Differences in Sex-Specific Frequency of Glucocerebrosidase Variant Carriers and Familial Parkinsonism

Abstract

Background:

While men and women with the LRRK2 G2019S variant appear to be equally likely to have Parkinson disease (PD), the sex-distribution among GBA variant carriers with PD, including limited to specific variant severities of GBA, is not well understood. Further, the sex-specific genetic contribution to PD without a known genetic variant is controversial.

Objectives:

To better understand sex differences in genetic contribution to PD, especially sex-specific frequencies among GBA variant carriers with PD (GBA PD) and LRRK2-G2019S variant carriers with PD (LRRK2 PD).

Methods:

We assess differences in the sex-specific frequency in GBA PD, including in subsets of GBA variant severity, LRRK2 PD, and idiopathic PD in an Ashkenazi Jewish cohort with PD. Further, we expand prior work evaluating differences in family history of Parkinsonism.

Results:

Both idiopathic PD (267/420 men, 63.6%) (p<0.001) and GBA PD overall (64/107, 59.8%) (p=0.042) were more likely to be men, while no difference was seen in LRRK2 PD (50/99, 50.5%) and LRRK2/GBA PD (5/10, 50%). However, among GBA PD probands, severe variant carriers were more likely to be women (15/19 women, 79.0%) (p=0.005), while mild variant carriers (44/70 men, 62.9%) (p=0.039) and risk-variant carriers (15/17 men, 88.2%) (p=0.001) were more likely to be men.

Conclusions:

Our study demonstrates that the male-sex predominance present in GBA PD overall was not consistent across GBA variant severities, and a female-sex predominance was present among severe GBA variant carriers. Thus, research and trial designs for PD should consider sex-specific differences, including across GBA variant severities.

Introduction

There is a consistent sex discrepancy in the prevalence of Parkinson disease (PD)^1–4, with disease occurring approximately 1.4 times more frequently in men.⁵ Differences in sex-related frequencies are not well understood, nor are the differential contributions of protective and deleterious factors between men and women.^{6,7,8–15,16,17} These include longer exposure to potentially protective endogenous or exogenous sex hormones in women.^11–15 It has been postulated that men have a higher relative contribution of non-genetic risk factors than women ⁷, including greater exposure to occupational environmental toxins.⁸ Differences may also be attributable to sex-specific genetic factors. ^18,19 However, evaluation of GWAS data did not identify disparate heritability estimates of autosomal dominant loci between men and women with PD²⁰, and data are conflicting regarding whether family history is increased in women with PD.^7,21,22

Differences in sex-related frequency of disease occur in some, but not all genetic forms of PD, and this may depend not only on the gene but also the particular variant.²³ However, the evidence is not consistent across studies, including across separate meta-analyses. For example, the male/female ratio in PD associated with the LRRK2 G2019S variant (LRRK2 PD) consistently approximates 50:50^{21,24–27,15}. A landmark meta-analysis evaluating men and women carrying LRRK2 variants overall, and separately by G2019 and G2385 variants, demonstrated that both sexes shared similar odds for developing PD, lending support to the hypothesis that the genetic load in LRRK2 PD outweighs sex-specific protective or deleterious factors. ²⁸ A follow-up meta-analysis studying sex differences in LRRK2 PD demonstrated a higher prevalence of PD in women carrying LRRK2 variants that was attributed to the G2019S variant but not the G2385 variant ²⁹. The distribution of sex in PD associated with GBA variants (GBA PD), the leading genetic contributor to PD, is controversial. While most have shown a male predominance, or more equal sex distribution, a subsequent meta-analysis demonstrated a higher prevalence of women among GBA PD in North America and Europe, but not in Asia or Oceana. ^30–32,33 It has separately been suggested that the sex distribution in GBA PD may also vary depending on the specific GBA variant under study.²³

To better understand sex differences in genetic contribution to PD and among genetic subgroups, we assess the sex distribution among carriers of GBA variants and/or LRRK2 G2019S, as well as potential gene associated sex differences in family history in a cohort comprising GBA PD, LRRK2 PD, and PD without genetic variants.

