Article Text

This article has a correction. Please see:

Download PDFPDF

Original article
Physical workload and risk of low back pain in adolescence
  1. Paula Mikkonen1,
  2. Eira Viikari-Juntura2,
  3. Jouko Remes2,
  4. Tuomo Pienimäki3,
  5. Svetlana Solovieva2,
  6. Simo Taimela4,
  7. Paavo Zitting5,
  8. Markku Koiranen5,
  9. Päivi Leino-Arjas2,
  10. Jaro Karppinen1,2
  1. 1Institute of Clinical Sciences, Department of Physical and Rehabilitation Medicine, University of Oulu, Oulu, Finland
  2. 2Finnish Institute of Occupational Health, Oulu and Helsinki, Finland
  3. 3Social Insurance Institution, Oulu, Finland
  4. 4Department of Public Health, University of Helsinki, Helsinki, Finland
  5. 5Institute of Health Sciences, University of Oulu, Oulu, Finland
  1. Correspondence to Dr Paula Mikkonen, Department of Physical and Rehabilitation Medicine, University of Oulu, Oulu University Hospital, Box 21, Oulu 90029, Finland; paula.mikkonen{at}ppshp.fi

Abstract

Objectives To evaluate the role of physical workload in low back pain (LBP) among adolescents.

Methods Working history and physical workload factors at 18 years were assessed for 1984 members of the Northern Finland Birth Cohort 1986. The associations between work characteristics and LBP were analysed by multinomial logistic regression. Those with and without LBP at 18 years of age were compared in two subsamples. The incidence of LBP was studied among the 986 subjects without LBP at 16 years of age. Persistence of LBP was studied among the 728 subjects with LBP at 16 years of age. Latent class analysis (LCA) was used to form natural clusters of workload factors and their associations with LBP were investigated using log-binomial regression.

Results 753 (75%) subjects without LBP at 16 years of age had been working during the 2-year follow-up period. The average duration of work was 6.2 months. In adolescent girls, working regularly or irregularly and duration of work exposure were associated with incident LBP. Of specific physical workload factors, only awkward trunk postures were associated with incident LBP in both genders (RR 1.2 in girls and 1.7 in boys). The work exposure patterns in adolescent girls and boys were different. In the LCA, subjects in a cluster with high exposure to awkward trunk postures or an overall physically demanding job had a higher likelihood of incident LBP in both genders (RR 1.3–1.9). None of the specific workload factors or clusters was associated with persistent LBP.

Conclusions Physical workload factors constitute a risk for LBP even in adolescents.

  • Low back pain
  • workload
  • adolescent
  • workload
  • epidemiology
  • back disorders
  • musculoskeletal
  • back disorders
  • statistics
  • public health
  • polymorphisms
  • mathematical models
  • genetic susceptibility
  • osteoarthritis
  • physiology
  • sickness absence
  • rehabilitation
  • organisation of work
  • longitudinal studies
  • genetic susceptibility
  • fitness for work
  • smoking
  • physical work
  • stress
  • disability

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

What is known about this subject

  • Physical work environmental factors have been shown to associate with low back pain (LBP) in adults; however information of a possible association in adolescents is missing.

What this study adds

  • In this cohort study the duration and type of work exposure were associated with LBP in adolescents during a time course of 2 years.

  • High exposure to awkward trunk postures or an overall physically demanding job increased the risk of incident LBP.

Policy implications

  • Physical work seems to increase the risk of LBP even in adolescents.

