Cannabis frequency not linked to appetite or body measures

The effects of cannabis use on gastrointestinal function are unclear. We conducted a preliminary study on whether cannabis use frequency was associated with appetite, eating behavior, and anthropometrics phenotypically and after controlling for genetic and environmental factors via a co-twin control design. We first tested the phenotypic associations between cannabis use frequency (average number of days per week) with appetite, eating behavior, anthropometrics, and diet in simple unadjusted models and then re-estimated those associations in models adjusting for alcohol and nicotine use frequency as covariates. Significant phenotypic effects were next analyzed with multilevel models to test these associations between- and within-twin pairs. Between-family effects capture shared familial variance (additive genetics 'A' and common environment 'C') and within-twin pair effects capture residual variance after accounting for shared components (unique environment 'E'). Participants included 90 adults (40 complete twin pairs) who use cannabis at least weekly. In unadjusted analyses, higher cannabis use frequency was associated with lower appetite (β = -0.199, p = .046), eating fewer meals per day (β = -0.234, p = .007), and eating a less nutritious diet (β = -0.237, p = .025); no phenotypic effects remained significant after adjusting for alcohol and nicotine use frequencies. Multi-level models showed significant between-family effects between higher cannabis use frequency and eating fewer meals per day (β = -0.318, p = .015) and eating a less nutritious diet (β = -0.296, p = .028); for all outcomes the within-twin effects were much smaller in magnitude compared to between-family effects, and non-significant. Our results do not suggest a causal association between cannabis use frequency with eating behavior or anthropometrics.