Methods

Subjects:

Consecutive probands with PD who reported Ashkenazi Jewish ancestry were invited to a study of genetics of Parkinson research at Mount Sinai Beth Israel/Mount Sinai (New York, USA), including a subset previously reported.⁷ Participants provided written informed consent, and the study was approved by the Institutional Review Board.

All participants were genotyped for LRRK2 G2019S and the 11 most common GBA variants in the Ashkenazi Jewish population (N370S, L444P, 84GG, IVS2+1, V394L, D409H, A456P, R496H, RecNciI, E326K, T369M,) using the Tag-It™ Mutation Detection Kit (Luminex Molecular Diagnostics, Toronto, ON, Canada) according to the manufacturer’s instructions. Multiplex polymerase chain reaction was used to amplify the regions around the target genes. These regions were subjected to allele-specific primer extension, hybridized to specific Luminex® beads via Universal Tags, and sorted on a Luminex® 100 IS platform (Luminex Corporation, Austin, TX, USA). Genotyping was then completed using the Tag-It™Data Analysis Software (Luminex Molecular Diagnostics), and as previously described.^34,35,36

Family history of parkinsonism in first-degree relatives of probands was determined by self-report through family history screen (standardized family history questionnaire or clinical research screen), or by pedigree review.⁷

Data Sharing:

Data requests from qualified investigators for purposes of replicating procedures and results can be made to the corresponding author for a subset of de-identified data, for which consent for sharing was obtained.

Analysis:

Differences in sex distribution, in the cohort overall, and in probands based on their LRRK2 G2019S and GBA variant status, were determined using one-sample tests of equality of proportions. Odds-ratios were estimated using logistic regression for the association between reported sex and genetic status. GBA variants were subcategorized into three groups based on the variant “severity” and the association between the GBA variant and Gaucher disease (GD), as severe (when biallelic causing neuronopathic GD [84GG, IVS2+1, L444P, RecNcil, V394L]), mild (associated with GD type 1 when biallelic mild or together with a severe variant [N370S, R496H]), and risk-variant (not causing Gaucher disease when biallelic, [E326K, T369M]).³⁷ Screening did not detect any carriers of GBA D409H, and GBA A456P was only present among RecNcil carriers. To determine whether family history of parkinsonism in first-degree relatives differed by sex overall, and separately based on LRRK2 and GBA variant status, sex-specific history of parkinsonism in a first-degree relative was analyzed using unadjusted logistic regression to estimate odds-ratios and standard errors, as well as adjusted for proband age at time of family history collection and age at onset of PD. A sensitivity analysis limited to the history of parkinsonism in a parent was also performed. All analyses were performed using Stata statistical software version 16 (StataCorp, TX).

Results:

Demographic, clinical and family history in first-degree relatives was available for 636 probands with PD, including: 420 idiopathic PD, 99 LRRK2 PD, 107 GBA PD, and 10 LRRK2/GBA PD. Among the 636 PD probands, 250 were women (39.3%) compared with 386 men (60.7%). Family history was obtained by family history screen (n=454) and pedigree review (n=182). 609 of the probands had complete information for the adjusted analysis.

Sex distribution

Sex distribution within groups:

Both idiopathic PD (267/420 men, 63.6% men, p<0.001) and GBA PD overall (64/107 men, 59.8%, p=0.042) were more likely to be men. In contrast, there was no difference in sex-specific distribution in LRRK2 PD (50/99 men, 50.5%) (p=0.920) and LRRK2/GBA PD (5/10 men, 50%) (p=1).

However, among GBA PD probands, severe GBA variant carriers were more likely to be women (15/19 women, 79.0% women, p=0.005), while it was the reverse association with mild GBA variant carriers (44/70 men, 62.9% men, p=0.039), and variant GBA carriers (15/17 men, 88.2% men, p=0.001), who were both more likely to be men.