Introduction

Musculoskeletal symptoms have become more frequent among adolescents in recent years.1 As low back pain (LBP) in adolescence predicts continuing symptoms in adulthood, such an increase may lead to a higher burden of low back disorders on future society.2 The prevalence of LBP increases considerably between 15 and 20 years of age.3 LBP in adolescence (10–19 years by the World Health Organization definition) has been linked to female gender,3 4 rapid growth spurt,5 early degeneration of intervertebral discs,6 overweight,7 smoking,8 high level of leisure time physical activity9 and psychosocial risk factors.10–12

An association between physical work in general and back pain in adolescents has been proposed,13 but only one longitudinal study has been conducted to date.14 Furthermore, we are not aware of studies on specific physical workload risk factors for LBP in adolescents. In adults, both the physical15 and psychosocial work environment16 have been shown to be associated with LBP. There is strong evidence that whole body vibration, manual materials handling and twisting and bending of the trunk increase the risk of LBP.17

According to a high school survey in the USA, 60% of the students were working sometimes during the study period and 2% experienced a work-related back injury each year, mostly sprains and strains.18 Changes in the labour market and the need for financing studies have increased the proportion of those with fixed-term and part-time jobs in adolescents and students at the beginning of the 21st century in Finland.19 20 The aim of this study was to explore the role of physical workload factors in the development of LBP in a cohort of adolescents aged 16 years at baseline and followed up for 2 years. Specifically, we wanted to evaluate whether the duration of work exposure or a specific type of physical workload factor, individually or in combination, has an effect on LBP. Analyses were made separately for those without and those with LBP at baseline—that is, both the incidence and the persistence of LBP were studied.

Materials and methods

Study population

The study population was drawn from the Northern Finland Birth Cohort 1986 (NFBC 1986), which includes data from birth to 18 years on children born in Northern Finland between 1 July 1985 and 30 June 1986 (n=9479, expected date of birth; n=9432 liveborn children). Information on LBP, health behaviour and the family's socioeconomic situation was acquired at the age of 16 years with a questionnaire (n=9215, addresses known). This survey is considered as the baseline in the analyses of this paper. The response rate to the questionnaire was 78% (n=7182) while 74% (n=6795) took part in a health examination. The current study population (Oulu Back Study, OBS) represents a subcohort of the NFBC 1986 living within 100 km from the city of Oulu (n=2951). Follow-up survey data on the occurrence of LBP and work history of the adolescents were collected between 2003 and 2004 when the cohort members were, on average, 18 years old. The response rate was 67% (1987/2951). Three subjects who were employed at 16 years and had unspecified previous work exposure were excluded, so the final study population consisted of 1984 adolescents.

Assessment of LBP

LBP was assessed identically at baseline and follow-up with the question: “Have you had any pain or ache in your low back area during the past 6 months?” The response options were: “No”, “Yes but I have not consulted a physician, physiotherapist, nurse, or other health professional because of my LBP” (‘Reporting LBP’) and “Yes, I have consulted a physician, physiotherapist, nurse or other health professional because of my LBP” (‘Consultation for LBP’). The question about back pain was supported by a manikin in which the low back area was shown as a shaded area between the lower ribs and glutaeal folds. We dichotomised the outcome to ‘No pain’ versus ‘Any LBP’ (Reporting LBP and Consultation for LBP combined) when the effects of physical work exposures on LBP were evaluated.

The subjects (n=986) who had no LBP at baseline formed the study population for the incidence of LBP. Among them, those with any LBP at the 2-year follow-up were compared with those who were still without symptoms. To study the persistence of LBP, those who had any LBP at baseline (n=728) formed the study group. Those who reported symptoms during the previous 6 months at the 2-year follow-up were compared with those who remained symptom-free.

Assessment of physical workload factors

Workload was assessed retrospectively at the age of 18 years for the previous 2-year period. For the evaluation of their main activity, the respondents were enquired to answer (No or Yes) to the following items: (1) I study in a high school; (2) I study in a vocational education school; (3) I study elsewhere; (4) I am in a regular full-time job; (5) I am working but not in a regular full-time job; (6) I am unemployed; (7) I am on maternity or nursing leave; and (8) I am not studying or at work for some other reason. The respondents were requested to specify their responses for items 3, 5 and 8 with open questions. Occupational title was also requested.

