Coffee, tea and dementia risk: Why the amount you drink could matter

By Dr. Liji Thomas, MDReviewed by Lauren HardakerSep 3 2026

Why age, genetics, dose, and even how coffee or tea is prepared could help explain caffeine’s complex relationship with brain health.

Study: Midlife window: moderate caffeine consumption and reduced dementia risk. Image Credit: Oksana Karuna / Shutterstock

A perspective article published in the journal npj Dementia reviewed epidemiological and mechanistic evidence to suggest that moderate caffeine consumption during midlife may be associated with a lower risk of dementia later in life.

Background

Caffeine is among the most popular psychoactive substances to be consumed worldwide. It is an adenosine receptor antagonist that enhances brain network complexity and neural activity. It can enhance neural processing while reducing cognitive load. However, these effects may vary with age and with genetic factors, such as CYP1A2 variants that determine the rate of caffeine metabolism.

Dementia is a rapidly growing public health challenge as the global population is graying. In the absence of curative treatment, preventive interventions remain paramount. The role of caffeine as a protective factor remains unproven, as prior studies have reported conflicting findings. The authors suggest that this may be due in part to differences in bioavailability induced by genetic variations in caffeine metabolism rates and cultural differences in preparation methods.

Study characteristics

The current perspective synthesizes evidence from a randomized controlled study on cerebral gray matter and blood flow responses to caffeine intake, several prospective observational studies, and meta-analyses. It was a targeted, non-systematic synthesis rather than a systematic review or new statistical analysis.

Moderate intake showed strongest protective associations

The evidence reviewed in the perspective suggested that the strongest associations with reduced dementia risk were with moderate caffeine consumption. For instance, one prospective cohort study showed that dementia risk was reduced by 18% overall with 2-3 cups of caffeinated coffee per day, comparing the highest versus lowest quartiles of coffee consumption, and by 14% with 1-2 cups of tea per day.

The same study reported a 35% lower dementia risk among participants aged 75 years or younger. The researchers observed non-linear dose-response patterns, with moderate doses associated with the greatest benefits. These associations remained significant regardless of APOE4 carrier status.

Another prospective cohort study showed that coffee consumption was negatively associated with dementia only among those with the rs762551 C allele of the CYP1A2 gene, which is associated with slower caffeine metabolism. In this subset, four or more cups of coffee per day were associated with about a 55% reduction in risk.

No protection was observed among fast metabolizers, and conversely, their risk tended to increase at high doses.

The researchers note that prolonged follow-up, electrophysiological data, and structural imaging, are all required to help define the biological mechanisms underlying these findings.

UK Biobank data on 8,715 adults aged 60-85.2 years showed that moderate coffee and tea intake was associated with a slower decline in certain areas of cognition, including fluid intelligence, which involves reasoning and problem-solving. At higher intakes, tea continued to correspond to a slower decline in fluid intelligence, while at or above 4 cups a day, coffee intake is associated with a more rapid increase in visual memory errors.

Meta-analyses have also linked tea consumption with lower dementia risk, while the landmark CAIDE study reported a 65% reduction in late-life dementia risk with moderate coffee intake in middle age. The perspective authors write:

“This suggests a critical window of opportunity during which the brain is most responsive to the neuroprotective effects of caffeine.”

Potential protective mechanisms

Coffee has been reported to reduce the perceived cognitive load and boost the ability to design procedures and strategies. Caffeine increases confidence in one’s ability to achieve a task and may optimize the use of attention when encountering a complex informational situation.

With age, adenosine A1 receptors become sparser. Caffeine blocks adenosine A1 and A2A receptors, influencing signal transmission and potentially reducing the accumulation of the dementia-associated protein β-amyloid.

In addition, adenosine antagonism increases overall excitation, shifting the system away from an inhibition-dominant state. The researchers describe the role of caffeine as shifting the neural network “toward a state of criticality, a regime of maximal information processing.” The result is a more dynamically complex state of neural network activity.

Caffeine produces a strong shift towards complexity in young adults, but produces a weaker impact on cortical activity in middle age. This is especially clear during rapid eye movement (REM) sleep. This may be related to age-related changes in adenosine receptor density.

The antioxidant properties of caffeine and other beverage constituents, including tea polyphenols, may also contribute to protective associations via reduced oxidative stress and neuroinflammation, as well as synaptic stability. These may also vary with age, genetic context, metabolic differences, and dose.

On the other hand, some mechanistic studies have linked caffeine exposure with a reduction in cerebral gray matter volume. In a small randomized study conducted during chronic sleep restriction, caffeine altered gray matter responses in a manner associated with adenosine A1 receptor availability. Separate Mendelian randomization research, which uses genetic differences to probe potentially causal relationships, linked genetically predicted higher coffee consumption with lower gray matter volume.

Adding to this apparent paradox, coffee consumption is associated with a lower risk of mortality. The authors postulate that caffeine-related gray matter reduction could reflect an efficiency-promoting mechanism involving experience-dependent neural pruning, although they emphasize that this remains speculative.

South Asian considerations

South Asians, such as those in India and Nepal, drink tea with milk and sugar, consuming 2-4 cups per day. The interaction of milk proteins and sugars with green tea catechins may reduce the bioavailability of the latter. Moreover, the CYP1A2 polymorphism profile of this population remains to be explored and requires large-scale prospective studies.

Notably, South Asian populations face a disproportionately high dementia risk linked in part to diabetes, hypertension, and dyslipidemia.

Limitations

The authors advocate caution in interpreting the epidemiological and observational evidence for associations between caffeine intake and dementia risk.

The dementia-risk evidence was largely observational, precluding causal inferences. Self-reporting of caffeine intake may introduce measurement errors and recall bias in exposure assessment.

Moreover, multiple other sources of confounding persist, including socioeconomic status, diet quality, physical activity, and cognitive activity. Caffeine sources vary widely in the dose of caffeine delivered and the method of preparation, making it difficult to identify protective dose thresholds. Reverse causality also cannot be ruled out, as early changes prior to dementia diagnosis may alter caffeine habits.

The influence of biological sex on caffeine metabolism, hormonal effects, and dementia risk remains to be explored. Finally, there are no randomized controlled trials examining the protective effect of caffeine against dementia.

Conclusions

While there appear to be associations between moderate caffeine intake in midlife and a lower risk of dementia in later life, these need to be validated by further research that accounts for confounding factors.

Moreover, the authors note that midlife habits, reduced adenosine receptor density, and functional brain complexity interact in determining cognitive performance. This limits the extent to which dietary observations, such as those on coffee consumption, can be directly translated into recommendations for reducing dementia risk.

“Ultimately, while caffeine provides short-term cognitive benefits and supports energy regulation, excessive consumption may impair sleep hygiene and alter the brain’s complexity.”

Caffeine intake may have varying effects on dementia risk depending on the stage of life, the other components of the beverage, and the genetic background. For instance, the relationship needs to be further explored in the South Asian context, adjusting for the cultural use of milk and sugar in caffeinated beverages and the routine consumption of tea as a daily staple. The metabolic and genetic profile of this population also needs to be taken into account.

Such studies could help determine whether moderate caffeine consumption during the midlife window has a meaningful role in reducing dementia risk and identify the populations most likely to benefit.