The Association Between Self-Reported Cigarette Smoking and Spinal Pain is Not Explained by Serum Cotinine Levels

Abstract

The aim of this study was to check if self-reported smoking is still associated with back pain above and beyond its association with cotinine, to test the hypothesis that the association of self-reported cigarette smoking with back pain is due to residual confounding. Secondary analyses of population-based cross-sectional data pertaining to 4470 adults were conducted. In multivariate analyses examining the associations of self-reported smoking with several spinal pain outcomes (neck pain, low back pain, low back pain with pain below knee, self-reported diagnosis of arthritis/rheumatism, and related limitations), further adjustment for serum cotinine concentrations was made. Self-reported cigarette smoking was associated with neck pain (adjusted Odds Ratio (aOR) Regular smokers vs. Non-smokers: 1.44; 95% Confidence Interval (CI): 1.14-1.82), low back pain (aOR: 1.48; 95% CI: 1.24-1.78), low back pain with pain below knee (aOR: 1.98; 95% CI: 1.42-2.76), as well as arthritis/rheumatism (aOR: 1.33; 95% CI: 1.03-1.71), and related functional limitations (P < .05). Further adjustment for serum cotinine concentrations brought about little change in the ORs or beta coefficients. These results do not support the hypothesis that serum cotinine concentrations explain the well-known relationship between cigarette smoking and spinal pain.