Neuropsychiatric characteristics of GBA-associated Parkinson disease

Abstract

Mutations in GBA1 are a well-established risk factor for Parkinson disease (PD). GBA-associated PD (GBA-PD) may have a higher burden of nonmotor symptoms than idiopathic PD (IPD). We sought to characterize the relationship between GBA-PD and neuropsychiatric symptoms. Subjects were screened for common GBA1 mutations. GBA-PD (n=31) and non-carrier (IPD; n=55) scores were compared on the Unified Parkinson Disease Rating Scale (UPDRS), Montreal Cognitive Assessment (MoCA), Beck Depression Inventory (BDI), and the State-Trait Anxiety Index (STAI). In univariate comparisons, GBA-PD had a greater prevalence of depression (33.3%) versus IPD (13.2%) (p<0.05). In regression models controlling for age, sex, disease duration, motor disability, and MoCA score, GBA-PD had an increased odds of depression (OR 3.66, 95% CI 1.13–11.8) (p=0.03). Post-hoc analysis stratified by sex showed that, among men, GBA-PD had a higher burden of trait anxiety and depression than IPD; this finding was sustained in multivariate models. Among women, GBA-PD did not confer greater psychiatric morbidity than IPD. These results suggest that GBA1 mutations confer greater risk of neuropsychiatric morbidity in PD, and that sex may affect this association.

1. Introduction

Biallelic mutations in the glucosylceramidase beta 1 (GBA1) gene, which encodes the lysosomal enzyme ß-glucocerebrosidase (GCase), cause Gaucher disease. In the last two decades, both monoallelic and biallelic GBA1 mutations have also been firmly established as strong genetic risk factors for Parkinson disease (PD) in many populations (1). The GCase pathway provides a novel focus for PD therapeutics through multiple potential mechanisms, including GCase activators, chaperones and gene therapy to increase the level and/or activity of GCase (2). As trials of promising candidate agents are imminent, it is important to better understand the phenotype of PD associated with GBA1 mutations (GBA-PD).

The motor features of GBA-PD often resemble typical idiopathic Parkinson disease (IPD); however, converging evidence suggests that GBA-PD may encompass a greater burden of nonmotor symptoms. GBA-PD patients may experience more autonomic dysfunction (3), neuropsychiatric disturbances (3, 4), and cognitive impairment (5), with a greater tendency to develop dementia compared to IPD (6–8). GBA1 mutations have also been strongly associated with dementia with Lewy bodies (DLB) (1, 7, 9), and GBA-PD may have a greater burden of cortical Lewy bodies than IPD (4, 10–12). Among neuropsychiatric features, anxiety and depression, which are highly prevalent in PD overall, may be even more prominent in GBA-PD (3). The presence and severity of these features may be relevant to clinical trial design, especially in randomization schema, but also as potential secondary endpoints. Therefore, we sought to further assess these features in a cohort of GBA-PD subjects.

2. Subjects and Methods

2.1 Subjects

Participants were ascertained from a tertiary care center’s Parkinson disease population who were enrolled in research on the genetics of PD in Ashkenazi Jews (AJ) (13, 14): consecutive AJ patients with PD were invited to participate in research, and all who consented underwent genetic screening and research assessments. Of the 86 participants, 41, all GBA1 non-carriers, were also cross-enrolled in the Michael J. Fox Foundation (MJFF) Ashkenazi Jewish LRRK2 Consortium. The diagnosis of PD was made according to UK PD Society Brain Bank criteria, except that participants were not excluded if they had a family history of PD (13, 15). DNA was screened for nine of the most common GBA1 mutations among Ashkenazim (N370S; L444P; 84GG; IVS2+1G➔A; V394L; del55bp; D409H; R496H; and A456P, a subset of whom were subsequently retested for the RecNciI allele (16)) (as previously described (17)), as well as the LRRK2 G2019S mutation (18, 19). GBA1 carriers were classified as heterozygous, homozygous or compound heterozygous. LRRK2 G2019S carriers, including those who also carried GBA1 mutations, were excluded from the current analysis. Those subjects without GBA1 or LRRK2 mutations were categorized as idiopathic PD (IPD). The study protocol was approved by the institutional review board at Mount Sinai Beth Israel.

