightand vertical anthropometric features (height,trunk and shoulder height, leg length) weretaken. These features have been reported inthe literature as potentially related to spinalsymptoms (Fairbank et al., 1984; Olsen etal., 1992; Salminen et al., 1992; Ebrall,1994; Burton et al., 1996). Pubertal stageinformation was not collected. The standardmethod of rating pubertal stages requiressubjects either to disrobe or self-report thestage of development of secondary sexualcharacteristics such as breast/penissize/development and pubic hair growth pat-terns (Tanner et al., 1976). These data wereinappropriate for collection from adoles-cents in the school setting and thus pre-cluded collection of biological ageinformation in the present study.

Preparation for data collection

This study involved both staff and final-yearphysiotherapy students in the School ofPhysiotherapy, University of South Australia.Senior researchers were involved in design-ing the study, obtaining ethical approval,devising and testing the measures and mea-surement sheets, managing the project, as

well as analysing and reporting the results.The physiotherapy students were involved inone of four measurement stations (anthropo-metric, questionnaire, posture and schoolbagweight). Each student completed an intensivetraining period with the senior researchers,and was required to demonstrate high intra-and inter-tester reliability in all measures rel-evant to their station prior to commencingthe study (evidenced by Pearson’s r statistics>0.9 and p values from paired Student’s t-tests >0.05). Measurement precision waschecked throughout the study by the seniorresearchers to ensure ongoing rigour in datacollection and recording.

Subject selection

In February 1998, all 48 metropolitan statehigh schools in Adelaide were indepen-dently invited by the Department of Educa-tion, Training and Employment (SouthAustralia) to participate in the study. Thefirst 12 schools to respond were enrolled inthe study, this being the maximum numberable to be tested as a result of time andresource constraints. The participatingschools were located in all regions of metro-politan Adelaide thus offering a geographi-cally and socio-economically representativesample of students. Each school was askedto allocate one class of students from eachof its five high school year levels and torequest permission from parents of everychild in the class to participate in the study.There were at least four classes per yearlevel in each school. The total studentnumber in the 60 classes selected was 2075.One class only was tested on each day so asto disrupt school routine as little as possible.Schools chose the day on which a particularclass was tested. This allocation wasbelieved to randomly distribute variability instudents’ data.

Steele et al.

PRI 6(2) crc 18/5/01 4:10 pm Page 96

Data collection

Standing height

Subjects’ standing height was measured incm, to one decimal place, by use of a steeltape measure attached to a wall, at 90° to thefloor. Two pieces of tape on the floor stan-dardized subject positioning — one 5 cmfrom (and parallel to) the wall, and the otherperpendicular to the first piece of tape (andlevel with the tape measure on the wall). Tominimize error in height measurement, hatsand high hairstyles were removed where pos-sible. Subjects stood barefooted and in astandard position, determined by the tape onthe floor and verbal instructions to ‘Standwith your heels on the back of the line paral-lel to the wall and on either side of the per-pendicular line. Stand up straight with yourshoulders back, hands by your side, chintucked and eyes looking straight ahead. Keepstill while the measurement is taken’.

A T-square was positioned against thewall and on the top of subjects’ heads tomeasure standing height accurately. Theprevalence and extent of scoliosis was notmeasured. Scoliosis is reported to be morecommon in girls than boys, ranging from0.06% to 1.7% in the adolescent population,therefore this population was not reported on(Liu and Huang, 1996; Stirling et al., 1996;Soucacos et al., 1997). This small proportionwas believed to be unlikely to influence theadolescent prevalence of low back pain, norputative relationships with low back pain.

Weight

Weight was measured with subjects barefoot,standing quietly on electronic scales (Met-tler) which were accurate to 0.01 kg.

Sitting trunk and shoulder height

A stool was placed against the wall at thebase of the tape measure positioned next to

the one used for standing height measure-ment. Subjects sat with their backs againstthe wall and hips and knees flexed to 90°.Subjects rested their hands on their thighsand were instructed to ‘Sit up straight withyour shoulders back, hands by your knees,chin tucked and eyes looking straight ahead.Keep still while the measurement is taken’.This provided us with the same postural pre-sentation as for standing height. Measure-ments were taken from the base of the seat inboth instances.