Sex distribution between groups:

A greater proportion of LRRK2 were women compared with idiopathic PD (LRRK2 PD vs. idiopathic PD 49.5% vs. 36.4, p=0.017), and GBA PD (LRRK2 vs GBA 49.5% vs 40.2%, p=0.18), though the latter comparison was not statistically significant (Table 1, Figure 1).

Table 1:

Summary of sex distribution among each major PD group, and separately among GBA-variant severities

Among PD groups
Cohort	N	Women (N, % cohort)	Men (N, % cohort)	p-value
Idiopathic PD	420	153 (36.4%)	267 (63.6%)	<0.001
LRRK2 PD	99	49 (49.5%)	50 (50.5%)	0.920
LRRK2/GBA PD	10	5 (50.0%)	5 (50.0%)	1.00
GBA PD	107	43 (40.2%)	64 (59.8%)	0.042
Among GBA-variant severities
Severe GBA PD	19	15 (79.0%)	4 (21.1%)	0.005
Mild GBA PD	70	26 (37.1%)	44 (62.9%)	0.377
Risk-variant GBA PD	17	2 (11.8%)	15 (88.2%)	0.002

Within group summary of sex distribution in PD overall and in the different genetic groups. Although idiopathic PD and GBA PD are more likely to be men, LRRK2 PD and LRRK2/GBA PD were not. Further, carriers of a severe GBA variant were more likely to be women, unlike carriers of a mild GBA variant, who were more likely to be men. One GBA PD among men was screened as part of an outside study and they were excluded from the variant-specific analysis. Verification of GBA variant status for this participant using our cohort’s methods is ongoing.

Figure 1: Sex Distribution of PD overall and in IPD, LRRK2 and GBA, including by GBA variant. 1A. Comparison of the frequency of LRRK2 and GBA variants (overall) among men and women.

Note the greater proportion of women vs. men with LRRK2 PD (20% vs. 10.8%), similar proportion of GBA PD in both (17.6% vs. 17.3%), and subsequent smaller proportion of idiopathic PD comprising women vs men (62.4% vs. 72.0%) (p=0.057). 1B. Distribution of sex by PD groups. Note the overall greater percentage of men with GBA PD vs. women with GBA-PD (p=0.042), unlike LRRK2 PD (p=0.920), and similar to IPD (p<0.001) 1C. Sex frequency in GBA-PD separated by variant severity, as defined by severe, mild or risk-variant.³⁷ GBA variants include the major variants (both severe and mild variants), which, when bi-allelic decrease activity of the glucocerebrosidase enzyme and cause Gaucher disease (GD), but when monoallelic increase the risk of synucleinopathies including PD, PD with dementia (PD-D), and dementia with Lewy bodies (DLB); and risk-variants, which when biallelic, do not cause Gaucher disease, but are associated with increased risk of PD and DLB.^37,56,73 Similar to 1B where GBA variants overall are associated with greater risk in men, there is significant increase in men with mild variants and risk-variants. In stark contrast, however, there is greater frequency of severe variants in women (p=0.001).

Odds ratios comparing GBA PD to idiopathic PD and LRRK2 PD, and separately comparing categories of GBA variants to each other, idiopathic PD and LRRK2 PD are shown in Table 2.