2.2 Assessments

In-depth assessment included a structured history related to Parkinson disease (age of onset, disease duration, initial symptoms, present symptoms), complete Unified Parkinson’s Disease Rating Scale (UPDRS) (20), Hoehn and Yahr disease staging (H&Y) (21), Schwab and England (S&E) activities of daily living scale, and formal cognitive and psychiatric assessments. Cognition was assessed by Montreal Cognitive Assessment (MoCA). A cutoff score of <26 was chosen to indicate cognitive impairment (22). Depression was assessed using the Beck Depression Inventory-II (BDI) (23, 24). A cutoff score of ≥14 was chosen for depression (24). The Spielberger State-Trait Anxiety Inventory (form Y) (STAI) (25, 26) was used to measure anxiety. The STAI employs two subscales, one assessing “state” anxiety (asking respondents to answer questions based on how they are feeling at the time of assessment), and one assessing “trait” anxiety (asking respondents to answer questions based on how they feel in general). A cutoff score of ≥55 for the STAI State subscale (STAI-S) was used to define clinical anxiety in this study (27).

Dopaminergic medication doses were converted to levodopa equivalent daily dose (LEDD) (28). Psychiatric medications were recorded by medication class, including selective serotonin reuptake inhibitors, serotonin-norepinephrine reuptake inhibitors, tricyclic antidepressants, nonspecific monoamine oxidase inhibitors, mirtazapine, bupropion, benzodiazepines, typical antipsychotic drugs, atypical antipsychotic drugs other than clozapine or quetiapine, and clozapine or quetiapine.

2.3 Analysis

Demographic variables, UPDRS subscores (parts I–IV), H&Y stage, S&E score, and raw scores from MoCA, BDI and STAI were treated continuously; MoCA, BDI and STAI-state scores were also treated dichotomously based on their cutoff scores. Univariate comparisons were performed using Student’s t-test or Wilcoxon ranksum as appropriate (STATA 12, Statacorp, TX). Four primary models were performed: linear regressions assessing the relationship between 1) continuous BDI score and mutation status (GBA-PD or IPD), 2) continuous STAI trait score and mutation status, and 3) continuous STAI state score and mutation status; and 4) a logistic regression model assessing presence of depression as determined by cutoff score of BDI ≥14. A parallel model for STAI-state score ≥55 was not performed as few individuals met threshold. All models included covariates of age, sex, duration of disease, disease motor severity (UPDRSIII), and MoCA.

3. Results

Subjects included 55 IPD, and 31 GBA1 mutation carriers with PD (GBA-PD), including 29 heterozygotes (24 with N370S mutations, 2 with 84GG, 2 with R496H, and 1 with L444P), and 2 compound heterozygotes (N370S/84GG; N370S/RecNciI). Demographics and group characteristics are shown in Table 1.

Table 1.