Using the same T-square as for standingheight, trunk height measurement wasrecorded to the top of subjects’ heads, andshoulder height was recorded to the top ofsubjects’ right acromioclavicular joint. Mea-surements were made only to one side of thebody (the right). It is acknowledged that thisdid not allow the researchers to assess theeffect of shoulder girdle asymmetry on lowback pain, and later studies by this team haverecorded this measurement bilaterally. How-ever, there is little in the literature to suggesta relationship between low back pain anddiscrepancies in upper trunk anthropometry,other than that found in the scoliotic popula-tion. It is believed that this proportion wouldhave been minimal in the present studysample.

Leg length

Subjects lay supine on a gymnastic mat,with shoes and socks removed, with theirbody positioned symmetrically with legs inline with trunk, and knees extended. Ablock of wood 8.5 cm wide was placedbetween the ankles to ensure standard dis-tancing between feet, and subjects wereinstructed to keep their feet in that positionwith toes pointing towards the ceiling toprevent hip rotation or abduction. The dis-tance between the most anterior tip of the


PRI 6(2) crc 18/5/01 4:10 pm Page 97

98

anterior superior iliac crest and the mostprominent part of the ipsilateral medialmalleolus of the ankle on both legs wasmeasured, using the one flexible tape mea-sure. For the purpose of the present paper,the average leg length was investigated.

Subischial height

This value was calculated after measure-ments were taken as the difference betweentotal height and trunk height (Pheasant,1995).

Spinal symptoms

Subjects completed a questionnaire, whichsought among other things, reports of anyevent of low back pain, which had occurredin the previous two weeks. There is no onestandard way of measuring low back pain.Previous studies consider lifetime preva-lence of low back pain (Fairbank et al.,1984; Balague et al., 1988; Olsen et al.,1992; Salminen et al., 1995; Burton et al.,1996), a set time frame prior to testing (Bal-ague et al., 1988; Olsen et al., 1992; Nissi-nen et al., 1994; Salminen et al., 1995), orpoint prevalence (Fairbank et al., 1984;Ebrall, 1994; Burton et al., 1996). A two-week period before testing was chosenbecause good reliability of recall of painfrom recent illness has been established foradolescents (Andrasik et al., 1985). More-over, recent work in adolescent injury recallin a recent time frame (one week) placesconfidence in the accuracy of adolescentresponses (Grimmer et al., 2000). Noattempts were made in the present study todescribe the quality, severity or manage-ment of low back pain because of lack ofvalid and reliable indicators of these attrib-utes (Wilson et al., 1999). The question-naire used has been reported in fullelsewhere (Grimmer and Williams, 2000).

Analysis

Data were analysed in strata of chronologi-cal age (in years) and gender. Spinal devel-opment is considered to be closelycorrelated with both features (Pheasant,1995). This assumption was tested by deter-mining the size of correlations between ageand each anthropometric variable for boysand girls. The prevalence of low back painwas reported as the percentage of boys andgirls in each year of age, who reported lowback pain in the previous two weeks. Meandifferences in vertical anthropometrybetween boys and girls with and withoutlow back pain were determined by use ofindependent Student’s t-tests. Logisticregression models were constructed to testeffects on the prevalence of low back pain,of each anthropometric measure, age andgender. In each model, the anthropometricvariable and age were applied in their con-tinuous forms. No interaction terms weretested because of the lack of a priorihypotheses about differences in effectacross the adolescent population. An alphalevel of 5% was applied in all instances todetect significant effects.

RESULTS

Subjects

Data were obtained from 1269 subjects,1260 of whom provided complete data (650girls and 610 boys). Overall, the samplerepresented 60.4% of the eligible studentcomplement available in the participatingschools. Potential subjects were lostbecause parents or the child declined con-sent, consent forms were not returned ontime, class timetables changed on the day oftesting rendering students unavailable, orstudents were absent from school on the dayof testing. Most of the absentees related toclass excursions, and as far as could beascertained from school records, there were

Steele et al.

PRI 6(2) crc 18/5/01 4:10 pm Page 98

less than 20 students overall who were absentfor illness. There were no differencesbetween schools in response to the study,suggesting that socio-economic status orgeographical location had no effect on com-pliance. Subjects’ ages ranged from 11 yearsto 19 years, with the majority of subjectsaged 12–17 years. For analysis purposes the11-year-olds (n = 5) were combined with the12-year-olds, and the one 19-year-old and theseven 18-year-olds were combined with the17-year-olds. The study sample comprised55 11–12-year-olds (30 girls, 25 boys); 27613-year-olds (139 girls, 137 boys); 281 14-year-olds (131 girls, 150 boys); 259 15-year-olds (132 girls, 127 boys); 252 16-year-olds(143 girls, 109 boys); and 137 17–18-year-olds (75 girls, 62 boys).

