Time zero alignment in target trial emulation of VMAT2 inhibitors versus anticholinergics for tardive dyskinesia.

Parkinson's disease (PD) is a common neurodegenerative disorder characterized by tremor, rigidity, bradykinesia, and postural instability. Age is the strongest risk factor, but potentially modifiable lifestyle exposures, including smoking and alcohol use, may also influence susceptibility. Alcohol is widely consumed, and clarifying its long-term association with PD risk could identify a feasible target for prevention strategies. Epidemiologic evidence linking alcohol use to PD has been inconsistent. Alcohol use disorder has been associated with higher PD risk in some studies, whereas others have reported null, inverse, or U-shaped associations, with findings varying by beverage type, sex, and drinking pattern. In addition, prospective cohorts have not consistently reproduced inverse associations observed in case–control studies, raising concerns about recall and selection bias and incomplete control of correlated behaviors, particularly smoking. Pooled analyses and meta-analyses further suggest that any overall association may be weak and heterogeneous across populations and study designs. A central methodological challenge is the long latency between alcohol exposure and PD onset or diagnosis. If alcohol-related neurobiologic injury accrues over decades, conventional analyses that average or summarize exposure across the life course may dilute or obscure latency-dependent effects. Mechanistic studies support the plausibility of cumulative harm, implicating oxidative stress, neuroinflammation, and dopaminergic toxicity with sustained alcohol exposure. Under this scenario, analyses that do not align age at initiation with the exposure-to-onset interval may yield apparently protective associations that reflect survivor bias, reverse causation (e.g., changes in drinking during prodromal stages), or exposure misclassification.