Summary of Demographics, Clinical Features and Primary Outcomes

	GBA-PD	IPD	p
n	31	55	—
Age	65.6 ±12.5	68.0 ± 11.4	n.s.
Men (n=54)	61.1 ± 13.0	68.2 ± 11.8	n.s.
Women (n=32)	70.5 ± 10.2	67.5 ± 10.8	n.s.
Women (%)	15 (48.4)	17 (30.9)	n.s.
Age of onset (yrs)	57.0 ± 12.7	59.7 ± 11.4	n.s.
men	52.6 ± 12.4	60.9 ± 11.9	0.03
women	61.7 ± 11.7	56.9 ± 9.8	n.s.
Duration of PD(yrs)	8.6 ± 5.7	8.3 ± 7.3	n.s.
men	8.5 ± 6.0	7.3 ± 5.6	n.s.
women	8.7 ± 5.6	10.6 ± 10.0	n.s.
Current Antidepressant use, n (%)	10 (32.3)	20 (36.4)	n.s.
men	4 (25.0)	12 (31.6)	n.s.
women	6 (40.0)	8 (47.1)	n.s.
Current Antipsychotic use (%)*	2 (7.41)	4 (8.16)	n.s.
men	2 (14.3)	3 (8.8)	n.s.
women	0	1 (6.7)	n.s.
UPDRS I	2.4 ± 2.0	2.4 ± 1.8	n.s.
men	2.9 ± 2.2	2.6 ± 1.9	n.s.
women	2.0 ± 1.7	1.9 ± 1.6	n.s.
UPDRS II	10.1 ± 7.0	9.4 ± 5.6	n.s.
men	10.4 ± 7.0	9.6 ± 6.1	n.s.
women	9.7 ± 7.2	8.9 ± 4.5	n.s.
UPDRS III	16.7 ± 8.7	20.4 ± 13.2	n.s.
men	18.6 ± 9.7	21.0 ± 14.1	n.s.
women	14.5 ± 7.1	19.0 ± 11.5	n.s.
UPDRS IV	2.5 ± 3.0	2.0 ± 2.3	n.s.
men	2.9 ± 3.6	1.8 ± 1.9	n.s.
women	2.1 ± 2.4	2.5 ± 2.9	n.s.
H&Y	2.3 ± 1.1	2.2 ± 0.9	n.s.
men	2.1 ± 1.0	2.2 ± 0.8	n.s.
women	2.4 ± 1.2	2.2 ± 1.0	n.s.
S&E ADL	79.1 ± 22.6	81.1 ± 18.2	n.s.
men	81.0 ± 13.7	80.3 ± 21.4	n.s.
women	77.3 ± 29	82.9 ± 8.3	n.s.
MoCA	24.5 ± 4.6	24.8 ± 4.9	n.s.
men	25.3 ± 3.3	24.3 ± 5.4	n.s.
women	23.6 ± 5.9	25.9 ± 3.6	n.s.
n (%) < 26	14 (51.9)	24 (44.4)	n.s.
men	7 (46.7)	19 (51.4)	n.s.
women	7 (58.3)	5 (29.4)	n.s.
BDI	11.0 ± 7.9	9.1 ± 8.4	n.s.
men	11.7 ± 9.4	6.9 ± 4.6	n.s.
women	10.3 ± 6.0	13.9 ± 12.2	n.s.
n (%) ≥ 14	10 (33.3)	7 (13.2)	<0.05
men	6 (37.5)	2 (5.6)	0.007
women	4 (28.6)	5 (29.4)	n.s.
STAI	—	—	—
State	34.6 ± 13.2	35.9 ± 12.1	n.s.
men	33.0 ± 13.0	32.8 ± 10.1	n.s.
women	36.3 ± 13.6	42.8 ± 13.6	n.s.
n (%) ≥ 55	1 (3.6)	3 (5.5)	n.s.
men	1 (7.14)	0	n.s.
women	0	3 (17.7)	n.s.
Trait	38.0 ± 13.0	35.4 ± 11.9	n.s.
men	39.9 ± 13.9	31.7 ± 10.0	0.04
women	36.1 ± 12.8	43.4 ± 12.1	n.s.

Results expressed as mean +/− SD.

Significant p values (<0.05) are represented; nonsignificant p values (≥0.05) denoted as “n.s.”

^*Note that all subjects taking antipsychotic medications were taking either quetiapine or clozaril.

3.1 Primary analysis

3.1.1 Univariate comparisons

GBA-PD and IPD did not differ by age, sex, age at onset, UPDRS I–IV, H&Y, and S&E. MoCA total and cutoff scores, BDI total score, the total state and trait STAI scores, and the STAI state cutoff score did not differ between groups. However, GBA-PD had higher prevalence of depression (BDI score ≥ 14) (33.3%) vs. IPD (13.2%) (p<0.05) (Table 1).

