A retrospective cohort study of adolescent idiopathic scoliosis. A comparison group of persons without scoliosis was also selected randomly from the general population.
To estimate the association between level of cigarette smoking and the prevalence and severity of back pain.
A postal questionnaire was used to elicit information on smoking histories, a variety of indices of low back pain, and potential confounding factors. The association between smoking and back pain was estimated separately for men and women in the cohort and in the comparison group using ordinal regression models.
The questionnaire was completed by 1287 women and 184 men who had adolescent idiopathic scoliosis and by 1130 women and 621 men in the comparison population who did not have scoliosis. Statistically significant associations between back pain and current cigarette smoking were found in the two groups of women and men with scoliosis, but not among men selected from the general population. In the three former groups, proportional odds ratios comparing current smokers to persons who never smoked ranged from 1.4 to 1.9. Among current smokers, the prevalence of back pain increased with cigarette consumption, and the proportional odds ratios ranged from 1.2 to 1.8 per 10 pack-years (no. of cigarettes smoked per day × no. of years/20). In these three groups, intensity, frequency, and duration of episodes of back pain also were found to increase with smoking consumption.
The finding that smokers have more frequent episodes of back pain may imply that smoking exacerbates back pain, and the observation that stronger associations between back pain and smoking were found in the scoliosis cohort suggests that smoking may have a greater impact onpersons with damaged spines.
From the *Division of Clinical Epidemiology, Royal Victoria Hospital, Montreal, †Epidemiology and Biostatistics Unit, INRS-Institut Armand-Frappier, Université du Québec, Laval, the Departments of ‡Epidemiology, Biostatistics and Occupational Health, and §Medicine, McGill University, Montreal, ∥Public Health Department, Hôpital Maisonneuve-Rosemont, Montreal, and ¶Hôpital Ste-Justine, Montreal, Quebec, Canada.
Supported by the National Health Research and Development Program of Health Canada, project number 6605-4166-55, CAFIR Université de Montréal, and Le Fonds de la recherche en santé du Québec, project numbers 871352 and 901397. Dr. Goldberg gratefully acknowledges support from the National Health Research and Development Program, Health Canada, through a National Health Research Scholar Award and support from les Fonds de la recherche en santé du Québec (FRSQ). Dr. Mayo appreciates receipt of a Health Scientist Career award from the FRSQ.
Acknowledgment date: February 17, 1998.
First revision date: August 14, 1998.
Acceptance date: November 3, 1998.
Address reprint requests to
Mark Goldberg, PhD
Epidemiology and Biostatistics Unit
University of Quebec
531, boulevard des Prairies
Laval, Quebec, Canada H7V 1B7
Device status category: 1.