The results of this study suggest that cases reported a more than doubling of the risk of exposure to occupational driving ≥4 h/day (adjusted OR 2.39, 95% CI 1.08–5.27) in unstratified analyses. Stratification of the analyses by age provided additional insight as, amongst those aged < 75 years, we found that cases were more than 3-fold likely to report exposure to heavy manual work in the agricultural/fishing industries (adjusted OR 3.47, 95%CI 1.29–9.29). Spondylolisthesis cases were more likely to report exposure to driving ≥4 h/day in the stratified analyses but these did not attain statistical significance. However, spondylolisthesis was associated with an almost doubling of risk of reporting sedentary work ≥2 h/day amongst those aged ≥75 years (OR 1.95, 95% CI 1.06–3.57). Additionally, we found that there was a negative association with spondylolisthesis (OR 0.36, 95% CI 0.15–0.82) amongst those aged < 75 years who reported that they climbed up stairs and slopes ≥1 h/day and a similar negative association was seen with walking ≥3 km/day amongst those aged ≥75 years (OR 0.55, 95% CI 0.29–1.02).
The finding that the associations of spondylolisthesis with agricultural/fisherman become attenuated at older ages (≥ 75 years) is not surprising. The retirement age in many Japanese industries is < 65 years so that, in this older cohort, many participants would have stopped work some years prior to their X-rays. It is clear from epidemiological studies that age, gender and hormonal factors play a role in the pathogenesis in spondylolisthesis and it is likely that the impact of occupational exposures overall will relatively diminish as that of other exposures (leisure-time physical activities) increase as years since retirement increases [1, 8].
The results need to be considered alongside some limitations. First, this is a cross-sectional study so that causal attribution cannot be made. Secondly, the participants in this study were sampled from the general population but not at random. We investigated their representativeness by comparing the study population with the general population of Japan for a key risk factor for osteoarthritis, body mass index (BMI). We found that the mean BMI of the participants was not significantly different from that of the general population of Japan (males: 23.71 (3.41) vs. 23.95 (2.64) kg/m2; females: 23.06 (3.42) vs. 23.50 (3.69) kg/m2). However, the study participants reported a lower prevalence of smoking and alcohol use than that seen in the Japanese general population, suggesting participants might live healthier lifestyles. This may limit the generaliseability of these findings. We also cannot rule out a possible selection bias as volunteers needed to be sufficiently healthy to participate and undergo spinal radiographs and this may have limited the possible involvement of elderly institutionalised adults. As spondylolisthesis is a common cause of impaired mobility in older people and immobility may lead to institutionalisation, this may have created a bias, but if so, the effect would have been to reduce the estimated prevalence of spondylolisthesis. The impact of this on the study results would however, only be biased if we believe that those who were employed in any specific occupation were more likely to be institutionalised than those who had worked in others, which seems unlikely. Occupational exposures were obtained by direct inquiry rather than being inferred from job title. Of course the information is dependent upon recall, but the subjects were unaware of their radiographic findings when they were recalling their occupation and occupational exposures so that a systematic bias is unlikely.
It is a strength of this study that all X-rays were assessed by one highly-trained orthopaedic surgeon (YI). In addition, considerable efforts were made to guarantee the reliability of the readings, including inter-observer and intra-observer studies with a sample of 5% of the X-rays, both of which suggested a very good level of reliability (kappa > 0.8 in both studies).
