Translational Psychiatry www.nature.

com/tp

ARTICLE OPEN

Tea consumption and risk of incident dementia: A prospective

cohort study of 377 592 UK Biobank participants
✉ 1,2,3 ✉
He-Ying Hu1,4, Bang-Sheng Wu2,4, Ya-Nan Ou1, Ya-Hui Ma1, Yu-Yuan Huang2, Wei Cheng 3
, Lan Tan1 and Jin-Tai Yu

© The Author(s) 2022

As a widely consumed beverage, tea boasts diverse health benefits. Herein, we aimed to investigate the association between tea
consumption and dementia risk. We conducted a prospective cohort study with 377 592 UK Biobank participants during a 9-year
follow-up. Cox regression models adjusted for age, sex, ethnicity, Townsend deprivation index, education, body mass index, lifestyle
factors, dietary factors and apolipoprotein E4 status were used to examine the association of tea consumption with dementia risk.
Subgroup analyses stratified by age, sex and forms of dementia (Alzheimer’s disease [AD] and vascular dementia [VD]) were
performed. Moreover, the restricted cubic splines were used to calculate the nonlinear relationship between daily dosage of tea
and dementia risk. After adjustment for all covariates, tea drinkers were 16% (95% confidence interval: 8–23) less likely to develop
dementia compared with non-drinkers. Moderate consumption (1–6 cups/day) of tea exerted significant protective effects.
Subgroup analyses showed that mid-aged participants or males benefited more from tea consumption. Moreover, moderate
1234567890();,:

drinkers had a 16–19% lower hazard of AD and a 25–29% lower hazard of VD. Furthermore, a U-shaped association between tea
consumption and dementia risk was shown (Pnon-linearity = 7E−04), and the consumption of around three cups per day showed the
strongest protective effect. Within 3 cups/day, drinking one extra cup of tea per day brought a 6% reduction of incidence. In
conclusion, moderate consumption of tea was significantly associated with a reduced risk of dementia, suggesting that tea
consumption could be a modifiable lifestyle factor for dementia.

Translational Psychiatry (2022)12:171 ; https://doi.org/10.1038/s41398-022-01923-z

INTRODUCTION pathology [16–18]. Mechanistic studies revealed that tea biomo-

Dementia is an important public health concern currently, with lecules (e.g., epigallocatechin gallate [EGCG], the predominant
around 55 million cases worldwide and an incidence of more than polyphenol in green tea) could invoke extensive cellular pathways
10 million new cases per annum [1]. Dementia has become a of antioxidants and activities of neurorescue, which might prevent
major cause of disability, dependency as well as mortality among memory deficits [11, 12, 19]. Also, tea biomolecules were found to
older people, and has posed substantial burden on patients, their have anti-inflammatory properties, thereby halting the progres-
carers, families as well as the society [2, 3]. Dementia is composed sion of cognitive decline [20, 21]. Tea was reported to play a
of Alzheimer’s disease (AD), which may contribute to 60–70% of neuroprotective role in the cognitive decline not only resulting
cases, vascular dementia (VD), which may contribute to 25% of from aging, but also from ischemia-reperfusion [22]. Previous
cases, and other forms of dementia [1, 4]. The development of studies suggested that tea had positive impacts (e.g., vasodilative
dementia is associated with genetic and environmental factors. functions) on cerebral blood vessels [23]. It is worth noting that
Among the environmental factors, diet is a potentially modifiable coffee was found to mitigate the pathological progression of
lifestyle factor for preventing dementia [4–7]. dementia as well as slow cognitive decline in recent animal
Tea is a widely consumed beverage around the world and it studies, and tea shared some molecules with coffee [24, 25]. Taken
contains various kinds of biomolecules such as polyphenols. Tea together, it is plausible to hypothesize that tea may reduce the
intake has been found to be associated with the prevention of incidence of dementia.
various diseases, such as stroke, cardiovascular diseases and The relationship between tea consumption and dementia has
neurodegenerative diseases [8–10]. Recently, accumulating epi- not been well established yet. Although most studies in healthy
demiological studies and systemic reviews have suggested that aging showed that tea was protective against dementia
tea intake can suppress brain aging and ameliorate cognitive [11, 14, 26, 27], some studies reported no significant association
dysfunction [11–14]. A population-based study has found that [28]. The credibility of these findings is easily compromised by the
green tea can mitigate the pathological changes of AD [15]. AD cross-sectional designs, insufficient follow-up periods, small
mouse models showed that the main component of black tea sample sizes or different reference groups. A recent cohort study
could inhibit the deposits of amyloid in the brain and reduce Aβ reported that tea consumption was associated with lower risks of

1
Department of Neurology, Qingdao Municipal Hospital, Qingdao University, Qingdao, China. 2Department of Neurology and Institute of Neurology, Huashan Hospital, Shanghai
Medical College, Fudan University, Shanghai, China. 3Institute of Science and Technology for Brain-Inspired Intelligence, Fudan University, Shanghai, China. 4These authors
contributed equally: He-Ying Hu, Bang-Sheng Wu. ✉email: [email protected]; [email protected]

