coffee and tea help cognitive function

 A recent epidemiological study found that the incidence of decreased cognitive function and the risk of dementia were less common in those drinking caffeinated coffee or tea (see cognitive benefit coffee tea JAMA2026 in dropbox, or doi:10.1001/jama.2025.27259)

 

Details:

-- two prospective US cohort studies, one including female participants in the Nurses’ Health Study (NHS, n = 86,606, with data from 1980-2023) and male participants from the Health Professionals Follow-up Study (HPFS, n = 45,215, with data from 1986-2023) who did not have cancer, Parkinson disease, or dementia at study entry

-- These studies employed a rigorously validated assessment of food and beverage intake through food frequency questionnaires over these many years, with dietary intake collected and updated every 2 to 4 years

-- dementia was identified via death records that included 98% complete data through state records, the National Death Index, next-of-kin confirmation, postal authorities, and biennial self-reported physician diagnoses of Alzheimer’s disease (AD) or other types of dementia

 

-- overall, 131,821 participants were included

-- NHS baseline characteristics:

    -- mean age 46, married 72%, education level 70% registered nurses, 20% bachelor’s degree, 10% graduate degree

    -- current smoking 7%, duration of physical activity 1.7 hours/week

    -- BMI 27, total energy intake 1700 kcal/d, red meat 0.7 servings/d; Alternative Healthy Eating Index (AHEI) 39

        -- AHEI is a dietary scoring system that includes healthy foods consisting of vegetables, fruits, whole grains, nuts, omega-3's, polyunsaturated fatty acids, and the unhealthy ones of sugar-sweetened drinks, red/processed meats, trans fats, and sodium; alcohol intake (which was scored differently), with total score up to 110. higher scores are associated with decreased risk of heart disease, cancer and all-cause mortality

    -- medical history: hypercholesterolemia in 59%, hypertension in 49%, family history of dementia in 19%, diabetes in 10%

    -- meds/vitamins: multivitamins 60%, postmenopausal hormones 47%, aspirin 44%, antidepressants 8%

    -- of the above, comparing those in the lowest to highest quartile of caffeinated coffee intake: lower marriage rate, higher BMI, more smoking, higher total energy intake, more red meat intake, higher levels of all 4 of the medical history diagnoses and medications/vitamins

    -- beverage consumption range comparing lowest to highest caffeinated coffee intake: caffeine 19 g/d to 657 g/d; caffeinated coffee 0 to 4.5 cups/d, decaf coffee 0.1 to 0 cups/d,  tea 0 cups/d in each group, alcohol 0 to 1.1 g/d, sugar-sweetened beverages 0.9 to 0.4 servings/d

 

-- HPFS baseline characteristics:

    -- mean age: 54, 72% married

    -- current smoking 3%, duration of physical activity 3.8 hours/week

    -- BMI 26, total energy intake 1900 kcal/d, red meat 0.7 servings/d; Alternative Healthy Eating Index (AHEI) 44

    -- medical history: hypercholesterolemia in 47%, hypertension in 39%, family history of dementia in 15%, diabetes in 8%

    -- meds/vitamins: multivitamins 54%, aspirin 56%, antidepressants 4%

    -- of the above, comparing those in lowest to highest quartile of caffeinated coffee intake:  lower marriage rate, more smoking, higher total energy intake, more red meat intake, higher BMI, higher levels of all 4 of the medical history diagnoses and medications/vitamins

    -- beverage consumption range comparing lowest to highest caffeinated coffee intake: caffeine 19 g/d to 379 g/d; caffeinated coffee 0 to 2.5 cups/d, decaf coffee 0.1 to 0 cups/d,  tea 0 cups/d in each group, alcohol 2 to 9.3 g/d

 

-- of note, there was no differentiation in the databases for caffeinated vs noncaffeinated teas 

 

-- Primary outcome: dementia as associated with coffee and tea consumption

-- Secondary outcomes:

    -- subjective cognitive decline assessed by a questionnaire-based score (range, 0-7; higher scores indicate greater perceived cognitive decline; cases defined as those with a score >=3)

           --objective cognitive function assessed only in the NHS cohort using telephone-based neuropsychological tests such as the Telephone Interview for Cognitive Status (TICS) score (range: 0-41) and by a measure of global cognition (a standardized mean z score for all 6 administered cognitive tests).

