Abstract
Although marijuana use is commonly associated with increased appetite and the likelihood of weight gain, research findings in this area are mixed. Most studies, however, report cross-sectional associations and rarely control for such important predictors as physical activity, socioeconomic status, and alcohol and other drug use.
Using data from Waves III (N = 13,038) and IV (N = 13,972) of the National Longitudinal Survey of Adolescent Health, we estimate fixed-effects models to more rigorously study the relationships between marijuana use and body mass index over time. Our analyses include numerous sensitivity tests using alternative estimation techniques and at Wave IV we investigate the relationship between marijuana use and an alternative measure of body size (waist circumference).
Results show that daily female marijuana users have a BMI that is approximately 3.1% (p<0.01) lower than that of non-users, whereas daily male users have a BMI that is approximately 2.7% (p<0.01) lower than that of non-users.
The present study indicates a negative association between marijuana use and BMI. Uncovering a negative association between marijuana use and weight status is a valuable contribution to the literature, as this result contradicts those from some previous studies, which were unable to address time-invariant unobserved heterogeneity.
Future theory-based research is necessary to explore the metabolic and behavioral pathways underlying the negative associations between marijuana use and BMI. A broader understanding of such mechanisms along with causal estimates will be most helpful to both policymakers and clinicians.