3.1.2 Multivariate comparisons

In regression models (Tables 2a–d), adjusting for age, sex, PD duration, UPDRS III, and MoCA, GBA-PD had increased odds of depression compared to IPD (OR 3.66, 95% CI 1.13–11.8) (p=0.03) (Table 2b). This finding was sustained in additional models adjusting for antidepressant use and for antipsychotic use (data not shown). There was no difference between GBA-PD and IPD in STAI state, trait, or BDI total scores in models adjusting for age, sex, PD duration, UPDRS III, and MoCA (tables 2a, c, d); these results also did not differ when adjusting for antidepressant use and antipsychotic use (data not shown). There was no difference in STAI state and trait scores specifically, even when adjusting for BDI cutoff score (data not shown). However female sex was a significant covariate in these models (for STAI-state, ß= 8.71, 95% CI = 3.55,13.9, p<0.01; for STAI-trait, ß = 7.76, 95% CI = 2.35, 13.2, p <0.01) (tables 2c and 2d).

Table 2.

Regression Models Describing Association with Depression and Anxiety

a: Linear regression model - BDI continuous score*
n = 77	Coefficient	95% CI	p
Age	0.00	−0.14, 0.14	0.96
Sex (F)	4.36	−0.17, 8.90	0.06
PD duration	0.23	−0.12, 0.58	0.20
UPDRS III	0.03	−0.16, 0.22	0.75
GBA PD	1.35	−2.20, 4.90	0.46
MoCA	−0.23	−0.61, 0.14	0.22

b: Logistic regression model evaluating presence of depression (BDI cutoff score)*
n = 77	OR	95% CI	p
Age	1.03	0.97, 1.09	0.42
Sex (F)	2.81	0.79, 10.0	0.11
PD duration	0.99	0.91, 1.08	0.90
UPDRS III	1.00	0.95, 1.06	0.87
GBA PD	3.66	1.13, 11.8	0.03
MoCA	0.94	0.83, 1.07	0.39

c: Linear regression model – STAI state continuous score*
n = 77	Coefficient	95% CI	p
Age	0.17	−0.07, 0.41	0.16
Sex (F)	8.71	3.55, 13.9	<0.01
PD duration	0.11	−0.30, 0.52	0.60
UPDRS III	−0.17	−0.48, 0.15	0.30
GBA PD	0.22	−5.31, 5.76	0.94
MoCA	−0.17	−0.84, 0.50	0.62

d: Linear regression model – STAI trait continuous score*
n = 76	Coefficient	95% CI	p
Age	0.10	−0.13, 0.33	0.39
Sex (F)	7.76	2.35, 13.2	<0.01
PD duration	0.16	−0.26, 0.58	0.46
UPDRS III	0.03	−0.29, 0.34	0.86
GBA PD	4.87	−0.55, 10.3	0.08
MoCA	0.01	−0.62, 0.64	0.98

^*The significance of the terms did not change when antipsychotic and antidepressant use were included in the models (data not shown)

Neuropsychiatric characteristics of GBA-associated Parkinson disease

3.2 Post-hoc analysis by sex

As female sex was significant in the multivariate models for STAI (see 3.1.2), post hoc analysis with stratification by sex was performed.

3.2.1 Univariate comparisons

Comparisons between IPD and GBA-PD, separated by sex, are shown in Table 1. In comparison to men with IPD, men with GBA-PD had significantly higher rates of depression as determined by BDI cutoff score, and higher STAI-trait scores. Men with GBA-PD also had a younger age of onset than men with IPD. Also of note, when men with GBA-PD were compared to women with GBA-PD, there was no difference in any of the neuropsychiatric measures; however, when men with IPD were compared to women with IPD, women with IPD had higher BDI scores (men 6.8 ± 4.6, women 13.9 ± 12.2) (p=0.03) and a greater frequency of depression as assessed by BDI cutoff score (men 5.6%, n=2, women 29.4%, n=5) (p=0.02). Compared to men with IPD, women with IPD also had higher STAI-state scores (men 32.8 ± 10.1, women 42.8 ± 13.6) (p=0.01), frequency of state anxiety as assessed by STAI-state cutoff (men 0%, women 17.7%, n=3) (p < 0.01), and STAI-trait scores (men 31.7 ± 9.9, women 43.4 ± 12.2) (p < 0.01).

