A Cross‐Sectional Comprehensive Assessment of the Profile and Burden of Non‐motor Symptoms in Relation to Motor Phenotype in the Nigeria Parkinson Disease Registry Cohort

ABSTRACT

Background

Data on non‐motor symptoms (NMS) in black Africans with Parkinson's disease (PD) are sparse.

Objective

To describe the profile of NMS in the Nigeria PD Registry (NPDR) cohort and explore the relationship between NMS and PD motor phenotype.

Methods

We conducted a cross‐sectional study of the frequency and burden of NMS, based on the non‐motor symptoms scale (NMSS) and the Chaudhuri method respectively in our cohort. Baseline demographics, disease characteristics (Hoehn and Yahr stage, MDS‐UPDRS total score and Part III motor score), motor phenotype (based on Stebbin et al's algorithm), and levodopa equivalent daily dose (LEDD) were documented.

Results

Data are presented for 825 PD whose mean age at study was 63.7 ± 10.1 years, female sex—221 [26.8%] while median PD duration was 36 months. PD phenotypes included tremor‐dominant 466 (56.5%), postural instability and gait disorder (PIGD) 259 (31.4%), and indeterminate 100 (12.1%). 82.6% were on treatment (median LEDD of 500 mg/24 hours). 804 (97.5%) endorsed at least 1 NMS. The median NMSS score was 26.0 while subscores for urinary and sexual function domains were significantly higher in males (P < 0.05). PIGD‐PD had more frequent NMS and higher frequency of severe/very severe NMSS burden (P = 0.000 for both). Nocturia and fatigue were the most prevalent NMS overall and across motor subtypes. PIGD phenotype and total UPDRS scores were the independent determinants of NMSS scores (P = 0.000).

Conclusion

The profile and burden of NMS, and association with motor subtype in our black African cohort is largely similar to descriptions from other populations.

Parkinson's disease (PD) is recognized as a neurodegenerative disease characterized by distinctive core motor features and an extensive array of non‐motor features. ¹ Olfactory dysfunction, mood disorders (e.g., anxiety and depression), cognitive impairment and dementia, autonomic dysfunction, sleep disorders, fatigue and pain comprise the spectrum of well‐documented non‐motor symptoms (NMS). NMS are consequential in PD and can antedate or overlap with the emergence and progression of motor features, and also fluctuate with dopaminergic therapy. ² , ³ NMS contribute significantly to overall burden and severity of PD and are predictors of poor quality of life, risk of institutionalization, and are predominant problems with advancing disease. ⁴ , ⁵

The prevalence of NMS in published literature differs, reaching 99% in some reports, and is a reflection of the variability in study methodologies, population studied, and the specific NMS. ⁶ The identification, diagnosis and treatment of NMS is still an unmet need in the management of PD globally and particularly in underserved populations. Certain disease characteristics or phenotypes may portend a higher NMS burden. Disease duration, motor severity, motor subtype, and ethnic or genetic variability have been implicated as factors that may influence the observed dissimilarities in occurrence of NMS in PD. ⁶ , ⁷ Furthermore, differences in pathobiology including putative routes of spread and predilection of pathology and preferential neurochemical abnormalities have been postulated to define clinically recognizable “subtypes” of NMS characterized by a predominance of specific NMS. ⁸ , ⁹ There is a dearth of literature on NMS in persons of black African ancestry with PD. Our study objective was to describe the profile of NMS and explore the association of NMS with motor subtype in Nigerians with PD from a nationally representative and moderately sized cohort participating in the Nigeria Parkinson Disease Registry. ¹⁰

Methods

Design and Setting

This was a cross‐sectional, descriptive study conducted through the Nigeria PD Research network. The network hosts the Nigeria Parkinson Disease Registry (NPDR) and is comprised of participating neurologists at 32 tertiary hospitals across the 6 regions of the country. The establishment of this network and registry has been described previously. ¹⁰

Sampling and Participants

The study protocol was approved by institutional Health Research Ethics Committees (HREC) and the National Health Research Ethics Committee (NHREC). Consecutively presenting and consenting persons with PD were recruited into the registry and a comprehensive linked database beginning in November 2016. PD cases fulfilled all United Kingdom Parkinson's Disease Society (UKPDS) Brain Bank criteria (excluding the exception of the presence of a first degree relative with PD) and the Movement Disorder Society (MDS) clinical diagnostic criteria for PD. ¹¹ , ¹² Demographic information and clinical disease characteristics were obtained using a standardized proforma. All participating neurologists had prior training in the use of the MDS‐Unified Parkinson's Disease Rating Scale (UPDRS) and study related scales.