Results:

 -- incident dementia during up to 43 years of follow-up (median 36.8 years; IQR, 28-42 years): 11,033 dementia cases (7975 in NHS cohort and 3058 in HPFS cohort):

    -- adjusting for potential confounders and pooling both cohort results, and comparing the fourth [highest] quartile of consumption with the first [lowest] quartile:

        -- higher caffeinated coffee intake was significantly associated with lower dementia risk:

            -- first quartile (lowest caffeinated coffee intake): 330 per 100,000 person-years

            -- second quartile: 298 per 100,000 person-years

            -- third quartile: 229 per 100,000 person-years

            -- fourth quartile: 141 per 100,000 person-years

        -- in adjusted model for the pooled cohort, this all translates to:

            -- caffeinated coffee: 141 in quartile 4 vs 330 cases in quartile 1 per 100 000 person-years, 18% decrease, adjusted hazard ratio aHR 0.82 (0.76-0.89), p<0.001

            -- decaffeinated coffee: no significant difference

-- prevalence of subjective cognitive decline in pooled cohort:

    -- caffeinated coffee intake:

        -- decline of 9.5% for the first quartile of intake; 7.8% in the fourth quartile

        -- in adjusted model comparing highest vs lowest quartiles: 15% decrease, aHR 0.85 (0.78-0.93), p<0.001

     -- tea intake: 

         -- lower in the adjusted model of the pooled cohort (7.8% vs 9.5%, respectively); prevalence ratio, a 14% decrease, aHR 0.86 [ 0.80 to 0.93]), p<0.001.

        -- other subjective scores comparing the highest tea intake tertile vs lowest:

            -- TICS score: mean difference 0.16 (0.08 to 0.25), p=0.001

            -- verbal memory score: mean difference 0.05, (0.03 to 0.07), p<0.001

            --global score: mean difference 0.04 (0.02 to 0.06), p<0.001

    -- decaffeinated coffee intake:

        -- in adjusted model comparing highest with lowest tertile of intake, 16% increase, aHR 1.16 (1.08-1.24), p<0.001

        -- verbal memory score 0.03 units lower in highest vs lowest tertile of intake, mean difference -0.03 (-0.05 to -0.091), p=0.01

-- NHS cohort, :

    -- higher caffeinated coffee intake was associated with better objective cognitive performance

    -- compared with participants in the lowest quartile, those in the highest quartile had:

        --  higher mean in the objective TICS score (mean difference 0.11 [0.01 to 0.21]), p=0.03

        --  higher mean global cognition score (mean difference, 0.02 [−0.01 to 0.04]), though this association was not statistically significant (P = .06)

    -- higher intake of tea showed similar associations with these cognitive outcomes

    -- decaffeinated coffee intake was not associated with lower dementia risk or better cognitive performance

 

  -- caffeine intake: their results of high caffeine intake were very consistent with the results of high caffeinated coffee intake

 

-- overall, there was a dose-response analysis that showed a nonlinear inverse associations of caffeinated coffee and tea intake levels with dementia risk and subjective cognitive decline:

    -- this graph of documented incident dementia reveals the rapid decline with up to one cup/d, continued decline to the 2-3 cups/day group, but plateauing thereafter

 

 A recent epidemiological study found that the incidence of decreased cognitive function and the risk of dementia were less common in those drinking caffeinated coffee or tea (see cognitive benefit coffee tea JAMA2026 in dropbox, or doi:10.1001/jama.2025.27259)

 

Details:

-- two prospective US cohort studies, one including female participants in the Nurses’ Health Study (NHS, n = 86,606, with data from 1980-2023) and male participants from the Health Professionals Follow-up Study (HPFS, n = 45,215, with data from 1986-2023) who did not have cancer, Parkinson disease, or dementia at study entry

-- These studies employed a rigorously validated assessment of food and beverage intake through food frequency questionnaires over these many years, with dietary intake collected and updated every 2 to 4 years

-- dementia was identified via death records that included 98% complete data through state records, the National Death Index, next-of-kin confirmation, postal authorities, and biennial self-reported physician diagnoses of Alzheimer’s disease (AD) or other types of dementia

 

-- overall, 131,821 participants were included

-- NHS baseline characteristics:

    -- mean age 46, married 72%, education level 70% registered nurses, 20% bachelor’s degree, 10% graduate degree

    -- current smoking 7%, duration of physical activity 1.7 hours/week

    -- BMI 27, total energy intake 1700 kcal/d, red meat 0.7 servings/d; Alternative Healthy Eating Index (AHEI) 39

        -- AHEI is a dietary scoring system that includes healthy foods consisting of vegetables, fruits, whole grains, nuts, omega-3's, polyunsaturated fatty acids, and the unhealthy ones of sugar-sweetened drinks, red/processed meats, trans fats, and sodium; alcohol intake (which was scored differently), with total score up to 110. higher scores are associated with decreased risk of heart disease, cancer and all-cause mortality