Received: 10 November 2021 Revised: 18 February 2022 Accepted: 18 February 2022

 H.-Y. Hu et al.
2
dementia, VD, poststroke dementia and stroke [29]. Evidence from continuous); education (higher [with college/university degree or other
human clinical trials suggested that moderate (1–6 cups/day) professional qualification]/lower); body mass index (BMI) (continuous);
rather than excessive consumption of tea could bring diverse lifestyle factors including typical sleep duration (continuous), smoking
health benefits [8, 30]. Therefore, we performed a prospective status (never/previous/current) and alcohol status (never/previous/cur-
cohort study to examine (1) the relationship between daily dosage rent); dietary factors including total consumption of vegetables (contin-
uous), fruit (continuous) and fish (continuous); apolipoprotein E4 (APOE4)
and dementia risk; and (2) the effect modification by age, sex as carrier status (carrier/non-carrier status as defined by genetic information).
well as the effects of tea consumption on separate outcomes (e.g., To investigate whether age or sex modified the effects of tea consumption
AD and VD) using large-scale data from 377 592 older adults from on dementia, the subgroup analyses stratified by age (midlife [≤65 years
the UK Biobank (UKB). old] and late-life [>65 years old]) and sex were conducted, respectively.
Additionally, AD and VD were considered as separate outcomes and the
subgroup analyses stratified by the form of dementia (AD/VD) were
METHODS performed.
Study population To explore the non-linear effects of tea consumption on incident
This is a large-scale prospective cohort study of participants enrolled in the dementia, the restricted cubic splines with five knots were introduced in
UKB, which recruited more than 500,000 participants aged 37–73 years the fully adjusted model using continuous measures of tea consumption.
between 2006 and 2010. The participants attended 1 of 22 assessment Sensitivity analyses were performed after restricting participants to
centers across the United Kingdom, and completed the touchscreen and those with a follow-up time of ≥4 years. In addition, given that underlying
nurse-led questionnaires. Data on clinic, gene, and risk factor were dementia might change the dietary habits before diagnosis, we repeated
obtained at baseline. Since recruitment, the participants have been the analyses after excluding incident dementia cases occurring during the
followed for clinical outcomes including dementia via hospital inpatient first year of follow-up in order to limit the possibility of reverse causality.
records, death certificates and primary care records. The UKB study We also performed sensitivity analyses after excluding individuals with a
received approval from the National Health Service (NHS) NorthWestMul- history of stroke at baseline. P < 0.05 was considered to be statistically
ticenter Research Ethics Committee. All participants provided written significant. R version 4.0.5 [31] and SPSS 25.0 were used for statistical
informed consent according to the Declaration of Helsinki. In this study, we analyses and figure preparation.
excluded participants with prevalent dementia at baseline, and those with
missing data on tea consumption, leaving 417 085 participants included.
RESULTS
Tea consumption Participant characteristics
Data on tea consumption, a self-reported item, were collected at baseline At baseline, a total of 377,592 participants from the primary care
as part of the UKB touchscreen questionnaire, where participants were dataset were included in this study (Fig. 1). Their mean age was
asked ‘How many cups of tea do you drink a day? (including black and 58.49 (SD 6.83) years; 204,980 participants (54.3%) were women.
green tea)’ (UKB Data-field ID:1488), with open-ended responses to the Around 85.1% of the participants reported consuming tea. During
question. Participants were divided into six categories according to tea a median follow-up of 9.09 years (interquartile range [IQR]
consumption, including non-drinkers as well as drinkers consuming 1–2 7.08–10.64), 5122 incident dementia events were recorded.
cups/day, 3–4 cups/day, 5–6 cups/day, 7–8 cups/day, and ≥9 cups/day. Baseline characteristics of participants stratified by the occurrence
of dementia were shown in Table 1, and the missing data for
Dementia diagnoses certain variables were recorded in Supplementary Table 1.
The diagnoses of all-cause dementia (ACD) were obtained using hospital Demented cases were older, more likely to be APOE4 carriers,
inpatient records from the Hospital Episode Statistics for England, the more economically deprived, less educated, more likely to smoke,
Scottish Morbidity Record data for Scotland, and the Patient Episode and less likely to be addicted to alcohol. Demented cases also had
Database for Wales. Additional cases were detected through linkage to higher BMIs, longer sleep, and more frequent consumption of
death register data from NHS Digital. Diagnoses were recorded using the
International Classification of Diseases (ICD-9 and ICD-10) coding system. vegetables, fruit as well as fish. More males than females were
Individuals with ACD were identified as having a diagnosis for AD (code diagnosed with dementia in the study population. However, there
331.0 in ICD-9 and codes F00 and G30 in ICD-10), VD (codes 290.4 in ICD-9 was no significant difference in the consumption of tea between
and codes F01 and I67.3 in ICD-10), or other dementia classifications (codes demented cases and non-demented participants (the proportion
290, 291.2, 294.1, 331.0-331.2 and 331.5 in ICD-9; codes A81.0, F02, F03, of non-drinkers 14.8% vs. 15.5%, P = 0.163). Participant character-
F05.1, F10.6, G31.0, G31.1, and G31.8in ICD-10). Moreover, dementia istics across six categories according to the daily consumption of
diagnoses were retrieved from primary care data using read codes (version tea were shown in Supplementary Table 2. In addition,
2 [Read v2] and version 3 [CTV3 or Read v3]). Follow-up visits began on characteristics of participants stratified by age or sex and
date of attending assessment center, and participants were followed up to
characteristics of participants stratified by the occurrence of AD
the earliest incident dementia diagnosis, date of death, the last data
collection date by the general practitioner, or the last time of hospital
or VD were shown in Supplementary Tables 3, 4, and 5,
inpatient admission, whichever occurred first. respectively.

Association between tea consumption and dementia

Statistical analysis Results from Cox proportional hazard regression models were shown
Before analyses, individuals aged < 45 years at baseline were excluded
since they had a low risk of incident dementia. Then we excluded in Supplementary Table 6 and Fig. 2. There was a significant
individuals who reported tea consumption of more than 15 cups a day association between tea consumption and a decreased risk of
(4 standard deviations [SD] above the mean). Baseline sociodemographic, incident dementia after controlling for age, sex and ethnicity (model
lifestyle, and main dietary characteristics in participants stratified by 1). Even in the further-adjusted model (model 2), the protective effect
dementia status (incident dementia and no incident dementia) were remained significant. Generally, HRs for incident dementia were 0.819
summarized. (95% CI: 0.760–0.884) in model 1 and 0.841 (95% CI: 0.767–0.921) in
Cox proportional hazard regression models were used to examine the model 2 for those who reported consumption of tea compared with
association of tea consumption with incident dementia, with non- the non-drinkers. As for different categories of tea consumption,
consumption as reference and the duration of follow-up as the timescale participants with moderate tea consumption (1–6 cups/day) had a
(N = 377,592). Hazard ratios (HRs) with 95% confidence intervals (CIs) were
reported for all analyses. Two models were fitted in our analyses. Model 1 lower hazard of dementia compared with the non-drinkers, with the
was adjusted for age at baseline (continuous), sex (female/male) and specific results as follows: model 1, HR = 0.823 (95% CI:0.751–0.902)
ethnicity (white/non-white). In model 2, the following covariates were for 1–2 cups/day, HR = 0.792 (95% CI: 0.727–0.862) for 3–4cups/day,
additionally considered: Townsend deprivation index (TDI) (combining HR = 0.803 (95% CI: 0.732–0.880) for 5–6 cups/day; model 2, HR =
information on social class, employment, car availability and housing; 0.857 (95% CI: 0.767–0.959) for 1–2 cups/day, HR = 0.801 (95% CI:

Translational Psychiatry (2022)12:171

 H.-Y. Hu et al.
3

Fig. 1 Flowchart of participant selection. TDI Townsend deprivation index, BMI Body mass index, APOE4 apolipoprotein E4.

