Coffee, adora2a, and cyp1a2: the caffeine connection in parkinsons disease

European Journal of Neurology 2011, 18: 756–765 Coffee, ADORA2A, and CYP1A2: the caffeine connection inParkinsonÕs disease R. A. Popata, S. K. Van Den Eedenb, C. M. Tannerc, F. Kameld, D. M. Umbache, K. Marderf,g,R. Mayeuxf,g, B. Ritzh, G. W. Rossi,j, H. Petrovitchi,j, B. Topola, V. McGuirea, S. Costellok,A. D. Manthripragadah, A. Southwickl, R. M. Myersl,* and L. M. NelsonaaDivision of Epidemiology, Department of Health Research and Policy, School of Medicine, Stanford University, Stanford, CA; bDivision ofResearch, Kaiser Foundation Research Institute, Oakland, CA; cThe ParkinsonÕs Institute, Sunnyvale, CA; dEpidemiology Branch, National Institute of Environmental Health Sciences, NIH, Research Triangle Park, NC; eBiostatistics Branch, National Institute of Environmental Health Sciences, NIH, Research Triangle Park, NC; fDepartment of Neurology, College of Physicians and Surgeons, Columbia University, New York, NY; gThe Gertrude H Sergievsky Center and the Taub Institute, College of Physicians and Surgeons, Columbia University, New York, NY; hDepartment of Epidemiology, School of Public Health, University of California, Los Angeles, CA; iVeterans Affairs Pacific Islands Health Care System, Honolulu, HI; jThe Pacific Health Research Institute, Honolulu, HI; kDepartment of Environment Science, School of Public Health, University of California, Berkeley, CA; and lDepartment of Genetics, School of Medicine, Stanford University, See editorial by Mellick and Ross, on page page 671.
Background and purpose: In 1-methyl-4-phenyl 1,2,3,6-tetrahydropyridine animal models of ParkinsonÕs disease (PD), caffeine protects neurons by blocking the aden- osine receptor A2A (ADORA2A). Caffeine is primarily metabolized by cytochrome P450 1A2 (CYP1A2). Our objective was to examine whether ADORA2A and CYP1A2 polymorphisms are associated with PD risk or modify the caffeine–PD association.
Methods: ParkinsonÕs Epidemiology and Genetic Associations Studies in the UnitedStates (PEGASUS) included five population-based case–control studies. One labora- tory genotyped four ADORA2A and three CYP1A2 polymorphisms in 1325 PD cases and 1735 age- and sex-matched controls. Information regarding caffeine (coffee)consumption and other lifestyle factors came from structured in-person or telephoneinterviews. Odds ratios (OR) and 95% confidence intervals (CI) were estimated usinglogistic regression.
Results: Two ADORA2A polymorphisms were inversely associated with PD risk –rs71651683, a 5¢ variant (adjusted allelic OR = 0.51, 95% CI 0.33–0.80, permutation-adjusted P = 0.015) and rs5996696, a promoter region variant (adjusted OR for ACand CC genotypes compared with the AA wild-type genotype were 0.76 (95% CI 0.57–1.02) and 0.37 (95% CI 0.13–1.01), respectively (permutation-adjusted P fortrend = 0.04). CYP1A2 polymorphisms were not associated with PD risk; however,the coffee–PD association was strongest among subjects homozygous for either vari-ant allele rs762551 (Pinteraction = 0.05) or rs2470890 (Pinteraction = 0.04).
Conclusion: In this consortium study, two ADORA2A polymorphisms were inverselyassociated with PD risk, but there was weak evidence of interaction with coffee con-sumption. In contrast, the coffee–PD association was strongest among slow metabo-lizers of caffeine who were homozygous carriers of the CYP1A2 polymorphisms.
Correspondence: R. A. Popat, PhD, Department of Health Research Coffee drinking has been associated with lower risk of and Policy, HRP Redwood Building, Room T209, Stanford University ParkinsonÕs disease (PD) in several case–control and School of Medicine, Stanford, CA 94305-5405, USA (tel.: 650 cohort studies. A recent meta-analysis showed that 498 5206; fax: 650 725 6951; e-mail: rpopat@stanford.edu).
coffee drinkers had a 30% reduction in PD risk com- *Present Address: HudsonAlpha Institute for Biotechnology, 601Genome Way, Huntsville, AL, USA.
