Association of coffee consumption and striatal volume in patients with Parkinson's disease and healthy controls

Abstract Background Mounting studies have demonstrated that coffee consumption significantly reduces the risk of developing Parkinson's disease (PD). However, there have been few investigations about the role of chronic coffee consumption in nigrostriatal structural neurodegeneration in PD. We aimed to investigate whether chronic coffee consumption is associated with the change in striatal volume in PD. Methods In this study, 130 de novo patients with PD and 69 healthy controls were enrolled from the Parkinson's Progression Markers Initiative cohort. Patients with PD and healthy controls were, respectively, divided into three subgroups, including current, ever, and never coffee consumers. Then, striatal volume was compared across the three subgroups. Correlation analyses were performed to assess the relationship between cups consumed per day and striatal volume. Furthermore, we included the factors that may have influenced nigrostriatal dopaminergic neurons in multiple linear regression analyses to identify significant contributing factors to striatal volume in each investigated striatal region. Results Current coffee consumers had decreased striatal volume compared with ever consumers in controls but not patients with PD. Furthermore, the correlation analyses revealed that cups per day were negatively correlated with striatal volume in current consumers of patients with PD and controls. In addition, multiple linear regression analyses showed that current coffee consumption remained as an independent predictor of a decrease in striatal volume in controls. Conclusions Our study showed that chronic coffee consumption was negatively correlated with striatal volume. In addition, our study showed that chronic coffee consumption was associated with the change in striatal volume in current—rather than ever coffee consumers, which suggests that the chronic effects of caffeine on striatal morphology may fade and even compensate after quitting coffee. Our study provides evidence for the effect of chronic coffee consumption on striatal volume in human brain in vivo.


| INTRODUC TI ON
Parkinson's disease (PD) is the second most common neurodegenerative disorder that affects 2%-3% of elderly people >65 years old worldwide. 1 With the advent of the aging age and the extension of human life expectancy in modern society, the number of PD cases is expected to double by 2030. 2 PD is characterized by bradykinesia, resting tremor, rigidity, and postural instability. 1 Neuronal loss in the substantia nigra that causes a deficit of striatal dopamine and intracellular protein (α-synuclein) accumulation that leads to the nigrostriatal pathway impairment are the neuropathological hallmarks of PD. 3 Major contributors to PD include both genetic risk factors 4 and environmental exposure to toxicants such as pesticides, heavy metals, and organic solvents. 5 Nevertheless, both retrospective and prospective epidemiological studies have demonstrated coffee consumption significantly reduces the risk of developing PD. [6][7][8][9] Caffeine is one of the most widely consumed psychoactive substances, with an average consumption of about 200-250 mg/day/person, as a standard cup of coffee contains 100 mg of caffeine. 10 After consumption, caffeine is rapidly absorbed through the gastrointestinal tract to the blood and then to the brain, 11 causing alertness and reducing fatigue. 12 The action of caffeine on the dopaminergic system is responsible for enhancing motor activity and exerting an antidyskinetic effect. 13,14 On the one hand, several studies demonstrated that chronic coffee consumption was associated with decreased functional connectivity 15 and decreased metabolism. 16 On the other hand, several recent studies revealed an inverse association between coffee consumption and brain volume using the data from UK Biobank. 17,18 PD is characterized by the degeneration of dopaminergic neurons in the nigrostriatal pathway. 19 However, there have been few investigations about whether chronic coffee consumption is associated with the change in striatal volume.
The primary objective of this study was to investigate whether chronic coffee consumption could be associated with the change in striatal volume. Therefore, we systematically assessed the baseline striatal volume of current, ever, and never coffee consumers in a large sample of de novo patients with PD and healthy controls enrolled from the Parkinson's Progression Markers Initiative (PPMI) cohort. Herein, we hypothesized that coffee consumption might be correlated with decreased striatal volume.

| MATERIAL S AND ME THODS
This study was in accordance with the approval of the Medical Ethics Committee of all PPMI sites involved and all participants were informed consent forms.