3.2.2 Multivariate models

When limited to men, in models adjusting for age, PD duration, UPDRS III and MoCA, GBA-PD did not differ from IPD in STAI-state anxiety scores; however, GBA-PD had increased trait anxiety scores on average compared with IPD (OR, 95% CI) (8.48, 1.53–15.44) (p=0.02). Men with GBA-PD were also more likely to meet BDI cutoff for depression (OR, 95% CI) (9.22, 1.44–59.19) (p=0.02). These relationships were maintained even when antidepressant use and antipsychotic use were included in the models. Among women, GBA-PD and IPD did not differ in state or trait anxiety scores, or in BDI scores. The differences between men and women with IPD for both STAI (p<0.01) and depression (p<0.05) persisted in multivariate analyses controlling for age, PD duration, UPDRS III and MoCA (data not shown).

4. Discussion

While depression and anxiety are frequent features of PD overall (29–45), our data support that depression is even more common in GBA-PD than IPD (3). Further, we raise questions regarding a potential sex effect, with men with GBA-PD exhibiting higher rates of anxiety and depression than men with IPD.

Neuropsychiatric symptoms are frequent features in PD associated with both monoallelic (3, 46) and biallelic GBA1 mutations (19). Clarifying the psychiatric phenotype of GBA-PD is important not only for clinical care and planning of clinical trials, but also in order to identify features in (motorically) asymptomatic GBA1 mutation carriers that may be associated with development of PD and DLB, or that may suggest early neurodegeneration (47, 48). Clinically significant depression may be present in 2.7% to 89% of PD patients (30) and a population-based study found an incidence rate among PD patients of nearly 26 cases per 1000 person-years, twice the incidence rate in the non-PD population (35). Anxiety risk in PD has been studied less extensively than depression, although many studies support a prevalence of 25–35% (31, 37, 43, 49–51), with some reporting a frequency as high as 50–70% (32, 52). Risk of depression and anxiety in PD may be affected by sex (32–34, 37), age at onset (32, 53), and disease duration (35, 49, 54). The impact of PD-related genes on the risk for psychiatric complications has also been explored, though less extensively. LRRK2 G2019S mutation carriers, for example, appear to have comparable rates of depression to non-carrier PD patients (13, 55), although carriers may have a higher burden of pre-morbid mood disorders (55). Others have suggested that rates of nonmotor symptoms in general, including depression, are comparable between genetic and non-genetic PD (56). An early description of GBA-PD found depression as a presenting symptom of GBA-PD in around 8.5%, comparable to the rate in an IPD comparison cohort (57); however, that study did not use a validated rating scale to detect depression. A subsequent study (3) employed a more standardized approach to nonmotor symptoms, and showed that GBA-PD had significantly higher levels of anxiety, depression, apathy, and sleep disruption than IPD. Another study, however, did not demonstrate a significant difference in BDI scores between GBA-PD and IPD among early-onset Parkinson disease subjects (5).

Our observed increased frequency of depression in GBA-PD, as measured by a categorical cutoff in the Beck Depression Inventory, supports clinical observations and a study by Brockman et al. (3) that GBA-PD subjects do have higher rates of psychiatric morbidity; however, unlike the study from Brockmann et al., in which continuous BDI scores were significantly worse among GBA-PD, in our study the continuous BDI scores only trended toward being worse in the GBA-PD group compared to IPD. In addition, unlike the former study, we did not observe worse cognition among GBA-PD overall, as assessed by MoCA. Of interest, 70% of GBA-PD subjects assessed in the Brockman study had GBA1 mutations classified as severe compared with only 13% in our study. This discrepancy might explain the cognition differences and difference in raw BDI scores between the two studies, as severity of mutations has been associated with worse disease, in particular earlier age of onset (57). Further, in a more recent study (58), there was no baseline difference in MoCA scores between GBA-PD and an IPD group matched for sex and disease duration, although there was a more rapid decline in MoCA performance over time among GBA-PD.