pared to non-drinkers [1]. The biological basis of the European Journal of Neurology Ó 2011 EFNS putative neuroprotective effect of caffeine is not com- of the study populations are presented in Table 1 and pletely understood; however, caffeine has been shown other details, including the research diagnostic criteria to protect neurons in the 1-methyl-4-phenyl 1,2,3, [17,18], are summarized in Table S1. The pooled data 6-tetrahydropyridine (MPTP) neurotoxin model of PD included 1325 PD cases and 1735 age- and sex-matched by blocking the adenosine A2A receptor (ADORA2A) [2–5]. Hence, polymorphisms in ADORA2A, the genethat encodes the ADORA2A receptor, might mediate the caffeine–PD association. Caffeine is primarilymetabolized in the body by cytochrome P450 1A2, an Data were collected by structured in-person or tele- enzyme encoded by the gene CYP1A2 [6,7]. Therefore, phone interviews. Data for each subject on the fol- polymorphisms in CYP1A2 may affect caffeine avail- ability and, thereby, modify caffeine effects on PD risk.
investigators of the five component studies: date of Previous studies in ethnically homogeneous popula- diagnosis or reference date, sex, self-reported race/eth- tions composed primarily of non-Hispanic Whites [8] or nicity, date of birth, family history of PD, smoking Asians [9,10] have evaluated the role of ADORA2A and history, and caffeine consumption. Race/ethnicity was CYP1A2 variants on caffeine–PD association, but did self-reported according to one of the following cate- gories: Hispanic White, non-Hispanic White, Asian, or We used information from five population-based African-American. The Human Subjects Committees at studies to evaluate whether variations in ADORA2A or the various institutions approved the study, and CYP1A2 were associated with PD risk and whether the informed consent was obtained from all cases and caffeine–PD association was modified by these genetic Methodology for ascertaining caffeine exposure dif- fered slightly among component studies and is brieflysummarized in Table S1. The Columbia University studies did not ascertain information regarding caffeineconsumption; therefore, caffeine–genotype interactions analyses included 925 cases and 1249 controls. Ques- This consortium study [ParkinsonÕs Epidemiology and tions pertaining to caffeine use from the other four Genetics Association Studies in the United States studies allowed the construction of the following (PEGASUS)] combined DNA and risk factor data from exposure measures: broad category of consumption five population-based case–control studies, of which, (ever/never) and average number of 6-oz cups con- two were nested within cohorts [11–16]. Characteristics sumed daily. Because the average amount of caffeine Table 1 Characteristics of ParkinsonÕs disease (PD) cases and controls in the consortium study The totals for the variables may not equal the number of cases and controls because of missing values; FAME, Farming and MovementEvaluation; HASS, Honolulu Asia Aging Study; PEAK, Parkinsonism Epidemiology at Kaiser; PEG, ParkinsonÕs disease Epidemiology andGenetics; aAge is age of diagnosis for PD cases; bOne or more first-degree relatives with PD.
Ó 2011 The Author(s)European Journal of Neurology Ó 2011 EFNS European Journal of Neurology 18, 756–765 per cup is highest in coffee, we evaluated genotype– plate were analyzed by automated allele calling soft- caffeine interactions separately for caffeinated coffee, ware (ABI Prism 7900 HT Sequence Detection System tea, and sodas, and only present results for genotype– 2.1; Life Technologies Corporation, Carlsbad, CA, USA) and reviewed by a skilled operator. Laboratorypersonnel were blinded to the identity and case–controlstatus of the samples. For quality control purposes, a 15% repeat set of redundant genotypes was tested along Component studies provided the consortium a DNA with a small number of samples with known genotypes.
sample from each of their subjects. ADORA2A and The Ôno callÕ rate was very low (<1% of samples), and CYP1A2 were sequenced by the Stanford Human thus, we are confident that we analyzed only high- Genome Center in 24 patients with early-onset PD randomly selected from the Parkinsonism Epidemiol-ogy at Kaiser (PEAK) case–control study. Functional regions of both genes were resequenced, including theexons, intron–exon junctions, and regions within Each component study sent interview data and data 500 bp of the 5¢ and 3¢ UTR regions [19,20]. Variants documentation to Stanford University. For statistical occurring at polymorphic frequencies (minor allele analyses, we used SASÒ statistical software (SAS frequency >1%) were identified and polymorphisms Institute, Cary, NC, USA) [21]. We evaluated whether were prioritized for genotyping based on function, genotype distributions for control subjects were in location, and frequency, with emphasis given to vari- Hardy–Weinberg equilibrium (HWE) among each ants affecting protein sequence and function (i.e., racial/ethnic group separately with chi-square or Fish- exonic variants producing nonsense and missense erÕs exact tests. We designated the minor allele based on changes) and variants affecting gene expression or white, non-Hispanic subjects and used it for all eth- mRNA stability (i.e., variants located in the promoter nicities, even when the designated minor allele was the region, 5¢UTR, 3¢ UTR, splice-site, and intron–exon more frequent allele in these other ethnic groups.