    -- medical history: hypercholesterolemia in 59%, hypertension in 49%, family history of dementia in 19%, diabetes in 10%

    -- meds/vitamins: multivitamins 60%, postmenopausal hormones 47%, aspirin 44%, antidepressants 8%

    -- of the above, comparing those in the lowest to highest quartile of caffeinated coffee intake: lower marriage rate, higher BMI, more smoking, higher total energy intake, more red meat intake, higher levels of all 4 of the medical history diagnoses and medications/vitamins

    -- beverage consumption range comparing lowest to highest caffeinated coffee intake: caffeine 19 g/d to 657 g/d; caffeinated coffee 0 to 4.5 cups/d, decaf coffee 0.1 to 0 cups/d,  tea 0 cups/d in each group, alcohol 0 to 1.1 g/d, sugar-sweetened beverages 0.9 to 0.4 servings/d

 

-- HPFS baseline characteristics:

    -- mean age: 54, 72% married

    -- current smoking 3%, duration of physical activity 3.8 hours/week

    -- BMI 26, total energy intake 1900 kcal/d, red meat 0.7 servings/d; Alternative Healthy Eating Index (AHEI) 44

    -- medical history: hypercholesterolemia in 47%, hypertension in 39%, family history of dementia in 15%, diabetes in 8%

    -- meds/vitamins: multivitamins 54%, aspirin 56%, antidepressants 4%

    -- of the above, comparing those in lowest to highest quartile of caffeinated coffee intake:  lower marriage rate, more smoking, higher total energy intake, more red meat intake, higher BMI, higher levels of all 4 of the medical history diagnoses and medications/vitamins

    -- beverage consumption range comparing lowest to highest caffeinated coffee intake: caffeine 19 g/d to 379 g/d; caffeinated coffee 0 to 2.5 cups/d, decaf coffee 0.1 to 0 cups/d,  tea 0 cups/d in each group, alcohol 2 to 9.3 g/d

 

-- of note, there was no differentiation in the databases for caffeinated vs noncaffeinated teas 

 

-- Primary outcome: dementia as associated with coffee and tea consumption

-- Secondary outcomes:

    -- subjective cognitive decline assessed by a questionnaire-based score (range, 0-7; higher scores indicate greater perceived cognitive decline; cases defined as those with a score >=3)

           --objective cognitive function assessed only in the NHS cohort using telephone-based neuropsychological tests such as the Telephone Interview for Cognitive Status (TICS) score (range: 0-41) and by a measure of global cognition (a standardized mean z score for all 6 administered cognitive tests).

Results:

 -- incident dementia during up to 43 years of follow-up (median 36.8 years; IQR, 28-42 years): 11,033 dementia cases (7975 in NHS cohort and 3058 in HPFS cohort):

    -- adjusting for potential confounders and pooling both cohort results, and comparing the fourth [highest] quartile of consumption with the first [lowest] quartile:

        -- higher caffeinated coffee intake was significantly associated with lower dementia risk:

            -- first quartile (lowest caffeinated coffee intake): 330 per 100,000 person-years

            -- second quartile: 298 per 100,000 person-years

            -- third quartile: 229 per 100,000 person-years

            -- fourth quartile: 141 per 100,000 person-years

        -- in adjusted model for the pooled cohort, this all translates to:

            -- caffeinated coffee: 141 in quartile 4 vs 330 cases in quartile 1 per 100 000 person-years, 18% decrease, adjusted hazard ratio aHR 0.82 (0.76-0.89), p<0.001

            -- decaffeinated coffee: no significant difference

-- prevalence of subjective cognitive decline in pooled cohort:

    -- caffeinated coffee intake:

        -- decline of 9.5% for the first quartile of intake; 7.8% in the fourth quartile

        -- in adjusted model comparing highest vs lowest quartiles: 15% decrease, aHR 0.85 (0.78-0.93), p<0.001

     -- tea intake: 

         -- lower in the adjusted model of the pooled cohort (7.8% vs 9.5%, respectively); prevalence ratio, a 14% decrease, aHR 0.86 [ 0.80 to 0.93]), p<0.001.

