Association between Coffee Consumption and Incident Risk of Disabling Dementia in Elderly Japanese: The Ohsaki Cohort 2006 Study

Abstract

Epidemiological studies of the association between coffee consumption and dementia have yielded inconsistent results. Therefore, we investigated the association between coffee consumption and incident risk of dementia in an elderly Japanese population. 23,091 subjects aged ≥65 y living in Ohsaki City, northeastern Japan, responded to the baseline survey in 2006. Of these, we analyzed 13,137 subjects who gave informed consent and were not disabled at baseline. The outcome was the incidence of disabling dementia defined by usage of the Long-term Care Insurance database. We used the Cox proportional hazards regression model for multivariate analysis. During 5.7 y of follow-up period, we identified 1,107 cases of incident dementia. Overall, coffee consumption was significantly associated with a lower risk of incident dementia. The multivariate-adjusted HRs for the incidence of dementia according to coffee consumption categories (never, occasionally, 1-2 cups/d, and ≥3 cups/d) were 1.00, 0.73 (95% CI, 0.62–0.86), 0.72 (95% CI, 0.61–0.84), and 0.82 (95% CI, 0.65–1.02; p for trend = 0.009), respectively. In addition, this significant inverse association was more remarkable among women, non-smokers, and non-drinkers. Coffee consumption is significantly associated with a lower risk of incident dementia.

Keywords

Coffee cohort dementia elderly Japan

INTRODUCTION

With the aging of the world population, dementia has become a growing burden. In Japan, the number of those demented in 2012 was estimated to be more than 4 million, and by 2025 it is assumed that this will increase to more than 7 million [1]. Globally, the number of those living with dementia in 2015 is estimated to be 47.5 million, and is projected to reach 75.6 million by 2030 and 135.5 million by 2050 [2].

In order to establish a strategy to prevent or delay cognitive decline, epidemiological studies have searched for modifiable lifestyle factors associated with dementia. A recent report has indicated that seven factors (diabetes, midlife hypertension, midlife obesity, physical inactivity, depression, smoking, and low educational attainment) have a large impact on dementia [3]. Besides these factors, it remains essential to seek nutritional factors that are protective against cognitive impairment. Here, we focused on coffee because it is assumed to be protective against neurodegenerative disease owing to the effect of its constituents such as caffeine. Previous reports have suggested that caffeine acts as a nonselective A1 and A2 adenosine receptor antagonist on cholinergic neurons [4]. However, epidemiologic studies of the association between coffee consumption and dementia risk have yielded inconsistent results [5]. Therefore, we aimed to investigate the association between coffee consumption and the incident risk of dementia in elderly Japanese.

METHODS

Study population

We used data from the Ohsaki Cohort 2006 Study, the details of which have been described elsewhere [6]. In brief, the source population for the baseline survey comprised all 31,694 residents aged 65 y or older in Ohsaki City, Miyagi Prefecture, northeastern Japan, on 1 December 2006. During the baseline period from 1 to 15 December 2006, each resident was given a baseline questionnaire inquiring about dietary habits as well as history of diseases, smoking status, alcohol drinking status, education status, body weight and height, time spent walking per day, psychological distress (Kessler 6-Item Psychological Distress Scale, K6) [7 –10], social support [11], and participation in community activities. Valid responses were received from 23,091 individuals, and these formed the study cohort (Fig. 1). Of these, we excluded 6,333 persons who declined our access to their Long-term Care Insurance (LTCI) information, 2,102 persons who had already been certified as disabled prior to follow-up conducted from 1 April 2007, 62 persons who had died or emigrated from the study district between the start of the baseline survey and follow-up (1 December 2006 to 30 March 2007), 188 persons whose LTCI information was not available, 4 persons who failed to answer the question about cognitive status in the LTCI information section, and 1,265 persons who did not answer the question about coffee consumption frequency in the baseline food frequency questionnaire (FFQ). Consequently, 13,137 subjects were investigated in the present study.