boundaries). In all samples, we genotyped four We used unconditional logistic regression analyses to ADORA2A and three CYP1A2 single nucleotide poly- estimate odds ratios (ORs) and 95% confidence inter- morphisms (SNPs) on PEGASUS samples (Table 2).
vals (CIs) for allelic and genotypic associations with PD PCR primers and TaqMan probes were designed risk. To evaluate the risk associated with an increasing based on the NCBI DNA sequence and purchased from number of copies of the variant allele for a given ABI (Applied Biosystems, Foster City, CA, USA). PCR polymorphism, we conducted a test of trend. All esti- assays were run in TaqMan Universal Master Mix mates were adjusted for sex, age, study site, and race/ (Applied Biosystems). Fluorescence data files from each Table 2 Polymorphic variants in the adenosine receptor A2A (ADORA2A) and cytochrome P450 1A2 (CYP1A2) genes genotyped in PEGASUSsubjects PEGASUS, ParkinsonÕs Epidemiology and Genetics Association Studies in the United States; ars5751876 and rs3032740 in strong linkagedisequilibrium (DÕ = 0.997 and r2 = 0.98); bP < 0.01 for Hardy–Weinberg Equilibrium chi-square (HAAS study); cAA = fast caffeinemetabolizers, CA or CC = 0 slow caffeine metabolizers; drs2470890 and rs2472304 were in strong LD (DÕ = 0.993, r2 = 0.986).
European Journal of Neurology Ó 2011 EFNS European Journal of Neurology 18, 756–765 We excluded subjects who identified their race/eth- not observe effect modification by sex for caffeine–PD nicity as other (n = 18) and subjects whose genotyping associations; hence, all our genotype–caffeine interac- assay results could not be called (n = 58), leaving 1325 tion-related analyses combined men and women and cases and 1735 controls for analysis. For analyses of the adjusted for sex as a covariate. Inverse associations were newly discovered SNPs, we also excluded the 24 early- also observed with caffeinated tea (adjusted OR = 0.81, onset cases in the discovery sample. Information 95% CI 0.67–0.96). Among tea drinkers, PD risk regarding variants in monogenic genes was only avail- decreased by 7% per cup of average daily consumption; able for the PEAK case–control study (578 cases, 630 however, this estimate was not statistically significant controls; 39.6% of all subjects in the PEGASUS con- (adjusted OR = 0.93, 95% CI 0.83–1.03). Consump- sortium). In a sub-analysis, the risk estimates were tion of caffeinated soda was not associated with PD risk unchanged when we excluded PEAK subjects who (adjusted OR = 1.00, 95% CI 0.82–1.22).
carried any of the known pathogenic variants inmonogenic genes (PARKIN, a-synuclein, DJ1, PINK1, and LRRK2). Because we did not have the informationto exclude possible monogenic cases of Parkinsonism in The four ADORA2A SNPs we selected were in HWE the majority of subjects, we conducted our primary among non-Hispanic White, African-American, and Hispanic controls (Table 2). The Asian subgroup from We evaluated whether polymorphisms in ADORA2A HAAS was not in HWE at P < 0.01 (rs5751876, and CYP1A2 were effect modifiers of the caffeine–PD rs3032740, and rs5996696); however, no substantial associations (ever/never and average cups consumed differences in the ADORA2A-PD associations were among ever-drinkers, separately for caffeinated coffee, observed after excluding these samples. Therefore, tea, and sodas). We evaluated effect measure modifi- genotypic associations for ADORA2A SNPs include cation on a multiplicative scale by testing the signifi- subjects from all five studies (Table 3).
cance of the interaction terms in the logistic regression SNPs rs5751876 and rs3032740 were in strong linkage model using the likelihood ratio chi-square test, which disequilibrium (DÕ = 0.997 and r2 = 0.98) in all racial/ compares the model with the interaction term to the ethnic groups; hence, further discussion will be limited to rs3032740, which has functional relevance as it shown We used a permutation-based approach to adjust to reduce protein expression [23]. The deletion for P-values for multiple testing [22]. We randomly per- rs3032740, identified as the variant among White con- mutated the case–control status of subjects within trols (non-Hispanic and Hispanic), was more frequent strata defined by sex, race/ethnicity, and site. For each than the Tins among African-Americans and Asians.
