Introduction Parkinson’s disease (PD) is the second most common neurodegenerative disorder. Metal exposure has been suggested as a possible environmental risk factor by many epidemiological studies, but results have been inconsistent. Additionally, existing reviews on metal exposure and PD risk lack careful screening for study design and quality, especially on the exposure assessment.

Objectives We aimed to synthesize the literature on metal exposure and PD risk by examining the quality of the overall study and exposure assessment method.

Methods We conducted a systematic review on observational studies of metal levels from biological matrices, and dietary and occupational/environmental sources among PD patients and controls. We searched the PubMed/MEDLINE, EMBASE and Cochrane databases up to July 2020. Metal species included manganese, iron, copper, lead, mercury, aluminum, calcium, selenium, zinc, magnesium, cadmium, chromium and nickel, and the outcome was idiopathic PD. We applied an adapted Newcastle-Ottawa Scale (NOS) and a previously established exposure assessment rating to evaluate each individual study. We then performed meta-analyses with random-effects model.

Results 80 case-control studies were included, of which 69 were graded as low or moderate quality. The majority of case-control studies were hospital-based and applied biomonitoring approaches to quantify metal levels after disease diagnosis. Studies on copper, iron, manganese and zinc were most prevalent. Meta-analyses showed no significant PD risk for these metals and heterogeneity among studies was substantial. Furthermore, 5 cohort studies were retained, but the population source, metal exposure and follow-up period were heterogeneous.

Conclusion The level of evidence on metal exposure and PD risk is limited and no consensus can be drawn from the literature. Reverse causality cannot be ruled out by existing biomonitoring studies. Studies assessing metal levels before disease onset are needed to improve our understanding of the role of metals in the etiology of PD.