| Participants
Data used in this study were all obtained from the PPMI database tional, international, multicenter cohort study in a large cohort. The study of the PPMI database aimed to identify clinical, imaging, genetic, and biospecimen biomarkers of PD progression. Study protocols and manuals are available on the PPMI website.
In all, 202 participants with magnetic resonance imaging (MRI) scans and questionnaires about coffee consumption were enrolled in this study. The coffee consumption questionnaires included several questions ranging from "cfqa1" to "cfqa5." "cfqa1" referred to the question "In your lifetime, have you ever regularly drunk caffeinated coffee, that is, at least once per week for 6 months or longer?" "cfqa3" referred to the question "Do you currently drink caffeinated coffee?" "cfqa5day" referred to the question "During the time you were regularly drinking caffeinated coffee, on average, about how many cups per day did you drink?" In this study, if "cfqa1 = yes" and "cfqa3 = yes," the participants were regarded as current consumers; if "cfqa1 = yes" and "cfqa3 = no," the participants were regarded as ever consumers; and if "cfqa1 = no" and "cfqa3 = no," the participants were regarded as never consumers.
None of the participants were current coffee consumers (cfqa3 = yes) whoever had not consumed coffee (cfqa1 = no). One participant with poor quality of T1-weighted images, one participant with image format conversion error, and one participant who did not know/prefer not to answer the "cfqa3" were excluded. Finally, 130 de novo patients with

| MRI acquisition and volumetric analysis of striatum
All baseline high-resolution 3D T1-weighted imaging (T1WI) data were obtained at PPMI imaging centers, which were acquired according to a standardized protocol. At baseline, all PD patients were drug-naïve. Before preprocessing, MRI images of raw DICOM format were reviewed and converted into the Neuroimaging Informatics Technology Initiative (NII) format. All NII images were preprocessed and analyzed using the CAT12 toolbox (Computational Anatomy Toolbox; http://dbm.neuro.uni-jena.de/cat/) implemented in SPM12 (http://www.fil.ion.ucl.ac.uk/spm/softw are/spm12/). CAT12 served as the platform for preprocessing the structural MRI data. For processing and analysis steps, preset parameters in accordance with a standard protocol (http://www.neuro.uni-jena.de/cat12/ CAT12 -Manual.pdf) were used, applying default settings unless indicated otherwise. Then, the volume of each striatal subregion (caudate and putamen) and whole striatum were estimated according to the Neuromorphometrics and Cobra atlas, respectively. Finally, to correct for volume differences due to different head sizes, the volume of each striatal region was normalized by the total intracranial volume (TIV) of each individual, respectively: normalized striatal volume = striatal volume/individual TIV × 10 3 (cm 3 ).

| Statistical analysis
To assess the clinical and striatal volume characteristics among subgroups in patients with PD and controls, respectively, one-way analysis of variance (ANOVA) or t-test was used to compare continuous variables that did follow a normal distribution, the Kruskal-Wallis test, or Mann-Whitney U test were used to compare continuous variables that did not follow a normal distribution, and the χ 2 test and Fisher's exact test were used for categorical variables. The Kolmogorov-Smirnov test was used to assess the normality of the continuous variables. To determine whether there was a relationship between coffee cups consumed per day and striatal imaging characteristics, partial correlation coefficients were calculated for each investigated striatal region after adjusting for age and gender in the subgroups of current and ever consumers, respectively. Furthermore, we included the factors that may have influenced nigrostriatal dopaminergic neurons in multiple linear regression analyses to identify significant contributing factors to the volume in the investigated striatal regions. In Model 1, the independent variables included age, gender, coffee consumption (current vs. ever/never consumers), and smoking history (current vs. ever/never smokers) for PD patients and controls. Moreover, Model 2 included the independent variables from Model 1, as well as age at onset, disease duration, H-Y stages, and MDS-UPDRS part III that were known to be associated with dopaminergic density in PD patients. Bonferroni correction was made to adjust for multiple comparisons of striatal volume characteristics. The corrected significance level is 0.05 divided by the total number of comparisons (the total number of the investigated striatal regions) provided for either patients with PD or healthy controls, that is, 0.05/6 = 0.0083 for striatal volume comparisons. Post hoc tests of one-way ANOVA were conducted using the least significant difference (LSD) method. All statistical analyses were performed using SPSS Statistics 26 software (IBM Corporation, New York).

| Demographic and clinical characteristics
In this study, a total of 130 patients with PD and 69 controls were finally included. Baseline demographic and clinical characteristics of three subgroups were summarized for patients with PD and controls in Table 1. There were no significant differences in age and gender among the three subgroups of patients with PD and controls.
Education years were lower in PD-NC than in PD-EC (p = 0.006) and PD-CC (p = 0.032). In addition, MDS-UPDRS (specifically the part I, part II, part III, and total score), H-Y stages, MOCA, GDS, SCOPA, STAI, QUIP, UPSIT, and ESS were also compared across the PD or controls subgroups, and these clinical characteristics showed no significant difference.