of 10 000 permuted data sets, we used logistic regres- After adjustment for age, sex, race/ethnicity, and site, we sion to compute an age-, sex-, race-, and site-adjusted did not find an overall association of rs3032740 geno- per allele effect estimate for each polymorphism. The types with PD risk (Table 3), and associations were resulting empirical P-value distribution of 10 000 min- similar across racial/ethnic groups (Table S2).
imum P-values was used to estimate multiple compar- The frequency of variant allele for rs71651683 was 1.1% in cases and 2.1% in controls (adjusted allelicOR 0.51, 95% CI 0.33–0.80, permutation-adjustedP = 0.015). The variant allele was only present in Whites (non-Hispanic and Hispanic) and a few African- The five case–control studies were similar in some American control subjects (9.1%). Because no cases demographic characteristics but differed in others carried two copies of the variant allele, only genotypic (Table 1). Mean age was fairly similar across the studies; associations involving heterozygotes were estimable, however, Honolulu Asia Aging Study (HAAS) subjects and genotype–coffee interactions could not be evalu- were older. Subjects from the PEAK, Farming and Movement Evaluation (FAME), and PEG studies were The ADORA2A promoter variant, rs5996696, was primarily White, HAAS subjects were all Asians, and the inversely associated with PD risk (3.7% cases, 5.6% Columbia University study was comprised of 28% controls; adjusted allelic OR 0.70, 95% CI 0.54–0.91).
Hispanics. History of caffeinated coffee consumption Compared to subjects homozygous for the wild-type was associated with a 28% reduced risk of PD (adjusted allele (AA), the adjusted OR for PD risk among sub- OR = 0.72, 95% CI 0.58–0.88); and, among coffee jects with one (AC) or two copies of the variant allele drinkers, the risk decreased 12% with each one cup (CC) were 0.76 (95% CI 0.57–1.02) and 0.37 (95% CI increase in daily average consumption (adjusted 0.13–1.01), respectively (permutation-adjusted P-value OR = 0.88, 95% CI 0.83–0.94; data not shown). We did Ó 2011 The Author(s)European Journal of Neurology Ó 2011 EFNS European Journal of Neurology 18, 756–765 Table 3 Genotype frequency (%), adjusted odds ratios (OR), and 95% confidence intervals (CI) for the association between adenosine receptorA2A (ADORA2A) and cytochrome P450 1A2 (CYP1A2) polymorphisms and ParkinsonÕs disease in PEGASUS NE = not estimable; PEGASUS, ParkinsonÕs Epidemiology and Genetics Association Studies in the United States; aAdjusted for age, sex, site,and race/ethnicity; bAdjusted for age, sex and site; crs5751876 and rs3032740 were in strong linkage disequilibrium (DÕ = 0.997 and r2 = 0.98); dPfor trend = 0.01, permutation-adjusted P for trend = 0.04; eP for trend = 0.03, permutation-adjusted P for trend = 0.1; frs2470890 andrs2472304 were in strong linkage disequilibrium (DÕ = 0.993, r2 = 0.986).
The coffee (ever/never)–PD association was similar association of rs762551 genotypes with risk of PD among ever-drinkers, the inverse association with daily SNPs rs2470890 and rs2472304 were in strong linkage number of cups of coffee was strongest among those disequilibrium (DÕ = 0.993, r2 = 0.986); hence, further homozygous for the deletion (adjusted OR = 0.70, discussion will be limited to rs2470890, the exonic var- 95% CI 0.55–0.86, Pinteraction = 0.08, Table 4). Results iant. The allele ÔCÕ for rs2470890, identified as the vari- were similar when coffee–genotype interactions were ant based on non-Hispanic Whites controls, was the restricted to non-Hispanic Whites only (Table S3). No more frequent allele among the other race/ethnic groups interactions of ADORA2A genotypes were observed (Table 2). We did not find an overall association of with caffeinated tea or soda (data not shown).
rs2470890 genotypes with risk of PD among non-His-panic Whites, African-Americans, and Asians (Table 3and Table S2). However, among Hispanic subjects with one (TC) or two copies (CC) of the variant allele, the All three CYP1A2 SNPs were in HWE within every adjusted OR for PD risk were 1.67 (95% CI 0.8–3.4) and ethnic group. For rs762551, homozygous wild-type 2.1 (95% CI 1.0–4.3), respectively (P for trend = 0.05, carriers (AA) are rapid caffeine metabolizers, and het- permutation-adjusted P-value for trend = 0.2).