All questions on workload were posed separately for the current job, previous jobs lasting at least 12 months and previous jobs lasting less than 12 months. For each job, information was requested on the duration (months) and weekly working hours and, for the current job, the number of workdays per week. For each job the respondents were asked to estimate its physical strenuousness according to the following categories: (1) sedentary work with limited walking (eg, office work at a desk, industrial sewer, watchmaker); (2) fairly light work with considerable walking but no lifting or carrying heavy objects (eg, office work with walking, shop assistant, light industrial work, supervisor); (3) fairly strenuous work with walking and lifting heavy objects or climbing stairs or uphill (eg, carpenter, cattle tender, engineering work and heavier industrial work); and (4) very strenuous work with lifting or carrying heavy objects such as shovelling, digging or hammering (eg, forest work, heavy farm work, heavy construction and industrial work).

In addition, the respondents were asked to report the presence (No or Yes) of specific physical exposures for the current job and the longest previous job held for at least 12 months and the longest job of jobs held for <12 months. The specific average exposures were: (1) kneeling or squatting for ≥1 h/day; (2) driving a car, tractor or other motor vehicle for ≥4 h/day for >3 months a year; (3) manually lifting, carrying or pushing objects heavier than 5 kg (medium weight objects) for >2 times/min for ≥2 h/day; (4) manually lifting, carrying or pushing objects heavier than 20 kg (heavy weight objects) for ≥10 times/day; (5) work with hands above the shoulder level for ≥1 h/day; (6) work in an awkward trunk posture—that is, working standing or on the knees in a position leaning forward without support (eg, car repair, dentist or dental nurse) for ≥1 h/day; and (7) standing or walking for ≥5 h/day. The selected exposures may theoretically cause increased biomechanical stress on the lumbar spine and are frequently considered as potential risk factors for musculoskeletal diseases related to workload.15

Assessment of covariates

Leisure time physical activity, smoking, body mass index (BMI) and the family's socioeconomic status at 16 years of age were considered as possible confounders. Socioeconomic status was considered as it can be regarded as a proxy for possible early influences on health behaviours, mental distress and engagement in physical work.

Leisure time physical activity was assessed with the question: “How often do you participate in brisk physical activity outside school hours for at least 20 min at a time?” Brisk physical activity was defined as physical activity causing at least some sweating and shortness of breath. The activity frequency was divided into four categories: once a month or less often; more than once a month but less than twice a week; 2 or 3 times a week; and >3 times a week. Smoking status was determined at 16 years of age based on responses to questions: “Have you ever smoked in your life?” and “Are you smoking now?” The assessment included the duration and frequency of smoking and the number of cigarettes smoked. Smokers were defined as those who smoked regularly (5–7 days/week) and non-smokers as those who did not smoke at all or who smoked occasionally. Measured weight and height at 16 years with the accuracy of 0.1 kg and 1 cm (measurement in health examination or self-reported for those who did not participate) were used to calculate BMI (kg/m2). Socioeconomic status was assessed from the questionnaire sent to the parents. It was based on a 5-item work and education questionnaire: (1) higher clerical employees; (2) self-employed; (3) lower clerical employees; (4) workers; and (5) students, pensioned citizens, unemployed or unknown. The father's response was prioritised in the assessment.

Statistical methods

The associations between workload variables and incidence and persistence of LBP were analysed with log-binomial regression analysis stratifying by gender. Risk ratios (RR) and their 95% CIs were adjusted for smoking (non-smokers as reference), BMI (middle tertile as reference), level of leisure time physical activity (2–3 times a week as reference) and the family's socioeconomic status (upper level administrative as reference) at the age of 16 years. The inclusion of BMI or socioeconomic status did not affect the results and they were removed from the final analysis. The duration of work exposure (months) had a skewed distribution and a logarithmic transformation was performed per unit of SD.