Assessment of coffee consumption

We collected information about the frequency of coffee consumption at baseline through the FFQ. Subjects were asked to select from the following choices: never, occasionally, 1-2 cups/d, 3-4 cups/d, or ≥5 cups/d. We then classified these into four categories by combining the upper two categories into a single category “≥3 cups/d” due to the small sample size of 555 subjects for ≥5 cups/d. In addition, this frequency categorization was more suitable for the moderate coffee consumption in our rural study region. The typical volume of one cup of coffee in the studied region was considered to be 150 mL. We had previously reported the validity and reproducibility of this FFQ by calculating the Spearman’s coefficient of correlation with the four 3-d food records during four seasons within the year, and correlation between the two FFQs 1-y apart, respectively. For the results pertaining to coffee consumption, the validity was 0.70, and the reproducibility was 0.72 [12].

Assessment of other factors

We considered the following variables to be potential confounders: age at baseline, body mass index (BMI) calculated as self-reported weight in kg divided by the square of height in meters (<18.5, 18.5–24.9, ≥25.0 kg/m², missing), a history of stroke, hypertension, myocardial infarction, diabetes mellitus, arthritis, osteoporosis, and fracture (yes, no), cigarette smoking (never, past, current smoker, missing), alcohol consumption (never, past, current, missing), green tea consumption (never, occasionally, 1-2 cups/d, 3-4 cups/d, ≥5 cups/d, missing), education level (age when last graduated; ≤15, 16–18, ≥19 y of age, missing), walking duration (<0.5 h/d, 0.5-1 h/d, >1 h/d, missing), psychological distress score (<13, ≥13, missing) [8], whether or not subjects perceived feelings of social support for all 5 categories (yes, no, missing), and whether or not subjects participated in any of 4 community activities (yes, no, missing). The extent of social support was measured by asking the following questions: is there someone to whom you can talk when in trouble, whom you can consult when ill, who helps you with your daily housework, who can take you to a hospital when ill, and who can take care of you if you become bedridden? Each question required either a yes or no answer. Similarly, the extent of participation in community activities was measured by asking the frequency of participation in the following activities: 1) neighborhood associations, 2) sports, exercise, or leisure, 3) volunteering for activities related to nonprofit organizations, and 4) any other type of social gatherings. The prepared answers to these questions were never, a few times/y, 1 time/mo, 2-3 times/mo, 1 time/wk, 2-3 times/wk, and ≥4 times/wk.

Case ascertainment

As the primary outcome, we defined the incidence of “disabling dementia” using the two steps of certification for LTCI in Japan. LTCI is a national mandatory nursing insurance implemented in April 2000 to support the daily activities of those aged ≥65 y and in need of long-term care services provided by municipalities [13 –16]. First, as we described previously [17], we confirmed the incidence of “disability” when eligible persons were certified as disabled by the Municipal Certification Committee using two information sources: the computerized assessment of functional disability processed from a standardized questionnaire by the Ministry of Health, Labour, and Welfare, and the Doctor’s Opinion Paper reported by the primary physicians, mostly general practitioners, or psychiatrists [18]. The questionnaire includes 74 items, composed of 6 dimensions: motor function, activity of daily living, cognitive function, mental and behavioral disorders, adaptation to social life, and use of medical procedures. The Doctor’s Opinion Paper gives information about the major disease that has caused disability. A previous study has reported that this LTCI certification was significantly associated with subjects’ Mini-Mental State Examination (MMSE) score [19]. Several epidemiologic studies have already utilized this LTCI certification as a measure of incident disability [20 –22].

Secondly, of those disabled, we ascertained the incidence of “disabling dementia” ranked II and over on the Dementia Scale (Degree of Independence in Daily Living for Elderly with Dementia) in the Doctor’s Opinion Paper, which is the most common cutoff point used in the LTCI system and previous outcome measures [23 –25]. The Dementia Scale consists of six ranks: rank 0, not demented; rank I, with mild cognitive dysfunction but no dementia-related symptoms or behavioral disturbance; rank II, moderately demented with dementia-related behavioral disturbance; and through rank V as severely demented with the need for medical treatment [26]. A previous study has shown that the Dementia Scale is well correlated with MMSE scores (R² = 0.990) [27].

We obtained annually in December a data set including information on LTCI certification, death, or emigration based on an agreement with Ohsaki City Government about the secondary use of statistical data regulated by the law on Epidemiologic Research and Privacy Protection.

Ethics

At baseline, we considered each subject’s response to the questionnaire as their consent to participate in the baseline survey and follow-up study. Also, we confirmed LTCI certification information for those who submitted written consent along with the questionnaires. The study protocol was reviewed and approved by the Ethics Committee of Tohoku University Graduate School of Medicine.