erozygotes (AC) and homozygotes (CC) are slow caf- For the rs762551 polymorphism, the effect of coffee feine metabolizers [24,25]). We did not find an overall consumption (ever versus never) was strongest among European Journal of Neurology Ó 2011 EFNS European Journal of Neurology 18, 756–765 Ó 2011 The Author(s)European Journal of Neurology Ó 2011 EFNS European Journal of Neurology 18, 756–765 subjects homozygous for the variant allele (adjusted groups, a finding that is consistent with two other OR = 0.33, 95% CI 0.16–0.68, Pinteraction = 0.05; reports that did not find associations of rs3032740 Table 4). Similarly, for the exonic variant rs2470890, (or rs5751876, a SNP is strong linkage disequilibrium the coffee–PD association was strongest among carriers (LD) with rs3032740) with PD risk [8,9].
of two copies of the variant allele (adjusted OR = 0.43, Metabolism by CYP1A2 is the primary pathway for 95% CI 0.27–0.69; Pinteraction = 0.04). Among ever the conversion of caffeine to paraxanthine. For the coffee drinkers, a one 6-oz cup increase in coffee con- most frequently studied intronic variant, rs762551, we sumption was associated with an approximately 18% expected the risk of PD to be lower among slow reduction in PD risk among heterozygotes (TC) and metabolizers (AC or CC) compared to fast metabolizers homozygous variants (CC) for rs2470890 compared to (AA) as the former would have higher caffeine levels only a 5% reduction in PD risk among homozygous [24,25,27] resulting in greater neuroprotection. How- wildtypes (Pinteraction = 0.015, Table 4). When analysis ever, consistent with other reports [8,10], in our study, was restricted to non-Hispanic whites only the results slow metabolizer status did not by itself render any were similar (Table S3); however, the power for geno- protection against risk of PD. The other CYP1A2 SNPs type-coffee (ever/never) interactions was reduced.
genotyped, rs2470890 (exon) and rs2472304 (intron), No interactions of CYP1A2 polymorphisms were were in strong LD; their associations with PD risk have observed with caffeinated tea or soda (data not shown).
not been previously reported. Interestingly, the ÔCÕ allelefor rs2470890, the minor allele among non-Hispanicwhites, was the more common allele among African- Americans, Asians, and Hispanics. We observed an We report two interesting and novel findings in this increased PD risk associated with the ÔCÕ allele among consortium study that comprised five US case–control Hispanics, but the permutation-adjusted per allele effect studies of PD. First, a polymorphism in the promoter was not statistically significant at alpha = 0.05; hence, region of ADORA2A (rs5996696) was associated with a this finding should be interpreted with caution, espe- 30% decreased risk of PD. Second, a newly identified cially as the functional impact of this exonic variant is polymorphism (rs71651683) in the 5¢ transcription start region of ADORA2A was associated with a 49% Pooled analysis from the five case–control studies decreased risk of PD. The associations of the 5¢ and supported the inverse association of caffeinated coffee promoter ADORA2A variants with PD risk have not consumption with PD risk. A primary objective of this been previously reported. Because these associations study was to evaluate whether the coffee–PD associa- remain after adjusting the P-values for multiple com- tion was modified by ADORA2A or CYP1A2 poly- parisons, they are less likely to represent false-positive morphisms. Because variants that would result in a non-functioning ADORA2A receptor would probably In advance of the study, we hypothesized that any not be influenced by caffeine, we hypothesized that caffeine would be more protective among homozygous expression or function of the receptor would be pro- carriers of the wild-type allele. However, our findings tective. This hypothesis was based on findings from do not support this hypothesis. For the two ADORA2A animal models of PD: knockout mice with non-func- tioning ADORA2A receptor showed protection against rs5751876, although the coffee–genotype interaction MPTP toxicity, and the effect was similar to those was stronger with cups consumed than with ever/never related to receptor blockade by caffeine or a pharma- consumption, neither provided convincing evidence of cologic agent (e.g., KW-600) [2–4]. While the functional interaction. These results are consistent with two other importance of rs5996696 and rs71651683 ADORA2A reports that did not find any effect modification of SNPs is not currently known, they are likely to reduce caffeine–PD association with these SNPs [8,9]. We were protein expression by affecting transcription [26].
unable to adequately evaluate interactions of rs5996696 Therefore, our finding that these two ADORA2A SNPs and rs71651683 ADORA2A polymorphisms with coffee are inversely associated with PD is consistent with the consumption because the variant allele frequencies for role of the ADORA2A receptor in caffeine-associated these SNPs were relatively small (<6%).