Measurements

PD severity and motor severity were assessed using the MDS‐UPDRS total and Part III scores respectively. Disability was evaluated using the Hoehn and Yahr (H&Y) scale, and further characterized as mild (H&Y 1–2), moderate (H&Y 3) and severe (H&Y 4–5). ¹³ PD was categorized into motor sub‐types (tremor‐dominant (TD), postural instability and gait difficulty (PIGD) and indeterminate (I)) using the method described by Stebbins et al. ¹⁴ The frequency and severity of NMS was evaluated with the non‐motor symptom scale (NMSS). ¹⁵ The NMS burden was graded (using the method proposed by Chaudhuri et al, which is based on the NMSS scores) into levels: 0 (none, score 0); 1 (mild, score1–20); 2 (moderate, score 21–40); 3 (severe, score 41–70); and 4 (very severe, score ≥ 71). ¹⁶ PD was also categorized by treatment status (dopaminergic therapy vs. dopaminergic therapy‐naïve). For those on treatment, levodopa equivalent daily dose (LEDD) in 24 hours was calculated. ¹⁷

Data Management and Statistical Analyses

Study data were managed using Research Electronic Data Capture (REDCap) and analyzed using IBM Statistical Package for the Social Sciences (SPSS) version 22 (IBM Corp., Armonk, NY). ¹⁸

Demographic and clinical variables were analyzed using parametric and nonparametric tests as appropriate. Quantitative data presented are expressed as mean ± SD (normally distributed data) or median and interquartile range (IQR) (non‐normally distributed data). Categorical variables are expressed as numbers and percentages. Inter‐group comparison of summary data for continuous variables was conducted with Analysis of Variance (ANOVA), while categorical variables were compared using Chi‐square (Χ²) test. Inter‐group analyses of non‐normally distributed data (disease duration, LEDD and NMSS) were conducted using the Mann–Whitney U and Kruskall‐Wallis tests.

Post‐hoc tests were conducted using the Tukey test for comparison of mean age among PD motor phenotypes, the Z test for pairwise testing of significant intergroup comparison of NMS burden and motor phenotype and Tamhane's T2 for comparison of mean non‐motor domain scores among PD phenotypes. We conducted correlation analyses using Spearman's rho to explore the association between the NMSS burden (total scores) and PD‐related characteristics. To explore the determinants of NMS severity, we used multiple linear regression analysis. P value was conventionally presumed significant when <0.05.

Results

Cohort Characteristics

A total of 839 persons with PD were enrolled in the Nigerian PD registry at the time of the data analyses for this report (October, 2020). Fourteen ¹⁴ records were excluded from this analysis on account of incompleteness of assessments (missing data), leaving a total of 825 PD records. The mean age of PD at study was 63.7 ± 10.1 years (median 65; range of 23–95 years). Overall, 682 (82.6%) persons with PD were on dopaminergic therapy. There was no significant difference in disease severity although MDS‐UPDRS total score was non‐significantly higher in the subgroup on dopaminergic treatment (64.3 ± 30.6 vs. 61.6 ± 29.6; P = 0.34). Table 1 displays the baseline demographic, clinical and treatment characteristics of the participants.

TABLE 1.

