Characterizing mild cognitive impairment in prodromal Parkinson’s disease: A community‐based study in China

Abstract Objective The International Parkinson and Movement Disorder Society (MDS) has published research criteria for prodromal Parkinson's disease (pPD), which includes cognitive impairment as a prodromal marker. However, the clinical features of mild cognitive impairment (MCI) in pPD remain unknown. Our study aimed to evaluate the frequency and clinical features of mild cognitive impairment of pPD in the elderly in China. Methods The cross‐sectional community‐based study recruited 2688 participants aged ≥50 years. Subjects were diagnosed with pPD according to the MDS criteria. Overall, 39 pPD and 22 healthy controls underwent comprehensive clinical and neuropsychological assessment. MCI was also diagnosed by the MDS criteria. Next, we investigated the relationship between clinical factors and cognition. Results Among the 2,663 dementia‐free and Parkinson disease (PD)‐free participants, 55 met the criteria for pPD (2.1%) and 23 pPD met the criteria for MCI. Memory, attention/working memory, and executive function were the most frequent impaired domains, and amnestic MCI multidomain phenotype was the most frequent MCI subtype (69.57%) in pPD. Additionally, correlation analysis revealed that the global cognitive performance was negatively related to UPDRS‐III score (r = −0.456, p = 0.004). Conclusion MCI, specifically impairment in memory, attention/working memory, and executive domain, is present at the prodromal stage of PD. In addition, cognitive performance is correlated with motor symptoms in pPD. Our results reflect that cognitive profile, combined with motor symptoms, can help clinicians to identify individuals with pPD early, as those would be the optimal candidates for neuroprotective therapy.


| INTRODUC TI ON
Parkinson's disease (PD) is a neurodegenerative disorder and whose diagnosis criteria include motor symptoms as the core feature of the disease. The motor symptoms are defined as bradykinesia, rest tremor, or rigidity. 1 Several studies demonstrated that 40%-60% of dopaminergic neurons had already degenerated by the time motor symptoms met the criteria for PD. [2][3][4] The neurodegenerative period, when nonmotor symptoms or mild parkinsonian signs are present, without the classical motor symptoms, is defined as prodromal PD (pPD). 5 In addition, PD is becoming strongly recognized as a multisystemic disorder with both motor and non-motor symptoms, and certain non-motor features present decades prior to achieving a clinical diagnosis. 6 One of the key points of the current research in PD is the identification and characterization of its prodromal period. Prior longitudinal studies, including large population-based cohort, have investigated the risk factors for pPD. [7][8][9] Unfortunately, studies in this field are limited, as longitudinal study is time-consuming, costly, and poor timeliness. Besides, due to the low incidence rate of PD and different definitions of pPD, some studies have focused on Rapid eye movement (REM) sleep behavior disorder (RBD) patients and individuals with different genetic mutations who are subsequently diagnosed with PD, but whose features may differ from individuals with typical sporadic PD. [10][11][12] Furthermore, based upon non-motor markers, the International Parkinson and Movement Disorder Society (MDS) published research criteria for pPD, 5 with high specificity. 13 Thus, in order to better understand the risk factors of PD in East China population, our group established a prospective, community population-based pPD cohort in Jiangsu province.
It is interesting to note that the percentage of memory decline was high in the pPD group, based on our preliminary research. Mild cognitive impairment (MCI), an intermediate state between normal cognitive aging and early dementia, 14 was present in newly diagnosed PD patients. 15,16 In a retrospective study, 5.5% of PD patients complained about experiencing non-specific "cognitive impairment" prior to their diagnosis. 17 In addition, previous studies have suggested that cognitive impairment is related to incidence of future PD. 7,8,18,19 Meanwhile, the MDS recently updated the diagnosis of pPD, and added global cognitive defect as a prodromal marker. 20,21 However, the majority of studies, which use questionnaires or limited cognitive tests, have failed to evaluate the clinical characteristics of MCI among individuals that are at risk for PD in the general elderly population. 7,17,22 The objectives of this present study were to (1) investigate the individuals presumed to be at an increased risk of PD in a communitydwelling elderly population of East China, and (2) to examine the cognitive profile of individuals with prodromal PD. We hope that our study is able to provide important proof to investigate the features of individuals with pPD in a community-dwelling elderly population of East China.

| Study population
This study recruited 2,688 volunteers from approximately nine districts across Nanjing City between March 2017 and August 2020.
Overall, 25 volunteers were excluded due to incomplete information, and 2663 participants comprised the sample of the community survey. We found 64 possible/probable prodromal PD patients (i.e., ≥30% probability of pPD) in the community survey. Next, we invited them to the hospital by telephone for further examination. However, two possible pPD and six probable pPD participants quit the study.
In further hospital-based assessment, seven participants were diagnosed as possible pPD, and 49 were diagnosed as probable pPD. But 10 of probable pPD refused or had insufficient knowledge to carry out neuropsychological assessment. Finally, 39 pPD (i.e., probable pPD) were divided into either a pPD group with mild cognitive impairment (pPD-MCI) or a pPD group with normal cognition (pPD-NC) (See Figure 1). Inclusion criteria in this study were residents who had lived in the local community for longer than 1 year and between the age of 50-80 years. Exclusion criteria were as follows: (1)

| Standard protocol approvals, registrations, and patient consents
The study was granted approval by the Medical Ethics Committee of the Second Affiliated Hospital of Zhejiang University College Southeast University-Nanjing Medical University, Grant/Award Number: 2018DN0031 profile, combined with motor symptoms, can help clinicians to identify individuals with pPD early, as those would be the optimal candidates for neuroprotective therapy.