Correlation between age andanthropometric features

There were stronger correlations betweenchronological age and anthropometric fea-tures for boys than girls, suggesting thatanthropometric growth occurs more lin-early with chronological age for boys thangirls. The size of these correlations isreported in Table 1. Scatterplots of anthro-pometric measures and age conf irmedgender differences for all anthropometricvariables, where for every anthropometricvariable, there was a consistent linear pat-tern for boys, but for girls there were twodistinct but less strong clusters, whichsuggested a different pattern of female

growth — one group spurting around 14years and the other spurting later, at 16–18years.

Prevalence of low back pain

A higher percentage of girls than boysreported recent low back pain in each agegroup, with this difference being significantin all but the 12- and 13-year-olds. Reportsof recent low back pain were provided by:

• 23.3% girls (n = 7) and 12% boys(n = 3) in the 12-year-old age group.


• 25% girls (n = 32) and 14.7% boys(n = 22) in the 14-year-old age group.


• 37.8% girls (n = 54) and 26.7% boys(n = 29) in the 16-year-old age group.


Standing height

There was no consistency in standing heightdifferences for students with and withoutlow back pain, within or across age strata orgender. The only significant difference wasobserved for boys aged 16 years, where theback pain sufferers were taller by 3 cm onaverage. The mean standing height of sub-jects with, and without, low back pain isreported in Table 2.


TABLE 1: Correlation between age and anthropometric variables for boys and girls (age range 12–17 years)

Height Weight Trunk height Leg length Shoulder height

Girls 0.04 0.06 0.09 0.006 0.08

Boys 0.38 0.19 0.36 0.25 0.27

PRI 6(2) crc 18/5/01 4:10 pm Page 99

100

Trunk and shoulder height

As for standing height, there was no unifor-mity in trunk and shoulder height differ-ences for girls or boys with and without lowback pain. Signif icant differences wereobserved at p <0.05 for the 15-year-old girlsfor trunk and shoulder height, and for 16-year-old boys for trunk height, where in allinstances the low back pain sufferers hadlonger trunks than asymptomatic subjects.The mean trunk and shoulder height of sub-jects with and without, low back pain is alsoreported in Table 2.

Weight

There were no patterns between weight andlow back pain. Only girls aged 14 and 16

years, with and without low back pain,demonstrated signif icant differences inweight; however, the 14-year-olds with lowback pain were signif icantly lighter(p <0.05), whereas the 16-year-olds withlow back pain trended towards being signif-icantly heavier (p = 0.07). Mean weight forgirls and boys with and without low backpain is reported in Table 3.

Subischial height

A signif icantly smaller mean subischialheight measurement was found for 12-year-old girls with low back pain, compared withtheir asymptomatic peers. Similar to allother height measurements, no consisten-cies with reports of low back pain were

Steele et al.

TABLE 2: Mean standing height, trunk height and shoulder height (cm) (± standard errors (SE)) for girlsand boys, with and without low back pain

Girls Boys

With LBP Without LBP With LBP Without LBP

Standing height12 years 155.8 (3.2) 160.7 (1.4) 160.2 (5.5) 154.1 (2.2)13 years 159.9 (1.1) 159.3 (0.6) 163.9 (1.7) 161.3 (0.8)14 years 161.9 (1.1) 162.7 (0.6) 166.8 (1.6) 167.5 (0.6)15 years 164.2 (0.9) 162.9 (0.7) 174.3 (1.2) 172.6 (0.7)16 years 163.8 (1.2) 162.9 (0.7) 177.2 (1.1)** 174.1 (0.8)**17 years 163.7 (1.2) 164.8 (1.0) 178.2 (2.0) 178.4 (0.9)