Demographic and clinical characteristics of PD

	All Participants	Males	Females	P value
	n = 825	n = 604	n = 221
Mean age at study (yr)	63.7 ± 10.1	63.5 ± 10.2	64.8 ± 9.6	0.25
Mean age at onset of PD (yr)	59.6 ± 10.4	59.3 ± 10.3	60.5 ± 10.5	0.83
Disease duration (mo)				0.75*
Mean ± SD	47.5 ± 45.8	47.2 ± 47.8	48.4 ± 53.3
Median (IQR)	36.0 (−0.0)	36.0 (38.0)	36.0 (44.0)
Young‐onset PD (≤ 50 yr) n (%)	39 (4.7)	32 (5.3)	7 (3.2)	0.20
Dopaminergic Treatment
n (%)	682 (82.6)	500	182	0.89
Mean LEDD (mg)	530.8 ± 253.9	542.6 ± 255.0	498.5 ± 248.6	0.07*
Median LEDD (IQR)	500 (375)	375 (500)	375 (500)
Hoehn & Yahr Stage
Median (IQR)	2.0 (1.0)	2.0 (1.0)	2.0 (1.0)	0.41*
MDS‐UPDRS total score
Mean ± SD	63.8 ± 30.4	63.7 ± 30.4	64.1 ± 30.5	0.77
Median (IQR)	60.0 (39.5)	61 (38.0)	60.0 (42.5)
MDS‐UPDRS Part III (motor) sub‐score
Mean ± SD	40.5 ± 19.3	40.3 ± 19.1	41.0 ± 19.8	0.65
NMSS score (total)
Mean	34.9 ± 33.5	35.4 ± 33.5	33.6 ± 33.5	0.40*
Median (IQR)	26.0 (33.0)	26.0 (33.8)	26.0 (31.5)
NMS burden
None	21 (2.5)	15 (2.5)	6 (2.7)
Mild	336 (40.7)	244 (40.4)	92 (41.6)
Moderate	221 (26.8)	159 (26.3)	62 (28.1)
Severe	153 (18.5)	114 (18.9)	39 (17.6)	0.92**
Very severe	94 (11.4)	72 (11.9)	22 (10.0)

^* Non‐parametric test (Mann–Whitney U) for the P values.

^** X² test used for comparison.

LEDD, Levodopa Equivalent Daily Dose; MDS‐UPDRS, International Parkinson and Movement Disorders Society Unified Parkinson Disease Rating Scale; NMSS, Non‐motor Symptom Scale; IQR, Inter‐quartile range.

Non‐motor Symptom Profile

Overall, 804 (97.5%) persons with PD endorsed at least 1 NMS. The 5 most frequent NMS were nocturia (59.4%), fatigue (55.6%), constipation (46.7%), memory issues (45.9%) and daytime somnolence (41.5%). There was no sex difference in the total NMSS scores and NMS burden overall. The NMS domain scores are presented in Table S1 and show significantly higher scores by sex (male) only for domains 7 (urinary) and 8 (sexual function) (P = 0.01 and 0.00 respectively). The NMSS score for PD on dopaminergic treatment was significantly higher (36.1 ± 38.8) than the score for dopaminergic treatment‐naïve PD (29.1 ± 25.7; median scores of 26 and 20 respectively (P = 0.02).

Motor Phenotype and Non‐motor Symptom Profile

The TD motor subtype was present in 56.5% (466 of 825), while the PIGD and Indeterminate phenotypes occurred in 31.4% (259/825) and 12.1% (100/825) respectively. Age at onset and age at study were higher in the PIGD phenotype, who also had worse disability (H&Y) and worse PD severity and motor severity (MDS‐UPDRS total and Part III scores) (P < 0.05, Table 2). Post‐hoc analysis (Tukey) revealed that the significant mean differences in age (at onset and at study) were between indeterminate and TD and between indeterminate and PIGD subtypes only (P = 0.025 and 0.001).

TABLE 2.

Comparison of demographic and clinical characteristics based on PD motor phenotype

