Health care workers are generally characterized by having a high physical work load and high prevalence of overweight [9]‐[10]. However, a causal relationship between these risk factors and development of LBP remains debatable. A main reason for the debate is the different methodological designs of previous studies. For example, several studies have included health care workers with existing LBP in their analyses [11-13]. Given that studies have shown that earlier history of LBP is a strong predictor of future LBP [14]‐[15], this may have contributed to the inconsistent findings. Other methodological aspects, like cross-sectional designs and study populations with varying durations of exposure to physical work demands at baseline may also be related to different conclusions. Thus, prospective cohort studies determining the risk for developing LBP due to being overweight or obese and to have moderate or high physical work load are necessary for optimizing preventative strategies.

Table ​Table11 shows characteristics of the study population. Small, but statistically significant differences at first follow-up between respondents with low, moderate and high physical work load were observed for BMI, age, social support at work and influence at work. Significant differences at first follow-up between respondents with normal weight, overweight and obese were observed for age and influence at work.

Table ​Table22 shows the results of the prospective analyses of the association between physical work load, body weight and the risk of developing LBP. In the unadjusted analyses (Model I), compared with health care workers with low physical work load, those with high physical work load had a significantly higher risk for developing LBP 1–7days (OR 2.02), as well as LBP over 7days (OR 2.16) than health care workers with low physical work load. After adjustment for possible confounders (Model II), the health care workers with high physical work load had a 78% increased risk for LBP 1–7days (OR 1.78), as well as LBP over 7days (OR 1.78) compared to their colleagues with low physical workload. In both the unadjusted and adjusted analyses, overweight and obese health care workers did not have a significant increased risk for developing LBP compared to their normal weighted colleagues.

Our study is the first to investigate the effect of overweight and obesity on the development on LBP among newly educated female health care workers without prior history of LBP. However, another study by Van Nieuwenhuyse et al. look into the effects of physical characteristics measured in physical examinations on the development of LBP among health care workers and distribution companies with no or little LBP history [24]. That study finds a significant effect of obesity on later development of LBP among asymptomatic health care workers. However, their results are not directly comparable to ours because their analysis did not control for work environment factors. Although overweight and obesity do not seem to be risk factors for developing LBP among health care workers, they were shown to be related with the development of hypertension, coronary heart disease, type 2 diabetes and stroke [25]. Notably, 41% of female health care workers were overweight in our study population. Thus, from a public health perspective, initiatives for reducing overweight and obesity among health care workers are recommended.

Female health care workers with high physical work load had 78% increased risk for developing LBP compared to their colleagues with low physical work load. This finding validates previous observations among health care workers [11-13,26], although neither of these included newly educated female workers without prior history of pain. Our unique design provides strong evidence that high physical work load increases the risk for the development of LBP among health care workers. The strength of our study is the baseline population without prior exposure to health care work and without a life-time history of LBP, followed prospectively for two years into their working life after graduation. This is particularly important given that a previous study has indicated that LBP and disability due to LBP during baseline is a significant indicator for dropout in the following two years [9]. As shown in Table ​Table1,1, there was no difference among respondents with normal weight, overweight and obese concerning their estimated physical workload. As many respondents may have been overweight or obese for several years, a concern is the selection of overweight and obese individuals without LBP in the analysis. An additional analysis showed no differences in prevalence of LBP between normal weight (60%) and overweight (58%) respondents at baseline, whereas obese showed a significantly increased prevalence of LBP (65%, P<0.001). However, the cross-sectional nature of this baseline correlation does not reveal anything about causality. Because previous episodes of LBP have been shown to be a risk factor for new episodes of LBP [14,27], the exclusion of respondents with prior history of LBP strengthens the validity of our findings. Self-administered questionnaires have been criticized for not providing accurate information on exposure to physical work load. However, persons with LBP may tend to perceive their physical workload as higher than those without LBP [28,29], bringing the results that include those with previous LBP to question [11-13,26,30]. As the respondents included in our study had no prior history of LBP, a systematic bias on the reported physical work load is not very likely.