Trunk height12 years 81.4 (1.6) 82.8 (0.8) 84.9 (2.7) 79.6 (1.1)13 years 83.2 (0.6) 82.8 (0.3) 84.2 (1.0) 82.7 (0.4)14 years 84.9 (0.8) 85.5 (0.3) 83.7 (1.2) 85.6 (0.4)15 years 86.7 (0.4)** 85.3 (0.4)** 89.7 (0.6) 88.7 (0.5)16 years 86.4 (0.6) 86.2 (0.3) 91.6 (0.7)** 89.9 (0.4)**17 years 86.3 (0.7) 86.2 (0.6) 91.5 (1.0) 92.2 (0.4)

Shoulder height12 years 53.8 (1.1) 54.4 (0.6) 54.8 (2.4) 51.1 (0.9)13 years 54.3 (0.5) 54.6 (0.4) 54.9 (0.9) 54.1 (0.4)14 years 55.5 (0.7) 56.4 (0.3) 55.1 (0.7) 56.0 (0.4)15 years 57.8 (0.6)** 55.9 (0.3)** 58.9 (0.5) 68.7 (0.6)16 years 57.6 (0.6) 56.7 (0.3) 60.5 (0.8) 59.3 (0.4)17 years 57.1 (0.5) 57.9 (0.5) 60.6 (0.7) 61.4 (0.5)

LBP = low back pain.Differences at p < 0.05 marked by **.

PRI 6(2) crc 18/5/01 4:10 pm Page 100

found across, or within age groups. Meansubischial height measurements arereported in Table 4.

Mean leg length

On examination of mean leg length and lowback pain, the 16-year-old boys were again

highlighted, with the low back pain suffer-ers displaying signif icantly longer leglength than their asymptomatic colleagues.No other consistencies were found across orwithin strata. Mean leg length is reportedfor subjects with and without low back painin Table 4.


TABLE 3: Mean weight (kg) (± standard errors (SE)) for girls and boys, with and without low back pain

Girls Boys


12 years 51.5 (4.5) 52.2 (2.0) 57.5 (9.7) 48.9 (2.5)13 years 54.7 (2.1) 53.4 (0.8) 55.7 (2.4) 54.3 (1.2)14 years 54.9 (2.1)** 60.3 (1.4)** 57.7 (2.4) 58.6 (1.0)15 years 61.6 (1.8) 60.4 (1.2) 66.6 (2.3) 64.5 (1.3)16 years 62.5 (1.7)** 58.7 (0.9) (p = 0.07) 68.1 (1.8) 65.0 (1.1)17 years 60.5 (1.9) 63.8 (2.1) 70.7 (4.4) 71.9 (1.8)


TABLE 4: Mean subischial height, and leg lenght (cm) (± standard errors (SE)) for girls and boys, with andwithout low back pain

Girls Boys


Subischial height12 years 74.4 (1.7)** 77.9 (0.9) (p = 0.07) 75.3 (2.8) 74.4 (1.4)13 years 75.3 (0.7) 76.4 (0.4) 79.8 (0.9) 78.6 (0.4)14 years 77.5 (0.7) 77.4 (0.5) 84.5 (1.1) 83.9 (0.5)15 years 77.1 (0.8) 76.7 (0.5) 85.6 (0.7) 84.2 (0.5)16 years 77.4 (0.8) 78.5 (0.7) 86.7 (1.3) 86.2 (0.8)17 years 77.4 (1.7) 78.5 (0.7) 86.7 (1.3) 86.2 (0.8)

Leg length12 years 83.8 (2.1) 86.9 (0.9) 84.2 (4.1) 81.8 (1.2)13 years 86.3 (0.8) 85.7 (0.4) 87.3 (1.0) 87.0 (0.5)14 years 87.2 (0.7) 87.3 (0.5) 89.7 (1.1) 90.1 (0.4)15 years 87.5 (0.8) 87.3 (0.5) 92.5 (0.9) 92.2 (0.5)16 years 86.5 (0.8) 86.4 (0.5) 94.6 (0.7)** 92.6 (0.5)**17 years 87.6 (0.8) 87.8 (0.8) 94.6 (1.3) 94.8 (0.8)


PRI 6(2) crc 18/5/01 4:10 pm Page 101

102

Logistic regression modelling

The lack of overall effect of any anthropo-metric measure on adolescent low back painwas confirmed by the output of logisticregression models, where the odds of lowback pain changing incrementally with anyof our measures used was approximately1.0. In every model, however, there was asignificant confounding effect of increasingage, and of being female. This confirms theneed to examine anthropometric effects onadolescent low back pain per year of age,and per gender.