Statistical analyses

We counted the person-years for each subject from 1 April 2007 until the date of incident dementia, date of death, date of emigration from the study district, incident functional disability without dementia, or the end of study period (30 November 2012), whichever occurred first. In this analysis, deaths without LTCI certification were censored.

Statistical differences in baseline characteristics among the four categories of coffee consumption frequency were tested using ANOVA for continuous variables and the chi-square test for categorical variables. We then used the multivariate-adjusted Cox proportional regression model to calculate the hazard ratio (HR) and 95% confidence interval (CI) of incident disabling dementia for each coffee consumption category, considering the lowest one (for those who never consumed coffee) as reference. Dummy variables for missing data on categorical covariates were created and included in multivariate analyses. The p-values for linear trend were also calculated by considering the coffee consumption categories as a continuous variable: 1 for the lowest category up to 4 for the highest. We also conducted analyses stratified by several factors that are thought to be risk factors for cognitive impairment.

In this study, we had no baseline data on subjects who were cognitively impaired but not certified as disabled. Therefore, these subjects were included in the baseline survey, leading to a possibility of reverse causality. In an attempt to exclude those suspected of cognitive impairment, we further conducted sensitivity analysis by using alternative data in the “Kihon Checklist”, which we also inquired at the baseline survey. This checklist contains the following three binary questions: (1) people around you say that you repeat the same thing and have become forgetful, (2) you do not make phone calls by yourself, and (3) you find yourself not knowing today’s date. The elderly was suspected of subjectively-based cognitive decline if they answered yes to at least one of the three questions. This checklist was developed by the Ministry of Health, Labor, and Wealth, and had been officially used throughout the country. However, the data is subjectively-based and its validation is limited compared with the Clinical Dementia Rating [28]. Therefore, we used this data only for the sensitivity analysis as exclusion criteria, with the aim to examine the possibility of reverse causality. We used SAS version 9.3 (SAS Institute, Cary, North Carolina) for all analyses. All tests were two-tailed, and differences at p <0.05 were considered statistically significant.

RESULTS

During the 5.7 y of follow-up, only 124 subjects were lost due to emigration from the district without being certified as disabled (follow-up rate; 99.1% ). Among a total of 65,249 person-years accrued, 1,107 cases of incident disabling dementia occurred (incidence rate; 8.4% ), and 870 subjects died due to any cause without certification of disability. Table 1 (see also Supplementary Table 1) shows the baseline characteristics of the total subjects according to the frequency of coffee consumption. As shown, subjects with a higher frequency of coffee consumption tended to be male, younger, and obese, to have lower rates of hypertension and diabetes mellitus, to smoke and drink currently, to have a higher education level, to walk less, and to be less distressed.

Table 2 shows the association between coffee consumption and the incident risk of disabling dementia. As shown in Model 3, there was a significant inverse association between coffee consumption and incident disabling dementia after adjustment for covariates. The HRs for the incidence of dementia according to coffee consumption categories (never, occasionally, 1-2 cups/d, and ≥3 cups/d) were 1.00, 0.73 (95% CI, 0.62–0.86), 0.72 (95% CI, 0.61–0.84), and 0.82 (95% CI, 0.65–1.02; p for trend = 0.009), respectively. Furthermore, in Model 4, there was a stronger significant inverse association after exclusion of 311 cases of incident disabling dementia during the first 2 y of follow-up (p for trend = 0.008). Analysis with additional exclusion of 486 subjects who died within the first 2 y of follow-up showed similar HRs (95% CIs) with Model 4 (data not shown). On the other hand, when analyzed without combining the upper two categories, we observed a U-shaped association (Supplementary Table 2). We also conducted analysis with a more severe endpoint of Rank III and over on the Dementia Scale, resulting in no significant association (Supplementary Table 3).

For sensitivity analysis, we analyzed 8,121 subjects after exclusion of 5,016 subjects who subjectively reported cognitive decline in the Kihon Checklist. The HRs (95% CIs) according to coffee consumption categories (never, occasionally, 1-2 cups/d, and ≥3 cups/d) were 1.00, 0.61 (0.46–0.80), 0.63 (0.48–0.83), and 0.77 (0.54–1.09; p for trend = 0.083), respectively (Table 3). Also, there was a significant inverse association for coffee consumption after additional exclusion of 81 cases of incident disabling dementia within the first 2 years of follow-up (p for trend = 0.022, also shown in Table 3). Accordingly, the inverse association of coffee consumption had remained even after consideration of reverse causality.