Table 2:

Regression analysis of sex distribution among each major PD group and separately among GBA-variant severities

Model comparing odds of being a woman among PD groups
Cohort	OR (95% CI)	p-value
LRRK2 PD vs Idiopathic PD	1.71 (1.10–2.66)	0.017
LRRK2 PD vs LRRK2/GBA PD	0.98 (0.27–3.60)	0.976
LRRK2/GBA PD vs Idiopathic PD	1.75 (0.50–6.12)	0.385
GBA PD vs Idiopathic PD	1.17 (0.76–1.81)	0.473
GBA PD vs LRRK2 PD	0.69 (0.39–1.19)	0.180
GBA PD vs LRRK2/GBA PD	0.67 (0.18–2.46)	0.548
Model comparing odds of being a woman including GBA-variant severities
Cohort	OR (95% CI)	p-value
Severe GBA PD vs Idiopathic PD	6.54 (2.13–20.07)	0.001
Severe GBA PD vs LRRK2 PD	3.83 (1.19–12.34)	0.025
Severe GBA PD vs LRRK2/GBA PD	3.75 (0.71–19.71)	0.118
Severe GBA PD vs Mild GBA PD	6.35 (1.90–21.17)	0.003
Severe GBA PD vs Risk-variant GBA PD	28.12 (4.46–177.46)	<0.001
Mild GBA PD vs Idiopathic PD	1.03 (0.61–1.74)	0.909
Mild GBA PD vs LRRK2 PD	1.66 (0.88–3.10)	0.112
Mild GBA PD vs LRRK2/GBA PD	0.59 (0.16–2.24)	0.439
Mild GBA PD vs Risk-variant GBA PD	4.43 (20.94–0.94)	0.060
Risk-variant GBA PD vs Idiopathic PD	0.23 (0.05–1.03)	0.055
Risk-variant GBA PD vs LRRK2 PD	0.14 (0.03–0.92)	0.010
Risk-variant GBA PD vs LRRK2/GBA PD	0.13 (0.02–0.92)	0.040

Two models estimating odds ratios, comparing odds of being a woman between groups. As was expected, LRRK2 PD were more likely to be women than idiopathic PD (OR, SE: 1.71, 1.10–2.66; p=0.017). Although the odds of being a woman were not different between GBA PD and idiopathic PD (OR, SE: 1.17, 0.76–1.81; p=0.473) or GBA PD and LRRK2 PD (OR, SE: 0.69, 0.39–1.19; p=0.180), severe GBA PD were more likely to be women than both idiopathic PD (6.54, 2.13–20.07; p=0.001), LRRK2 PD (3.83, 1.19–12.34; p=0.025), and even mild GBA PD (6.35, 1.90–21.17; p=0.003).

Of note, severe GBA variant carriers were more likely to be women than idiopathic PD (OR (95% CI) = 6.54 (2.13–20.07), p=0.001), LRRK2 PD (3.83 (1.19–12.34), p=0.025), as well as mild GBA variant carriers (6.35 (1.90–21.17), p=0.003) (Table 2).

Family history of parkinsonism

Among idiopathic PD, female probands were more likely to report a family history of parkinsonism in a first-degree relative compared to male probands (22.5% of women vs 15.4% of men) (OR (SE) =1.70 (0.43), p=0.038) (Figure 2). In the genetic forms, there was no difference in the likelihood to report a family history of parkinsonism in a first-degree relative between men and women: among LRRK2 PD (40.8% of women vs. 30.6% of men, OR (SE)=1.23 (0.51), p=0.622), GBA PD (18.6% of women vs. 18.8% of men, OR (SE)=0.99 (0.50), p=0.985), or LRRK2/GBA PD (20.0% of women vs. 40.0% of men, OR (SE)=0.38 (0.54), p=0.497). There was also no difference when separately subdividing GBA PD by the severity of variants. All results were maintained when adjusting for proband age at time of family history collection and age at onset of PD.

Figure 2. Proportion of participants, overall and by PD group, with a family history of PD, stratified by sex.

