Tobacco smoking and thyroid function: a population-based study.

BACKGROUND The association between tobacco smoking and thyroid function is incompletely understood. METHODS In a cross-sectional, population-based study conducted between August 15, 1995, and June 18, 1997, of 20 479 women and 10 355 men without previously known thyroid disease, we calculated the geometric mean serum concentration of thyrotropin and the prevalence of hypothyroidism and hyperthyroidism among current, former, and never smokers. RESULTS Among women, the mean thyrotropin level was lower in current (1.33 mIU/L; 95% confidence interval [CI], 1.29-1.36 mIU/L) and former smokers (1.61 mIU/L; 95% CI, 1.56-1.65 mIU/L) compared with never smokers (1.66 mIU/L; 95% CI, 1.63-1.70 mIU/L). Similarly, among men, the mean thyrotropin level was lower in current (1.40 mIU/L; 95% CI, 1.36-1.44 mIU/L) and former smokers (1.61 mIU/L; 95% CI, 1.57-1.66 mIU/L) compared with never smokers (1.70 mIU/L; 95% CI, 1.66-1.75 mIU/L). In former smokers, thyrotropin levels increased gradually with time since smoking cessation (P for trend < .001). Among current smokers, moderate daily smoking was associated with higher thyrotropin levels than heavier smoking. In women, the prevalence of overt hypothyroidism was lower in current smokers compared with never smokers (odds ratio, 0.60; 95% CI, 0.38-0.95), whereas the prevalence of overt hyperthyroidism was higher among current smokers (odds ratio, 2.37; 95% CI, 1.34-4.20). The associations related to subclinical thyroid dysfunction were similar to those for overt thyroid disease. CONCLUSIONS These findings indicate that smoking is negatively associated with hypothyroidism but positively associated with hyperthyroidism. The associations with smoking cessation suggest that smoking may have reversible effects on thyroid function. Notably, we report for the first time, to our knowledge, a lower prevalence of overt hypothyroidism among current smokers.