Table 4 shows analysis stratified according to gender, age, BMI level, smoking and alcohol consumption status, as well as interactions between coffee consumption categories and these factors, which were tested through addition of cross-product terms to the multivariate model; the p-interaction was considered as the p-value for the cross-product term. The results demonstrated a discrepancy in the risk for incident disabling dementia between men and women with significant interaction. We found a significant inverse association between coffee consumption and the incident risk of disabling dementia among women with HRs according to coffee consumption categories (never, occasionally, 1-2 cups/d, and ≥3 cups/d) of 1.00, 0.64 (95% CI, 0.52–0.79), 0.65 (95% CI, 0.53–0.81), and 0.74 (95% CI, 0.54–1.01; p for trend = 0.003), respectively. On the other hand, there was no such association among men, with a p-value for interaction of 0.044.

Similarly, the results for never and ever smokers differed with significant interaction. There was a significant inverse association among never smokers with HRs according to coffee consumption categories (never, occasionally, 1-2 cups/d, and ≥3 cups/d) of 1.00, 0.69 (95% CI, 0.56–0.86), 0.73 (95% CI, 0.58–0.91), and 0.78 (95% CI, 0.56–1.08; p for trend = 0.030), respectively. In contrast, there was no such association among ever smokers, with a p-value for interaction of 0.044.

In addition, the results for never and ever alcohol drinkers also differed with significant interaction. There was a significant inverse association among never drinkers with HRs according to coffee consumption categories (never, occasionally, 1-2 cups/d, and ≥3 cups/d) of 1.00, 0.63 (95% CI, 0.50–0.79), 0.63 (95% CI, 0.51–0.79), and 0.65 (95% CI, 0.47–0.90; p for trend = 0.001), respectively. Meanwhile, there was no such association among ever drinkers, with a p-value for interaction of <0.001.

The results of analysis stratified according to age and BMI level showed no significant interactions. There were inverse associations for individuals aged both <75 y and ≥75 y. Similar inverse associations were seen for individuals with BMI values of both <25 and ≥25.

DISCUSSION

We observed a significant inverse association between coffee consumption and the incident risk of disabling dementia. A strong inverse association remained even after exclusion of incident disability within the first 2 y of follow-up.

In addition, these significant inverse associations were more remarkable among women, non-smokers, and non-drinkers in consideration of the significant interactions. The favorable effect observed in women relative to men may be explained by the larger bioavailability of caffeine due to lower caffeine metabolism in women [29]. Also, the favorable effect seen in never smokers relative to ever smokers may be explained by the fact that caffeine has a shorter half-life in smokers [30]. A significant inverse association was seen only among those aged ≥75 years at baseline. This may have been due to the longer exposure of these subjects to caffeine. With regard to stratification by BMI level, the inverse association was more pronounced among subjects with BMI <25. Thus, obesity can be assumed as a risk factor. However, though not significant, there was an inverse association in the obese group. This finding suggests a probable protective effect of coffee consumption against vascular inflammation due to obesity. Finally, although a systematic review has reported an association between moderate alcohol consumption and a lower risk of cognitive impairment [31], a null association for ever-drinkers was found in our study. Therefore, more information about the amount of alcohol consumption should have been obtained for analysis.

Up to now, epidemiologic studies of the association between coffee consumption and dementia have yielded inconsistent results. Nine cohort studies had been published up to 2010; three studies demonstrated a significant inverse association with dementia, Alzheimer’s disease, and cognitive impairment [32 –34], one longitudinal study demonstrated a reduction in cognitive decline score [35], and the others demonstrated no association [36 –40]. A meta-analysis of only these nine cohort studies conducted in 2010 concluded that there was no association, with a risk ratio of 0.93 (95% CI, 0.83–1.04, I² = 0.0% ) [41]. Since then, three cohort studies have been published, and all suggested no association [42 –44]. Accordingly, these results were additionally included in a recent meta-analysis [45] and a systematic review [46], but again no significant association was demonstrated.