As the black signifies individuals with parkinsonism in a first-degree relative, the grey area nonetheless includes a subset that likely have a major genetic contribution that is yet to be determined. As anticipated, because of lower penetrance in GBA PD compared to LRRK2 PD, the proportion with first degree family member with parkinsonism is greater in the LRRK2 PD (38.4%) than GBA PD (18.7%) (p=0.002). As a smaller proportion of men have a first-degree family history of parkinsonism, there is a larger contribution of other factors including additional genetic, epigenetic, environmental, hormonal and structurally different factors.^19,68,69 We cannot exclude that there is also an undetermined “protective factor” portion that is also responsible for the smaller percentage of non-first-degree family member cases in the women. We postulate that the differential lipid response to inflammation that may be responsible for sex differences in GBA variant PD may play a role in idiopathic PD as well. ^57–61

Because the penetrance of PD associated variants increases with age³⁰, and siblings are younger and less likely to have reached older ages than the parents, we performed a sensitivity analysis limiting the history of parkinsonism to the parents of the probands. Results from the overall model were maintained such that among idiopathic PD, women were more likely to report a family history of parkinsonism in a parent compared to men (19.6% vs. 11.6%, OR (SE)=1.86 (0.52), p=0.027). There was no difference found between men and women in the likelihood of parkinsonism in a parent among LRRK2 PD (28.6% vs. 32.0%, OR (SE)=0.85 (0.37), p=0.711), GBA PD (11.6% vs. 14.1%, OR (SE)=0.80 (0.48), p=0.715), or LRRK2/GBA PD (20.0% vs. 40.0%, OR (SE)=0.38 (0.54), p=0.497). Lastly, while idiopathic PD were more likely to report a history of parkinsonism in a father (43/61, 70.5%) compared to a mother (19/61, 31.2%) (p=0.002), there was no difference in maternal or paternal risk of parkinsonism among LRRK2 PD or GBA PD (Table 3).

Table 3:

Differences in reported history of paternal and maternal parkinsonism between PD groups

	Parkinsonism in parent N (%)	Parkinsonism in mother N (%)	Parkinsonism in father N (%)	p
Idiopathic PD	61/420 (14.5%)	19/61 (31.2%)	43/61 (70.5%)	0.002
LRRK2 PD	30/99 (30.3%)	16/30 (53.3%)	16/30 (53.3%)	1
GBA PD	14/107 (13.1%)	5/14 (35.7%)	9/14 (64.3%)	0.268
LRRK2 GBA PD	3/10 (30.3%)	2/3 (66.7%)	1/3 (33.3%)	0.531

While idiopathic PD were more likely to report a history of parkinsonism in a father (43/61, 70.5%) compared to a mother (19/61, 31.2%) (p=0.002), there was no difference in maternal or paternal risk of parkinsonism among LRRK2 PD or GBA PD

No difference in age at onset was found between men and women in any PD group. Similarly, proband age at time of family history collection was not different between men and women (p=0.443).

Discussion:

Our study demonstrates that although male-sex predominance was present in GBA PD overall, it was not consistent across GBA variant types and was even reversed for severe variants. This supports that sex differences in GBA PD may be variant dependent.²³ Most surprising was our finding that severe-GBA variant carriers were most likely to be women (15/19), while mild-GBA variant carriers and GBA risk-variant carriers (comprised of E326K and T369M carriers) were more likely to be men (15/17). We also replicated most previous reports demonstrating a greater frequency of men in GBA PD overall when considering all GBA variants together^33,38, although some have also shown that GBA PD may have similar sex ratios between men and women,^30–32 or even greater frequency of women among certain risk variants.³⁹