DISCUSSION

Response rate

The present study sample is believed to bethe largest to date in Australia in studies thatexamined the relationship between adoles-cent low back pain and anthropometric mea-sures. Although the present study samplewas generated from a small number of vol-unteer metropolitan high schools in Ade-laide, it is believed that it provides defensiblebaseline data that may well reflect any groupof adolescents. Subjects were drafted into thestudy by independent allocation by teachersof one class per year level in each school. Webelieved that this randomly assigned variabil-ity in responses to recent events of low backpain, and also variability in anthropometricmeasures. The same reasons for non-compli-ance were found in each school setting, andthere was no reason to suspect that responseswould have differed in students who did notparticipate in the present study. This is par-ticularly relevant for reports of low backpain, as very few potential subjects wereabsent from school on the day of testing dueto illness.

Prevalence of low back pain

Significantly more girls reported recent lowback pain than boys, supporting findings by

Balague et al. (1988), but disputing those ofBurton et al. (1996) (who suggest a higherprevalence of young males with low backpain than young females) and Olsen et al.(1992) (who suggest that there is no differ-ence). A retrospective two-week view wastaken in order to capture the prevalence ofrecent low back pain. This, and the dichoto-mous nature of the low back pain preva-lence question, may well be criticized asproviding an insensitive disease measure-ment. Moreover, the potential for recallerrors may have attenuated, or inflated, theresults. However, we had confidence in ado-lescents’ ability to recall at least one recentevent of low back pain (Andrasik et al.,1985; Grimmer et al., 2000), and we wereloath to undertake more extensive data col-lection where there were few more validand reliable measures of the nature and fre-quency of low back pain for the adolescentpopulation.

Relationship between anthropometry andlow back pain

The present study highlights not onlygender and age differences between anthro-pometric features and low back pain butalso the lack of consistency of f indingsacross, and within, age strata. It may wellbe argued that this lack of consistencydiminishes the value of our findings, andthat the significant findings may have hap-pened by chance. However, we believe thatour study had sufficient power and minimalselection or sampling bias, and used stan-dard measurement protocols to minimizemeasurement error. Thus our findings sup-port the need to consider each adolescentchronological year of age separately withrespect to associates of low back pain. Thegrowth rates proposed by Tanner et al.(1976) indicate the variability with whichadolescent skeletal development occurs.

Steele et al.

PRI 6(2) crc 18/5/01 4:10 pm Page 102

Although biological age data would haveallowed pubertal stages to be estimatedmore accurately than is possible withchronological age, current ethical andschool-based environmental restrictionspotentially constrain the type of data thatcan be collected in large-scale studies onadolescents in Western countries. Thus, it isimportant to use the data that are readilyavailable in order to understand anthropo-metric differences in young people report-ing back pain, and to flag whether changesin anthropometry are likely to reflect incipi-ent long-term spinal problems.

When modelling the anthropometricdata with reports of recent low back pain,and controlling for the effect of age, acohort concept was imposed on the cross-sectional data. That is, a global modellingapproach was used on data that had beencollected in a short time period from youngpeople ranging in age from 11 years to 18years, to predict associations that may welldiffer in chronological age strata (proxyingdevelopmental (pubertal) stages). Thisapproach provides an estimate only, forassociations that it is acknowledged wouldbe obtained more appropriately by use of alongitudinal study approach, where data iscollected from the same group of youngpeople repeatedly over time. Thus, the useof logistic regression models in the presentdataset at best provides an overview of therelationship between vertical anthropome-try, low back pain and age, for boys andgirls. The potential for oversimplificationof the relationship between these variablesis particularly evident in the light of thedifferent correlation patterns for boys andgirls between age and the anthropometricmeasures. Notwithstanding, the stronglysignif icant f indings from the logisticregression modelling of the effect of ageand gender highlight the importance ofseparate consideration of each year of age

of boys and girls when attempting to under-stand anthropometric effects on adolescentlow back pain.