A number of reasons for the discrepancy between the results of our study and previous studies can be considered. First, our sample size was larger than those of previous studies, which involved between about 500 and 5,000 subjects. Even the most influential study [38] included in the meta-analysis [41], including 4,197 men and 2,820 women, found no significant association for either gender. Second, our multivariate analysis included as a covariate the psychological distress status of individuals. It has been reported that this factor is both inversely associated with coffee consumption [47] and a risk factor for dementia [48]. Third, we also included social support or community activities for individuals as covariates, because it has been reported that these factors are inversely related to dementia risk [49]. Besides, elderly Japanese tend to consume coffee more frequently in a social setting, and this would have impacted on our results.

With regard to our endpoint, we could not differentiate between dementia subtypes because these are not required upon application for LTCI certification. However, previous studies have reported that coffee consumption is associated with a lower risk of most types of dementia, including Alzheimer’s disease [33] and vascular dementia [50]. One possible mechanism for this is the protective effect of caffeine against neurodegeneration through its action as a nonselective A1 and A2 adenosine receptor antagonist on cholinergic neurons [4]. Also, coffee consumption is associated with a lower risk of diabetes and hypertension, which would lead to protection from arterial sclerosis and inflammation [51]. Notifications are also needed that our subjects were disabled in addition to having dementia. However, in AGES, a large cohort in Japan, the incidence rate of functional disability and disabling dementia was 3.82 and 1.93 per 100 person-years, respectively, during 2003–2013 [52]. In comparison, our incidence rate for functional disability and disabling dementia during the 5.7 y of follow-up was 3.82 and 1.70 per 100 person-years, respectively. Therefore, our results can possibly be generalized to the Japanese population. Further discussion will be needed with regard to comparison with cohorts in other countries.

Our study had several limitations. First, we did not ask the individuals about the frequency of coffee consumption in mid-life or after baseline. In addition, we did not exclude those with cognitive impairment among those not certified as disabled at baseline due to the lack of data for neuropsychological tests. These limitations might have led to reverse causality due to changes in coffee consumption habits from unhealthy conditions in mid-life. In order to minimize this reverse causality, we conducted further analysis with exclusion of incident cases of disabling dementia during the first 2 y of follow-up. As a result, a stronger inverse association was revealed. In addition to examine the possibility of reverse causality, the sensitivity analysis with exclusion of those who subjectively reported cognitive decline in the Kihon Checklist showed a similar inverse association. Therefore, our results would suggest the beneficial effect of coffee consumption. Second, we did not ask whether the subjects consumed caffeinated or decaffeinated coffee, although we consider caffeinated coffee to be much more popular in Japan. Third, we did not measure occupational level or premorbid intellectual function, which might have impacted on the findings. Furthermore, we lacked data on medications that can be protective against neurodegeneration, such as nonsteroidal anti-inflammatory drugs [53], or those that affect caffeine metabolism.

In conclusion, we have found that coffee consumption is significantly related to a lower risk of disabling dementia in this elderly Japanese population.

Footnotes

ACKNOWLEDGMENTS

We are grateful to all the participants and staff who participated in the Ohsaki Cohort 2006 Study.

This research was supported by a grant-in-aid from the Ministry of Health, Labour and Welfare, Health and Labor Sciences research grants, Japan (Comprehensive Research on Cardiovascular and Life-Style Related Diseases: H26-Junkankitou [Seisaku]-Ippan-001).

Authors’ disclosures available online ().

Appendix

References

Ministry of Health, Labour, and Welfare of Japan (2015) A comprehensive strategy for the promotion of dementia Measures - towards a community friendly to the elderly with dementia, etc.- (A New Orange Plan) (in Japanese), http://www.mhlw.go.jp/stf/houdou/0000072246.html, Accessed on Aug 9, 2015

Alzheimer’s Disease International (2013) Policy brief for G8 heads of government: The global impact of dementia 2013-2050, http://www.alz.co.uk/research/G8-policy-brief, Accessed on Aug 9, 2015.

Norton

, Matthews

, Barnes

, Yaffe

, Brayne

(2014) Potential for primary prevention of Alzheimer’s disease: An analysis of population-based data. Lancet Neurol 13, 788–794.

Arendash

, Schleif

, Rezai-Zadeh

, Jackson

, Zacharia

, Cracchiolo

, Shippy

, Tan

(2006) Caffeine protects Alzheimer’s mice against cognitive impairment and reduces brain beta-amyloid production. Neuroscience 142, 941–952.