In order to better address the impact of sex on PD prevalence in genetic subtypes, meta-analyses, including in both Ashkenazi Jewish and non-Ashkenazi Jewish populations, as well as a diversity of mutation types, are necessary. Our study may be divergent from other studies for several reasons. Heterogeneity in cohorts and the frequency of genotypic variants across different ethnic groups^40,41, as well as differences in GBA screening methods, may have contributed to the diverse reporting of sex distribution in GBA PD overall.²⁸ European cohorts are more likely to have a greater proportion of severe variant carriers than, for example, Ashkenazi Jewish cohorts in whom the mild variants are most frequent.^42,43 GBA screening methodologies, notoriously difficult because of a pseudogene, have changed over time, as has our knowledge of variants considered pathogenic.⁴⁴ Case ascertainment decisions, specifically whether variant cases (i.e. GBA E326K and GBA T369M) are included as GBA PD, or excluded as GBA DLB, for example, and clinical characteristics, such as whether subjects with primarily dementia versus PD are screened, may lead to significant cohort differences including sex-related differences. Although we did not have systematic data on dementia in family members, our finding that the male sex predominance varied across GBA variant types suggests that the population studied and GBA variants screened likely contribute to some of the phenotypic heterogeneity. A recent meta-analysis in GBA carriers suggests men are more likely to develop DLB compared to women (OR, [95% CI]: 1.60 [0.93, 2.74]), though the increase was not statistically significant (p=0.09).⁴⁵ To better discern risk-specific sex differences in these sub-groups, inclusion of participants with PD dementia and DLB in future study is warranted. Alternatively, our sex difference may be due to chance and sample size.

The intriguing finding of female predominance among severe GBA carriers will be most important if replicated in further work, but merits discussion as the sex differences in PD are poorly understood. The finding raises questions about possible pathophysiologic differences in expression of the variants.^46,47 Severe variants confer the greatest genetic risk of GBA related PD as they have higher penetrance, as well as earlier age of onset.^37,48 As such, one might postulate there would be fewer male/female differences. Indeed this is the case for LRRK2 PD, which has a higher penetrance than mild variant or risk-variant GBA, approximating 25% by the age of 80, and no sex-related differences in penetrance.⁴⁹ We speculate that the sex ratio reversal we observed for severe-GBA variants could be a consequence of a difference in disease expression, but one that encompasses not only PD, but DLB. The spectrum of GBA related synucleinopathy may be thought of as ranging from DLB with particularly early cortical Lewy body deposition, to Parkinson Disease with dementia (PD-D), with cortical Lewy bodies later in the clinical course, to the mildest, Parkinson Disease, with a lower cortical Lewy body burden.^{23,31,50–54} We propose that men with severe variants may be more likely to develop DLB, with more widespread cortical pathology, while those with mild and risk variants, will more often develop PD-D and PD. In contrast, the women with severe GBA variants may be more likely to develop PD-D and PD, while women with mild and risk-variant variants may be slightly less likely to develop PD than men.

Support for this sex-related differential expression hypothesis includes the observations that GBA variants are also the leading genetic risk factor for DLB, that sex-related differences in GBA variants have been observed both in GBA PD and GBA related DLB²³, and that there is a propensity for dementia and more widespread synuclein pathology in men.⁵⁵ There is also greater risk for dementia in male GBA carriers with parkinsonism, including carriers of the severe GBA variant, L444P.²³ Further, in one pathologic and clinical study of DLB, GBA variants were present in 36% of pure DLB cases, and the great majority, 90%, of GBA variant carriers were men.⁵⁶ It is thus possible that our findings point not to the lack of expression of severe GBA variants in men, but to a varied phenotype of DLB rather than PD. Hence the “missing” severe GBA carriers may not be controls, but rather may have DLB, and the sex-differences described in our study may reflect the preferential sampling of parkinsonism patients (PD and PD-D) from our movement disorders clinic. It is likely that DLB and severe PD-D would instead have presented to the cognitive/behavioral clinic. Had we also evaluated DLB patients, we may have found that the male severe GBA variant carriers clustered in the DLB cohort. However, we did not include DLB in our cohort, and in order to test this hypothesis, future studies are needed that systematically recruit both DLB and PD.

Further, others have reported sex differences that depend on the particular variant, but in a different direction than ours.²³ Straniero et al, found that the GBA T369M carriers in their DLB cohort were more likely to be women, whereas in our sample, probands carrying a GBA T369M variant were more likely to be men (7/7 GBA PD, 100%). This highlights that in order to determine the validity of our hypothesis it will be important not only to look at variant severity, but at the sex distribution among GBA PD with severe, mild, and risk-variant GBA variants.

