Smoking and weight status: Understanding the relationship of cigarette smoking trajectories through adolescence and weight status in young adulthood in the US

Citation

Patel, Minal & Kaufman, Annette (2016). Smoking and weight status: Understanding the relationship of cigarette smoking trajectories through adolescence and weight status in young adulthood in the US. 2016 Add Health Users Conference. Bethesda, MD.

Abstract

Introduction: Adolescent cigarette smoking has steadily declined since 1999 while childhood obesity rates have tripled since the 1980s. Few studies have looked at the relationship of smoking and weight in youth and young adulthood. This study examines the influence of smoking trajectories beginning in adolescence on weight status in young adulthood. Methods: The study sample was drawn from Add Health Waves I-IV (N=13,361) excluding Wave IV pregnant women and active military. Repeated-measures latent class analyses generated the independent variable of smoking trajectory using smoking status in Waves I-IV and age of initiation. Weight status at Wave IV was primarily measured through self-reported body mass index (BMI) and waist circumference (WC) as a secondary measure. Covariates included self-reported gender, race/ethnicity, BMI, and parental household income (Wave I) and educational attainment (Wave IV). Weighted bivariate analyses assessed the relationship between smoking trajectories and BMI at each wave. Weighted multivariate linear regression models tested the relationship of smoking trajectories and weight status at Wave IV, controlling for demographic covariates. Results: Four distinct smoking trajectories were generated: nonsmokers (44%), early establishers (23%), late establishers (21%), and experimenters (12%). The average BMI at Wave IV was 28.4. In bivariate analyses compared to nonsmokers, early and late establishers had a significantly lower BMI at Wave III, and all smoking trajectories had a significantly lower BMI at Wave IV. In an adjusted multivariate regression model predicting BMI at Wave IV, all smoking trajectories had a significantly lower BMI than nonsmokers [early establishers: β=-1.27, CI: -1.56, -0.98; late establishers: β=-0.84, CI: -1.16, -0.52; & experimenters: β=-0.63, CI: -0.93, -0.34; (p<0.05)]. An adjusted regression model predicting WC showed similar trends [early establishers: β=-2.20, CI:-3.05, -1.36, late establishers: β=-1.08, CI: -2.06, -0.10, & experimenters: β=-1.05, CI: -1.85, -0.24; (p<0.05)] compared to nonsmokers. Discussion: This is one of the first studies to address the relationship of smoking trajectories and BMI at the national level. Clinical implications need to be further examined and account for the relative disease risk of changes in BMI and WC in relation to changes in smoking status. Future studies may consider lifestyle factors related to obesity such as physical activity and eating behaviors.

URL

https://addhealth.cpc.unc.edu/wp-content/uploads/docs/events/2016%20Add%20Health%20Users%20Conference%20Abstracts_2016_06_16.pdf

Reference Type

Conference proceeding

Book Title

2016 Add Health Users Conference

Author(s)

Patel, Minal
Kaufman, Annette

Year Published

2016

City of Publication

Bethesda, MD

Reference ID

6341