Introduction: This study investigated the systemic response of serum bone alkaline phosphatase (SBAP) and urinary N-telopeptide (UNTX) to tobacco exposure and environmental tobacco smoke (ETS) and the possible effect modification (and variability) of this response by racial/ethnic origin. Methods: Data (n=5411) were obtained from the National Health and Nutrition Examination Survey, with data analysis done on adults aged ≥ 20years. Outcome variables were SBAP and UNTX. Independent variable was tobacco exposure measured using serum cotinine levels and adjusted for covariates. Generalized linear models were used to explore associations. Results: A percentage increase in log transformed serum cotinine was associated with a 0.005 percentage increase in log transformed SBAP (CI: 0.002, 0.008) and 0.02 percentage increase in log transformed UNTX (CI: −0.01, 0.04) with interaction between cotinine and race/ethnicity (p=0.01). Stratifying by race/ethnicity, tobacco exposure was associated with significant decreases in UNTX among non-Hispanic Whites–0.008(−0.014, −0.002) and Mexican Americans −0.014 (−0.025, −0.002) only. Categories of serum cotinine were associated with a monotonic increase in SBAP (p for trend <0.001) and monotonic non-linear decrease in UNTX (p for trend>0.05). Conclusions: Tobacco and environmental tobacco exposure are associated with SBAP and increased bone formation. The response of UNTX to these exposures is modified by race/ethnicity with non-Hispanic Whites and Mexican-Americans less sensitive to the resorptive effects of tobacco exposure on bone.

Introduction: This study aimed to examine the association of smoking status with homocysteine levels and to determine whether the association is modified by oestradiol or cholesterol. Methods: Data (N = 4580) were obtained from National Health and Nutrition Examination Survey 2003–2004 with analysis done in 2018 on adults aged ≥20 years. The outcome was homocysteine; smoking status was the exposure variable and categorized as current, former or never smoker. Generalized linear models were used to examine the associations between smoking status and homocysteine levels, while assessing the impact of oestradiol and cholesterol. Results: After adjusting for age, sex, ethnicity, education and income level, homocysteine levels did differ by smoking status ((current smokers versus never smokers: β: 0.18 CI: 0.00, 0.36), (former smokers: β: 0.10 CI: –0.09, 0.28)). The addition of oestradiol as an interaction term in adjusted models was associated with a 16.6% increase in homocysteine levels when compared to models without the interaction term. Oestradiol but not cholesterol did moderate the association between smoking status and homocysteine levels. Discussion and conclusions: Homocysteine levels did differ across smoking status after adjusting for confounders. Oestradiol did moderate the relationship between homocysteine and smoking status.