For the CYP1A2 rs76551variant, we hypothesized A previous study showed that rs3032740 reduces that the inverse coffee–PD association would be stron- protein expression [23]; therefore, we expected the ger among slow metabolizers compared to rapid presence of this variant to be protective for PD. How- metabolizers who carry two copies of the wild-type allele.
ever, we did not find any suggestion of a protective We did observe that the coffee–PD association was effect of this polymorphism in any of the race/ethnicity strongest among subjects homozygous for the variant European Journal of Neurology Ó 2011 EFNS European Journal of Neurology 18, 756–765 allele; however, it was somewhat weaker for hetero- variants appeared to modify the protective effects of zygotes, who are also considered physiologically to be slow metabolizers. Furthermore, although the interac-tion was statistically significant at the alpha = 0.05 level, interaction P-values were not adjusted for multi-ple comparisons and hence must be interpreted with Funding was provided to the PEGASUS genetic con- caution. Similar to our results, Tan et al. [10] found sortium by the Michael J Fox Foundation for Parkin- that among Asian subjects, the caffeine–PD association sonÕs Research. Additional funding to individual was also stronger in slow compared to fast metabolizers investigators for the original studies was provided by: (OR 0.19 vs. 0.40); however, the caffeine–genotype NIH NS R01-31964 and Tobacco-Related Disease interaction was not statistically significant in multivar- Research Fund Grants and 8RT-0131 and 11RT-0237 iable analysis [10]. Fascheris et al. [8] did not find any (Dr. Lorene Nelson,); NIH R01-NS32527 (Drs. Richard effect of rs762551 variant on caffeine–PD association; Mayeux and Karen Marder), NIA PO1 AG07232 (Dr.
however, in their study, caffeinated coffee consumption Richard Mayeux); NIH ES10544 and UES12078, pilot funding from SCEHSC # 5P30 ES07048, the ParkinsonÕs For the CYP1A2 exonic variant rs2470890, subjects Disease Association (Dr. Beate Ritz); United States homozygous for the variant allele also showed the Department of the Army DAM.D.17-98-1-8621, NIA strongest coffee–PD inverse association. The func- NO1-AG-4-2149, NHLBI NO1-HC-05102, and VA tional significance of this synonymous variant is not Medical Research funds (Dr. G. Webster Ross); NIEHS known, and it is possible that it has no effect on 01-ES10803 and U54-ES12077 (Drs. Caroline Tanner protein structure or function. A possible explanation and Freya Kamel). The FAME study was supported in for the minimal modification of the coffee–PD asso- part by the Intramural Research Program of the NIH, ciation by CYP1A2 polymorphisms might be that National Institute of Environmental Health Sciences, paraxanthine, the primary metabolite produced from and National Cancer Institute (Division of Cancer Epidemiology and Genetics). The information in this ADORA2A receptor in vitro, and preliminary studies in study does not necessarily reflect the position or the mice show that, like caffeine, paraxanthine can also policy of the government and no official endorsement Our consortium study had several strengths. The five constituent case–control studies in the consortium were methodologically rigorous and included careful selec-tion of well-characterized cases, a majority of whom The authors declare no financial or other conflict of were newly diagnosed with PD, as well as population- or community-based controls. For genotype–PD asso-ciations, we used a permutation approach to adjust P-values for multiple comparisons, thereby minimizingtype I error. Our study had some limitations as well.
Additional Supporting Information may be found in Although we included subjects from diverse racial/eth- nic groups, we did not have sufficient numbers in all Table S1. Description of study populations in the subgroups (e.g., African-Americans, n = 95) to esti- ParkinsonÕs Epidemiology and Genetics Association mate genotypic effects with precision or to have suffi- Studies in the United States (PEGASUS).
cient power to evaluate caffeine–genotype interactions.