        -- other subjective scores comparing the highest tea intake tertile vs lowest:

            -- TICS score: mean difference 0.16 (0.08 to 0.25), p=0.001

            -- verbal memory score: mean difference 0.05, (0.03 to 0.07), p<0.001

            --global score: mean difference 0.04 (0.02 to 0.06), p<0.001

    -- decaffeinated coffee intake:

        -- in adjusted model comparing highest with lowest tertile of intake, 16% increase, aHR 1.16 (1.08-1.24), p<0.001

        -- verbal memory score 0.03 units lower in highest vs lowest tertile of intake, mean difference -0.03 (-0.05 to -0.091), p=0.01

-- NHS cohort, :

    -- higher caffeinated coffee intake was associated with better objective cognitive performance

    -- compared with participants in the lowest quartile, those in the highest quartile had:

        --  higher mean in the objective TICS score (mean difference 0.11 [0.01 to 0.21]), p=0.03

        --  higher mean global cognition score (mean difference, 0.02 [−0.01 to 0.04]), though this association was not statistically significant (P = .06)

    -- higher intake of tea showed similar associations with these cognitive outcomes

    -- decaffeinated coffee intake was not associated with lower dementia risk or better cognitive performance

 

  -- caffeine intake: their results of high caffeine intake were very consistent with the results of high caffeinated coffee intake

 

-- overall, there was a dose-response analysis that showed a nonlinear inverse associations of caffeinated coffee and tea intake levels with dementia risk and subjective cognitive decline:

    -- this graph of documented incident dementia reveals the rapid decline with up to one cup/d, continued decline to the 2-3 cups/day group, but plateauing thereafter

 

 

 

-- so, the most pronounced associated differences were observed with intake of approximately 2 to 3 cups per day of caffeinated coffee, or 1 to 2 cups per day of tea, or 300 mg/d of caffeine (no graphs for tea or decaf coffee); greater increases in caffeine did not seem to matter

-- in a sensitivity analyses:

    -- this study also found that the above relationship between caffeine and cognitive health was independent of taking metformin, aspirin, and lipid-lowering and antihypertensive meds

    -- though subgroups prone to AD (those with APOE4 genotype, Polygenic risk score for AD, smoking or increased BMI) are at higher risk of AD, they were at higher risk in this study, but they did not have a significant relative outcome difference.

    -- the benefit of caffeinated coffee and tea was more profound in those <=75yo:

        -- individuals <=75yo had adjusted 35% decrease in cognition: aHR 0.65 (0.56-0.76), p<0.001

        -- individuals >75yo had adjusted 19% decrease in cognition: aHR 0.81 (0.75-0.88), p<0.001

Commentary:

-- dementia is quite common, with Alzheimer’s dementia (AD) affecting more than 6 million people in the US and with projections that that would increase to 13 million by the year 2050.

-- dementia is a progressive problem, from mild cognitive impairment with measurable deficits on objected testing and ultimately to clinical dementia

-- as we know, current therapeutic alternatives for treating dementia are quite limited, which increases the importance of finding potentially modifiable risk factors. Of course, stopping smoking and maintaining a normal body weight are both potentially modifiable (more on this in the next blog).

-- In addition to this, there are quite impressive changes associated with caffeine that provide potential benefit, as found in the above large epidemiologic studies

-- coffee contains caffeine and polyphenols which reduce both oxidative stress and neuroinflammation and thereby might provide neuroprotection: for detailed analysis see https://www.mdpi.com/1422-0067/22/1/107

    -- there is also evidence that chronic caffeine exposure can affect amyloid beta plaques and tau tangle pathways (caffeine antagonizes adenosine A1 and A2A receptors, modulates synaptic transmission, and attenuates ABeta accumulates, with documentation of lowering ABeta levels, suppressing beta- and gamma-secretase activity, enhancing neural plasticity and stimulating mitochondrial function and prosurvival signaling pathways. caffeine may also lower brain proinflammatory cytokines and mitigate neuroinflammation

    -- and, caffeine has been linked to improved insulin sensitivity and vascular function, reduces the risk of type 2 diabetes, and both coffee and tea contain bioactive compounds including polyphenols, chlorogenic acid and cactechins that are antioxident and have vascular benefits by reducing oxidative stress and improving cerebrovascular function. tea also has epigallocatechin-3-gallate and L-theanine that further enhance neural relaxation (inducing a state of relaxed alertness) and neuroprotection; these all might protect against cognitive decline, in the setting of increased caffeine consumption

 

-- this if a study of 2 impressive cohorts with lots of granular data over many years, with frequent updating of that data, that found the following:

-- 2-3 cups of caffeinated coffee per day was associated with both objective and subjective decreases in cognitive decline

-- this finding was independent of the genetic risk factors found in some individuals (APOE4 genotype and AD polygenic risk score), though overall these genetic risk factors did predict worse outcomes (but still better with caffeine)