Variable	TD	PIGD	Indeterminate	P value
	N = 466	N = 259	N = 100
Males (%)	342 (73.4)	184 (71.0)	78 (78.0)	0.41
Mean age at study (yr)	62.7 ± 9.2	65.5 ± 10.6	64.9 ± 9.3	0.001*
Mean age at onset (yr)	58.8 ± 10.0	60.8 ± 11.3	60.7 ± 9.2	0.018*
Young onset PD, n (%)	23 (4.9)	15 (5.8)	1 (1.0)	0.15
H&Y (median, IQR)	2.0, 0.0	3.0, 1.0	2.0, 1.0	0.000**
Disease duration (mo)
Mean ± SD	45.2 ± 47.0	52.9 ± 45.8	44.5 ± 38.9	0.07**
Median (IQR)	(36.0, 48.0)	(36.0, 48.0)	(34.5, 45.8)
Dopaminergic treatment
Mean LEDD/24 hr (mg)	510.6 ± 234.4	563.9 ± 289.8	532.5 ± 225.2	0.1**
Median LEDD	375 (500)	375 (500)	375 (500)
MDS‐UPDRS scores
Mean Total score ± SD	58.0 ± 26.3	73.7 ± 34.5	65.2 ± 30.5	0.000**
Median, IQR	(56.5, 37.0)	(69.0, 46)	(58.5, 42.5)
Mean Part III (Motor) score ± SD	38.3 ± 17.7	44.1 ± 21.2	40.9 ± 19.9	0.02**
Median, IQR	(38.0, 27.3)	(41.0, 26.0)	(37.5, 28.5)
NMS Absent (n, %)	17 (3.6)	3 (1.2%)	1 (1)	0.07
NMSS total score
Mean ± SD	28.9 ± 27.2	44.7 ± 39.7	37.6 ± 36.4
Median (IQR)	20.0 (27.3)	36.0 (43.0)	27.5 (37)	0.000**
NMSS burden, n (%)
None	17 (3.6)	3 (1.2)	1 (1)
Mild	217 (46.6)	79 (30.5)	40 (40)
Moderate	122 (26.2)	71 (77.4)	28 (28)	0.000***
Severe	74 (15.9)	62 (23.9)	17 (17)
Very severe	36 (7.7)	44 (17)	14 (14)

^* Tukey's post‐hoc test carried out as there was homogeneity of variance (equal variances assumed).

^** Kruskal‐Wallis test used for comparison as data was not normally distributed.

^*** Post‐hoc analysis carried out using Z‐test.

H & Y, Hoehn and Yahr; SD, standard deviation; IQR, interquartile range; NMS, non‐moptor symptoms; NMSS, non‐motor symptoms scale.

The distribution of NMS burden varied significantly among the motor phenotypes (P < 0.05, Table 2). However, post‐hoc testing using the Z‐test revealed that for a “mild” burden of NMS, proportions differed significantly between indeterminate and PIGD subtypes and indeterminate and TD subtypes but not between PIGD and TD subtypes. Within the “very severe” NMS burden category, there was a significant difference in frequency between TD and indeterminate and TD and PIGD subtypes but not between PIGD and indeterminate subtypes (X²–35.0, P = 0.000).

The domain specific scores and the frequency of each NMS were compared overall and by motor phenotype in Table S1 and Table 3. Total NMSS score was significantly poorer (higher) in the PIGD phenotype (followed by the indeterminate and TD phenotypes) (P = 0.000). Only domain 9 sub‐scores (miscellaneous domain) did not vary significantly among the 3 phenotypes (Table S1). Post‐hoc analyses for domains 1 to 8 however revealed there were no differences between the mean NMSS scores of the phenotypes in the following domains—perceptual problems/hallucination, attention/memory and the sexual function domains (Table S1). A significantly higher proportion of PD with PIGD subtype reported the presence of specific non‐motor symptoms with the exception of insomnia, loss of interest in surroundings, difficulty concentrating, memory, forgetfulness, altered sex, taste and smell and weight change (Table 3). Post‐hoc analyses (Z test) confirmed the significant differences in the distribution of NMS occurred among the PIGD and TD phenotypes only for 15 of 22 NMS (Table 3).

TABLE 3.

