Numerous observational studies have demonstrated a dose and duration-dependent reduced risk of PD among smokers. The underlying mechanisms remain misunderstood. In particular it is unclear to what extent these effects are due to nicotine itself versus prodromal disease-related changes in the urge to smoke (reverse causality).

This study involved 502,304 participants from the UK Biobank from 2006 to 2010. Smoking status was collected through self-reported questionnaires. Participants without PD at baseline but who went on to develop the disease in the 15-year follow-up were stratified into yearly subgroups. Logistic regression analyses using controls matched to age, sex, center, and recruitment year were performed for each subgroup. 

2813 participants developed PD during the 15-year follow-up.  Our findings confirmed a decreased risk of incident PD among current smokers (OR: 0.708, 95% CI: 0.588 -0.840). When stratified by time, this effect magnified with shorter prodromal intervals, displaying a significant linear trend  (p-value 0.043). On further analysis, the odds ratio trajectory over time adhered to a binomial distribution, optimally represented by a second-degree polynomial function (p-value 0.034). Employing a Piecewise Linear Regression model, we pinpointed the optimal inflection point at approximately 7 years preceding the PD diagnosis. 

Two distinct risk patterns among smokers were identified: a modest and stable baseline risk reduction, combined with a superimposed additional decrease starting around year -7. Several post-mortem and neuroimaging studies have suggested that substantia nigra degeneration may commence up to 7 years before a clinical diagnosis of PD. These findings suggest that while a baseline lifetime reduced PD risk among smokers is present, an additional component, possibly driven by reverse causality, emerges around seven years before diagnosis.