Latent class analysis (LCA) was used to obtain natural clusters—that is, groups in which the individuals had similar workload profiles.21 22 All workload exposures during the evaluated 2-year period were included, whether they had occurred simultaneously or at different time points. Using the LCA method with conditional independence assumption and Bayes' theorem, each individual's a posteriori probability in each class was calculated and he or she was then assigned to the latent class cluster with the highest a posteriori probability. The number of clusters was determined with two statistical diagnostics: Bayesian Information Criterion (BIC) and the Vuong-Lo-Mendell-Rubin likelihood ratio test (LRT).23 24 According to recent simulation studies, the model with the lowest value of BIC is the one to be preferred.24 The low p value of LRT indicates that the estimated model with ‘n’ clusters fits the data better than the model with ‘n–1’ clusters. The number of clusters was first determined based on the LRT (maximum clusters) and then checked for compatibility with the BIC measure. The statistical program M-Plus Version 523 was used for the LCA. The association between incident and persistent LBP and LCA clusters was investigated using log-binomial regression analysis. Adjusted risk ratios (RR) and their 95% CIs were calculated. The analyses were adjusted for smoking and leisure time physical activity and performed separately for girls and boys using SAS software (V.9.1, SAS Institute Inc).

Results

Characterisation of responders

At baseline the participants were living in two-parent families more often than non-respondents (boys: 81% vs 67%, p<0.001; girls: 77% vs 63%, p=0.001). Additionally, the participating girls had a slightly lower BMI and were more likely to be non-smokers than non-respondents at 16 years of age.25 The participating boys reported slightly better health status. At follow-up the participants were slightly slimmer, more physically active and more likely to be non-smokers than non-respondents, but the differences were not statistically significant. The participants and non-respondents did not differ with respect to LBP at either time point.

Prevalence of LBP

Altogether, 48% of the girls and 35% of the boys reported LBP at baseline when the subjects were 16 years of age. The prevalence of ‘Reporting LBP’ was 38% (43% in girls and 31% in boys) and the prevalence of ‘Consultation for LBP’ was 5% (5% in girls and 4% in boys). At 18 years of age the prevalence of ‘Reporting LBP’ in the study population was 50% (57% in girls and 42% in boys) while the prevalence of ‘Consultation for LBP’ was 6% (6% in girls and 5% in boys).

Incidence and persistence of LBP at 2-year follow-up

In adolescents without LBP at baseline, the incidence of ‘Reporting LBP’ was 29% in girls and 19% in boys and that of ‘Consultation for LBP’ was 2% in both genders. The persistence of ‘Reporting LBP’ was 53% in girls and 46% in boys and that of ‘Consultation for LBP’ was 19% in girls and 17% in boys.

Working history and physical workload factors

By 18 years of age, 814 girls (75%) and 640 boys (71%) had been working, 391 (20%) were employed (34 in full-time jobs), 1338 (67%) were studying, 90 (5%) were not working or studying, while data about their main activity was not available for 165 subjects (8%). Of the adolescents without LBP at baseline, 381 girls (75%) and 372 boys (74%) had been working. The mean (SD) duration of work was 6.2 (10.3) months (range 0–122). Girls had been working slightly longer than boys (6.4 (10.0) months (range 0–81) vs 5.9 (10.6) months (range 0–122). Physically very strenuous work was reported by 182 girls and 304 boys.

Infrequently reported specific exposures included driving (n=5) and manual handling of heavy objects (n=56) in girls and driving (n=84) in boys. Standing or walking was the most frequent exposure in both genders (495 girls and 396 boys), followed by kneeling (n=352), awkward trunk postures (n=321) and working with hands above shoulder level (n=218) in girls, and manual handling of medium weight objects (n=254), kneeling (n=251) and awkward trunk postures (n=233) in boys.

In girls the following exposures were more frequently seen in combinations than alone: kneeling or squatting, manual handling of medium weight or heavy objects, working hands above shoulder level, awkward trunk postures and standing or walking ≥5 h/day. In boys the exposures more frequently seen in combinations than alone were: kneeling or squatting, driving, manual handling of medium weight or heavy objects, working with hands above shoulder level, awkward trunk postures and standing or walking ≥5 h/day.