Gelber

, White

(2014) Coffee and the risk of dementia and cognitive impairment. In Diet and Nutrition in Dementia and Cognitive Decline, 1st Edition, Martin

, Preedy

, eds, Academic Press, pp. 891–902.

Kuriyama

, Nakaya

, Ohmori-Matsuda

, Shimazu

, Kikuchi

, Kakizaki

, Sone

, Sato

, Nagai

, Sugawara

, Tomata

, Akhter

, Higashiguchi

, Fukuchi

, Takahashi

, Hozawa

, Tsuji

(2010) The Ohsaki Cohort 2006 Study: Design of study and profile of participants at baseline. J Epidemiol 20, 253–258.

Kessler

, Green

, Gruber

, Sampson

, Bromet

, Cuitan

, Furukawa

, Gureje

, Hinkov

, Hu

, Lara

, Mneimneh

, Myer

, Oakley-Browne

, Posada-Villa

, Sagar

, Viana

, Zaslavsky

(2013) Screening for serious mental illness in the general population with the K6 screening scale: Results from the WHO World Mental Health (WMH) survey initiative. Int J Methods Psychiatr Res 19(Suppl 1), 4–22.

Furukawa

, Kawakami

, Saitoh

, Ono

, Nakane

, Nakamura

, Tachimori

, Iwata

, Uda

, Nakane

, Watanabe

, Naganuma

, Hata

, Kobayashi

, Miyake

, Takeshima

, Kikkawa

(2008) The performance of the Japanese version of the K6 and K10 in the World Mental Health Survey Japan. Int J Methods Psychiatr Res 17, 152–158.

Furukawa

, Kessler

, Slade

, Andrews

(2003) The performance of the K6 and K10 screening scales for psychological distress in the Australian National Survey of Mental Health and Well-Being. Psychol Med 33, 357–362.

10.

Kessler

, Andrews

, Colpe

, Hiripi

, Mroczek

, Normand

, Walters

, Zaslavsky

(2002) Short screening scales to monitor population prevalences and trends in non-specific psychological distress. Psychol Med 32, 959–976.

11.

Muraoka

, Ikichi

, Ihara

(1996) The physical and psychological and social background factor of elderly depression in the community (Japanese). Ronen Seishin Igaku Zasshi 7, 397–407.

12.

Shimazu

, Tsubono

, Kuriyama

, Ohmori

, Koizumi

, Nishino

, Shibuya

, Tsuji

(2005) Coffee consumption and the risk of primary liver cancer: Pooled analysis of two prospective studies in Japan. Int J Cancer 116, 150–154.

13.

Ikegami

(1997) Public long-term care insurance in Japan. JAMA 278, 1310–1314.

14.

Imai

, Fujii

, Fukuda

, Nakao

, Yahata

(2008) Health-related quality of life and beneficiaries of long-term care insurance in Japan. Health Policy 85, 349–355.

15.

Ministry of Health, Labor, and Welfare of Japan (2008) Long-term care, health and welfare services for the elderly, http://www.mhlw.go.jp/english/policy/care-welfare/care-welfare-elderly/index.html, Accessed on Aug 9, 2015

16.

Tsutsui

, Muramatsu

(2005) Care-needs certification in the long-term care insurance system of Japan. J Am Geriatr Soc 53, 522–527.

17.

Tomata

, Watanabe

, Sugawara

, Chou

, Kakizaki

, Tsuji

(2013) Dietary patterns and incident functional disability in elderly Japanese: The Ohsaki Cohort 2006 Study. J Gerontol A Biol Sci Med Sci 69, 843–851.

18.

Moriyama

, Tamiya

, Kamimura

, Sandoval

, Luptak

(2014) Doctors’ Opinion Papers in Long-term Care Need Certification in Japan: Comparison between clinic and advanced treatment hospital settings. Public Policy Adm Res 4, 31–37.

19.

Arai

, Zarit

, Kumamoto

, Takeda

(2003) Are there inequities in the assessment of dementia under Japan ’s LTC insurance system?. Int J Geriatr Psychiatry 18, 346–352.

20.

Tomata

, Kakizaki

, Nakaya

, Tsuboya

, Sone

, Kuriyama

, Hozawa

, Tsuji

(2012) Green tea consumption and the risk of incident functional disability in elderly Japanese: The Ohsaki Cohort 2006 Study. Am J Clin Nutr 95, 732–739.

21.

