Background: Sick leave due to long-term musculoskeletal pain is an increasing public health problem, particularly among women, in Western countries. Studies have shown that reduced work ability and well-being are common in this population, but do they really matter for return to work (RTW)?

Purpose: The purpose of the present study was to compare work ability and well-being over time, among sick-listed women with long-term pain in neck/shoulders or back who had RTW and who remained on sick leave.

Methods: A postal survey was sent to 600 women receiving time-loss benefits according to the Swedish Social Insurance registers. Inclusion criteria: (i) age between 18 and 65 years, (ii) Sick leave: ≥ 1 month of their ≥ 50% service, (iii) pain in neck/shoulders or back: ≥ 3 months, and (iv) understanding Swedish. Exclusion criteria: (i) rheumatoid arthritis, (ii) multiple sclerosis, (iii) stroke, (iv) cancer, (v) Parkinson, (vi) bipolar disease, (vii) schizophrenia, (viii) pregnancy. After 12 months, a follow-up survey was sent to the 208 women who answered the survey at baseline. At the follow-up, 64 women had RTW and 41 remained on sick leave. The Work Ability Index (WAI) and the Life Satisfaction questionnaire (LiSat) were used to assess work ability and well-being, respectively. To determine whether the groups (RTW and non-RTW) differed at baseline for WAI and LiSat, respectively, independent t-tests were performed. If significant differences were found, the baseline values were included as a covariate in univariate analysis of variance with change in WAI and LiSat, respectively, as outcome and groups as independent variable. The level of significance was set to p < 0.05.

 Results: For WAI, the groups differed at baseline (t= -7.54, p < 0.001). In the univariate analysis of covariance, the groups changed significantly over time (F= 28.47, p < 0.001) after adjusting for baseline values. The group that returned to work improved more than the group that remained on sick leave. For LiSat, there was no difference between groups at baseline (t= -1.58, p = 0.47). In the univariate analysis of variance, the groups changed significantly over time (F= 9.30, p = 0.003). The group that returned to work improved, whereas the group that remained on sick leave showed the opposite pattern.

Conclusion: It appears that work ability and well-being do matter for RTW. The results for well-being in particular highlight the need to intervene against prolonged sick leave in this population.