The optimal number of clusters was determined to be three among girls and four among boys (minimum BIC values 5784.4 and 5812.7, p=0.0327 and p=0.0007 in the LRT test for girls and boys, respectively). Among girls, cluster 1 consisted of subjects who were not exposed to any specific workload (‘Reference’). Subjects in cluster 2 were exposed to kneeling, working with hands above the shoulder level, awkward trunk postures and standing or walking (‘Awkward trunk postures’). Subjects in cluster 3 were exposed to kneeling, manual handling (medium or heavy objects), working hands above shoulders, awkward trunk posture and standing or walking (‘Physically demanding job’). Among boys, cluster 1 consisted of subjects who were not exposed to any specific workload (‘Reference’). Subjects in cluster 2 were exposed to kneeling, working with hands above shoulders, awkward trunk postures and standing or walking (‘Awkward trunk postures’). Subjects in cluster 3 were exposed to manual handling (medium or heavy objects) and standing or walking (‘Manual handling’), while subjects in cluster 4 were exposed to all seven workloads (‘Physically demanding job’). Table 1 shows the distribution of physical workload factors in each cluster.

Table 1

Distribution of physical workload factors in each latent class analysis (LCA) cluster

Physical workload factors and incidence and persistence of LBP

In girls, working regularly or irregularly increased the risk of incident LBP (RR 1.48, 95% CI 1.23 to 1.79; table 2). The duration of work exposure was associated with incident LBP in both genders (RR per increment of 1 month 1.15, 95% CI 1.03 to 1.29 in girls; RR 1.12, 95% CI 0.97 to 1.29 in boys). Of specific physical workload factors, the risk of incident LBP was increased in girls exposed to manual handling of medium weight objects (RR 1.30, 95% CI 1.04 to 1.62) and awkward trunk postures (RR 1.23, 95% CI 1.03 to 1.48) and, in boys, the risk of incident LBP was increased in those exposed to kneeling (RR 1.74, 95% CI 1.37 to 2.20), working with hands above shoulder level (RR 1.32, 95% CI 1.00 to 1.75) and awkward trunk postures (RR 1.69, 95% CI 1.33 to 2.15; table 3).

Table 2

Risk of incident* low back pain according to the main activity and duration of work exposure during follow-up (n=986)

Table 3

Risk of incident* low back pain according to physical workload factors during follow-up (n=986)

In both genders, belonging to ‘Awkward trunk postures’ or ‘Physically demanding job’ clusters was associated with an increased likelihood of incident LBP (in girls: RR 1.28, 95% CI 1.05 to 1.55 and RR 1.35, 95% CI 1.05 to 1.74, respectively; in boys: RR 1.85, 95% CI 1.38 to 2.48 and RR 1.55, 95% CI 1.14 to 2.12, respectively; table 4).

Table 4

Risk of incident* low back pain (LBP) according to clusters of physical workload factors during follow-up

Persistent LBP was not associated with working regularly or irregularly, duration of work or with specific physical load factors in either gender (data not shown). No associations were observed between LCA clusters and persistent LBP (table 5).

Table 5

Risk of persistent* low back pain (LBP) according to clusters of physical workload factors during follow-up

Discussion

Main findings

In this 2-year follow-up study among adolescents we observed that, in girls at baseline, working regularly or irregularly and duration of work exposure were associated with incident LBP. Furthermore, in both genders, awkward trunk postures and being in a cluster with high exposure to awkward trunk postures or an overall physically demanding job were associated with incident LBP. The magnitude of the associations was comparatively low for both individual workload factors and the clusters. None of the specific workload factors or clusters was associated with LBP in adolescents at baseline.