-- the overall findings are consistent with other studies: for example, an "umbrella review" of caffeine confirmed a moderate 10% decreased dementia risk, with risk of 0.90 (0.82-0.97), p<0.00001 (https://pmc.ncbi.nlm.nih.gov/articles/PMC11668367/), though these studies were much less sophisticated than the current one. a UK Biobank study found that coffee and tea consumption in patients with hypertension was associated with decreased dementia (https://europepmc.org/article/med/39256489)

    -- as an aside, the NHS/HPFS article above did mention this UK Biobank article, but the reference they cited was an article on older Japanese people. i mention this because it is unfortunately not so uncommon to have incorrect references even in top-of-the-line journals..... (ie, if a journal article is relying on a really important reference, best to check it out...)

-- the relationship they found in the current study, with a non-linear threshold effect of caffeine benefit, does have a plausible explanation: "the absorption, transportation, metabolism and storage of caffeine and other bioactive compounds in coffee and tea have physiological limits", with the CYP1A2 system being saturated at higher doses

    -- also, more coffee intake above the 2-3 cup threshold might interfere more with sleep or increase anxiety, and that might nullify a caffeine cognitive benefit

    -- there is also a coffee threshold in terms of the modest non-linear benefit for cardiovascular disease (https://www.ahajournals.org/doi/10.1161/CIRCULATIONAHA.113.005925)

-- it should be noted that the actual cognitive benefit was not so dramatic, both for objective as well as subjective findings of cognitive benefit by the TICS score of only 0.11 points and the global cognition z score

    -- but, as with all conditions that are highly prevalent, even small benefits will positively affect lots of people

-- it is unclear why those <75yo did better than the older group. ?because the older group has irreversible neural changes that limit the cognitive benefit??

 

-- the fact that the result on cognitive outcomes is similar in those taking just caffeine and those taking caffeinated coffee supports a specific role of the caffeine in coffee in the improved cognitive outcomes; the lack of objective evidence that decaffeinated coffee provides cognitive benefit also reinforces this conclusion. there is the caveat here that coffee and tea seem to have accessory neuroprotective mechanisms noted above that might improve the effect of just caffeine alone in cognitive health

-- this, of course to me, raises the issue of whether the other very important xanthine (methylxanthine) in chocolate is also associated with less decline of cognitive health. Studies have suggested this to be the case: https://pubmed.ncbi.nlm.nih.gov/27163823/ . and there is evidence that the darker the chocolate the better. see https://gmodestmedblogs.blogspot.com/2014/10/chocolate-and-memory-this-time.html

-- the many healthy benefits of coffee have been the topic of many clinical articles. https://gmodestmedblogs.blogspot.com/2019/01/coffee-and-decreased-mortality.html reviews some of the evidence and has reference to other blogs. And these blogs bring up that the actual coffee benefit may vary by the method of coffee-making

Limitations:

-- there were very substantial baseline differences in those patients in the lowest quartile of caffeinated coffee intake: those drinking the least caffeinated coffee had less social support (fewer married, who overall have less social support), more smoking, higher BMI, consumed more food, more red meat, and had more medical comorbidities and took more meds, and therefore had a higher risk of developing dementia independent of caffeine. though this study reported a multi-variable adjustment for their results, this statistical adjustment is much less accurate when there are so many variables involved.

-- we do not have information about the caffeine content of the tea consumed, though tea by itself appeared to have the same benefit as caffeinated coffee (though tea generally less caffeine than coffee)

    -- since the study also found that caffeine by itself was associated with the benefits of caffeinated coffee, that would suggest that there was a very high percent of the tea consumption being caffeinated

        -- but, as noted above, there are other potential benefits of tea beyond caffeine that could explain some of the increased cognitive benefit even in noncaffeinated tea

    -- it would still be helpful to know for sure what the benefit of noncaffeinated tea is through a new study

-- these 2 cohorts followed in this study are pretty amazing in terms of the amount of data accumulated and assessed, but the nurses and doctors involved do not really reflect a group generalizable to the overall population, with their having many fewer comorbidities, fewer meds, higher health literacy, and having an overall healthier lifestyle

-- we do not have granular information about why the patients who switched from caffeinated to noncaffeinated coffee in this study did so. did they switch out of the caffeinated group because of underlying medical conditions. and perhaps these underlying conditions put them at a higher risk of dementia, and this distorted the results. or perhaps some of the patients on caffeinated coffee elected to change to noncaffeinated coffee because of early cognitive decline and attribution of that to the caffeine. and if these were common problems, the results analyzed above would have a substantial bias (confounding by indication)

-- also, as noted above, the benefits of caffeinated coffee may vary by the method of coffee preparation, information not found in this study

-- since the two epidemiological studies used in this study were not randomized controlled trials, one cannot project causality to the results. there could well have been important residual confounding that affected the results

-- the reporting of dementia for this study may not have been fulling accurate (some from self-reported information), undercutting the specificity of the results

so, here is another study that attests to the health benefits of caffeinated coffee and tea, this one focusing on decreasing cognitive decline.

