Coffee, tea and more

Following 402,260 US adults for 14 years, coffee drinkers at 4 to 5 daily cups had 10 to 15% lower all-cause mortality than non-drinkers, with benefits visible from 1 cup and gone by the extremes [1]. Large European and multi-ethnic cohorts replicated the association [7].

An umbrella review spanning more than 200 meta-analyses found coffee associated with lower mortality, cardiovascular disease, and several cancers, with the largest benefit near 3 to 4 cups, and crucially the same pattern in decaffeinated coffee [4]. Coffee is the largest polyphenol source in many diets.

In 40,530 Japanese adults, 5+ daily cups of green tea meant 16% lower all-cause mortality over 11 years, driven by cardiovascular deaths [2]. In roughly 500,000 UK Biobank adults, 2+ daily cups of black tea meant 9 to 13% lower mortality [3]. Yerba mate carries a similar polyphenol load, though without cohorts this size behind it.

The mortality benefit survived a modest spoon of sugar in UK Biobank coffee drinkers [8], but blended and heavily sweetened coffee drinks carry 30 to 60 g of sugar, the territory where sugar-sweetened beverages track with 21% higher mortality at 2+ daily servings [9]. A note on milk: it neither helps nor hurts. A small acute study found milk casein blunted tea’s short-term vascular effect [10], but in the half-million-person UK Biobank cohort the mortality benefit held regardless of whether drinkers added milk or sugar [3]. Take your tea or coffee with milk without a second thought; it is the sugar dose that decides which curve you are on.

In a randomized crossover trial, 32 oz of energy drink raised blood pressure and prolonged the corrected QT interval for hours, while a matched dose of plain caffeine did not, implicating the other ingredients [5]. Energy drinks have no cohort evidence of a longevity benefit, and most carry a full sugary-drink load on top. A note on sugar-free cans: they dodge the sugary-drink curve, but they do not inherit coffee’s benefit. That benefit tracks the polyphenols (decaf shows it, item 2), which energy drinks lack, and sugar-free versions still raised blood pressure like plain caffeine in crossover trials [11] while the QT signal came from caffeine-matched comparisons, so the sugar was never the whole problem. Best read: roughly neutral, like diet soda, with an unstudied additive stack on top.

Caffeine taken 6 hours before bed still cut objectively measured sleep by more than an hour, and 3 hours before bed was worse [6]. With a 5 to 6 hour half-life, a 4pm coffee is still a quarter-dose at 2am. Short sleep carries its own mortality curve [12], easily large enough to erase coffee’s benefit.

Regulators converge on 400 mg of caffeine a day (about 4 cups of coffee) as the safe adult ceiling, and 200 mg in pregnancy [13, 14]. Separately, the IARC classifies drinking any beverage above 65°C as probably carcinogenic to the esophagus, a finding driven by scalding-hot mate studies; the mate itself was cleared [15].