Frequency of non‐motor symptoms by motor phenotype

	All	TD	PIGD	Indeterminate
	N = 825	N = 466	N = 259	N = 100
	n (%)	n (%)	n (%)	n (%)	P value
Cardiovascular
1. Light‐headedness	310 (37.6)	145 (31.1)	121 (46.7)	44 (43.6)	0.00*
2. Falls	76 (9.2)	22 (4.7)	45 (17.4)	9 (9.0)	0.00* , **
Sleep/fatigue
3. Daytime somnolence	342 (41.5)	171 (36.7)	129 (49.8)	42 (42.0)	0.003* , **
4. Fatigue	460 (55.6)	238 (51.1)	165 (63.7)	57 (57.0)	0.004* , **
5. Insomnia	297 (36.0)	156 (33.5)	100 (38.6)	41 (41.0)	0.21
6. RLS	130 (15.8)	62 (13.3)	54 (20.8)	14 (14.0)	0.03* , **
Mood/cognition
7. Loss Interest surroundings	220 (26.7)	98 (37.8)	98 (37.8)	24 (24.0)	0.00*
8. Loss of interest new activity	267 (32.4)	121 (26.0)	110 (42.5)	36 (36.0)	0.00* , **
9. Nervousness	226 (27.4)	101 (21.7)	96 (37.1)	29 (29.0)	0.00* , **
10. Depression	303 (36.7)	136 (29.2)	131 (50.6)	36 (36.0)	0.00*
11. Flat mood	253 (30.7)	113 (24.2)	110 (42.5)	30 (30.0)	0.00* , **
12. Difficulty experiencing pleasure	291 (35.3)	139 (29.8)	113 (43.6)	39 (39.0)	0.001* , **
Perceptual problems/hallucinations
13. Hallucinations	97 (11.8)	37 (7.9)	46 (17.8)	14 (14.0)	0.00* , **
14. Delusions	78 (9.5)	30 (6.4)	38 (14.7)	10 (10.0)	0.001* , **
15. Double vision	116 (14.1)	42 (9.0)	58 (22.4)	16 (16.0)	0.00* , **
Attention/memory
16. Difficulty concentrating	223 (27.0)	114 (24.5)	79 (30.5)	30 (30.0)	0.17
17. Memory	379 (45.9)	207 (44.4)	125 (48.3)	47 (47.0)	0.59
18. Forgetfulness	260 (31.5)	138 (29.6)	94 (36.3)	28 (28.0)	0.13
Gastrointestinal tract
19. Drooling daytime	246 (29.8)	106 (22.7)	107 (41.3)	33 (33.0)	0.00* , **
20. Dysphagia	124 (15.0)	52 (11.2)	62 (23.9)	10 (10.0)	0.00*
21. Constipation	385 (46.7)	193 (41.4)	144 (55.6)	48 (48.0)	0.001* , **
Urinary
22. Urgency	263 (31.9)	126 (27.0)	102 (39.4)	35 (35.0)	0.002* , **
23. Urine frequency	253 (30.7)	116 (24.9)	101 (39.0)	36 (36.0)	0.00* , **
24. Nocturia	490 (59.4)	244 (52.4)	181 (69.9)	65 (65.0)	0.00* , **
Sexual function
25. Altered sex	329 (39.9)	174 (37.3)	112 (43.2)	43 (43.0)	0.24
26. Problem having sex	316 (38.3)	164 (35.2)	108 (41.7)	44 (44.0)	0.10
Miscellaneous
27. Pain	312 (37.8)	163 (35.0)	118 (45.6)	31 (31.0)	0.01*
28. Taste and smell	325 (39.4)	183 (39.3)	107 (41.3)	35 (35.0)	0.55
29. Weight change	204 (24.7)	102 (21.9)	73 (28.2)	29 (29.0)	0.10
30. Excessive sweating	150 (18.2)	72 (15.5)	60 (23.2)	18 (18.0)	0.04* , **

Bold font highlights the domains of the NMSS.

^* Significant at the 0.05 level.

^** Post‐hoc testing (Z test) revealed significant difference was between the tremor dominant (TD) and the postural instability and gait disorder (PIGD) phenotypes for those non‐motor symptoms (P < 0.05).

Spearman's correlation was used to analyze the strength of the association between the NMSS score and PD related clinical characteristics. Only age was not significantly correlated with NMSS, P > 0.05. (Table 4). While there was no correlation between the age of onset cohort‐wide and the motor subtypes, the age at study did correlate with the NMSS (r_{s =} 0.076, P = 0.000).

TABLE 4.

Correlation between NMSS scores and PD‐related clinical characteristics in relation to motor phenotype

Variables	All PD		Tremor‐dominant		PIGD		Indeterminate
	(n = 825)		n = 466		n = 259		n = 100
	rs	P	rs	P	rs	P	rs	P
Age at onset (yr)	0.014	0.68	0.020	0.33	−0.006	0.46	−0.126	0.11
Age at study (yr)	0.076*	0.000	0.063	0.09	0.049	0.22	−0.017	0.43
Disease duration (mo)	0.207*	0.000	0.157*	0.001	0.235*	0.000	0.239	0.009
H & Y stage	0.306*	0.000	0.244*	0.000	0.280*	0.000	0.238*	0.009
Total UPDRS	0.506*	0.000	0.469*	0.000	0.505*	0.000	0.463*	0.000
Part I MDS‐UPDRS	0.655*	0.000	0.588*	0.000	0.687*	0.000	0.752*	0.000
Part II MDS‐ UPDRS	0.494*	0.000	0.455*	0.000	0.542*	0.000	0.452*	0.000
Part III MDS‐UPDRS	0.358*	0.000	0.335*	0.000	0.371*	0.000	0.315*	0.000

r_s is Spearman's rho.