0.721–0.890) for 3–4 cups/day, HR = 0.859 (95% CI: 0.769–0.960) for pronounced associations were found in males, and the consump-
5–6 cups/day. However, no significant difference was found in tion of 1–6 cups/day led to a 19 to 22% reduction in dementia
dementia risk between individuals who consumed more than 6 cups incidence in the fully-adjusted model (see Supplementary Fig. 2).
of tea per day and non-drinkers, with the specific results as follows: Results from the subgroup analyses stratified by the form of
model 1, HR = 0.910 (95% CI: 0.802–1.032) for 7–8 cups/day, HR = dementia (AD/VD) were summarized in Supplementary Table 9. In
0.952 (95% CI: 0.817–1.108) for ≥9 cups/day; model 2, HR = 0.891 model 1, the consumption of 1–6 cups/day was associated with a
(95% CI: 0.766–1.037) for 7–8 cups/day, HR = 0.946 (95% CI: lower incidence of AD. In model 2, participants who consumed
0.788–1.136) for ≥9 cups/day. 1–4 cups of tea a day were 16 to 19% less likely to develop AD, but
Results of subgroup analyses stratified by age were shown in the HR for 5–6 cups/day was not significant (Fig. 3). In terms of
Supplementary Table 7. In midlife, there were significant associa- incident VD, a similar picture to ACD was seen in both adjusted
tions of tea consumption with a reduced risk of dementia in both models. Moderate consumption (1–6 cups/day) led to a 25 to 29%
two models, and the dementia incidence was approximately reduction in VD incidence (Fig. 3).
17–26% lower among individuals with tea consumption of 1–6
cups per day in the fully-adjusted model. However, this association Non-linear relationship between tea consumption and
disappeared in the late-life after controlling for all covariates (see dementia
Supplementary Fig. 1). In Fig. 4, we used restricted cubic splines to flexibly model and
Results of subgroup analyses stratified by sex were shown in visualize the association of tea-consumption with dementia. After
Supplementary Table 8. Females who consumed 3–4 cups of tea a adjusting for all the covariates, the risk of incident dementia
day were less likely to develop dementia in both two models (HR = substantially decreased until it reached bottom at around 3 cups/
0.800 [95% CI: 0.687–0.933] in the fully-adjusted model). More day, and increased thereafter (P for non-linearity = 7E−04). Within

Translational Psychiatry (2022)12:171

 H.-Y. Hu et al.
4
Table 1. Baseline characteristics of participants stratified by the occurrence of dementia.
Variables All participants No incident dementia Incident dementia P-value
(n = 377,592) (n = 372,470) (n = 5122)
Age (mean ± SD), years 58.49 ± 6.83 58.41 ± 6.82 64.55 ± 4.39 <0.001b
Age group (years), n (%)
Midlife 311,107 (82.4) 308,598 (82.9) 2509 (49.0) –
Late-life 66,485 (17.6) 63,872 (17.1) 2613 (51.0) –
Female, n (%) 204,980 (54.3) 202,580 (54.3) 2400 (46.8) <0.001c
White, n (%) 358,908 (95.1) 354,036 (95.0) 4872 (95.1) 0.823c
Whether reported consuming tea, n (%) 0.163c
Non-consumption, n (%) 55,935 (14.8) 55,141 (14.8) 794 (15.5)
Consumption, n (%) 321,657 (85.1) 317,329 (85.1) 4328 (84.4)
APOE4 carrier status, where reporteda, n <0.001c
(%)
Carrier 93,049 (28.5) 9075 (28.2) 1974 (45.2) –
Non-carrier 233,914 (71.5) 231,523 (71.8) 2391 (54.8) –
Townsend deprivation index (median, −2.20 (−3.67, 0.43) −2.20 (−3.67, 0.42) 1.89 (−3.52, 1.31) <0.001d
IQR), where reporteda
Education, where reporteda, n (%) <0.001c
Without college degree 184,074 (62.0) 181,946 (62.0) 2128 (68.4) –
With college degree 112,722 (38.0) 111,741 (38.0) 981 (31.6) –
2
Body mass index (mean ± SD), kg/m , 27.60 ± 4.81 27.60 ± 4.81 27.81 ± 4.95 0.002b
where reporteda
Typical sleep duration (mean ± SD), 7.16 ± 1.13 7.16 ± 1.13 7.29 ± 1.38 <0.001b
hours, where reporteda
Smoking status, where reporteda, n (%) <0.001c
Current 38,360 (10.2) 37,817 (10.2) 543 (10.7) –
Previous 137,212 (36.5) 135,018 (36.4) 2194 (43.2) –
Never 200,486 (53.3) 198,140 (53.4) 2346 (46.2) –
Alcohol status, where reporteda, n (%) <0.001c
Current 345,956 (91.7) 341,592 (91.8) 4364 (85.4) –
Previous 14,316 (3.8) 13,944 (3.7) 372 (7.3) –
Never 16,882 (4.5) 16,508 (4.4) 374 (7.3) –
Vegetable consumption (median, IQR), 4 (3, 6) 4 (3, 6) 5 (3, 7) <0.001d
times/ week, where reporteda
Fruit consumption (median, IQR), times/ 3 (2, 4) 3 (2, 4) 3 (2, 5) <0.001d
week, where reporteda
Fish consumption (median, IQR), times/ 2 (1, 3.5) 2 (1, 3.5) 2 (1.5, 4) <0.001d
week, where reporteda
a
Some patients had missing data for these variables. The missing data were not reported here.
b
Comparisons between groups were performed via the t-test, cchi-square test, dMann-Whitney U-test.
SD Standardized deviation, IQR Interquartile range, APOE4 apolipoprotein E4.

3 cups/day, the HR for dementia was 0.943 (95% CI: 0.907–0.981) consumption of 2 to 3 cups per day (AD: P for non-linearity =
per cup increase, indicating an approximately 6% reduced risk for 0.046, Supplementary Fig. 5A; VD: P for non-linearity = 0.019,
one extra cup a day. Supplementary Fig. 5B).
A similar U-shaped relationship was observed in midlife, and the
risk of dementia reached bottom at around 3 cups of tea per day Sensitivity analyses
(P for non-linearity = 8E−04, Supplementary Fig. 3A). We didn’t When we restricted participants to those with a follow-up time of
observe any nonlinear relationship in the late-life (see Supple- ≥4 years (n = 355 114), the HRs were similar to the main results in
mentary Fig. 3B). The non-linear relationships were observed in both of the adjusted models (see Supplementary Table 10).
both females and males. The largest reduction of dementia Sensitivity analyses after excluding dementia cases that occurred
incidence was associated with intake of around 3 cups per day for within the first 1 year follow-up (n = 57) to control the reverse
females (P for non-linearity = 0.006, Supplementary Fig. 4A) and causality showed that the HRs were of similar magnitude to the
intakes of 3–6 cups per day for males (P for non-linearity = 0.022, main results (see Supplementary Table 11). Exclusion of partici-
Supplementary Fig. 4B). Furthermore, the U-shaped associations of pants who had a history of stroke at baseline (n = 11 643) also did
tea-consumption with AD and VD were shown, and the largest not change these associations significantly (see Supplementary
reductions in both incidences were associated with tea Table 12).

