Clustering of motor and nonmotor traits in leucine-rich repeat kinase 2 G2019S Parkinson's disease nonparkinsonian relatives: A multicenter family study

Objectives: The objective of this study was to determine phenotypic features that differentiate nonparkinsonian first-degree relatives of PD leucine-rich repeat kinase 2 (LRRK2) G2019S multiplex families, regardless of carrier status, from healthy controls because nonparkinsonian individuals in multiplex families seem to share a propensity to present neurological features. Methods: We included nonparkinsonian first-degree relatives of LRRK2 G2019S familial PD cases and unrelated healthy controls participating in established multiplex family LRRK2 cohorts. Study participants underwent neurologic assessment including cognitive screening, olfaction testing, and questionnaires for daytime sleepiness, depression, and anxiety. We used a multiple logistic regression model with backward variable selection, validated with bootstrap resampling, to establish the best combination of motor and nonmotor features that differentiates nonparkinsonian first-degree relatives of LRRK2 G2019S familial PD cases from unrelated healthy controls. Results: We included 142 nonparkinsonian family members and 172 unrelated healthy controls. The combination of past or current symptoms of anxiety (adjusted odds ratio, 4.16; 95% confidence interval, 2.01-8.63), less daytime sleepiness (adjusted odds ratio [1 unit], 0.90; 95% confidence interval, 0.83-0.97], and worse motor UPDRS score (adjusted odds ratio [1 unit], 1.4; 95% confidence interval, 1.20-1.67) distinguished nonparkinsonian family members, regardless of LRRK2 G2019S mutation status, from unrelated healthy controls. The model accuracy was good (area under the curve = 79.3%). Conclusions: A set of motor and nonmotor features distinguishes first-degree relatives of LRRK2 G2019S probands, regardless of mutation status, from unrelated healthy controls. Environmental or non-LRRK2 genetic factors in LRRK2-associated PD may influence penetrance of the LRRK2 G2019S mutation. The relationship of these features to actual PD risk requires longitudinal observation of LRRK2 familial PD cohorts.