F10 Environmental modifiers of huntington’s disease: using propensity scores and outcome analyses to identify causal links

Background Although Huntington’s disease (HD) is a genetically inherited disorder, it is not clear how non-genetic factors such as environment and behavior may contribute to progression. Aims We used a causal inference approach to run simulated pseudo-randomized trials in a large (n=2,914) longitudinal dataset, to study the effects of exposure to non-genetic factors on progression of HD: education, employment status, tobacco use, alcohol use, and use of recreational and therapeutic drugs. Methods Each factor was investigated in isolation while controlling for nineteen others factors (including baseline severity of HD), to guarantee that groups were well balanced at baseline on all potential confounders using propensity score (PS) weights. Outcomes were compared several years later using doubly robust (DR) outcome models. The primary outcome was a composite HD severity measure that included assessments of motor, cognitive and functional abilities. Results We only found significant evidence that antidepressant use was detrimental to HD progression over time, compared with similarly matched individuals who were not taking antidepressants (effect size difference=0.13; 95% confidence interval=0.05,0.21). Conclusions This study is the first to examine the impact of non-genetic factors on HD using causal inference methods. We show that previously reported significant factors – including alcohol and recreational drug use – were no longer causally linked to HD progression after PS weighting. This highlights the important role PS weighting can play in examining non-genetic factors contributing to HD progression, and the caution needed when interpreting findings from studies that do not attempt to use such methods.