Although there is debate in the literatureregarding the nature of the relationshipbetween anthropometric measures and lowback pain, it is plausible that biomechanicalstresses on structures of the spine (resultingin pain) may be related to forces and leversunderpinning vertical spinal and leg dimen-sions. The adolescent growth and develop-ment literature suggests that usually, thespine completes its growth prior to the longbones in the legs (Tanner et al., 1976;Pheasant, 1995). Data captured during peri-ods of intense adolescent growth may wellhighlight opportunities for abnormalmechanical stresses on spinal structures.Adaptive muscular responses may ensue,and such imbalances, if unaddressed, maywell continue into adulthood as latentcauses of adult spinal problems. Withoutlongitudinal data, it would not be possibleto categorically attribute adolescent lowback pain to differences in anthropometricfindings, nor to retrospectively attributeadult low back pain to adolescent anthropo-metric anomalies. However, as the presentstudy highlights, the potential for a relation-ship between low back pain and anthropo-metric features is strong at different ages ofadolescence. Physiotherapists are well-placed to do more than treat the presentingsymptoms of young people with low backpain. Their knowledge of biomechanics andmuscle performance enables them to take alonger term view with respect to proactivelyidentifying and addressing musculoskeletalincongruities which may predispose thegrowing spine to pathobiomechanics(Sahrmann, 1988; Bonjour et al., 1991; Jullet al., 1993; Hodges and Richardson, 1997).Physiotherapists are thus in a position toidentify these incongruities to patients, par-ents and medical practitioners, and to pre-


PRI 6(2) crc 18/5/01 4:10 pm 3

104

scribe exercise programmes to addressopportunities for muscle imbalance thatmay be associated with incongruities ofbony development.

CONCLUSION

The present study examined a large sampleof adolescents aged 11–18 years, and foundthat recent low back pain was commonlyreported, and by significantly more girlsthan boys. There were differences in therelationship between vertical anthropomet-ric measures and low back pain, for girlsand boys in different chronological agestrata. Differences in gender age strata sug-gest changing risk features for low backpain as adolescents mature. Longitudinalstudies should be undertaken to confirmand clarify relationships. Physiotherapiststreating young people with low back painneed to be aware of the potential forgrowth-related biomechanical aspects ofpresenting pain, and to be proactive in con-sidering strategies to minimize the potentialfor adolescent low back pain to become afeature of adulthood.

ACKNOWLEDGEMENTS

The authors wish to acknowledge the 1998 fourth-year cohort of physiotherapy students of the Univer-sity of South Australia, the Centre for Allied HealthResearch, and the staff and students of all of the highschools involved in the study.

REFERENCES

Andrasik F, Burke EJ, Attanasio V, Rosenblum EL.Child, parent and physician reports of child’sheadache pain: relationships prior to and follow-ing treatment. Headache 1985; 25: 421–425.

Balague F, Dutoit G, Waldburger M. Low back painin schoolchildren. Scandinavian Journal of Reha-bilitation Medicine 1988; 20: 175–179.

Balague F, Damidot P, Nordin M, ParnianpourM, Waldburger M. Cross-sectional study of theisokinetic muscle trunk strength among schoolchildren. Spine 1993; 18: 1199–1205.

Bonjour JP, Theintz G, Buchs B, Slosman D, RizzoliR. Critical years and stages of puberty for spinaland femoral bone mass accumulation during ado-lescence. Journal of Clinical Endocrinology andMetabolism 1991; 73: 555–563.

Biering-Sorensen F. Physical measurements as riskindicators for low back trouble over a one-yearperiod. Spine 1984; 9: 106–119.

Burton AK, Clarke RD, McClune TD, Tillotson KM.The natural history of low back pain in adoles-cents. Spine 1996; 21: 2323–2328.

Corrigan B, Maitland G. Practical Orthopaedic Medi-cine. London: Butterworths, 1986.

Deyo RA, Diehl AK. Psychosocial predictors of dis-ability in patients with low back pain. Journal ofRheumatology 1988: 15: 1557–1564.

Duggleby T, Kumar S. Epidemiology of juvenile lowback pain: a review. Disability and Rehabilitation1997: 19: 505–512.

Ebrall PS. Some anthropometric dimensions of maleadolescents with idiopathic low back pain. Jour-nal of Manipulative and Physiological Therapeu-tics 1994; 17: 296–301.

Fairbank JCT, Pynsent PB, Van Poortvliet JA,Phillips H. Influence of anthropometric factorsand joint laxity in the incidence of adolescentback pain. Spine 1984; 9: 461–464.

Grimmer KA, Williams MT, Gill TK. High SchoolStudents and Backpacks — A Cross-sectionalstudy. University of South Australia, 1999. ISBN0 86803 627 7.