Translational Psychiatry (2022)12:171

 H.-Y. Hu et al.
5

Fig. 2 Association between tea consumption and the risk of incident dementia. In both adjusted models, tea consumption was associated
with a reduced risk of dementia (model 1: HR = 0.819 [95% CI:0.760-0.884]; model 2: HR = 0.841 [95% CI: 0.767–0.921]). Moderate consumption
(1–6 cups/day) of tea exerted significant protective effects. P values were computed by Cox proportional hazard regression. Model 1 was
adjusted for age, sex and ethnicity. Model 2 was adjusted for age, sex, ethnicity, TDI, education, BMI, typical sleep duration, smoking status,
alcohol status, total consumption of vegetables, total consumption of fruit, total consumption of fish and APOE4 status. HR Hazard ratios, CI
Confidence interval, TDI Townsend deprivation index, BMI Body mass index, APOE4, apolipoprotein E4.

DISCUSSION from the small sample size (81 AD cases among 1836 individuals)
Leveraging data from 377 592 individuals from the UKB, this study and excessive category combination. By the way, although
investigated the association between tea consumption and inconsistent results were shown regarding coffee intake, most of
dementia over the 9-year follow-up. Our results showed that the studies showed a reduction for the risk of dementia [33–35]. It
moderate drinkers (1–6 cups/day) were less likely to develop ACD, should be noted that tea and coffee shared some molecules (e.g.,
AD, and VD after adjustment for covariates. Another major finding polyphenols and caffeine) which are considered to reduce
was the U-shaped association of tea consumption with dementia dementia risk. Compared with previous researches, this cohort
risk, and the optimal intake was around 3 cups per day. These study used a larger sample of non-demented older adults.
findings suggested tea to be a modifiable lifestyle factor for Sufficient cases of ACD, AD and VD were observed during a long
dementia prevention. follow-up time. Our study focused on the dosage-dependent
Previous studies yielded inconsistent results. A study among associations of tea consumption with ACD, AD and VD, obtaining
9375 Chinese older adults showed that participants who nonlinear relationships between daily dosage and dementia
consumed <2 cups/day, 2–4 cups/day, and ≥4 cups/day were incidences. Compared to Zhang et al.’s, our study considered
less likely to have cognitive impairment compared with non- APOE4 status as a covariate in the full-adjusted model and focused
drinkers, which supported our findings; however, the design was on AD in addition to VD. The age range of the total participants
cross-sectional and the reference selection was different from our (≥45 years old) was larger in this study, and the effect modification
analyses [14]. A longitudinal study among 1305 older adults also by age or sex was observed in subgroup analyses. When
found that tea consumption could reduce the risk of cognitive calculating the relationships between categories of tea consump-
decline compared with non-consumption over 5.3 years of follow- tion and dementia risk, we used more specific category
up [32]. The Ohsaki Cohort Study among 13,645 Japanese found combinations, and thus the dosage-dependent relationships were
that green tea consumption was associated with a lower risk of more convinced and were in accordance with the results of non-
incident dementia over 5.7 years of follow-up, which was linear analyses. Besides, we performed a great deal of sensitivity
consistent with our major finding [26]. In addition, Zhang et al. analyses to examine the stability of results.
conducted a large-scale cohort study and found that tea Although the mechanisms underlying the associations between
consumption was associated with lower risks of dementia as well tea consumption and dementia are unclear, some possibilities
as VD [29]. However, with tea consumption at low frequency (≤1 may explain these findings. Firstly, oxidative stress was suggested
times per week) as the reference, a longitudinal study found no to be involved in the pathogenesis of both AD and VD [36]. Recent
significant association of frequent tea consumption (≥3 times per vivo and vitro studies have focused on the antioxidant effects of
week) with AD [28]. The insignificant association probably resulted tea biomolecules [37, 38]. For example, EGCG could chelate the

Translational Psychiatry (2022)12:171

 H.-Y. Hu et al.
6

Fig. 3 Associations of tea consumption with risks of AD and VD in the fully-adjusted model. After adjusting for all covariates, participants
who consumed 1–4 cups of tea a day were 16–19% less likely to develop AD compared with non-drinkers. Besides, the tea consumption of
1–6 cups/day brought a 25 to 29% reduction in VD incidence. P values were computed by Cox proportional hazard regressions. Covariates
included age, sex, ethnicity, TDI, education, BMI, typical sleep duration, smoking status, alcohol status, total consumption of vegetables, total
consumption of fruit, total consumption of fish and APOE4 status. HR Hazard ratios, CI Confidence interval, AD Alzheimer disease, VD Vascular
dementia, TDI Townsend deprivation index, BMI Body mass index, APOE4 apolipoprotein E4.