The etiology of sex differences in GBA PD, and idiopathic PD as well, might be attributable in part to sexual dimorphism in sphingolipid metabolism, inflammation, and microgliosis, as well as the more widely studied differences in distribution of environmental and hormonal factors.^57–61 These factors may play a role in idiopathic PD without GBA variants, or may be exaggerated in those harboring GBA variants. As GBA-related effects may be mediated, at least in part by substrate accumulation of sphingolipids,⁶² we postulate that women may be less vulnerable to the additional sphingolipid accumulation than men.

Further, as glucocerebrosidase activity may be decreased in idiopathic PD⁶³, a group that also has a greater predominance of men, our study raises the question as to whether sex-related differences in sphingolipids and lipid metabolism might play a role not only in GBA related parkinsonism, but in idiopathic PD as well. Estrogen may also modulate lipid rafts and preserve neuronal membrane lipids.⁶⁴ Further, sex has an effect on monocyte gene expression in PD, with greater inflammatory response to lipopolysaccharide in women with PD,⁵⁷ and overall enhanced expression of interleukin 6, TNF-alpha, and interleukin 1 beta in astrocytes in response to lipopolysaccaride stimulation.⁶⁵ Several lines of evidence support that sex-hormones may also mediate the microglia response, and the differential immune response could thus mediate sex-dependent effects of inflammation^61,66,67. Thus, the degree to which lipid changes are associated with specific GBA variants and present in non-GBA variant PD warrants further examination.

As predicted, our study confirms that the male predominance of PD is lost in LRRK2 G2019S carriers compared with idiopathic PD.^21,24,25,27 This is in line with most other studies of major LRRK2 variants, and further supports that the genetic contributions in LRRK2 G2019S carriers outweighs other sex-related factors associated with development of disease.

Finally, in this larger follow-up analysis we continue to demonstrate that a greater proportion of women with idiopathic PD report family history of PD, leading us to postulate that the relative contribution of “extra-genetic” factors is greater among men than women (Figure 2). This is consistent with some,^7,21 but not all prior studies. These “extra-genetic” factors may include epigenetic influences, as well as environmental, hormonal and other contributors.^10,19,68,69

Our cohort has several limitations. Our sample size was smaller than others, and we did not have complete GBA sequencing. However, because we limited our cohort to Ashkenazi Jews, and were able to ascertain the major GBA variants in this population, it is likely that only a few additional GBA cases were missed.⁷⁰ Additionally, because we limited our analysis to the LRRK2 G2019S variant, we cannot evaluate the sex-related effect that might be seen with other LRRK2 variants. Further, for our investigation of family history in first-degree relatives, multiple sources of information (pedigree vs. family history screen) were utilized in order to determine the proband’s family history of parkinsonism. However, in sensitivity analyses adjusting for information source, all significant findings remained. We were also not able to clinically confirm the reported diagnosis of parkinsonism in many first-degree relatives. We also cannot exclude that the greater family history seen in female probands compared to male probands is explained by a sex-specific, differential recall bias of PD in a family member,⁷¹ though whether a sex-specific differential recall of family history of PD exists has not been consistently shown,⁷² and our study supports previous reports of a greater family history in women compared to men.^7,21 Lastly, it is a limitation that our results may not be generalizable to other populations, particularly those with different founder effects.

Conclusion

Our study highlights sex-related differences in PD, suggesting a proportionally greater non-genetic risk in men that appears to explain much of the sex difference, as well as a postulated sex-related expression difference in phenotype relative to GBA variants. As variants in GBA constitute the leading genetic etiology of PD and DLB, and clinical trials addressing GBA have begun, it will be essential to consider sex related differences in trial design, including allocation schema to treatment and placebo arms. Further study, including meta-analysis in Ashkenazi Jewish and non-Ashkenazi Jewish cohorts continues to be warranted.