^*Significant at the level of P < 0.01.

H & Y, Hoehn and Yahr; MDS‐UPDRS, Movement Disorders Society Unified Parkinson's Disease Rating Scale.

We also explored the determinants of NMS severity using multiple linear regression. Using a forced entry model, and making no apriori assumptions, we included 10 variables as follows—age at onset, age at study, sex, disease duration, dopaminergic treatment status (yes/no), LEDD, disease severity (MDS‐UPDRS total score) and each motor phenotype i.e., PIGD, TD and indeterminate. Sex, age at onset, age at study, LEDD, indeterminate and TD phenotypes were excluded (either because of non‐significant association with NMSS score (P > 0.05) in correlation analysis or due to multicollinearity (correlation between any 2 independent variables >0.7). The only independent determinants of NMSS in the final model that fit (F = 80.6, P = 0.000) were the disease severity (MDS‐UPDRS total score) and PIGD phenotype.The summary data are displayed in Table 5.

TABLE 5.

Univariate linear regression between multiple determinants and total non‐motor symptom score

Model 2	B	Standard error	95% C.I. (B) LB ‐ UB	β	P
Independent variable
Disease duration	0.02	0.25	−0.46–0.50	0.003	0.93
MDS‐UPDRS total score	0.55	0.03	0.48–0.62	0.50	0.000*
PIGD phenotype	6.04	2.19	1.74–10.35	0.084	0.006*
Dopaminergic treatment status	4.90	2.66	−0.32–10.13	0.055	0.066

Multiple linear regression Model 2 (sex, gender, age, tremor‐dominant and indeterminate phenotypes removed (see text for Model 1). Model fit with F = 80.6, P = 0–.000. Adjusted R² = 0.28 and standard error of 28.4.

^*β significant at the <0.01 level.

MDS‐UPDRS, Movement Disorders Society Unified Parkinson Disease Rating Scale; PIGD, postural instability gait disorder.

Discussion

This is the largest and most comprehensive report on non‐motor symptom phenomenology in persons with PD in sub‐Saharan Africa to date, and the first on the continent to explore classification of PD by motor subtype utilizing the most current algorithm based on the MDS‐UPDRS. ¹⁴ Our study is also the first in the region to explore the relationship between PD motor phenotype and NMS burden. We enrolled a nationally representative cohort of PD enabling a more robust exploration, and this is reflected in the similarity and comparability of the baseline demographic characteristics of participants in this report with previous studies. ¹⁹ , ²⁰ Firstly, we found that NMS are highly prevalent in our PD population consistent with the experience described in previous reports. ⁶ , ²⁰ , ²¹ , ²² The burden of NMS was predominantly mild–moderate, with the most frequent symptoms being nocturia and fatigue, both of which have been reported as being most common in several other studies. ⁴ , ²⁰ , ²² , ²³ , ²⁴ , ²⁵ Martinez‐Martin and colleagues studied ~900 ethnically diverse persons with PD from South America, USA, Asia and Europe and reported that nocturia, fatigue and drooling of saliva were the most frequent NMS in approximately two‐thirds of their cohort. The cohort had similar age at onset of PD in the 6th decade as reported in our study, but a longer duration of PD (approximately 3.5 years longer). The latter may have contributed to the higher frequency of drooling of saliva as may be seen with advancing disease. ²² Nocturia was also one of the most frequent NMS documented in more than half of the 215 Belgian PD studied by Crossiers et al. and second only to urgency using the NMS Quest to quantify NMS. ²⁶ Dribbling of saliva was reported in 40.8% of their PD cohort, higher than our finding of 29.8% which is more similar to the 31% reported in the PRIAMO study. ⁶

Our findings are consistent with reports from a longitudinal study in Singapore where sleep/fatigue and urinary domains were the most frequent NMS both at baseline and follow‐up. ²⁵