bivalent metal-ions and prevent oxidation resulting from reactive could cross the blood-brain barrier to reach the brain parenchyma
hydroxyl radicals [39]. EGCG was also found to scavenge the free [48, 49].
radicals, and thus alleviating the neuronal apoptosis as well as Amyloid precursor protein (APP) could play an important role in
promoting the neuronal differentiation [37, 38, 40–42]. Methyl- the deposition of Aβ, thus accelerating AD progression [50].
xanthines, a group of biomolecules derived from tea (especially Noteworthily, tea polyphenols including epicatechin (EC) and
black tea), included theophylline (1,3-dimethylxanthine) and epigallocatechin (EGC) were found to suppress APP processing by
caffeine. Caffeine was suggested to have significant antioxidant neutralizing excessively free iron via chelation reactions [51].
effects [43, 44]. In human clinical trials, an increase in plasma Besides, AD mouse models have showed reduced Aβ pathology in
antioxidant capacity after moderate intakes of green or black tea brain after a long-term oral delivery of tea polyphenols [16–18].
was observed [8, 30]. The antioxidative function might be a key Furthermore, previous studies in cognitively normal individuals
factor in the neuroprotective role of tea in brain diseases and showed that the components of green tea could alleviate
cognitive performance. Secondly, neuroinflammation has been abnormal tau metabolism and mitigate the detrimental effects
demonstrated to play an important role in the progression of of tau pathology, suggesting green tea as a potential prophylactic
dementia [45]. Animal studies found that EGCG could decrease the for AD [15].
expression of pro-inflammatory cytokines (e.g., interleukin [IL]-1β The association of green or black tea with VD was partly due to
and tumor necrosis factor [TNF]-α), increase the expression of anti- their protection against transient ischaemic attack (TIA) and
inflammatory cytokines (e.g., IL-10), inhibit astrocyte activation stroke, as suggested by population-based studies [23, 52, 53].
and promote microglial activation. Methylxanthines were also Besides, the protective effects of tea on VD might relate to its
found to increase the production of cerebrospinal fluid (CSF) as positive impacts on cerebral blood vessels. For example, tea was
well as promote the clearance of neurotoxins, which might suggested to have vasodilative functions by enhancing the
attenuate neuroinflammation and thereby provide biological releases of prostacyclin and NO, as well as reducing oxidative
plausibility for the prevention of dementia [20, 21]. Thirdly, l- stress [54]. Accumulating evidence has shown that tea consump-
theanine, a specific amino acid extracted from tea leaves, was tion could reduce blood pressure as well as serum lipid levels,
suggested to be an intervention for cognitive improvement [46]. improving cerebral perfusion and reducing the risk of VD [55–58].
In a randomized controlled trial (RCT) study, electroencephalo- Moreover, caffeine in tea was revealed to protect against ischemic
grams showed that participants treated by l-theanine had higher neuronal injury by blocking the receptors of neurotransmitter
levels of theta waves, interpreted to be an indicator of cognitive adenosine [59–61]. However, VD represents a heterogeneous
alertness, in the temporal, frontal, parietal, and occipital lobes after group of dementia with different pathophysiologies depending
3 h of reading compared with the placebo group [47]. Note- on the source (e.g., small vessel disease versus large ischemia) and
worthily, since the beneficial compounds in tea are small, they also shares some pathologies with AD [62]. Therefore, advances in

Translational Psychiatry (2022)12:171

 H.-Y. Hu et al.
7
More pronounced associations were observed in midlife in this
study. The associations in the late-life were attenuated possibly
due to confounding factors. Confounders such as BMI, alcohol
intake and dietary habits were associated with both tea
consumption and dementia, but these relations may differ in
magnitude and even direction between midlife and the late-life.
Most importantly, the pathological processes leading to dementia
start many years before the clinical manifestation of cognitive
decline, and the longitudinal changes related to tea consumption
in the brain may also begin earlier in adult life. A population-based
study found tea could mitigate AD pathologies in individuals aged
65 years or younger, which was consistent with our findings [15].
Therefore, tea consumption may be listed as one of the strategies
during midlife to prevent dementia. Moreover, males seemed to
benefited more from tea consumption than females in this study.
Previous evidence showed that brain structure and function
differed between males and females throughout the aging
process, and the sex differences might have an influence on the
protective effects of tea [66]. Previous cross-sectional studies
found that tea consumption had a protective effect against
amnestic MCI in males rather than females [67, 68]. Frequent tea
consumption was found to mitigate tau pathologies in males, but
not in females [15]. In the UKB, males were more likely to be tea
drinkers, and this could also make the protective effects of tea
Fig. 4 A non-linear relationship between tea consumption and more pronounced. The sex difference in the effects of
incident dementia. There was a U-shaped association between tea tea highlighted the importance of differential preventive mea-
consumption and incident dementia, and the consumption of
around three cups per day showed the strongest protective effect (P sures in meles and females.
for non-linearity = 7E-04). P-value was computed using restricted This study has several strengths. We had a large sample size. In
cubic splines functions in the Cox proportional hazard regression addition, we used two degrees of confounder-adjustment in the
model. The model was adjusted for age, sex, ethnicity, TDI, analyses to make the results more convincing. The results after
education, BMI, typical sleep duration, smoking status, alcohol adjusting for more extensive confounders (including education,
status, total consumption of vegetables, total consumption of fruit, TDI, BMI, lifestyle factors, dietary factors and APOE4 status) were
total consumption of fish and APOE4 status. HR Hazard ratios, CI stable. Besides, we discussed the potential mechanisms under-
Confidence interval, TDI Townsend deprivation index, BMI Body lying the effects of tea on brain actions in greater depth than
mass index, APOE4 apolipoprotein E4. previous studies. This study also has some limitations. Firstly, the
the classification of dementia subtypes may help estimate the dynamic changes in tea consumption during follow-up might
associations more consistently in the future. influence the dosage-dependent associations. However, tea
Additionally, our study revealed a non-linear association of tea consumption was only reported at baseline in this study. Secondly,
consumption with dementia and suggested daily consumption of we only had information on tea consumption by cup measures,
three cups to be strongest protective. Within three cups, the but no universal standard tea cup size was recognized. Therefore,
benefits of tea consumption increased with daily dosage. Previous the optimal consumption for preventing dementia identified in
studies on the dosage-dependent relationship between tea our analyses should be regarded as an estimation. Thirdly, the
consumption and dementia were scarce. A dose-response meta- kinds and concentrations of bioactive components varied among
analysis found that one cup of tea per day led to a 6% reduction in the processing methods of tea (such as black and green tea) [69].
the risk of dementia or mild cognitive impairment (MCI), and two However, there was no information about the type of tea, limiting
cups led to an 11% decrease [11]. However, the data extracted for our ability to examine the biomolecule-related mechanisms.
this meta-analysis were not enough to calculate a nonlinear Studies exploring the associations between different types of tea
relationship convincingly, because only three studies were and dementia are warranted in the future.
included, and two of them didn’t take the daily dosage of tea In conclusion, tea consumption was shown to be associated
into consideration. Importantly, as our results showed, increase in with a lower risk of dementia in the total participants. As a
consumption beyond 3 cups/day was not associated with potentially modifiable lifestyle factor, tea consumption could play
increased risk of harm, but the magnitude of the benefit was a pivotal role in the primary prevention for dementia. This study
reduced, and the estimates did not reach significance when daily would offer a relatively simple and low-cost solution to the
consumption was beyond 6 cups. One possible explanation was interventions of age-related cognitive decline or dementia.
that excessive caffeine in tea could disturb sleep and block the
anti-stress effects in humans [63, 64]. Also, a study on caffeinated
drinks showed that excessive intake of caffeine (>6 cups of drinks DATA AVAILABILITY
The datasets described in this manuscript are available from the UK Biobank with an
per day) might cause competitive binding to adenosine receptors approved protocol. External investigators can request the data and approval of use
and thus cause the morphological changes in brain [65]. More- on application to the UK Biobank (www.ukbiobank.ac.uk/).
over, estimates from higher intakes included smaller numbers of
participants, and this could result in the imprecision observed for
dementia at these levels of tea consumption. Anyhow, our REFERENCES
analyses indicated that the best intervention effect might be 1. Organization WH: Dementia fact Sheet. [Internet] Available from: http://www.
achieved at around three cups of tea per day in a future RCT study, hoint/mediacentre/factsheets/fs362/en/ (accessed 2 September 2021).
and this intervention would not result in significant harm to 2. Malik R, Georgakis MK, Neitzel J, Rannikmäe K, Ewers M, Seshadri S, et al. Midlife
participants. vascular risk factors and risk of incident dementia: Longitudinal cohort and