Table S2. Genotype frequency (%), adjusted odds Methodology for ascertaining caffeine exposure infor- ratios (OR), and 95% confidence intervals (CI) for the mation varied between studies; however, the methods were comparable enough to allow construction of rel- (ADORA2A) and cytochrome P450 1A2 (CYP1A2) evant caffeine-related variables for our analyses.
polymorphisms and ParkinsonÕs disease in PEGASUS This consortium study characterized ADORA2A and CYP1A2 SNPs in Whites (non-Hispanic and Hispanic), Table S3. Adjusted odds ratios (OR) and 95% con- Asians, and African-Americans. Two ADORA2A fidence intervals (CI) for the association between caf- SNPs, which have not been previously studied, were feinated coffee consumption and ParkinsonÕs disease in inversely associated with PD risk. While the results of our study do not support the hypothesis that the inverse coffee–PD association was modified by putative func- Please note: Wiley-Blackwell are not responsible for tional polymorphisms in ADORA2A, two CYP1A2 the content or functionality of any supporting materials Ó 2011 The Author(s)European Journal of Neurology Ó 2011 EFNS European Journal of Neurology 18, 756–765 supplied by the authors. Any queries (other than 16. Ross GW, Abbott RD, Petrovitch H, et al. Association of missing material) should be directed to the corre- coffee and caffeine intake with risk of Parkinson disease.
17. Langston JW, Widner H, Goetz D, et al. Core assessment program for intracerebral transplantations (CAPIT). Mov 18. Hughes AJ, Ben-Shlomo Y, Daniel SE, Lees AJ. What 1. Herna´n MA, Takkouche B, Caaman˜o-Isorna F, Gestal- features improve the accuracy of clinical diagnosis in Otero JJ. A meta-analysis of coffee drinking, cigarette ParkinsonÕs disease: a clinicopathologic study [published smoking, and the risk of ParkinsonÕs disease. Ann Neurol erratum In Neurology 1992;42:1436]. Neurology 1992; 42: 2. Chen JF, Xu K, Petzer JP, et al. Neuroprotection by 19. Sambrook J, Fritsch EF, Maniatis T. Molecular Cloning.
caffeine and A(2A) adenosine receptor inactivation in a A Laboratory Manual. New York: Cold Spring Harbor model of ParkinsonÕs disease. J Neurosci 2001; 21: RC143.
3. Xu K, Xu YH, Chen JF, Schwarzschild MA. CaffeineÕs 20. Gordon D, Abajian C, Green P. Consed: a graphical tool neuroprotection against 1-methyl-4-phenyl-1,2,3,6-tetra- for sequence finishing. Genome Res 1998; 8: 195–202.
hydropyridine toxicity shows no tolerance to chronic 21. SAS System Under Microsoft Windows, Release 9.1 caffeine administration in mice. Neurosci Lett 2002; 322: (TS2MO). Cary, NC: SAS Institute, Inc., 2006.
22. Troendle JF. Multiple comparisons between two groups 4. Kalda A, Yu L, Oztas E, Chen JF. Novel neuroprotection on multiple Bernoulli outcomes while accounting for by caffeine and adenosine A(2A) receptor antagonists in covariates. Stat Med 2005; 24: 3581–3591.
animal models of ParkinsonÕs disease. J Neurol Sci 2006; 23. Alsene K, Deckert J, Sand P, de Wit H. Association between A2a receptor gene polymorphisms and caffeine- 5. Fink JS, Kalda A, Ryu H, et al. Genetic and pharmaco- induced anxiety. Neuropsychopharmacology 2003; 28: logical inactivation of the adenosine A2A receptor atten- uates 3-nitropropionic acid-induced striatal damage.
24. Sachse C, Brockmoller J, Bauer S, Roots I. Functional significance of a C to A polymorphism in intron 1 of the 6. Butler MA, Iwasaki M, Guengerich FP, Kadlubar F.
cytochrome P450 1A2 (CYP1A2) gene tested with caf- Human cytochrome P450A (P-4501A2), the phenacetin feine. Br J Clin Pharmacol 1999; 47: 445–449.
O-deethylase, is primarily responsible for the hepatic 25. Han X-M, Ou-Yang D-S, Lu P-X, et al. Plasma caffeine 3-demethylation of caffeine and N-oxidation of carcino- metabolite ratio (17X/137X) in vivo associated with G- genic arylamines. Proc Natl Acad Sci 1989; 86: 7696–7700.
2964A and C734 polymorphisms of human CYP1A2.
7. Gu L, Gonzalez FJ, Kalow W, Tang BK. Biotransfor- Pharmacogenetics 2001; 11: 429–435.
mation of caffeine, paraxanthine, theobromine and the- 26. Kim SH, Kim YK, Park HW, et al. Adenosine deaminase ophylline by cDNA-expressed human CYP1A2 and and adenosine receptor polymorphisms in aspirin-intol- CYP2E1. Pharmacogenetics 1992; 2: 73–77.
erant asthma. Respir Med 2009; 103: 356–363.
