Drinking six or more cups of coffee a day is associated with smaller brain volume and a 53% increased risk for dementia, results of a large study suggest.
“With coffee intake, moderation is the key, and especially high levels of consumption may have adverse long-term effects on the brain,” study investigator Elina Hypponen, PhD, professor of nutritional and genetic epidemiology and director of the Australian Center for Precision Health at the University of South Australia, told Medscape Medical News.
“These new data are concerning, and there is a need to conduct further carefully controlled studies to clarify the effects of coffee on the brain.”
The study was published online June 24 in Nutritional Neuroscience.
Coffee is a potent nervous system stimulant and is among the most popular nonalcoholic beverages. Some previous research suggests it benefits the brain, but the investigators note that other research shows a negative or U-shaped relationship.
To investigate, the researchers examined data from the UK Biobank, a long-term prospective epidemiologic study of more than 500,000 participants aged 37 to 73 years who were recruited in 22 assessment center in the United Kingdom between March 2006 and October 2010.
During the baseline assessment, information was gathered using touchscreen questionnaires, verbal interviews, and physical examinations that involved collection of blood, urine, and saliva samples. An imaging substudy was incorporated in 2014, the goal of which was to conduct brain, heart, and body MRI imaging for 100,000 participants.
The investigators conducted analyses on disease outcomes for 398,646 participants for whom information on habitual coffee consumption was available. Brain volume analyses were conducted in 17,702 participants for whom valid brain imaging data were available.
Participants reported coffee intake in cups per day. Researchers grouped coffee consumption into seven categories: nondrinkers, decaffeinated coffee drinkers, and caffeinated coffee drinkers who consumed <1 cup/d, 1–2 cups/d, 3–4 cups/d, 5–6 cups/d, and >6 cups/d.
The reference category was those who consumed 1–2 cups/d, rather than those who abstained from coffee, because persons who abstain are more likely to be at suboptimal health.
“Comparing the health of coffee drinkers to the health of those choosing to abstain from coffee will typically lead to an impression of a health benefit, even if there would not be one,” said Hypponen.
The researchers obtained total and regional brain volumes from the MRI imaging substudy starting 4 to 6 years after baseline assessment. They accessed information on incident dementia and stroke using primary care data, hospital admission electronic health records, national death registers, and self-reported medical conditions.
Covariates included socioeconomic, health, and other factors, such as smoking, alcohol and tea consumption, physical activity, stressful life events, and body mass index.
The investigators found that there was a linear inverse association between coffee consumption and total brain volume (fully adjusted β per cup, -1.42; 95% CI, -1.89 to -0.94), with consistent patterns for gray matter, white matter, and hippocampal volumes.
There was no evidence to support an association with white matter hyperintensity (WMH) volume (β -0.01; 95% CI, -0.07 to 0.05).
Higher Consumption, Higher Risk
The analysis also revealed a nonlinear association between coffee consumption and the odds of dementia (P nonlinearity = 0.0001), with slightly higher odds seen with non-coffee drinkers and decaffeinated coffee drinkers and more notable increases for participants in the highest categories of coffee consumption compared to light coffee drinkers.
After adjusting for all covariates, the odds ratio of dementia among persons in the category of coffee intake was 1.53 (95% CI, 1.28 – 1.83). After full adjustments, the association with heavy coffee consumption and stroke was not significant, although “we can’t exclude a weak effect,” said Hypponen.
“For the highest coffee consumption group, the data support an association which may be anywhere from 0% to 37% higher odds of stroke after full adjustment,” she added.
People at risk for hypertension may develop “unpleasant sensations” and stop drinking coffee before a serious adverse event occurs, said Hypponen. In a previous study, she and her colleagues showed that those who have genetically higher blood pressure tend to drink less coffee than their counterparts without the condition.
“This type of effect might be expected to naturally limit the adverse effects of coffee on the risk of stroke,” said Hypponen.
The odds remained elevated for participants drinking >6 cups/d after accounting for sleep quality. There were no differences in risk between men and women or by age.