Translational Psychiatry (2022)12:171

 H.-Y. Hu et al.
8
Mendelian randomization analyses in the UK Biobank. Alzheimer’s Dement: J 29. Zhang Y, Yang H, Li S, Li WD, Wang Y. Consumption of coffee and tea and risk of
Alzheimer’s Assoc. 2021;17:1422–31. developing stroke, dementia, and poststroke dementia: A cohort study in the UK
3. Livingston G, Huntley J, Sommerlad A, Ames D, Ballard C, Banerjee S, et al. Biobank. PLoS Med. 2021;18:e1003830.
Dementia prevention, intervention, and care: 2020 report of the Lancet Com- 30. Rietveld A, Wiseman S. Antioxidant effects of tea: evidence from human clinical
mission. Lancet (Lond, Engl). 2020;396:413–46. trials. J Nutr. 2003;133:3285s–3292s.
4. Zhang H, Greenwood DC, Risch HA, Bunce D, Hardie LJ, Cade JE. Meat con- 31. R Core Team: R: A Language and Environment for Statistical Computing. In. R
sumption and risk of incident dementia: cohort study of 493,888 UK Biobank Foundation for Statistical Computing, Vienna, Austria. URL https://www.R-project.
participants. Am J Clin Nutr. 2021;114:175–84. org/; 2021.
5. Ros E. Can specific nutrients, foods, or dietary patterns modulate cognitive 32. Shirai Y, Kuriki K, Otsuka R, Kato Y, Nishita Y, Tange C, et al. Green tea and coffee
function in (older) adults? Latest evidence from randomized controlled trials. intake and risk of cognitive decline in older adults: the National Institute for
Current opinion in clinical nutrition and metabolic care. 2021;24:511–20. Longevity Sciences, Longitudinal Study of Aging. Public Health Nutr.
6. Kivipelto M, Mangialasche F, Ngandu T. Lifestyle interventions to prevent cognitive 2020;23:1049–57.
impairment, dementia and Alzheimer disease. Nat Rev Neurol. 2018;14:653–66. 33. Eskelinen MH, Ngandu T, Tuomilehto J, Soininen H, Kivipelto M. Midlife coffee
7. Bell IR. Diet and nutrition in Alzheimer’s disease and other dementias of late life. and tea drinking and the risk of late-life dementia: a population-based CAIDE
Explor (N. Y, NY). 2005;1:299–301. study. J Alzheimer’s Dis: JAD. 2009;16:85–91.
8. Cabrera C, Artacho R, Giménez R. Beneficial effects of green tea-a review. J Am 34. Eskelinen MH, Kivipelto M. Caffeine as a protective factor in dementia and Alz-
Coll Nutr. 2006;25:79–99. heimer’s disease. J Alzheimer’s Dis: JAD. 2010;20(Suppl 1):S167–74.
9. Shin S, Lee JE, Loftfield E, Shu XO, Abe SK, Rahman MS et al. Coffee and tea 35. van Gelder BM, Buijsse B, Tijhuis M, Kalmijn S, Giampaoli S, Nissinen A, et al.
consumption and mortality from all causes, cardiovascular disease and cancer: A Coffee consumption is inversely associated with cognitive decline in elderly
pooled analysis of prospective studies from the Asia Cohort Consortium. Int. J. European men: the FINE Study. Eur J Clin Nutr. 2007;61:226–32.
Epidemiol (in the press). https://doi.org/10.1093/ije/dyab161 (2021). 36. Luca M, Luca A, Calandra C. The role of oxidative damage in the pathogenesis
10. Keller A, Wallace TC. Tea intake and cardiovascular disease: an umbrella review. and progression of Alzheimer’s Disease and vascular Dementia. Oxid Med Cell
Ann Med. 2021;53:929–44. Longev. 2015;2015:504678.
11. Ran LS, Liu WH, Fang YY, Xu SB, Li J, Luo X, et al. Alcohol, coffee and tea intake 37. Pervin M, Unno K, Ohishi T, Tanabe H, Miyoshi N, Nakamura Y. Beneficial effects of
and the risk of cognitive deficits: a dose-response meta-analysis. Epidemiol green tea catechins on neurodegenerative diseases. Molecules (Basel, Switzer-
Psychiatr Sci. 2021;30:e13. land). 2018;23:1297.
12. Yasmeen H, Hasnain S. In vitro antioxidant effect of Camellia sinensis on human 38. Singh NA, Mandal AK, Khan ZA. Potential neuroprotective properties of
cell cultures. Pak J Pharm Sci. 2015;28:1573–81. epigallocatechin-3-gallate (EGCG). Nutr J. 2016;15:60.
13. Mancini E, Beglinger C, Drewe J, Zanchi D, Lang UE, Borgwardt S. Green tea 39. Kumamoto M, Sonda T, Nagayama K, Tabata M. Effects of pH and metal ions on
effects on cognition, mood and human brain function: A systematic review. antioxidative activities of catechins. Biosci, Biotechnol, Biochem. 2001;65:126–32.
Phytomedicine: Int J Phytother Phytopharmacol. 2017;34:26–37. 40. Soung HS, Wang MH, Tseng HC, Fang HW, Chang KC. (-)Epigallocatechin-3-gallate
14. Shen W, Xiao Y, Ying X, Li S, Zhai Y, Shang X, et al. Tea consumption and cognitive decreases the stress-induced impairment of learning and memory in rats. Neu-
impairment: A cross-sectional study among Chinese elderly. PloS one. 2015;10: rosci Lett. 2015;602:27–32.
e0137781. 41. Unno K, Nakamura Y. Green tea suppresses brain aging. Molecules (Basel, Swit-
15. Ma YH, Wu JH, Xu W, Shen XN, Wang HF, Hou XH, et al. Associations of green tea zerland). 2021;26:4897.
consumption and cerebrospinal fluid biomarkers of Alzheimer’s Disease pathol- 42. Rodríguez-Morató J, Xicota L, Fitó M, Farré M, Dierssen M, de la Torre R. Potential
ogy in cognitively intact older adults: The CABLE study. J Alzheimer’s Dis: JAD. role of olive oil phenolic compounds in the prevention of neurodegenerative