    -- and, equally important, though tangential to this study, is the likely benefit of chocolate (and the darker the better)

-- this study found that there was a threshold of benefit, with the maximum cognitive benefit being at 2-3 cups of caffeinated coffee or 2 cups of tea

-- the benefit was not numerically large for caffeinated coffee; however, given the huge numbers of people who develop significant cognitive decline, the population benefit would still be large

    -- and, the addition of the caffeine benefit with the other neuro-healthy endeavors (not smoking, decreasing overweight, controlling diabetes, etc) may well demonstrate additive benefit, eg eating well (https://gmodestmedblogs.blogspot.com/2015/11/mediterranean-diet-and-brain-volume.html, and exercise (https://gmodestmedblogs.blogspot.com/2020/01/exercise-and-cognitive-health.html) may decrease cognitive decline more with the addition of caffeinated coffee or tea

        -- ie, in the absence of rigorous data, decreasing the combination of modifiable dementia risk factors is likely to increase the benefit found with caffeine

— one concern that i have is that there are so many caffeinated beverages that people drink to achieve the beneficial effects of caffeine. these huge numbers of quite popular caffeinated soft drinks contain sugar or, even worse, have non-sugar sweeteners that are quite injurious to the gut microbiome: both sugar and artificial sweeteners have proinflammatory effects on the colonic microbiome (https://pmc.ncbi.nlm.nih.gov/articles/PMC7284805/. and we know that the microbiome is so important to the overall health of the human body. another article from the Framingham study found that non-sugar artificial sweeteners are associated with dementia: https://gmodestmedblogs.blogspot.com/2017/05/stroke-and-dementia-and-artificial.html.

-- so i think that part of our public health message should be that caffeinated coffee and tea are somewhat beneficial in decreasing the risk of dementia, but best by minimizing any drinks with added sugar or especially with non-sugar artificial sweeteners. And healthy eating, exercise, not smoking, maintaining a healthy body weight, controlling hypertension and diabetes all decrease the risk of dementia. And the more, the merrier

geoff

-----------------------------------

If you would like to be on the regular email list for upcoming blogs, please contact me at gmodest@bidmc.harvard.edu

to get access to all of the blogs:  go to http://gmodestmedblogs.blogspot.com/ to see the blogs in reverse chronological order

or you can just click on the magnifying glass on top right, then type in a name in the search box and get all the blogs with that name in them

 

-- so, the most pronounced associated differences were observed with intake of approximately 2 to 3 cups per day of caffeinated coffee, or 1 to 2 cups per day of tea, or 300 mg/d of caffeine (no graphs for tea or decaf coffee); greater increases in caffeine did not seem to matter

-- in a sensitivity analyses:

    -- this study also found that the above relationship between caffeine and cognitive health was independent of taking metformin, aspirin, and lipid-lowering and antihypertensive meds

    -- though subgroups prone to AD (those with APOE4 genotype, Polygenic risk score for AD, smoking or increased BMI) are at higher risk of AD, they were at higher risk in this study, but they did not have a significant relative outcome difference.

    -- the benefit of caffeinated coffee and tea was more profound in those <=75yo:

        -- individuals <=75yo had adjusted 35% decrease in cognition: aHR 0.65 (0.56-0.76), p<0.001

        -- individuals >75yo had adjusted 19% decrease in cognition: aHR 0.81 (0.75-0.88), p<0.001

Commentary:

-- dementia is quite common, with Alzheimer’s dementia (AD) affecting more than 6 million people in the US and with projections that that would increase to 13 million by the year 2050.

-- dementia is a progressive problem, from mild cognitive impairment with measurable deficits on objected testing and ultimately to clinical dementia

-- as we know, current therapeutic alternatives for treating dementia are quite limited, which increases the importance of finding potentially modifiable risk factors. Of course, stopping smoking and maintaining a normal body weight are both potentially modifiable (more on this in the next blog).