| Striatal volume characteristics
There was no significant difference of the volume in all striatal regions between patients with PD and controls (Table S1). The volume of each striatal region was further compared across three subgroups in patients with PD and controls, respectively. In patients with PD, there was no significant difference in the volume of any striatal region across the three subgroups. However, in controls, current consumers had a tendency to lower volume in each striatal region than ever/never consumers ( Figure 2, Table 2 in PD-CC ( Figure 3A,B, Table S2). In addition, cups per day were  Table S2).

| Independent contribution of coffee consumption to striatal volume
The results of striatal volume in multiple linear regression analysis of Model 1 for PD patients and controls are presented in Table 3.
In patients with PD, there was no significant association between current coffee consumption and the volume of any striatal region.

| DISCUSS ION
In the present study, we compared the striatal volume across three decreased to only 15% of control values without caffeine pretreatment. 33 Furthermore, another experiment demonstrated that delayed caffeine administration could also reduce the loss of nigral dopamine cell bodies and block the nigral neurodegenerative process in rats. 34 In the present study, we observed an inverse association between current coffee consumption and striatal volume in controls and between coffee cups consumed per day and striatal volume in current consumers of PD and controls. Recently, using the data from UK Biobank, several large-scale studies revealed an inverse association between coffee consumption and gray matter volume. 17,18 Moreover, the finding of this inverse association was TA B L E 2 Striatal volume between subgroups of PD patients and healthy controls, respectively.  potentiation via blocking A2AR may be partially responsible for the effects of chronic coffee consumption on striatal structural change. In addition, we speculate the decreased striatal volume may also be due to dehydration, change in neuronal numbers, and change in vascular compartments caused by coffee consumption.

PD-CC PD-EC PD-NC
Notably, as shown in Figure 2, the present study demonstrated that there was a significant decrease in striatal volume in current rather than ever consumers, while there was a tendency for striatal volume to increase in ever consumers. Our findings indicate that the chronic effects of caffeine on striatal morphology may also fade and even compensate after quitting coffee consumption.
In the present study, the influence of current coffee consumption on striatal volume was only detected in controls rather than patients with PD. Moreover, the correlation analysis revealed that cups per day were negatively correlated with volume in current consumers in four of the investigated striatal regions in controls but only in two of the investigated striatal regions in patients with PD. Furthermore, an inverse association was found between age and striatal volume in all of the investigated striatal regions in controls but not in patients with PD. One possible explanation for the observation that controls showed more significant results than patients with PD may be a statistical "floor" effect due to PD pathologies. In Bohnen et al.'s study, they examined associations between asymmetric hemispheric nigrostriatal dopaminergic denervation with DAT imaging and pegboard scores in patients with PD. 39 Their results showed there was no significant correlation between pegboard scores of the most affected arm and DAT binding of the most denervated striatum, whereas there was an inverse correlation between pegboard scores of the least affected arm and DAT binding of the least denervated striatum. It was suggested that these discrepant results were caused by a statistical "floor" effect due to more severe denervation of the striatum. 39 In addition, we observed an inverse association between age and striatal volume in controls. Consistent with our findings, a recent TA B L E 3 Factors associated with volume in each striatal subregion by a multiple linear regression analysis in PD patients and healthy controls. Future studies should investigate the acute effects of coffee consumption and thus explain whether acute coffee consumption influences striatal volume.

Striatal regions Variables
In conclusion, the present study showed that current coffee consumption was negatively correlated with striatal volume in controls but not patients with PD. The correlation analysis revealed that cups per day were negatively correlated with striatal volume in current consumers in both patients with PD and controls. In addition, our study firstly showed that there was a significant decrease in striatal volume in current rather than ever consumers, while there was a tendency for striatal volume to increase in ever consumers.
Our findings indicate that the chronic effects of caffeine on striatal morphology may also fade and even compensate after quitting coffee consumption. Our study provides evidence for the effect of chronic coffee consumption on striatal volume in human brain in vivo.

CO N FLI C T O F I NTER E S T S TATEM ENT
The authors have no conflict of interest to report.

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.

CO N S ENT TO PA RTI CI PATE
Written informed consent was obtained from all participants.