2020;77:411–21. diseases. Molecules (Basel, Switz). 2015;20:4655–80.
16. Cox CJ, Choudhry F, Peacey E, Perkinton MS, Richardson JC, Howlett DR, et al. 43. Poole R, Kennedy OJ, Roderick P, Fallowfield JA, Hayes PC, Parkes J. Coffee
Dietary (-)-epicatechin as a potent inhibitor of βγ-secretase amyloid precursor consumption and health: umbrella review of meta-analyses of multiple health
protein processing. Neurobiol aging. 2015;36:178–87. outcomes. BMJ (Clin Res ed). 2017;359:j5024.
17. Zeng YQ, Wang YJ, Zhou XF. Effects of (-)epicatechin on the pathology of APP/ 44. Azam S, Hadi N, Khan NU, Hadi SM. Antioxidant and prooxidant properties of
PS1 transgenic mice. Front Neurol. 2014;5:69. caffeine, theobromine and xanthine. Med Sci Monit: Int Med J Exp Clin Res.
18. Zhang Z, Wu H, Huang H. Epicatechin plus treadmill exercise are neuroprotective 2003;9:Br325–330.
against moderate-stage amyloid precursor Protein/Presenilin 1 mice. Pharma- 45. Kakutani S, Watanabe H, Murayama N. Green tea intake and risks for dementia,
cogn Mag. 2016;12(Suppl 2):S139–146. Alzheimer’s disease, mild cognitive impairment, and cognitive Impairment: A
19. Serafini M, Ghiselli A, Ferro-Luzzi A. In vivo antioxidant effect of green and black Systematic Review. Nutrients. 2019;11:1165.
tea in man. Eur J Clin Nutr. 1996;50:28–32. 46. Lavretsky H. Lifestyle medicine for prevention of cognitive decline: Focus on
20. Ohta A, Sitkovsky M. Methylxanthines, inflammation, and cancer: fundamental green tea. Am J Geriatr Psychiatry: Off J Am Assoc Geriatr Psychiatry.
mechanisms. Handbook Exp Pharmacol. 2011;(200):469–81. 2016;24:890–2.
21. Han ME, Kim HJ, Lee YS, Kim DH, Choi JT, Pan CS, et al. Regulation of cere- 47. Park SK, Jung IC, Lee WK, Lee YS, Park HK, Go HJ, et al. A combination of green tea
brospinal fluid production by caffeine consumption. BMC Neurosci. 2009;10:110. extract and l-theanine improves memory and attention in subjects with mild
22. Schimidt HL, Vieira A, Altermann C, Martins A, Sosa P, Santos FW, et al. Memory cognitive impairment: a double-blind placebo-controlled study. J Medicinal Food.
deficits and oxidative stress in cerebral ischemia-reperfusion: neuroprotective 2011;14:334–43.
role of physical exercise and green tea supplementation. Neurobiol Learn Mem. 48. Wong AD, Ye M, Levy AF, Rothstein JD, Bergles DE, Searson PC. The blood-brain
2014;114:242–50. barrier: An engineering perspective. Front Neuroeng. 2013;6:7.
23. Shen L, Song LG, Ma H, Jin CN, Wang JA, Xiang MX. Tea consumption and risk of 49. Franke H, Galla HJ, Beuckmann CT. An improved low-permeability in vitro-model
stroke: A dose-response meta-analysis of prospective studies. J Zhejiang Univ Sci of the blood-brain barrier: Transport studies on retinoids, sucrose, haloperidol,
B. 2012;13:652–62. caffeine and mannitol. Brain Res. 1999;818:65–71.
24. Arendash GW, Schleif W, Rezai-Zadeh K, Jackson EK, Zacharia LC, Cracchiolo JR, 50. Pimplikar SW, Ghosal K. Amyloid precursor protein: More than just neurode-
et al. Caffeine protects Alzheimer’s mice against cognitive impairment and generation. Stem cell Res Ther. 2011;2:39.
reduces brain beta-amyloid production. Neuroscience. 2006;142:941–52. 51. Wang YJ, Thomas P, Zhong JH, Bi FF, Kosaraju S, Pollard A, et al. Consumption of
25. Chu YF, Chang WH, Black RM, Liu JR, Sompol P, Chen Y, et al. Crude caffeine grape seed extract prevents amyloid-beta deposition and attenuates inflamma-
reduces memory impairment and amyloid β(1-42) levels in an Alzheimer’s mouse tion in brain of an Alzheimer’s disease mouse. Neurotox Res. 2009;15:3–14.
model. Food Chem. 2012;135:2095–102. 52. Chung M, Zhao N, Wang D, Shams-White M, Karlsen M, Cassidy A, et al. Dose-
26. Tomata Y, Sugiyama K, Kaiho Y, Honkura K, Watanabe T, Zhang S, et al. Green Tea response relation between tea consumption and risk of cardiovascular disease
Consumption and the Risk of Incident Dementia in Elderly Japanese: The Ohsaki and all-cause mortality: A systematic review and meta-analysis of population-
Cohort 2006 Study. Am J Geriatr Psychiatry: Off J Am Assoc Geriatr Psychiatry. based studies. Adv Nutr (Bethesda, Md). 2020;11:790–814.
2016;24:881–9. 53. Wang M, Bai Y, Wang Z, Zhang Z, Liu D, Lian X. Higher tea consumption is
27. Kitamura K, Watanabe Y, Nakamura K, Sanpei K, Wakasugi M, Yokoseki A, et al. associated with decreased risk of small vessel stroke. Clin Nutr (Edinb, Scotl).
Modifiable Factors Associated with Cognitive Impairment in 1,143 Japanese 2021;40:1430–5.
Outpatients: The Project in Sado for Total Health (PROST). Dement Geriatr Cogn 54. Bogdanski P, Suliburska J, Szulinska M, Stepien M, Pupek-Musialik D, Jablecka A.
Disord Extra. 2016;6:341–9. Green tea extract reduces blood pressure, inflammatory biomarkers, and oxida-
28. Dai Q, Borenstein AR, Wu Y, Jackson JC, Larson EB. Fruit and vegetable juices and tive stress and improves parameters associated with insulin resistance in obese,
Alzheimer’s disease: the Kame Project. Am J Med. 2006;119:751–9. hypertensive patients. Nutr Res (N. Y, NY). 2012;32:421–7.