Hozawa

, Sugawara

, Tomata

, Kakizaki

, Tsuboya

, Ohmori-Matsuda

, Nakaya

, Kuriyama

, Fukao

, Tsuji

(2013) Relationship between serum isoflavone levels and disability-free survival among community-dwelling elderly individuals: Nested case-control study of the Tsurugaya project. J Gerontol A Biol Sci Med Sci 68, 465–472.

22.

Kondo

, Kawachi

, Hirai

, Kondo

, Subramanian

, Hanibuchi

, Yamagata

(2009) Relative deprivation and incident functional disability among older Japanese women and men: Prospective cohort study. J Epidemiol Community Health 63, 461–467.

23.

Yamagishi

, Ikeda

, Moriyama

, Chei

, Noda

, Umesawa

, Cui

, Nagao

, Kitamura

, Yamamoto

, Asada

, Iso

(2014) Serum coenzyme Q10 and risk of disabling dementia: The Circulatory Risk in Communities Study (CIRCS). Atherosclerosis 237, 400–403.

24.

Chei

, Yamagishi

, Ikeda

, Noda

, Maruyama

, Cui

, Imano

, Kiyama

, Kitamura

, Asada

, Iso

(2014) C-reactive protein levels and risk of disabling dementia with and without stroke in Japanese: The Circulatory Risk in Communities Study (CIRCS). Atherosclerosis 236, 438–443.

25.

Ikeda

, Yamagishi

, Tanigawa

, Cui

, Yao

, Noda

, Umesawa

, Chei

, Yokota

, Shiina

, Harada

, Murata

, Asada

, Shimamoto

, Iso

(2008) Cigarette smoking and risk of disabling dementia in a Japanese rural community: A nested case-control study. Cerebrovasc Dis 25, 324–331.

26.

Health and Welfare Bureau for the Elderly of Japan (1993) Standardized assessment manual for the levels of Activity Daily Living deficiencies among elderly with dementia (Japanese), http://www.mhlw.go.jp/file/06-Seisakujouhou-12300000-Roukenkyoku/0000077237.pdf, Last updated Apr, 2015, Accessed on Aug 9, 2015.

27.

Hisano

(2009) The relationship between Revised Hasegawa Dementia Scale (HDS-R), Mini-Mental State Examination (MMSE) and Bed-fast Scale, Dementia Scale (Japanese). Jpn J Geriatr Psychiatry 20, 883–891.

28.

Meguro

(2012) The validity of the Basic Checklist in the old-old population – The Kurihara Project (Japanese). Jpn J Geriatr Psychiatry 23, 725–730.

29.

Carrillo

, a , Benitez

(1996) CYP1A2 activity, gender and smoking, as variables influencing the toxicity of caffeine. Br J Clin Pharmacol 41, 605–608.

30.

Benowitz

, Hall

(1989) Persistent increase in caffeine concentrations in people who stop smoking. BMJ 298, 1075–1076.

31.

Peters

, Peters

, Warner

, Beckett

, Bulpitt

(2008) Alcohol, dementia and cognitive decline in the elderly: A systematic review. Age Ageing 37, 505–512.

32.

Eskelinen

, Ngandu

, Tuomilehto

, Soininen

, Kivipelto

(2009) Midlife coffee and tea drinking and the risk of late-life dementia: A population-based CAIDE study. J Alzheimers Dis 16, 85–91.

33.

Lindsay

, Laurin

, Verreault

, Hebert

, Helliwell

, Hill

, McDowell

(2002) Risk factors for Alzheimer’s disease: A prospective analysis from the Canadian Study of Health and Aging. Am J Epidemiol 156, 445–453.

34.

Santos

, Lunet

, Azevedo

, de MendonÇa

, Ritchie

, Barros

(2010) Caffeine intake is associated with a lower risk of cognitive decline: A cohort study from Portugal. J Alzheimers Dis 20(Suppl 1), S175–S185.

35.

van Gelder

, Buijsse

, Tijhuis

, Kalmijn

, Giampaoli

, Nissinen

, Kromhout

(2007) Coffee consumption is inversely associated with cognitive decline in elderly European men: The FINE Study. Eur J Clin Nutr 61, 226–232.

36.

Tyas

, Manfreda

, Strain

, Montgomery

(2001) Risk factors for Alzheimer’s disease: A population-based, longitudinal study in Manitoba, Canada. Int J Epidemiol 30, 590–597.