An examination of the consumption of tea, which often contains caffeine, did not show an association with brain volume or the odds of dementia or stroke.
“We don’t know whether the difference between associations seen for coffee and tea intake reflects the difference in related caffeine intake or some other explanation, such as dehydration or effects operating through blood cholesterol,” said Hypponen.
Although reverse causation is possible, there’s no reason to believe that it is relevant to the study results. Genetic evidence suggests a causal role of higher coffee intake on risk for Alzheimer’s disease. In addition, results of a clinical trial support the association between higher caffeine intake and smaller gray matter volume, said Hypponen.
The mechanisms linking coffee consumption to brain volumes and dementia are not well established. However, Hypponen noted that caffeine has been used to induce apoptosis in cancer studies using glial cells.
“Furthermore, adenosine receptors, which mediate many of the effects of caffeine in the brain, have been suggested to influence the release of growth factors, which in turn can have an influence on astrocyte proliferation and angiogenesis in the brain,” she said.
Some types of coffee contain cafestol, which increases blood cholesterol and can have adverse effects though related mechanisms, said Hypponen.
The mechanism may also involve dehydration, which may have a harmful effect on the brain. The study suggested a correlation between dehydration and high coffee intake.
“Of course, if this is the case, it is good news, as then we can do something about it simply by drinking some water every time we have a cup of coffee,” she said.
Coffee contains antioxidants, and although previous studies have suggested it might be beneficial, this hypothesis is “too simplistic,” said Hypponen. “While coffee is not going to be all ‘bad’ either, there are a lot of controversies and suggestions about beneficial effects of coffee which may not be true, or at least do not reflect the full story.”
If the drinking of coffee is at least partly determined by an individual’s health status, then that would often lead to misleading conclusions in observational studies, said Hypponen.
“When one uses as a comparison people who already have poor health and who do not drink coffee because of that, coffee intake will by default appear beneficial simply because there are more people with disease among those choosing abstinence,” she said.
Before now, there was “very little evidence about the association between coffee intake and brain morphology,” and the studies that were conducted were relatively small, said Hypponen.
One of these smaller studies included a group of women aged 13 to 30 years. It found that coffee consumption was not associated with total brain volumes, but the findings suggested a U-shaped association with hippocampal volume; higher values were seen both for nondrinkers and the groups with higher consumption.
A small study of elderly patients with diabetes showed no evidence of an association with white matter volume, but there was a possible age-dependent association with gray matter volume.
The largest of the earlier studies had results that were very similar to those of the current study, suggesting that increasing coffee intake is associated with smaller hippocampal volumes, said Hypponen.
One of the study’s limitation included the fact that full dietary information was only available for a subsample and that factors such as dehydration were measured at baseline rather than at the time of brain MRI.
Another possible study limitation was the use of self-reported data and the fact that lifestyle changes may have occurred between baseline and MRI or covariate measurement.
In addition, the study is subject to a healthy-volunteer bias, and its implications are restricted to White British persons. The association needs to be studied in other ethnic populations, the authors note.
A Reason to Cut Back?
Commenting on the findings for Medscape Medical News, Walter Willett, MD, DrPH, professor of epidemiology and nutrition, Harvard T. H. Chan School of Public Health, Boston, Massachusetts, said the study is large and quite well done.
“It does raise questions about an increase in risk of dementia with six or more cups of coffee per day,” said Willett. “At the same time, it provides reassurance about lack of adverse effects of coffee for those consuming three or four cups per day, and little increase in risk, if any, with five cups per day.”
It’s not entirely clear whether the increase in risk with six or more cups of coffee per day represents a “true effect” of coffee, inasmuch as the study did not seem to adjust fully for dietary factors, high consumption of alcohol, or past smoking, said Willett.
The findings don’t suggest that coffee lovers should give up their java. “But six or more cups per day is a lot, and those who drink that much might consider cutting back a bit while research continues,” said Willett.
The study was supported by the National Health and Medical Research Council.
Nutr Neurosci. Published online June 24, 2021. Abstract