Translational Psychiatry (2022)12:171

 H.-Y. Hu et al.
9
55. Yarmolinsky J, Gon G, Edwards P. Effect of tea on blood pressure for secondary Science and Brain-Inspired Technology, and the 111 Project (No. B18015). WC was
prevention of cardiovascular disease: A systematic review and meta-analysis of supported by grants from the National Natural Sciences Foundation of China (No.
randomized controlled trials. Nutr Rev. 2015;73:236–46. 82071997), the Shanghai Rising-Star Program (No. 21QA1408700) and Natural
56. Cornelis MC, van Dam RM. Habitual coffee and tea consumption and cardio- Science Foundation of Shanghai (No. 18ZR1404400). This study was supported by
metabolic biomarkers in the UK Biobank: The role of beverage types and genetic grants from the National Natural Science Foundation of China (82071201, 91849126),
variation. J Nutr. 2020;150:2772–88. Shanghai Municipal Science and Technology Major Project (No.2018SHZDZX01),
57. Ninomiya T, Ohara T, Hirakawa Y, Yoshida D, Doi Y, Hata J, et al. Midlife and late- ZJLab, Shanghai Center for Brain Science and Brain-Inspired Technology, Tianqiao
life blood pressure and dementia in Japanese elderly: The Hisayama study. and Chrissy Chen Institute, and the State Key Laboratory of Neurobiology and
Hypertension (Dallas, Tex: 1979). 2011;58:22–28. Frontiers Center for Brain Science of Ministry of Education, Fudan University.
58. Emdin CA, Rothwell PM, Salimi-Khorshidi G, Kiran A, Conrad N, Callender T, et al.
Blood pressure and risk of vascular dementia: Evidence from a primary care registry
and a cohort study of transient ischemic attack and stroke. Stroke. 2016;47:1429–35. AUTHOR CONTRIBUTIONS
59. Nehlig A, Daval JL, Debry G. Caffeine and the central nervous system: mechan- HYH, LT, and JTY did the manuscript preparation and drafting. HYH, BSW, YNO, YHM,
isms of action, biochemical, metabolic and psychostimulant effects. Brain Res YYH, WC, and JTY did the data acquisition. HYH, BSW, and YNO did the data analysis
Brain Res Rev. 1992;17:139–70. and interpretation. LT and JTY are responsible for the study conception and design.
60. Lunt MJ, Ragab S, Birch AA, Schley D, Jenkinson DF. Comparison of caffeine- All authors have contributed to the manuscript by revising and editing critically for
induced changes in cerebral blood flow and middle cerebral artery blood velocity important intellectual content and given final approval of the version and agreed to
shows that caffeine reduces middle cerebral artery diameter. Physiological Meas. be accountable for all aspects of the work presented here.
2004;25:467–74.
61. Schwarzschild MA, Xu K, Oztas E, Petzer JP, Castagnoli K, Castagnoli N Jr, et al.
Neuroprotection by caffeine and more specific A2A receptor antagonists in
animal models of Parkinson’s disease. Neurology. 2003;61(11 Suppl 6):S55–61. COMPETING INTERESTS
62. Kalaria RN, Kenny RA, Ballard CG, Perry R, Ince P, Polvikoski T. Towards defining The authors declare no competing interests.
the neuropathological substrates of vascular dementia. J Neurol Sci.
2004;226:75–80.
63. Unno K, Noda S, Kawasaki Y, Yamada H, Morita A, Iguchi K, et al. Reduced stress ADDITIONAL INFORMATION
and improved sleep quality caused by green tea are associated with a reduced Supplementary information The online version contains supplementary material
caffeine content. Nutrients. 2017;9:777. available at https://doi.org/10.1038/s41398-022-01923-z.
64. Jahrami H, Al-Mutarid M, Penson PE, Al-Islam Faris M, Saif Z, Hammad L. Intake of
caffeine and its association with physical and mental health status among uni- Correspondence and requests for materials should be addressed to Lan Tan or Jin-
versity students in Bahrain. Foods (Basel, Switzerland). 2020;9:473. Tai Yu.
65. Pham K, Mulugeta A, Zhou A, O’Brien JT, Llewellyn DJ, Hyppönen E. High coffee
consumption, brain volume and risk of dementia and stroke. Nutritional Neurosci. Reprints and permission information is available at http://www.nature.com/
2021:1–12. reprints
66. Giedd JN, Raznahan A, Mills KL, Lenroot RK. Review: Magnetic resonance imaging
of male/female differences in human adolescent brain anatomy. Biol Sex Dif- Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims
ferences. 2012;3:19. in published maps and institutional affiliations.
67. Xu H, Wang Y, Yuan Y, Zhang X, Zuo X, Cui L, et al. Gender differences in the
protective effects of green tea against amnestic mild cognitive impairment in the
elderly Han population. Neuropsychiatr Dis Treat. 2018;14:1795–801.
68. Huang CQ, Dong BR, Zhang YL, Wu HM, Liu QX. Association of cognitive
impairment with smoking, alcohol consumption, tea consumption, and exercise Open Access This article is licensed under a Creative Commons
among Chinese nonagenarians/centenarians. Cogn Behav Neurol: Off J Soc Attribution 4.0 International License, which permits use, sharing,
Behav Cogn Neurol. 2009;22:190–6. adaptation, distribution and reproduction in any medium or format, as long as you give
69. Tong W, Yu J, Wu Q, Hu L, Tabys D, Wang Y, et al. Black tea quality is highly appropriate credit to the original author(s) and the source, provide a link to the Creative
affected during processing by its leaf surface microbiome. J Agric Food Chem. Commons license, and indicate if changes were made. The images or other third party
2021;69:7115–26. material in this article are included in the article’s Creative Commons license, unless
indicated otherwise in a credit line to the material. If material is not included in the
article’s Creative Commons license and your intended use is not permitted by statutory
ACKNOWLEDGEMENTS regulation or exceeds the permitted use, you will need to obtain permission directly
This study utilized the UK Biobank Resource under application number 19542. We from the copyright holder. To view a copy of this license, visit http://creativecommons.
would like to thank all the participants and researchers from the UK Biobank. JF was org/licenses/by/4.0/.
supported by National Key R&D Program of China (No.2019YFA0709502), National
Key R&D Program of China (No. 2018YFC1312904), Shanghai Municipal Science and
© The Author(s) 2022
Technology Major Project (No. 2018SHZDZX01), ZJ Lab, Shanghai Center for Brain

Translational Psychiatry (2022)12:171