37.

van Boxtel

, Schmitt

, Bosma

, Jolles

(2003) The effects of habitual caffeine use on cognitive change: A longitudinal perspective. Pharmacol Biochem Behav 75, 921–927.

38.

Ritchie

, Carrière

, de Mendonca

, Portet

, Dartigues

, Rouaud

, Barberger-Gateau

, Ancelin

(2007) The neuroprotective effects of caffeine: A prospective population study (the Three City Study). Neurology 69, 536–545.

39.

, Feng

, Niti

, Kua

, Yap

(2008) Tea consumption and cognitive impairment and decline in older Chinese adults. Am J Clin Nutr 88, 224–231.

40.

Laitala

, Kaprio

, Koskenvuo

, Räihä

, Rinne

, Silventoinen

(2009) Coffee drinking in middle age is not associated with cognitive performance in old age. Am J Clin Nutr 90, 640–646.

41.

Santos

, Costa

, Santos

, Vaz-Carneiro

, Lunet

(2010) Caffeine intake and dementia: Systematic review and meta-analysis. J Alzheimers Dis 20(Suppl 1), S187–S204.

42.

Gelber

, Petrovitch

, Masaki

, Ross

, White

(2011) Coffee intake in midlife and risk of dementia and its neuropathologic correlates. J Alzheimers Dis 23, 607–615.

43.

Noguchi-Shinohara

, Yuki

, Dohmoto

, Ikeda

, Samuraki

, Iwasa

, Yokogawa

, Asai

, Komai

, Nakamura

, Yamada

(2014) Consumption of green tea, but not black tea or coffee, is associated with reduced risk of cognitive decline. PLoS One 9, e96013.

44.

Mirza

, Tiemeier

, de Bruijn

, Hofman

, Franco

, Kiefte-de Jong

, Koudstaal

, Ikram

(2014) Coffee consumption and incident dementia. Eur J Epidemiol 29, 735–741.

45.

Kim

, Kwak

, Myung

(2015) Caffeine intake from coffee or tea and cognitive disorders: A meta-analysis of observational studies. Neuroepidemiology 44, 51–63.

46.

Panza

, Solfrizzi

, Barulli

, Bonfiglio

, Guerra

, Osella

, Seripa

, Sabbà

, Pilotto

, Logroscino

(2015) Coffee, tea, and caffeine consumption and prevention of late-lifecognitive decline and dementia: A systematic review. J Nutr Health Aging 19, 313–328.

47.

Lucas

, Mirzaei

, Pan

, Okereke

, Willett

, O’Reilly

ÉJ

, Koenen

, Ascherio

(2011) Coffee, caffeine, and risk of depression among women. Arch Intern Med 171, 1571–1578.

48.

Deckers

, van Boxtel

, Schiepers

, de Vugt

, Muñ oz Sánchez

, Anstey

, Brayne

, Dartigues

, Engedal

, Kivipelto

, Ritchie

, Starr

, Yaffe

, Irving

, Verhey

, Köhler

(2015) Target risk factors for dementia prevention: A systematic review and Delphi consensus study on the evidence from observational studies. Int J Geriatr. Psychiatry 30, 234–246.

49.

James

, Wilson

, Barnes

, Bennett

(2011) Late-life social activity and cognitive decline in old age. J Int Neuropsychol Soc 17, 998–1005.

50.

Vercambre

, Berr

, Ritchie

, Kang

(2013) Caffeine and cognitive decline in elderly women at high vascular risk. J Alzheimers Dis 35, 413–421.

51.

Andersen

, Jacobs

Jr , Carlsen

, Blomhoff

(2006) Consumption of coffee is associated with reduced risk of death attributed to inflammatory and cardiovascular diseases in the Iowa Women’s Health Study. Am J Clin Nutr 83, 1039–1046.

52.

Saito

, Kondo

, Ojima

, Hirai

, JAGES group (2015) Criteria for social isolation based on associations with health indicators among older people. A 10-year follow-up of the Aichi Gerontological Evaluation Study (Japanese). Nihon Koshu Eisei Zasshi 62, 95–105.

53.

Asanuma

, Miyazaki

, Ogawa

(2004) Neuroprotective effects of nonsteroidal anti-inflammatory drugs on neurodegenerative diseases. Curr Pharm Des 10, 695–700.