LAUSR

Sex (biological) or gender (self-defined) may affect different aspects of Parkinson’s disease (PD), including risk, onset of motor and nonmotor symptoms, diagnosis, disease course, and management. Numerous efforts have been made to study the reasons for these sex and gender differences, yet we are far from fully understanding them. Among the most prominent differences are risk and prevalence of PD; in 2016, worldwide PD prevalence peaked between 85 and 89 years, at 1.7% for men and 1.2% for women. The age-adjusted men/women ratio among patients with PD was 1.4, similar to the ratio reported in 1990, suggesting that the factors differentiating risk in men and women remained similar in the last three decades and possibly longer. Multiple hypotheses regarding the role of sex or gender in modifying the risk for PD have been suggested, including genetic, biological, and environmental exposure differences between men and women. Using the advent of Mendelian randomization (MR), Kusters et al. indirectly examined one of the more common hypotheses: that female sex hormones such as estrogen may reduce the risk for PD in women. This hypothesis is well supported by epidemiological, clinical, in vivo, and in vitro evidence, although conflicting data have also been reported. Proving or refuting this hypothesis, as well as other sexor gender-related hypotheses, is important for understanding the risk factors of PD, its underlying mechanisms, and potential treatments, including potentially different treatments for women and men with PD. MR, used in the study by Kusters and colleagues, is a method that examines potential causation between a potential risk factor and a trait, using genetic data from genome-wide association studies (GWASs). In brief, common genetic variants that are associated with the risk factor in question are being used as an instrument that represents the exposure to the risk factor. Then, the association of this instrument with the trait it is suspected to affect is being tested. This method, using genetic variants that are randomly assigned at conception (Mendel’s laws) and therefore not affected by external factors (hence the name Mendelian randomization, representing the notion that these genetic variants should be normally and randonly distributed in the population tested), is aimed to avoid confounders and reverse causation and other biases that might affect studies that will directly study the effect of the risk factor on the studied trait. However, MR can still have biases because it relies on previous GWASs that may have biases of their own, it may still have residual and undetected pleiotropy despite the tests and procedures to avoid it, and because it considers only the portion of the risk factor in question attributed by genetics. These and other challenges related to MR and its interpretation in neurodegenerative diseases have been recently reviewed. In this specific study by Kusters et al., the risk factors examined for causation in PD were age at menopause and age at menarche, which may represent the length of exposure to estrogen. Using two independent cohorts, stratified by sex, the authors suggest that age at menopause is causally associated with reduced risk for PD in women. Because the age of menarche may also affect the length of exposure to estrogen, it could be expected that it will also be causally associated with PD, yet no association was found. It is possible that the role of estrogen exposure is relevant only later © 2021 International Parkinson and Movement Disorder Society