Comparison with previous studies

This appears to be the first study to analyse the associations of specific exposures and physical workload clusters with the development of LBP in adolescents. Overall, there are few studies on the relationship of physical work with LBP in this age group. Some previous reports are in line with our findings. In a cross-sectional survey of children and teenagers with an average age of 13 years in a low-income area of Brazil, monotonous work and awkward postures were associated with back pain.26 However, these observations on child labour are not very comparable to adolescent work in a welfare state. Working during the school year seemed to be associated with the development of LBP and musculoskeletal pain in a Canadian cohort study of high school students with an average age of 14 years. There was no such association in the summertime interval.14 27 We did not ask if work took place during the school year or in the summertime.

In the study of high school students, the development of musculoskeletal symptoms was highest in white collar work, followed by blue-collar work and lowest in childcare work.27 In our study, the two most often reported occupations for girls were gardening/farm work and cleaner in the awkward postures cluster, and gardening/farm work and shop assistant in the physically demanding job cluster. Among boys, property maintenance/gardening and construction work were the most common occupations in both of these clusters. The Finnish National Work and Health Survey 2006 listed the occupations that were considered to be physically fairly or very demanding as estimated by the workers. The most demanding occupations included skilled agricultural/fishery work, elementary occupations, craft and related trades, service/shop work and plant/machine operators. These occupations also involve more manual lifting and carrying than the other occupations.28

In a Belgian cross-sectional study on young workers, first-ever episodes of LBP were common in the first year of employment. They were associated with seated work, flexion or rotation movements of the trunk and more than 3 years of work involving heavy lifting. The individuals were young adults working in tenured positions and mostly full-time. In our study, adolescents worked mostly in part-time jobs and had a shorter working history. Regardless of these differences, we also found an association between awkward trunk postures and incident LBP.29

The exposure to whole body vibration by driving a motor vehicle was not associated with LBP in this study. In Finland, those under 18 years of age are not permitted to have a driving licence. In this study, boys had been driving to some extent but the exposure to whole body vibration was nevertheless negligible and therefore we cannot draw any conclusions about the association between whole body vibration and LBP.

No positive trend between combined physical work load factors and LBP was found in adolescents who already had LBP at 16 years of age. This could be due to selection bias—that is, those with LBP may have sought physically lighter jobs. However, no clear evidence for this was found. Among boys, those with LBP at 16 years of age were less likely to work in property maintenance/gardening (p=0.001) but more likely to work in construction (p=0.012) than those free of pain at baseline (data not shown). Among girls, LBP at 16 years of age was not associated with subsequent work exposure. Previous LBP is known to predict future symptoms also in adolescence.2 30 Thus, in this group with early symptoms of LBP, the role of genetic or unknown environmental factors may be more important.

Significance of the study

A systematic review concluded that the evidence on an association between heavy physical work and LBP is conflicting as seven of the 12 studies reported no significant association.15 There is controversy regarding the work-relatedness of LBP due to occupational lifting and awkward postures.31–33 However, one review found moderate evidence for heavy physical work, although in the sensitivity analysis the association was attenuated.17 In this study, due to the young age of our cohort, the proportion of those with multiple exposures and those exposed to strenuous work was rather low. Even though the work exposure usually consisted of several separate episodes, we observed an association between especially awkward trunk postures and incident LBP.

There is controversy regarding the effect of physical loading on the lumbar spine; heavy physical loading at work is regarded as harmful whereas, in leisure time, physical activity is thought to be beneficial.34 35 The relationship between physical activity and LBP is thought to be U-shaped.36 A very high level of physical activity has been observed to provoke LBP37 and hospitalisations38 among adolescents. In this study, we found that exposure to physical work was associated with self-reported LBP independently of the level of physical activity.

The current study showed that even short-term exposure to physical work may be predictive of LBP in adolescence. We can only speculate on the mechanism of the association. Manual materials handling and awkward postures may lead to failure of lumbar tissues. Another possibility is that the pain symptoms reported do not reflect serious tissue pathology. A further follow-up of our study cohort might reveal whether the later occurrence or severity of LBP among those exposed early to physical workload significantly differs from those without such exposure.