-- In addition to this, there are quite impressive changes associated with caffeine that provide potential benefit, as found in the above large epidemiologic studies

-- coffee contains caffeine and polyphenols which reduce both oxidative stress and neuroinflammation and thereby might provide neuroprotection: for detailed analysis see https://www.mdpi.com/1422-0067/22/1/107

    -- there is also evidence that chronic caffeine exposure can affect amyloid beta plaques and tau tangle pathways (caffeine antagonizes adenosine A1 and A2A receptors, modulates synaptic transmission, and attenuates ABeta accumulates, with documentation of lowering ABeta levels, suppressing beta- and gamma-secretase activity, enhancing neural plasticity and stimulating mitochondrial function and prosurvival signaling pathways. caffeine may also lower brain proinflammatory cytokines and mitigate neuroinflammation

    -- and, caffeine has been linked to improved insulin sensitivity and vascular function, reduces the risk of type 2 diabetes, and both coffee and tea contain bioactive compounds including polyphenols, chlorogenic acid and cactechins that are antioxident and have vascular benefits by reducing oxidative stress and improving cerebrovascular function. tea also has epigallocatechin-3-gallate and L-theanine that further enhance neural relaxation (inducing a state of relaxed alertness) and neuroprotection; these all might protect against cognitive decline, in the setting of increased caffeine consumption

 

-- this if a study of 2 impressive cohorts with lots of granular data over many years, with frequent updating of that data, that found the following:

-- 2-3 cups of caffeinated coffee per day was associated with both objective and subjective decreases in cognitive decline

-- this finding was independent of the genetic risk factors found in some individuals (APOE4 genotype and AD polygenic risk score), though overall these genetic risk factors did predict worse outcomes (but still better with caffeine)

-- the overall findings are consistent with other studies: for example, an "umbrella review" of caffeine confirmed a moderate 10% decreased dementia risk, with risk of 0.90 (0.82-0.97), p<0.00001 (https://pmc.ncbi.nlm.nih.gov/articles/PMC11668367/), though these studies were much less sophisticated than the current one. a UK Biobank study found that coffee and tea consumption in patients with hypertension was associated with decreased dementia (https://europepmc.org/article/med/39256489)

    -- as an aside, the NHS/HPFS article above did mention this UK Biobank article, but the reference they cited was an article on older Japanese people. i mention this because it is unfortunately not so uncommon to have incorrect references even in top-of-the-line journals..... (ie, if a journal article is relying on a really important reference, best to check it out...)

-- the relationship they found in the current study, with a non-linear threshold effect of caffeine benefit, does have a plausible explanation: "the absorption, transportation, metabolism and storage of caffeine and other bioactive compounds in coffee and tea have physiological limits", with the CYP1A2 system being saturated at higher doses

    -- also, more coffee intake above the 2-3 cup threshold might interfere more with sleep or increase anxiety, and that might nullify a caffeine cognitive benefit

    -- there is also a coffee threshold in terms of the modest non-linear benefit for cardiovascular disease (https://www.ahajournals.org/doi/10.1161/CIRCULATIONAHA.113.005925)

-- it should be noted that the actual cognitive benefit was not so dramatic, both for objective as well as subjective findings of cognitive benefit by the TICS score of only 0.11 points and the global cognition z score

    -- but, as with all conditions that are highly prevalent, even small benefits will positively affect lots of people

-- it is unclear why those <75yo did better than the older group. ?because the older group has irreversible neural changes that limit the cognitive benefit??

 

-- the fact that the result on cognitive outcomes is similar in those taking just caffeine and those taking caffeinated coffee supports a specific role of the caffeine in coffee in the improved cognitive outcomes; the lack of objective evidence that decaffeinated coffee provides cognitive benefit also reinforces this conclusion. there is the caveat here that coffee and tea seem to have accessory neuroprotective mechanisms noted above that might improve the effect of just caffeine alone in cognitive health

-- this, of course to me, raises the issue of whether the other very important xanthine (methylxanthine) in chocolate is also associated with less decline of cognitive health. Studies have suggested this to be the case: https://pubmed.ncbi.nlm.nih.gov/27163823/ . and there is evidence that the darker the chocolate the better. see https://gmodestmedblogs.blogspot.com/2014/10/chocolate-and-memory-this-time.html

-- the many healthy benefits of coffee have been the topic of many clinical articles. https://gmodestmedblogs.blogspot.com/2019/01/coffee-and-decreased-mortality.html reviews some of the evidence and has reference to other blogs. And these blogs bring up that the actual coffee benefit may vary by the method of coffee-making

Limitations:

-- there were very substantial baseline differences in those patients in the lowest quartile of caffeinated coffee intake: those drinking the least caffeinated coffee had less social support (fewer married, who overall have less social support), more smoking, higher BMI, consumed more food, more red meat, and had more medical comorbidities and took more meds, and therefore had a higher risk of developing dementia independent of caffeine. though this study reported a multi-variable adjustment for their results, this statistical adjustment is much less accurate when there are so many variables involved.