K E Y W O R D S
cognition, Parkinson's disease, population-based study, prodromal of Medical and the Affiliated Brain Hospital of Nanjing Medical University. All participants were required to sign informed consent prior to participation.

| MDS research criteria and selection of markers
The MDS defined possible and probable pPD as having a probability score of 30%-80% and ≥80%, respectively. 5,23 Additionally, the MDS considered probable pPD met the criteria for prodromal PD (pPD). 5 We calculated the probability of pPD using available markers, as suggested by MDS and conducted by prior studies. 20,24 Several risk markers were used for the calculation, including male sex, occupational solvent exposure, regular pesticide exposure, coffee or tea use, non-smoking status, family history (including siblings with PD at age onset <50 years old or any first-degree relative with PD), and abnormal hyperechogenicity of the substantia nigra (SN). Additional prodromal markers were also assessed, including possible subthreshold parkinsonism, RBD (Polysomnographic-proven idiopathic RBD or positive RBD screen questionnaire), constipation, olfactory loss, excessive daytime somnolence, symptomatic hypotension, urinary dysfunction, severe erectile dysfunction, and depression/anxiety.

| Community-based assessment
All participants underwent a face-to-face interview, and all investigators that carried out the questionnaire received questionnaire application training. We collected socio-demographic information (i.e., gender, age, education levels, alcohol and tea use, and leisure activities) of the participants. Diagnosis of prior diseases (e.g., diabetes, hypertension, hyperlipidemia, stroke, coronary heart disease, and mood disorders) was obtained via self-report. We utilized a standardized structured questionnaire to evaluate the presence of a variety of environmental and lifestyle risk markers, as well as prodromal markers, during the initial interview. Standardized structured questionnaire, which included risk and prodromal markers, was obtained via self-report, and current use of medications (i.e., calcium channel blockers, beta-blockers, statins, antiplatelets, antidepressants, and benzodiazepines) were also recorded. Furthermore, we utilized the Sniffin' Sticks test 26 to screen possible/ probable pPD patients for olfactory dysfunction, and SN ultrasound to screen for abnormal hyperechogenicity of the SN.

| MCI diagnosis and cognitive status assessment
Mild cognitive impairment was diagnosed as per the MDS Task Force Level II diagnostic criteria, which provides an optimal, efficient neuropsychological test battery for diagnosis, 27 and applied assessment of the MCI subtypes. 15,28 The pPD and HC subjects were administered a formal, comprehensive neuropsychological battery. The neuropsychological battery, including memory, visuospatial function, language,   Binominal variables are represented as the number and percentage of their respective category. The differences of categorical variables among the three groups were analyzed by using the Chi-squared test.

| Statistical analyses
We set p < 0.05 as the threshold for statistical significance.
As previously described, 18,29 we transformed scores from cognitive tests into z scores using mean and SD values of the HC group.
Next, a domain z score was calculated by averaging individual neuropsychological test z scores (Table 1 shows domains and corresponding neuropsychological tests). A global z score based on the domain z scores was calculated in the same manner. For analytical purposes, the scores of some cognitive tests were reversed, and a higher global z score indicates better cognitive performance.
Logistic regression analyses (dichotomous variable) were conducted in order to explore the influence of potential confounding factors, including sex, the use of non-steroidal anti-inflammatory drugs (NSAID), and memory decline, on possible/probable pPD.
Partial correlation analysis (for normally distributed data) was performed to investigate the relationship between global cognition and clinical features in pPD, controlling the effects of age, years of education, gender, and HAMD.
Characteristics of the 2,663 subjects in the community population are detailed in Table 2. The proportion of females in possible/probable pPD group is lower compared with the other group, while the percentage of memory decline and NSAID use in possible/probable pPD group is higher compared with non-pPD group. Furthermore, the percentage of memory decline is noticeably high in possible/ probable pPD group. As expected, logistic regression analysis demonstrated that compared with those with less than 30% probability score for pPD, those with possible/probable pPD are three-fold more likely to experience memory decline (OR 3.372; 95% CI 1.709-6.651; p < 0.001).
Compared to the HC group, subjects in the pPD group performed worse in all cognitive tests, with the greatest differences for TMT-A, AVLT-delayed recall, and BNT scores (See Table S1).  which may point to overlapping dopaminergic network systems or pathologies. 32,33 In one pathological study, there was a remarkable reduction in dopaminergic neurons and terminals within the substantia nigra and putamen among subjects with minimal motor symptoms, who may represent pPD. 34   Although prospective studies are warranted, acknowledging the possible relationship of cognition and UPDRS-III score in prodromal PD can alert clinicians to look for motor symptoms in MCI patients, as well as for relevant non-motor symptoms and inform the probability of future PD. Furthermore, if possible, it can provide clues on how or when we may be able to intervene with neuroprotective or disease-modifying therapies.

CO N FLI C T O F I NTE R E S T
The author declares that there is no conflict of interest.

AUTH O R CO NTR I B UTI O N S
WGL, PYH, MMZ, LZ, WBZ, and PYX made substantial contributions to the conception or design of the work. CXP, YQL, JRR, LTL, and WGL made substantial contributions to the acquisition of the data.
CXP, YQL, JC, and WGL made substantial contributions to the analysis of the data. CXP, JC, and WGL made substantial contributions to the interpretation of the data. CXP drafted the work.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.