SGEM#509: I love the Java Jive & It Loves Me – Preventing Dementia with Coffee and Tea

Date: April 23, 2026  Guest Skeptic: Dr. Manrique Umaña McDermott is an attending physician specializing in Emergency Medicine based in San José, Costa Rica. He has a passion for medical education, is a renowned international speaker and serves as a faculty member in undergraduate Internal Medicine at UCIMED and postgraduate training programs in Emergency Medicine and Family and Community Medicine at the University of Costa Rica (UCR). You can follow him on X and Instagram at @umanamd.  Reference: Zhang et al. Coffee and Tea Intake, Dementia Risk, and Cognitive Function. JAMA 2026 March Case: A 47-year-old emergency physician presents to their primary care physician for a rare act of preventive care between a run of night shifts. She drinks 3 large coffees most workdays, switches to tea on post-nights when her hands are vibrating fast enough to start an IV at 20 paces and asks whether her caffeine habit is frying her brain or secretly protecting it. She has heard that coffee is either a miracle, a menace, or both, depending on which headline got posted in the group chat that week. Background: Coffee’s origin story reads like a case report from the annals of caffeinated discovery. Legend traces it back to Ethiopia, where a goat herder observed his animals behaving like over-caffeinated residents after nibbling on certain berries. From there, coffee spread through the Arabian Peninsula, where it was first cultivated and consumed in Yemen, eventually fueling the rise of coffeehouses. By the 17th century, coffee had reached Europe, where it was alternately praised as a miracle tonic and condemned as a suspicious stimulant. Over time, coffee became embedded in global culture, transitioning from a mystical brew to an industrial-scale commodity, and ultimately, a critical adjunct in emergency medicine workflow optimization. Costa Rica takes coffee seriously, arguably more seriously than most emergency departments (EDs) take shift coffee orders. Introduced in the late 18th century, coffee quickly became a cornerstone of the country’s economy and identity. The government actively promoted coffee cultivation, even offering farmers free land to grow it, resulting in a thriving industry based on small family farms rather than large plantations. Costa Rican coffee is renowned for its high quality, thanks to ideal growing conditions: volcanic soil, high altitude, and just enough rain to keep things interesting. The country even banned the production of low-quality coffee. Today, Costa Rica is a leader in sustainable coffee production. So, the next time you’re powering through a night shift, there’s a good chance your cognitive performance is being supported by carefully cultivated beans from a hillside in Central America. Emergency physicians do not need a pathophysiology lecture on caffeine; they need a fresh cup. Coffee is practically a staffing model, while tea is the civilized cousin, and both have long been part of the informal pharmacopeia of night shift survival. The real question is whether our specialty’s favourite legal liquid stimulant does anything beyond keeping our differential diagnoses alive until sunrise. Biologically, the hypothesis is plausible. Coffee and tea contain caffeine and other bioactive compounds, including polyphenols, that may influence oxidative stress, neuroinflammation, vascular function, and insulin sensitivity. These are all pathways that could plausibly matter for cognitive decline and dementia. But human studies have been inconsistent, and many older studies did not clearly distinguish between caffeinated and decaffeinated coffee.  Clinical Question: Is long-term intake of caffeinated coffee, decaffeinated coffee, or tea associated with incident dementia and cognitive outcomes? Reference: Zhang et al. Coffee and Tea Intake, Dementia Risk, and Cognitive Function. JAMA 2026 March Population: Adults from the Nurses’ Health Study (NHS) and Health Professionals Follow-up Study (HPFS). Excluded: People with cancer, Parkinson’s disease, or dementia at baseline; those with implausible total energy intake; and those missing caffeinated beverage intake data.  Exposure: Long-term intake of caffeinated coffee, decaffeinated coffee, and tea, assessed every 2 to 4 years with validated food frequency questionnaires (FFQs).  Comparison: Lower intake categories, especially the lowest quartile or tertile of consumption, depending on the beverage.  Outcomes: Primary Outcome: Incident dementia, identified via death records and physician diagnoses.  Secondary outcomes: Subjective cognitive decline and objective cognitive function; objective testing was assessed only in the NHS cohort, including a telephone interview for cognitive status (TICS) and composite cognitive measures.  