8. Facheris MF, Schneider NK, Lesnick TG, et al. Coffee, 27. Cornelis MC, El-Sohemy A, Kabagambe EK, Campos H.
caffeine-related genes, and ParkinsonÕs disease: a case– Coffee, CYP1A2 genotype, and risk of myocardial control study. Mov Disord 2008; 23: 2033–2040.
infarction. JAMA 2006; 295: 1135–1141.
9. Tan EK, Lu ZY, Fook-Chong SM, et al. Exploring an 28. Xu K, Xu Y-H, Chen J-F, Schwarzschild MA. Neuro- interaction of adenosine A2A receptor variability with protection by metabolites of caffeine may prolong its coffee and tea intake in ParkinsonÕs disease. Am J Med neuroprotective effect in the MPTP model of PD. Soc Genet B Neuropsychiatr Genet 2006; 141B: 634–636.
10. Tan EK, Chua E, Fook-Chong SM, et al. Association between caffeine intake and risk of ParkinsonÕs diseaseamong fast and slow metabolizers. Pharmacogenet 11. Van Den Eeden SK, Tanner CM, Bernstein AL, et al.
In addition to the main authors listed above, the fol- Incidence of ParkinsonÕs disease: variation by age, gender, lowing individuals also contributed to the study as and race/ethnicity. Am J Epidemiol 2003; 157: 1015–1022.
12. Kang GA, Bronstein JM, Masterman DL, et al. Clinical Stanford University School of Medicine (Stanford, characteristics in early ParkinsonÕs disease in a central CA; PEAK study): Dan Bloch, PhD. (Department of California population-based study. Mov Disord 2005; 20:1133–1142.
Health research & Policy) and Greer Murphy, M.D., 13. Mayeux R, Marder K, Cote LJ, et al. The frequency of PhD. (Department of Psychiatry); Shannon Brady, idiopathic ParkinsonÕs disease by age, ethnic group, and Amita Aggarwal and Xia Liu (Department of sex in northern Manhattan, 1988-1993. Am J Epidemiol Genetics, Stanford University School of Medicine, Stanford, CA and Stanford Human Genome Center, 14. Marder K, Logroscino G, Alfaro B, et al. Environmental risk factors for ParkinsonÕs disease in an urban multieth- Palo Alto, CA); Holly Tabor, PhD. (originally at nic community. Neurology 1998; 50: 279–281.
Stanford Human Genome Center, now at University 15. Alavanja MC, Sandler DP, McMaster SB, et al. The of Washington and Seattle ChildrenÕs Hospital), Eddie agricultural health study. Environ Health Perspect 1996; European Journal of Neurology Ó 2011 EFNS European Journal of Neurology 18, 756–765 Columbia University (New York, NY; Columbia Pacific Health Research Institute and Veterans Af- studies): Lorraine Clark PhD. (Taub Institute for fairs Pacific Islands Health Care System (Honolulu, HI; Research on AlzheimerÕs Disease & the Aging Brain, HAAS Study): Lon White M.D., Kamal Masaki, M.D.
Department of Pathology); Ming-Xin Tang PhD.
Division of Research, Kaiser Foundation Research (Taub Institute for Research on AlzheimerÕs Disease & Institute (Oakland, CA; PEAK study): Amethyst the Aging Brain, The Gertrude H. Sergievsky Center, Leimpeter, MS, Kathleen Albers, MPH, Allan Bern- College of Physicians & Surgeons).
National Institute of Environmental Health Sciences The ParkinsonÕs Institute (Sunnyvale, CA; FAME (Research Triangle Park, NC; FAME study): Dale study): Monica Korell, MS, Grace Bhudikhanok, PhD., Sandler, PhD., and Jane Hoppin, PhD. (National Sam Goldman, M.D., MPH, and William Langston, Institute of Environmental Health Sciences, National Institutes of Health, Research Triangle Park, NC).
UCLA (Los Angeles, CA; PEG study): Jeff Bronstein National Cancer Institute (Bethesda, M.D.; FAME M.D. and Yvette M. Bordelon M.D. (UCLA School of Medicine, Department of Neurology, Los Angeles, CA).
Ó 2011 The Author(s)European Journal of Neurology Ó 2011 EFNS European Journal of Neurology 18, 756–765

Source: http://research.hudsonalpha.org/Myers/wp-content/uploads/2012/05/Coffee-ADORA2A-and-CYP1A2-the-caffeine-connection-in-Parkinsons-disease.pdf