We found significantly worse urinary and sex‐related NMS in males. While reports of gender differences in occurrence of NMS in the PD population have not been homogenous, our report is similar to Sanchez‐Martinez CM et al, who reported the most common domains affected in males as urinary and sexual function (in addition to attention/memory and sleep/fatigue). ²⁷ A closer look at the individual NMS revealed drooling and sexual dysfunction as the most common symptoms. ²⁷ There was no gender difference in urinary symptoms in their cohort. This gender difference in urinary symptoms encountered in some studies could be attributed to underlying co‐morbidities such as prostate disease which is prevalent in older men and capable of aggravating urinary symptoms. Cultural sensitivities may impact on disclosure of sexual activity and/or complaints in post‐menopausal women. ²² , ²⁸

Secondly, the TD subtype was most frequent, followed by PIGD and then indeterminate forms. Persons with the PIGD subtype in our study were older at disease onset, had more disability, severe/advanced disease and longer disease duration, as has been alluded to in other studies. ¹⁹ , ²⁹ The higher frequency and greater burden of NMS in the PIGD subtype in our PD population is consistent with previous reports. ¹⁹ , ²⁶ , ²⁹ Although a longer duration of PD in persons with PIGD has been proposed as underlying the higher burden of NMS in some of the studies, Chaudhuri et al. proposed a “motor model” which surmises that a wider array of NMS was seen in non‐tremor dominant subtypes of PD regardless of the disease stage. ⁸ Sauerbier et al. on the other hand proposed a distinctive underlying neuropathological basis resulting in variable NMS burden per PD motor phenotype, suggesting that clinical expression of NMS results from rate and extent of Lewy Body pathology and degeneration rather than loss of a single neurotransmitter. ⁹ Three main routes of spread of pathology associated with certain NMS clusters have been suggested: brainstem route (sleep dysfunction and dysautonomia), olfactory‐limbic route (mood disorders, fatigue, pain and weight loss) and neocortical route (memory issues, apathy and anxiety). ⁹ However, there is insufficient clinical‐pathological correlation studies across populations to strengthen the proposition.

Thirdly, we specifically described the burden of each NMS (and symptom domain) per motor subtype, and noted that urinary frequency and fatigue were most frequent overall, and that specific categories of NMS (including the former) were significantly more prevalent in PIGD (including cardiovascular, most mood/cognition, perceptual problems, gastrointestinal complaints (including constipation), pain and excessive sweating). Indeed, PIGD phenotype and disease severity were the most convincing independent predictors of NMSS scores. The neurochemical substrate underlying urinary dysfunction and fatigue (serotonin and norepinephrine deficiency) may be more profound and more closely related to PIGD and perhaps partially explain the relatively higher burden in that phenotype. ⁹ , ²³ , ³⁰ PIGD may represent a more “aggressive” PD phenotype, including more rapid disease progression and wider spectrum of neuropathology. ³¹ The data from this current study support this assertion as we demonstrated a significantly higher median disease stage despite disease duration that was not significantly different. However, irrespective of motor phenotype, fatigue and urinary dysfunction were the most common NMS, each affecting more than half of the PD phenotypes. A recent meta‐analysis concluded that fatigue occurs early in PD and persists throughout the disease, and its prevalence does not differ among new and treated persons with PD. ³² Though the prevalence of fatigue has varied in published reports, it is generally accepted that fatigue occurs in about half of the PD population, in keeping with findings from this report. ³¹ , ³³

Urinary symptoms in PD have been attributed to bladder hyperreflexia related to dopamine deficiency in addition to other neurotransmitter dysfunction. ³³ While the pathophysiology underpinning the PIGD motor phenotype is not completely understood, a central deficiency in norepinephrine has been postulated to be responsible and this may explain the higher burden of depression in PIGD. ²⁹ , ³⁰ , ³⁴ , ³⁵ Overall, with reference to persons with PIGD manifesting a higher burden of NMS referable to most domains, our findings are largely consistent with reports from other populations. ¹⁹ , ²⁶ , ²⁹ The underlying mechanisms for the occurrence of the variety of NMS in PD and predilection in PIGD are undoubtedly complex, driven by a combination of factors related to the disease duration and severity and initiated by neuropathological changes. Although the occurrence and distribution of alpha‐synuclein and other pathologies outside the nigrostriatal dopaminergic system are crucial, as pointed out by Jellinger in a comprehensive report, the precise correlation with the underlying pathology remains unclear. ³⁶

Our report demonstrates the similarity in profile and burden of NMS in our black African cohort, re‐emphasizes the link to worse disease severity and PIGD motor phenotype, and reiterates the likelihood that non‐biological (e.g., under‐reporting and under‐recognition) rather than biological factors may underpin reports of lower NMS burden in persons of black African ancestry. ³⁷ Our findings support the recommendation to specifically enquire about the presence of NMS as the relationship to the spectrum of PD may not be apparent to persons living with the disease and thus preclude complaints during follow‐up.

