Background  and  purpose:  It has been suggested that trimethylamine N-oxide (TMAO) is related to Parkinson's disease (PD) in observational studies. However, the direction of this association is inconsistent. An exploratory Mendelian randomization study was conducted to investigate whether TMAO and its precursors have a causal relationship with PD.

Methods: Summary statistics were obtained for single nucleotide polymorphisms related to circulating levels of TMAO, betaine, carnitine and choline, and the corresponding data for the risk, age at onset and progression of PD from genome-wide association studies. Inverse-variance weighting was used as the primary method for effect estimation. The false discovery rate was applied to the correction of multiple testing. A p value of asso-ciation <0.05 but above the false discovery rate corrected threshold was deemed sug-gestive evidence of a possible association. A range of robust Mendelian randomization methods were used for sensitivity analysis.

Results: Suggestive evidence was observed of an inverse causal effect of TMAO on motor fluctuations (odds ratio [OR] 0.851, 95% confidence interval [CI] 0.731, 0.990, p= 0.037) and carnitine on insomnia (OR 0.817, 95% CI 0.700, 0.954, p=  0.010) and a positive causal effect of betaine on Hoehn–Yahr stage (OR 1.397, 95% CI 1.112, 1.756, p= 0.004), Unified Parkinson's Disease Rating Scale (UPDRS) III score (β=  0.138, 95% CI 0.051, 0.225, p=  0.002), motor fluctuations (OR 1.236, 95% CI 1.011, 1.511, p=  0.039), and cho-line on UPDRS IV (β=  0.106, 95% CI 0.026, 0.185, p=  0.009) and modified Schwab and England Activities of Daily Living Scale score (β= 0.806, 95% CI 0.127, 1.484, p= 0.020).

Conclusions: Our findings provide suggestive evidence that TMAO and its precursors have a causal effect on the progression of PD. Further investigation of the underlying mechanisms is required.