-- we do not have information about the caffeine content of the tea consumed, though tea by itself appeared to have the same benefit as caffeinated coffee (though tea generally less caffeine than coffee)

    -- since the study also found that caffeine by itself was associated with the benefits of caffeinated coffee, that would suggest that there was a very high percent of the tea consumption being caffeinated

        -- but, as noted above, there are other potential benefits of tea beyond caffeine that could explain some of the increased cognitive benefit even in noncaffeinated tea

    -- it would still be helpful to know for sure what the benefit of noncaffeinated tea is through a new study

-- these 2 cohorts followed in this study are pretty amazing in terms of the amount of data accumulated and assessed, but the nurses and doctors involved do not really reflect a group generalizable to the overall population, with their having many fewer comorbidities, fewer meds, higher health literacy, and having an overall healthier lifestyle

-- we do not have granular information about why the patients who switched from caffeinated to noncaffeinated coffee in this study did so. did they switch out of the caffeinated group because of underlying medical conditions. and perhaps these underlying conditions put them at a higher risk of dementia, and this distorted the results. or perhaps some of the patients on caffeinated coffee elected to change to noncaffeinated coffee because of early cognitive decline and attribution of that to the caffeine. and if these were common problems, the results analyzed above would have a substantial bias (confounding by indication)

-- also, as noted above, the benefits of caffeinated coffee may vary by the method of coffee preparation, information not found in this study

-- since the two epidemiological studies used in this study were not randomized controlled trials, one cannot project causality to the results. there could well have been important residual confounding that affected the results

-- the reporting of dementia for this study may not have been fulling accurate (some from self-reported information), undercutting the specificity of the results

so, here is another study that attests to the health benefits of caffeinated coffee and tea, this one focusing on decreasing cognitive decline.

    -- and, equally important, though tangential to this study, is the likely benefit of chocolate (and the darker the better)

-- this study found that there was a threshold of benefit, with the maximum cognitive benefit being at 2-3 cups of caffeinated coffee or 2 cups of tea

-- the benefit was not numerically large for caffeinated coffee; however, given the huge numbers of people who develop significant cognitive decline, the population benefit would still be large

    -- and, the addition of the caffeine benefit with the other neuro-healthy endeavors (not smoking, decreasing overweight, controlling diabetes, etc) may well demonstrate additive benefit, eg eating well (https://gmodestmedblogs.blogspot.com/2015/11/mediterranean-diet-and-brain-volume.html, and exercise (https://gmodestmedblogs.blogspot.com/2020/01/exercise-and-cognitive-health.html) may decrease cognitive decline more with the addition of caffeinated coffee or tea

        -- ie, in the absence of rigorous data, decreasing the combination of modifiable dementia risk factors is likely to increase the benefit found with caffeine

— one concern that i have is that there are so many caffeinated beverages that people drink to achieve the beneficial effects of caffeine. these huge numbers of quite popular caffeinated soft drinks contain sugar or, even worse, have non-sugar sweeteners that are quite injurious to the gut microbiome: both sugar and artificial sweeteners have proinflammatory effects on the colonic microbiome (https://pmc.ncbi.nlm.nih.gov/articles/PMC7284805/. and we know that the microbiome is so important to the overall health of the human body. another article from the Framingham study found that non-sugar artificial sweeteners are associated with dementia: https://gmodestmedblogs.blogspot.com/2017/05/stroke-and-dementia-and-artificial.html.

-- so i think that part of our public health message should be that caffeinated coffee and tea are somewhat beneficial in decreasing the risk of dementia, but best by minimizing any drinks with added sugar or especially with non-sugar artificial sweeteners. And healthy eating, exercise, not smoking, maintaining a healthy body weight, controlling hypertension and diabetes all decrease the risk of dementia. And the more, the merrier

geoff

-----------------------------------

If you would like to be on the regular email list for upcoming blogs, please contact me at gmodest@bidmc.harvard.edu

to get access to all of the blogs:  go to http://gmodestmedblogs.blogspot.com/ to see the blogs in reverse chronological order

or you can just click on the magnifying glass on top right, then type in a name in the search box and get all the blogs with that name in them