Low back pain in a general population : care seeking behaviour, lifestyle factors and methods of exposure assessment

Abstract

This thesis is based on the MUSIC-Norrtälje study, a population-based case-referent study of low-back pain with a two-year follow-up. Included in the study were 449 female and 342 male cases. The specific aims were to study their reasons for seeking care for low-back pain and to describe and follow the development of pain, disability, choice of caregiver and sick leave during the two-year period. A further aim was to study whether sports activities, body weight, and smoking influenced the risk of a new episode of low-back pain. Finally the present study thesis aimed at validating interview and questionnaire data concerning the physical exposures, work in a stooped position work and work with hands above shoulder level.

High disability (OR 7.4 for women and OR 4.9 for men) and high pain intensity (OR 3.7 for women and OR 1.7 for men) were strongly related to seeking care for low-back pain. Women not seeking care despite pain reported a more strained economic situation than those who sought care. Physical or psychosocial working conditions, life style factors, and psychosomatic complaints did not affect care-seeking behaviour.

Some improvements regarding pain and disability were reported after three months but not much after six and 24 months. During the follow-up, about 60 per cent of the cases paid a substantial number of visits to caregivers. The majority (70%) of the cases had not been on sick leave one single day during the two-year period.

Sports activities did not affect the risk of low-back pain among men. High intensity training 1-2 hours per week increased the relative risk of low-back pain for women (RR 1.6) compared to women who did not do sport. An increased risk of low-back pain was found for men with BMI > 30 (RR 2.2) but not for women. Smoking did not influence the risk of low-back pain.

Regarding validation of the interview, the correlation coefficients between interview responses concerning a "typical day" and observations during two working days was 0.79 for work in a stooped position and 0.65 for hands above shoulder level. The correlation coefficients between interview and questionnaire responses were 0.66 for work in a stooped position and 0.63 for hands above shoulder level.

In conclusion, the studies demonstrate that the most decisive factors for seeking care due to low-back pain were high disability and high pain intensity. Economic factors seemed to be important for women. Few of the cases became pain-free during the follow-up, but sick leave was rare. A certain amount of sports activities had a negative association with low-back pain for women. Smoking did not influence low-back pain. Interview data concerning work in a stooped position or work with hands above shoulder level are valid for exposure assessments. In addition, questionnaire data for those postures may be sufficiently accurate, at least at a dichotomous level.