There is disagreement in the literature as to how best to classify radiographic spondylolisthesis. Two principal types are proposed: lytic and degenerative. Lytic is associated with a pars defect of the vertebral arch and usually causes slippage at younger ages. Degenerative spondylolisthesis has been variously described but usually defines a slip occurring with an intact vertebral arch with/without associated advanced arthritis changes in the facet joints at the same level [8]. A systematic review of risk factors for spondylolisthesis found that the selection of cases for observational studies has differed widely [8], some studies restricting selection of cases to those with a slip at L4–5 but others including a slip at any level. Some, but not all, studies included anterior and posterior slips and some studies selected only symptomatic spondylolisthesis or surgical cases. Selection based upon symptoms was criticized in the review given the lack of conclusive evidence that spondylolisthesis causes low back pain. Given the current lack of clarity about case definition, we chose a case definition which is inclusive and consistent with that of other researchers [22,23,24]. In particular, it was not our intention to separate lytic from degenerative types of spondylolisthesis but rather to use this large population sample to investigate whether any type of occupational exposure was associated with the outcome, in order to generate hypotheses about causation and potentially develop preventive strategies. This is particularly important with occupational research where there is already an established tradition of putting in place preventive strategies to protect workers if an increased risk is established, often in the absence of information about mechanism of causation (e.g. asbestos and lung disease).
Low back pain associated with manual occupations has long been reported [25]. Degenerative changes at the cervical and lumbar spine have been described with working outdoors, heavy lifting and whole-body vibration [25,26,27]. An increased risk of disc degeneration on MRI scans has been reported amongst those who had undertaken the heaviest occupational lifting [28] and highest lifetime cumulative lifting load [29]. Moreover, increased rates of radiographic lumbar spondylosis were found amongst agricultural/forestry/fishery workers [26], and an increased risk of vertebral endplate sclerosis amongst concrete reinforcement workers [14]. In another recent analysis within the Wakayama Spine Study, we have shown an increased association of lumbar spinal stenosis with factory and construction work [submitted to Am J Ind Med]. The current study adds to a growing body of literature suggesting that the risk of spinal structural degenerative changes is increased in some occupations. The different occupational risk profiles warrant further investigation but might be explained by the selective mechanical effects of these exposures.
Spondylolisthesis has been previously reported in association with occupational driving and flying. Froom et al. reported a four-fold increased risk of lytic spondylolisthesis amongst helicopter pilots as compared with transport pilots and cadets [17]. Among Taiwanese taxi drivers, longer duration of exposure to driving was associated with an increased risk of spondylolisthesis [18]. Mariconda and colleagues found that, amongst 120 back pain patients with spondylolisthesis, occupational driving was the only factor associated with a greater degree of vertebral slip and other occupational activities (awkward posture, prolonged sitting and prolonged standing) were not [19]. It has been postulated that the mechanism by which driving/flying increase the risk of spondylolisthesis is through exposure to whole-body vibration. Indeed, animal studies found that exposure to continuous quantitative vibration diminished the proteoglycan content in the nucleus pulposus disrupting the integrity of the matrix and increasing the instability of the intervertebral disc [30]. The results of our study support the possibility of an association with occupational driving and of course workers in the agricultural/fishing industries may also be exposed to whole-body vibration in tractors or boats [31, 32].
The evidence about other occupational exposures and spondylolisthesis is conflicting. One study reported an increased risk of symptomatic spondylolisthesis amongst those with a heavy workload and undertaking manual handling of materials [19] but another found no association between risk of degenerative spondylolisthesis and the degree of lifting [33]. A study using the same radiographic definition of spondylolisthesis as in the current study found that, although the prevalence of individuals reporting that their ‘longest occupation involved physical labour’ was higher among men with a slip (10.2%) than those without (8.6%), the differences were not statistically significant [24]. It is interesting that we found an association with occupational sitting but only amongst older participants after stratification by age. Two other studies reported no association [19, 33] but differed importantly in their methodology: one [19] only included back pain patients attending outpatient clinics and willing to undergo MRI whilst the other [33] recruited > 4000 people to a population cohort but included a wider age range (22–93), so that the participants were on average 10 years younger than in the current study. More research is required to explore the impact of prolonged occupational sitting but, as sedentary work is becoming increasingly common with mechanization, this could be an important risk factor for spondylolisthesis. Importantly, our findings of a negative association with reported climbing flights of stairs or slopes ≥1 h/day and/or walking ≥3 h/day may suggest that, if replicated in longitudinal studies, being active during working hours may be beneficial in the long-term prevention of this condition.