Part-time or period-time jobs were quite common among adolescents (18% of the study population). It is known that almost one-fifth of the Finnish population aged <20 years of age are to some extent engaged in working, but two out of three of these are students. Part-time work has become more common among adolescents in recent years in Finland.39 At the same time, pain in the low back, neck and shoulders has become more common among Finnish adolescents.1 Further studies are needed among young workers.

Strengths and limitations of the study

The study population belongs to the NFBC and is a representative large sample of Finnish adolescents. The strength of this study is also the longitudinal design that enables control for previously known confounding factors at baseline.

A limitation of our study is that work exposure was assessed retrospectively. Having LBP at 18 years of age might have introduced some bias into the reporting of workload. However, the answers were generally fairly detailed and it seemed to be easy for adolescents to remember their working history. Previous studies of adolescents have mostly used either dichotomised ‘working’ versus ‘not working’ or a three-level (white collar, blue collar, childcare) workload classification, whereas we used more detailed questions about physical work type and specific work tasks. We even enquired about exposures to short-time jobs (<12 months). The number of subjects working in the study population was similar to former national surveys.20 39 A self-administered questionnaire has been proposed to be useful in classifying heterogeneous occupational tasks, but questionnaire-based assessment of workload is not sufficiently accurate to study quantitative exposure-effect relationships.40

To test the effect of multiple exposures on LBP, we empirically derived the patterns of co-occurring exposures by LCA and estimated the effects of these patterns on LBP. The adolescent girls and boys had been working in different types of jobs and the combinations of co-occurring exposures in their jobs were different. We adjusted for smoking and leisure time physical activity in all our analyses. BMI or socioeconomic status did not affect our risk estimates, and these variables were therefore removed from the final analyses. A limitation of derived patterns was that we were not able to specify whether the different work exposures occurred simultaneously or during different working periods.

The outcome was pain in the low back area during the past 6 months. We have used the terms ‘incidence’ and ‘persistence’ in our study but, strictly speaking, the term ‘incidence’ is misleading as the subjects who did not report LBP during the past 6 months at 16 years of age may have experienced pain at an earlier time point. Furthermore, as we analysed only two time points 2 years apart, we may not have captured all the ‘incident’ cases. Similarly, ‘persistence’—that is, LBP during the past 6 months at 16 and 18 years of age—is not comparable to chronic LBP. A further limitation is that we did not ask about the frequency of the pain episodes or pain intensity, and the questionnaire did not specifically differentiate menstruation pain from other types of LBP in girls, which may partly explain the higher prevalence of ‘Reporting LBP’ in girls. Finally, a limitation was that we did not have a measure of severity or activity limitation due to pain in our study. Instead, we asked if individuals had consulted a health professional because of their LBP.

One-third of the eligible population did not respond to the survey at the age of 18 years. The participants and non-respondents did not differ with respect to LBP. The respondents were more likely to be living in two-parent families and the responding girls were more often slimmer and more likely to be non-smokers than the non-respondents. It is possible that those who did not respond had a somewhat higher risk profile than the respondents, and we may have underestimated the true association of workload and LBP. A further limitation is that we did not assess psychological factors, which have been shown to influence LBP.16

Conclusions

In this cohort study, the duration of work exposure was associated with LBP in adolescents without LBP at baseline during a time course of 2 years. The work exposure patterns in girls and boys were different. Subjects in clusters with high exposure to awkward trunk postures or an overall physically demanding job had a higher likelihood of incident LBP in both genders. These associations were independent of possible confounding factors such as leisure time physical activity, smoking, BMI and the family's socioeconomic status. The results of this study suggest that physical workload factors constitute a risk for LBP even among adolescents.

References

Footnotes

  • Funding The Academy of Finland, grant numbers 200868 and 129504.

  • Competing interests None.

  • Patient consent Obtained.

  • Ethics approval Institutional ethics committee of Oulu University Hospital.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Linked Articles

  • Postscript
    BMJ Publishing Group Ltd