Leg-length discrepancy is associated with low back pain among those who must stand while working

Low back pain (LBP) is the most debilitating condition globally, presenting with severe socioeconomic and health-care consequences [ 1, 2]. The etiology of non-specific LBP is not well understood. Leg-length discrepancy (LLD) has been listed as one risk factor of LBP although its role as such has been questioned [ 3]. The reason for doubt is the contradictory data published as several studies have found a relation between LLD and LBP [ 4- 6] but on the other hand many have not found such an association [ 7- 10].

One explanation for the conflicting evidence on the role of LLD in LBP may be the diversity in measurement methods. Radiographic measurement of LLD has been traditionally regarded as the gold standard but radiation exposure is potentially harmful and therefore radiographic LLD measurements have become less popular. The ultrasound technique has been reported as reliable when compared to radiographic measurement [ 11]. Recently, we found that a modified ultrasound method was reliable and its agreement with radiographic method was excellent [ 12].

The aim of the current study was to evaluate the association between LLD and LBP using the novel ultrasound method in an occupational population consisting of meat cutters and customer service workers. We hypothesized that meat cutters, standing mostly in their work, would more likely experience LBP and pain-related consequences in case of LLD.

The study population consisted of 218 workers; 169 pork meat cutters (31 females and 138 males) and 50 customer service workers (41 females and 9 males). In all, 114 of the 169 meat cutters (68%; 26 females and 88 males) and 34 of the 50 customer service workers (68%; 30 females and four males) participated. Summer vacations were the main reason for refusing to participate in our study. There were no differences between the non-participants and participants with respect to gender, age, working years and sick leave days because of LBP (M50-54 in ICD-10) during the past year and the past five years (data not shown).

The distribution of LLD and characteristics of the total study population and according to the type of work are presented in Table  1. Fifty-five (49%) of the meat cutters and 15 (44%) of the customer service workers had LLD of at least 6 mm, while 16% and 15%, respectively, had LLD of at least 11 mm. LLD of at least 6 mm was more common among males (55%) than females (37%) and the gender difference was statistically significant (p < 0.05). LLD relative to leg length did not differ significantly between genders (p > 0.05).

In the total study population, LLD of 6 mm or more was associated with higher intensity of LBP during the past week and 3 months in the adjusted analyses (Table  2). As responses were logarithmized in the linear regression analysis, the interpretation of the estimates differs from the analysis using original scaling. The estimated effect of LLD > 5 mm, 0.53, indicates that the mean intensity of pain is approximately 53% greater compared to LLD ≤ 5 mm. Self-reported number of days with LBP was higher with those with LLD 6 mm or more but the probability of the days with LBP was not associated significantly with LLD in the adjusted analyses (Table  3). The sick leaves during past year were slightly longer among those with LLD 11 mm or more, but the differences were not statistically significant (Figure  1).

None of the gender interaction terms in multivariate analyses were statistically significant (p > 0.10). However, as males were significantly taller than females (p < 0.001), we repeated all multivariate analysis using relative LLD as a continuous dependent variable (data not shown). Similar results were obtained, as relative LLD was associated significantly with higher intensity of LBP during the past week (p < 0.001) and 3 months (p < 0.001) and with higher self-reported number of days with LBP (p < 0.01). No statistically significant associations between the probability of the days with LBP or sick leaves and relative LLD were found but the sick leaves were longer among meat cutters (data not shown).

There was no significant interaction effect between type of work and LLD regarding intensity or number of days of LBP (p > 0.10). However, both intensity of LBP and number of days of LBP were associated with LLD among meat cutters but not among customer service workers in stratified analysis (Figure  2). Meat cutters who had LLD of 6 mm or more had a significantly higher intensity of LBP during both the past week and past three months compared to those with LLD ≤ 5 mm, while among customer service workers no significant association was found.

In our study LLD (6 mm or more) was associated with a higher likelihood of low back symptoms among meat cutters with a standing and physically more demanding job.

