Pharmacoepidemiology
Statin Use and Dementia: A comparison of findings from different causal questions in UK Biobank Neal Jawadekar* Neal Jawadekar Adina Zeki Al Hazzouri
Prior studies conducted on the relationship between statins and dementia have been largely mixed. While randomized trials have failed to find a meaningful effect of statins on dementia risk, multiple longitudinal studies suggest statins may help protect against dementia. Amidst these findings, competing risk of death (i.e., loss-to-follow-up or censoring due to death) remains a pervasive problem across longitudinal dementia research which may bias results when left unaddressed. In this paper, we utilize the UK Biobank cohort to estimate the effects of statin initiation on dementia. We estimate the controlled direct effect (i.e., the direct effect of statins on dementia which does not go through death) as well as the total effect (capturing both the direct and indirect effects). We performed this analysis on a cohort of individuals aged 55+ by the end of UKB enrollment period in 2010, without a history of dementia, and no statin use in the 6 months prior to baseline. We estimated a controlled direct effect (absolute risk difference) of statin initiation (vs. no statin initiation) on 10-year dementia risk of +0.58% (+0.13% , +1.04%), and a total effect +0.39% (+0.23%, +0.55%). Even after accounting for the competing risk of death, our findings suggest that statin initiators have an increased risk of dementia.
