Pain in multiple sites and sickness absence trajectories: a prospective study among Finns

We studied the number of musculoskeletal pain sites as a predictor of sickness absence during a 7-year follow-up among a nationally representative sample (the Health 2000 survey) of occupationally active Finns 30 to 55years of age (3420 subjects who did not retire or die during the follow-up). Baseline data (questionnaire, interview, clinical examination by a physician) were gathered in 2000 to 2001 and linked with information from national registers on annual compensated sickness absence periods (⩾10workdays) covering the years 2002 to 2008. Pain during the preceding month in 18 body locations was inquired and combined into 4 sites (neck, upper limbs, low back, lower limbs). Demographic factors, BMI, smoking, leisure-time physical activity, sleep disorders, physical and psychosocial workload, and chronic diseases were assessed. Four distinct sickness absence trajectories emerged, labeled as Low (59% of the subjects), Ascending (21%), Mixed (11%), and High (9%). In multinomial logistic regression, the odds ratios (ORs) for belonging to the High vs. the Low trajectory increased with the number of pain sites, being 2.1 for single-site pain, 2.6 for 2 pain sites, 2.9 for 3 pain sites, and 4.1 for 4 pain sites, after adjustment for chronic diseases, demographic and lifestyle factors, and workload. The confidence intervals of the ORs did not include unity. The adjusted ORs for belonging to the Ascending trajectory were 1.1, 1.3, 1.7, and 1.7, respectively. As the number of pain sites was a strong independent predictor of work absenteeism, early screening of workers with multisite pain and interventions to support work ability seem warranted.