Among chronic LBP patients (n = 1309), almost every fifth had LLD of more than 9 mm compared to only four of 50 voluntary controls who had no history of LBP [ 4]. Among Finnish soldiers (n = 798) a higher prevalence of LLD was observed among those with LBP than without [ 5]. In a small-scale study, LBP patients (n = 10) with LLD of 10 mm reported increasing pain while standing for 20–30 minutes, followed by immediate relief upon sitting [ 6]. On the other hand, among factory workers (n = 594), workers with and without LBP had similar distributions of LLD up to 10 mm [ 10]. Similarly, no statistically significant difference in LLD was observed between chronic LBP patients (n = 70) and their age- and sex-matched controls [ 7].

It is possible that the contradictory results in previous studies are due to the differences in study populations and methodology used for measuring LLD. In most of the previous studies [ 4- 6, 10] LLD was assessed using erect-posture radiography, while some [ 7, 8] used even the clinical method, which is less reliable and valid than radiographic or ultrasound methods [ 15, 16]. The strength of our study is the validated accurate method to measure LLD [ 12]. Our laser-based ultrasound measuring method was accurate and easy to use, which makes it suitable for large-scale studies. In our study, the questionnaires were filled before the LLD measurements. The physiotherapists who measured LLD were not aware of workers’ low back symptoms or disease history. All meat cutters were working in the same department and performed similar physically demanding work tasks. Similarly, all customer service workers had identical work description. The characteristics of workers who participated in the study did not differ from those who declined.

In the current study, LLD was associated with low back symptoms after adjustments for confounding factors in the whole occupational sample. However, when occupational groups were analysed separately, the association was observed among meat cutters whereas among customer service workers from the same factory the effect was much smaller and not statistically significant. Yet the distribution of LLD was similar in both groups. Our findings suggest that LLD associated with standing and physically heavier job might render a higher risk for LBP than LLD in sedentary, physically less demanding office work. Among individuals who had LLD in our study, particularly lower extremity loading may have triggered LBP. LLD may cause abnormal biomechanical stress in the pelvis and lumbar spine during walking through asymmetrical loading. The exact mechanism by which LLD causes or augments LBP, however, is not clear. Although our study showed that both absolute and relative LLD were associated similarly with LBP, some anthropometric characteristics might weaken or strengthen the relationship between LBP and LLD.

In the statistical analyses, number of days with LBP had a skewed, highly dispersed distribution with large amount of zeros (zero inflated distribution); therefore Poisson regression could not be used. We considered different models available for zero inflated distribution and chose Hurdle model since some of the workers did not have LBP and hurdle model recognizes the possibility that there can be different factors contributing to probability to have LBP in the first place and to duration of LBP for those who have LBP.

The weakness of our study is the small sample size, especially in customer service workers. Consequently, we were not able to exclusively explore the role of working positions in the estimated associations and further studies should confirm the generality of our conclusions. Our data did neither allow for in-depth investigation of possible gender differences. As most meat cutters were males, a significant gender-specific mechanism behind the association between LLD and LBP, rather than standing or sitting, could potentially explain our findings. However, we believe that this is unlikely, because the associations were independent of leg length. A further limitation is that we did not measure exact time spent sitting or standing in the workplace. However, in the meat factory from which our study sample was drawn, customer service workers mostly sit and meat cutters mostly stand in their work, which result in greatly different overall time spent standing. We relied on self-reports of outcomes and the number of days with LBP during the last 12 months should only be interpreted as a rough proxy for LBP frequency due to long recall period. Finally, cross-sectional study design can only reveal an association but does not explain the interrelationships of different factors. Longitudinal studies are needed to assess whether LLD is a risk factor of LBP. Additional proof for the role of LLD in LBP would be provided by a randomised controlled trial showing a beneficial effect of LLD correction on LBP symptoms.

Our study found a significant association between LLD of 6 mm or more and low back symptoms. The association was apparent among meat cutters, who stand while working, but not among customer service workers, who mostly sit while working. We hope that future studies would assess whether our results can be replicated with larger sample size and in other populations using a similar reliable assessment method of LLD.