Type of Study: Prospective observational cohort study. Authors’ Conclusions: “Greater consumption of caffeinated coffee and tea was associated with lower risk of dementia and modestly better cognitive function.”  Quality Checklist for Observational Studies: Did the study address a clearly focused issue? Yes Did the authors use an appropriate method to answer their question? Yes  Was the cohort recruited in an acceptable way? Yes Was the exposure accurately measured to minimize bias? Yes Was the outcome accurately measured to minimize bias? Unsure Have the authors identified all-important confounding factors? Unsure Was the follow-up of subjects complete enough? Yes How precise are the results? Reasonably for the primary outcome. Do you believe the results? Yes Can the results be applied to the local population? Yes Do the results of this study fit with other available evidence? Yes Who funded the study? Multiple National Institutes of Health grants. Did the authors declare any conflicts of interest? Yes. Dr. Hu reported receiving research support from the Analysis Group; no other disclosures were reported. Result: The study included 131,821 middle-aged US health professionals from 2 cohorts: female nurses in the NHS and male health professionals in the HPFS. The mean age at baseline was 46.2 years in NHS and 53.8 years in HPFS, and 66% of the pooled sample was female. Follow-up was extraordinarily long, up to 43 years, with a median of 37 years. Higher coffee consumers tended to be younger, drink more alcohol, smoke more, and consume more total energy.  Key Result:  Higher caffeinated coffee intake was associated with lower dementia risk. Primary Outcome: Incidence of dementia HR 0.82 (95% CI 0.76 to 0.89) Secondary Outcomes? Higher caffeinated coffee intake was associated with lower prevalence of cognitive decline, 7.8% versus 9.5% in the highest versus the lowest quartile, prevalence ratio 0.85. In the NHS objective testing subgroup, higher caffeinated coffee intake was associated with a small increase in the telephone interview for cognitive status (TICS) score (mean difference 0.11), whereas the global cognition result (mean difference 0.02) did not reach conventional statistical significance. Tea showed similar patterns; decaf did not.  1. Residual Confounders: This is the biggest threat. Observational cohort studies are appropriate for exposure questions but inherently vulnerable to bias from prognostic differences between the exposed and the unexposed groups. Even after multivariable adjustment, coffee and tea drinkers may differ from lower-intake participants in ways that matter for cognition. These include sleep habits, education, work patterns, health-seeking behaviour, medication use, diet quality, and unmeasured social factors. The authors acknowledge residual confounding, including neuroactive drugs that were not fully captured across follow-up.  2. Reverse Causation: Even in a prospective cohort, early cognitive decline could influence beverage habits or the reliability of dietary self-report before formal dementia diagnosis. Someone drifting into cognitive decline may reduce coffee intake due to sleep disturbance, jitters, gastrointestinal symptoms, or a caregiver’s influence, making coffee appear protective when declining cognition changes the exposure. The authors conducted sensitivity analyses but still acknowledge that reverse causation cannot be fully excluded.  3. Imperfect Exposure Measurement: Repeated validated food frequency questionnaires (FFQs) are better than a single baseline diet snapshot, yet they cannot fully capture brew strength, cup size, preparation method, shifts between caffeinated and decaf products, or additives like sugar and cream. Mismeasurement of exposure can distort observed associations, and nutrition studies are especially vulnerable because the dose is harder to standardize than a prescription drug.  4. Outcome Ascertainment: The primary outcome relied on death records and self-reported physician diagnoses, with medical-record confirmation when available. That is pragmatic and probably better in health professionals than in the general population, but it still leaves room for missed cases and misclassification. The authors also could not isolate Alzheimer's disease specifically, which matters because all-cause dementia may lump together biologically distinct conditions with different relationships to caffeine.  5. Time-Varying Confounding & Imperfect Capture of Exposure Over Time: This combines nerdy point #1 and #3. Although the authors used repeated FFQs, beverage consumption is likely to change meaningfully over a 30- to 40-year follow-up. Participants may switch between caffeinated and decaffeinated coffee, change dose, or modify intake due to early (subclinical) cognitive decline, sleep issues, comorbidities, or physician advice. Even with cumulative averaging, this introduces time-varying confounding and exposure misclassification, which represents a major threat in longitudinal observational studies....