The association between BMI, smoking, drinking and thyroid disease: a cross-sectional study in Wuhan, China

Background There is no clear conclusion on the relationship between thyroid disease and obesity and lifestyle factors such as smoking and drinking. In this study, we analysed the association of body mass index (BMI), smoking and drinking with subclinical hypothyroidism (SHO) and thyroid nodules (TNs) with the results of a cross-sectional survey of urban residents in central China and discussed the potential mechanism linking these predictive factors and the two diseases. Methods This study included 1279 participants who were recruited from a Chinese community in 2011 and 2012. A questionnaire, laboratory examination and ultrasound diagnosis were conducted on these participants. Binary logistic regression analysis was used to analyse these factors. Results Overweight (BMI ≥ 25 kg/m 2 ) was closely related to SHO and TNs in univariate and multivariate logistic regression analyses. Smoking had a protective effect on SHO and TNs, while drinking had a protective effect on TNs in univariate logistic regression and multivariate logistic regression with some covariates, but there was no significant difference between smoking and drinking and the two kinds of thyroid diseases in multivariate logistic regression analysis with all the covariates. In subgroup analysis, BMI ≥ 25 kg/m2 was significantly associated with SHO in people with positive thyroid antibodies (odds ratio (OR) = 2.221, 95 % confidence interval (CI): 1.168–4.184, P = 0.015) and smokers (OR = 2.179, 95 % CI: 1.041–4.561, P = 0.039). BMI ≥ 25 kg/m2 was significantly associated with TNs in people over 60 years old (OR = 2.069, 95 % CI: 1.149–3.724, P = 0.015) and drinkers (OR = 3.065, 95 % CI: 1.413–6.648, P = 0.005). Drinking alcohol had a protective effect on TNs in smokers (OR = 0.456, 95 % CI: 0.240–0.865, P = 0.016) and people with BMI ≥ 25 kg/m 2 (OR = 0.467, 95 % CI: 0.236–0.925, P = 0.029). No significant association was found between smoking and the two thyroid diseases in different subgroups. Conclusions Obesity is a risk factor for both TNs and SHO, especially in elderly individuals and people with positive thyroid autoantibodies. Obesity and metabolic syndrome may be more associated with TNs than SHO. Smoking may have a protective effect on thyroid disease, while drinking may have a protective effect only on TNs.