Grimmer K, Jones D, Williams J. Prevalence of ado-lescent injury from recreational exercise: an Aus-tralian perspective. Journal of Adolescent Health2000.

Hodges P, Richardson CA. Contraction of the abdom-inal muscles associated with movement of thelower limb. Physical Therapy 1997; 77: 132–144.

Johnson R. Adolescent growth and development. In:Vaughan VC, Litt IF (eds). Child and AdolescentDevelopment. Clinical Implications. Philadel-phia: WB Saunders, 1990.

Jull G, Richardson C, Toppenberg R, Comerford M,Bui B. Towards a measurement of active musclecontrol for lumbar stabilisation. Australian Jour-nal of Physiotherapy 1993; 39: 187–193.

Klenerman L, Slade PD, Stanley IM, Pennie B, ReillyJP, Atchison LE et al. The prediction of chronic-ity in patients with an acute attack of low backpain in a general practice setting. Spine 1995; 20:478–484.

Kujala UM, Salminen JJ, Taimela S, Oksanen A,Jaakkola L. Subject characteristics and low backpain in young athletes and nonathletes. Medicine

Steele et al.

PRI 6(2) crc 18/5/01 4:10 pm Page 104

and Science in Sports and Exercise 1992; 24:627–632.

Liu SL, Huang DS. Scoliosis in China. A generalreview. Clinical Orthopaedics and RelatedResearch 1996; 323: 113–118.

Milgrom C, Finestone A, Lev B, Wiener M, FlomanY. Overexertional lumbar and thoracic back painamong recruits: a prospective study of risk factorsand treatment regimens. Journal of Spinal Disor-ders 1993; 6: 187–193.

Nissinen M, Heliovaara M, Seitsamo J, Alaranta H,Poussa M. Anthropometric measurements and theincidence of low back pain in a cohort of pubertalchildren. Spine 1994; 19: 1367–1370.

Olsen TL, Anderson RL, Dearwater SR, Kriska AM,Cauley JA, Aaron DJ et al. The epidemiology oflow back pain in an adolescent population. Amer-ican Journal of Public Health 1992; 82: 606–608.

Pheasant ST. Anthropometry and the design of work-spaces. In: Wilson JR, Corlett EN (eds). Evalua-tion of Human Work. London: Taylor & Francis,1995.

Sahrmann SA. Adult posturing. In: Kraus SL. TMJDisorders: Management of the CraniomandibularComplex. New York: Churchill Livingstone,1988.

Salminen JJ, Maki P, Oksanen A, Pentti J. Spinalmobility and trunk muscle strength in 15-year-oldschoolchildren with and without low-back pain.Spine 1992; 17: 405–411.

Salminen JJ, Erkintalo M, Laine M, Pentti J. Lowback pain in the young. A prospective three-yearfollow-up study of subjects with and without lowback pain. Spine 1995; 20: 2101–2108.

Soucacos PN, Soucacos PK, Zacharis KC, Beris AE,Xenakis TA. School screening for scoliosis. Aprospective epidemiological study in northwest-ern and central Greece. Journal of Bone and JointSurgery (American volume) 1997; 79:1498–1503.

Stirling AJ, Howel D, Millner PA, Sadiq S, SharplesD, Dickson RA. Late-onset idiopathic scoliosis inchildren six to fourteen years old. A cross-sec-tional prevalence study. Journal of Bone and JointSurgery (American volume) 1996; 78:1330–1336.

Tanner JM, Whitehouse RH, Marubini E, Resele LF.Adolescent growth in boys and girls. Annals ofHuman Biology 1976; 3: 109–126.

Wilson L, Hall H, McIntosh G, Melles T. Intertesterreliability of a low back pain classificationsystem. Spine 1999; 24: 248–254.

Address for correspondence: Mr Samuel Steele, c/oAssociate Professor Karen Grimmer, Centre forAllied Health Research, University of South Aus-tralia, City East Campus, North Terrace, Adelaide,SA 5000, Australia (E-mail: [email protected] or [email protected]).

(Submitted August 1999; accepted October 2000)


PRI 6(2) crc 18/5/01 4:10 pm Page 105

Vertical anthropometric measures and low back pain in adolescents

Documents

Transcript of Vertical anthropometric measures and low back pain in adolescents