Background: Current daily smokers have generally been found to have lower body mass index (BMI) as compared to never or former smokers, but this association may have reversed in recent times. Because of genetic influences on both smoking and BMI, familial confounding (genetics and shared environment) is important to take into account to get more information about the causality of this association.
Aims: We investigated whether familial factors confound the associations between smoking status and BMI and whether these associations have changed over time.
Methods: Data on 55,290 same-sex twin pairs, 18-69 years of age, with 80,712 paired tests from 21 twin cohorts representing 14 countries were derived from the CODATwins database. We analyzed pairs discordant for their smoking status. One concurrent measurement of BMI and smoking status for each individual per a 10-year period (1960-1969, 1970-1979, 1980-1989, 1990-1999, and 2000-2010) was allowed. Conditional (fixed-effect) linear regression models adjusted for age and age-squared were used.
Results: Current daily smokers had lower BMI (-0.37 kg/m2, 95%CI: -0.43, -0.31 in men, and -0.44: 95%CI: -0.52, -0.37 in women) and former smokers higher BMI (0.27, 95%CI: 0.21, 0.33 in men, and 0.16: 95%CI: 0.08, 0.23 in women) compared to their never-smoker co-twins. Compared to current daily smokers, former smokers had higher BMI (0.64, 95%CI: 0.59, 0.70 in men, and 0.60: 95%CI: 0.53, 0.67 in women). Similar trends were seen both in dizygotic and monozygotic pairs in both sexes across all 10-year periods except among former smokers. The BMI difference of former smokers compared to never-smokers attenuated in monozygotic pairs and was overall less clear in women.
Conclusions: Smoking is independently associated with lower BMI both in men and women, and this association has not changed over time. BMI change after smoking cessation may be confounded by genetics and other behavioral factors.
Ageing , Substance use: Alcohol, Nicotine, Drugs , Health (e.g., BMI, Exercise)