Limitations

We recognize that our study has certain limitations. We did not explore the chronology of development of NMS and thus the frequency and burden in the early stage of PD compared to later stage PD in individual participants is not distinguished. Such data would require a longitudinal study rather than the cross‐sectional design applied here.

Conclusion

NMS is prevalent in this cohort of PD of black African ancestry, with nocturia and fatigue being most prevalent in our cohort. NMS are also most frequently associated with the PIGD motor subtype and this is consistent with descriptions from other populations.

Author Roles

(1) Research project: A. Conception, B. Organization, C. Execution; (2) Statistical Analysis: A. Design, B. Execution, C. Review and Critique; (3) Manuscript Preparation: A. Writing of the first draft, B. Review and Critique; (4) Responsibility for final version.

O.O.O: 1A, 1B, 1C, 2A, 2B, 2C, 3A, 3B

K.W.W: 2C, 3B

B.E. 3A

S.A.A: 3A

F.I.O.: 3A

M.R.: 1B, 2C

N.U.O.: 1A, 1B, 1C, 2A, 2B, 2C, 3A, 3B

All Authors: 1C, 2B, 4

Disclosures

Funding Sources and Conflicts of Interest

There was no source of funding for this study and all authors have no conflicts of interest to declare.

Financial Disclosures for the Previous 12 Months

Dr. R. Akinyemi is supported by the following grants: US NIH/NHGRI (U01HG010273) and the UK Royal Society/African Academy of Sciences (FLR/R1?191,813). Dr. M. Rizig received funding from the University College London Grand challenges Small Grants (Award ID:177813 and the Michael J. Fox Foundation Genetic Diversity in Parkinson's Disease 2019 (Grant ID:17483). Prof. N. Okubadejo is supported by the Michael J. Fox Foundation Genetic Diversity in Parkinson's Disease 2019 (Grant ID:17483). The following authors report employment from the Federal Government of Nigeria: Oluwadamilola O. Ojo, Abiodun H. Bello, Sani A. Abubakar, Bertha C. Ekeh, Folajimi M. Otubogun, Temitope H. Farombi, Olaleye Adeniji, Francis I. Ojini, Frank A. Imarhiagbe, Yakub Nyandaiti, Morenikeji A. Komolafe, Michael B. Fawale, Gerald A. Onwuegbuzie, Yusuf Zubair, Uduak E. Williams, Funlola T. Taiwo, Shyngle I. Oyakhire, Nosakhare Osemwegie, Godwin O. Osaigbovo, Francis E. Odiase, Emmanuel E. Obehighe, Ernest O. Nwazor, Paul O. Nwani, Abiodun J. Kehinde, Cyril O. Erameh, Oluchi S. Ekenze, Franklin O. Dike, Salisu A. Balarabe, Mohammed W. Ali, John E. Akpekpe, Uchechi Agulanna, Christian E. Agu, Osigwe P. Agabi, Babatunde A. Ademiluyi, Akintunde A. Adebowale, Charles O. Achoru. Kolawole W. Wahab, Rufus O. Akinyemi, and Njideka U. Okubadejo report: employment, Federal Government of Nigeria; grants. Emmanuel U. Iwuozo reports employment from Benue State Government. Olajumoke O. Oshinaike, Olanike A. Odeniyi, Babawale Arabambi, and Oladunni V. Abiodun report employment from Lagos State Government. Yahaya O. Obiabo and Ohwotemu Arigbodi report employment from Delta State Government. Ifeyinwa Ani‐Osheku reports employment from Federal Capital Territory. Mie Rizig reports: University College London; grants.

Ethical Compliance Statement

The study was approved by the institutional research ethics committees of the participating centers and the National Health Research Ethics Committee of Nigeria. All participants provided written informed consent prior to inclusion in the registry and study. We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this work is consistent with those guidelines.

Supporting information

Table S1. Comparison of the domain specific scores by gender and motor phenotype