While we did not find an overall increase in anxiety in GBA1 carriers, female sex was significant in the multivariate models of anxiety, prompting stratified analyses by sex. In studies of IPD, female sex is sometimes reported as a risk factor for anxiety (32, 37, 43) and depression (33, 34, 36); in addition, among those with early, drug-naïve Parkinson disease, women may demonstrate higher rates of anxiety than men (59). It is of interest, then, that in our post-hoc analysis separating GBA-PD and IPD by sex, we found that men with GBA-PD had a greater burden of trait-anxiety and of depression than men with IPD, but that there was no difference between women with GBA-PD and women with IPD. Furthermore, in univariate analysis, men and women with GBA-PD had comparable rates of neuropsychiatric symptoms, while in the IPD group, women had a higher burden of both anxiety and depression, which is more consistent with prior literature on anxiety and depression in PD. While this finding might be attributable to stochastic differences because of sample size, it may also be due to a particular GBA1 effect in men, with mutations relatively increasing their risk of neuropsychiatric complications in PD. Alternatively, GBA1 mutations may somehow equalize the risk for anxiety and depression between men and women, eliminating whatever factors exist in IPD that either increase the risk for neuropsychiatric symptoms in women, or decrease it in men. It is of interest, too, that in our study, men with GBA-PD had a significantly earlier age of disease onset compared to men with IPD (52 years compared to 61 years); other studies(57, 60, 61) have identified an earlier age of onset for GBA-PD overall compared to IPD. Although these gender differences need to be re-assessed in a larger sample, our findings raise the question of whether and how sex may modify expression of PD among GBA1 mutation carriers (and, more broadly, among non-genetic IPD subjects). Male sex may increase the risk of developing PD among patients with Gaucher disease type 1 (62), although larger samples need to be studied. One recent study (63) found that, among those with dementia with Lewy bodies (DLB) both with and without GBA1 mutations, men were overwhelmingly represented in GBA-associated DLB, by a male:female ratio of 9:1; the ratio in DLB subjects without GBA1 mutations was 1.1:1. Also, whereas the sex distribution for many inherited parkinsonisms, including LRRK2 and parkin, does not demonstrate the typical increased frequency in men (56, 64), screens of GBA1 samples suggest that GBA-PD may have a more typical 60:40 male:female predominance (61). Whether sex may otherwise have an impact on the risk of PD, or on motor or nonmotor manifestations, among heterozygous GBA1 carriers remains an open area for investigation at the clinical and biological level.

The biological basis of increased neuropsychiatric symptoms among GBA-PD remains speculative. Anxiety and depression in PD may be related to Lewy body accumulation in brainstem structures such as the locus coeruleus and midbrain raphe (65, 66). GBA-PD subjects have greater hypoechogenicity in the midbrain raphe on transcranial sonography (3). This has been associated with depression in PD (67), and may reflect involvement of non-dopaminergic neurotransmitter systems that modulate mood. GBA-PD may have a greater burden of cortical Lewy bodies than IPD (4, 10), and anxiety and depression are frequent features of dementia with Lewy bodies (68–71), a disease that is also significantly more common in heterozygous GBA1 mutation carriers (7, 9). It is unclear whether other, subcortical brain structures, such as the locus coeruleus or midbrain raphe, may be similarly more vulnerable to Lewy body accumulation or neurodegeneration in GBA-PD than in IPD.

Our study has several potential limitations. While psychometric instruments to assess depression in PD, including the BDI, have been widely validated, there is debate about the choice of instruments to assess anxiety. The STAI Form Y was updated from the original Form X to replace items that overlapped between depression and anxiety (25), but Form Y has not been specifically validated in PD; this may have influenced our ability to detect anxiety among participants. Despite this limitation, Movement Disorder Society expert consensus guidelines have supported the use of both the STAI (26) and the BDI (24). A strength of the STAI is its assessment of both “state” and “trait” anxiety, reflecting the dynamic nature of anxiety, which is particularly relevant to PD, although our cross-sectional design did not fully capitalize on this strength. A recently-described, PD-specific anxiety rating scale, the Parkinson Anxiety Scale, ultimately may be a superior instrument for assessing anxiety in PD, both clinically and in research settings (72).

Another limitation relates to our participants, who were recruited from among Ashkenazi Jewish patients in a tertiary referral center; thus, our findings may not be representative of PD patients in the general patient population. While we did not match for age and sex, we did adjust for these factors in the regression models, which allowed detection of a potential sex effect. We also ascertained GBA1 mutation carriers by screening for common mutations in the Ashkenazi Jewish population, rather than by identifying mutations through complete gene sequencing; thus the number of GBA1 mutation carriers in our sample may have been underestimated. On the other hand, if as a result of screening rather than sequencing some of the carriers were misclassified as non-carriers, this would likely lead to an underestimation of the difference in neuropsychiatric symptoms among those classified as GBA-PD; as a result, the statistically significant differences that we did identify would thus be more likely to reflect a real gene effect.

Finally, our sample is relatively small, although it is comparable to or larger than the sample size in other studies assessing neuropsychiatric symptomatology in GBA-PD. As a result of our small sample size, our study was insufficiently powered to detect differences in all measures, so there may be additional differences between GBA-PD and IPD that we could not demonstrate with statistical significance. For example, several studies have identified a greater burden of cognitive impairment in GBA-PD compared to IPD (5); we could not demonstrate this. We would have required larger numbers to obtain statistically significant results for cognition: specifically, we would have required 300 GBA-PD subjects and 300 IPD subjects to achieve sufficient power to demonstrate a significant difference (p=0.05) in the observed frequency of cognitive impairment as defined by MoCA score < 26. Thus we do not assert that there is no difference in cognition between GBA-PD and IPD, but that larger numbers are required to confirm such a difference. The fact that we had sufficient power to detect significant differences in anxiety and depression suggests that there is likely a more robust difference in these features. Assessment in larger cohorts with PD specific scales that perform well in cognitively impaired individuals is warranted.

Mutated GCase provides an exceptional target for disease modifying therapies in GBA-PD (2). Designing clinical trials will require a nuanced understanding of the differences between GBA-PD and IPD and of the clinical and biological impact of GBA1 mutations on PD expression in different populations. Nonmotor symptom burden could reflect disease severity and indicate relative susceptibility of PD to disease-modifying therapies. Proper randomization or stratification schema in future trials might include nonmotor symptoms. Further, as discussions regarding pre-symptomatic trials are raised, it will be important to consider biomarkers in GBA-PD that may be present in carriers prior to phenoconversion. A recent study explored prodromal symptoms of PD, including depression, in GBA1 mutation carriers without PD; heterozygote carriers demonstrated a significant worsening of BDI scores over the 2-year study period compared to controls (47), suggesting that depression could be a marker of premotor disease progression, and thus a potential outcome measure in the assessment of presymptomatic, disease modifying therapies. Our data support the possibility that anxiety might also be worth exploring as a marker of disease severity or treatment response, in the motor or premotor phase of GBA-associated PD.

Highlights.

• GBA1 mutations confer greater susceptibility to depression in those with Parkinson disease.

• This relationship persists when controlled for age, sex, disease duration, motor disability, and cognitive performance.

• Men with Parkinson disease due to GBA1 mutations may be particularly susceptible to a greater burden of anxiety and depression compared to men with Parkinson disease without GBA1 mutations.