Caffeine

Of 7,211 urban working Indians who answered the caffeine-volume question in WHI 2026, 76.6% report at least one caffeinated drink on a typical workday. That figure has not moved — it was 76.3% in WHI 2025, on a smaller sample, and sits a few points under the global daily-consumption estimate of roughly 80%. Caffeine is the most widely consumed psychoactive substance in the world, and India's urban professionals are unremarkable in using it. The prevalence headline is saturated. It will read about the same in 2027.

The signal that does move sits inside the volume distribution.

Two cups is the modal workday. Three-quarters of the working population sits at two or fewer — a pattern the clinical literature treats as benign for most healthy adults. The reportable population is the 15.0% who drink three or more, one in five of everyone who drinks caffeine at all. This is the band where intake stops tracking enjoyment and starts tracking something else.

The volume rises with age; the intensity rises faster#

Any-caffeine consumption climbs gently with age — 73.0% in the 20s to 86.9% in the 50s. Heavy use climbs far more steeply.

The heavy-use rate more than doubles between the youngest and oldest working bands. The deterioration is gradual through the 30s, then accelerates after 40 — the same shape the Sleep page found for short sleep, which sets in after 40 rather than in the 30s. Tolerance is the obvious mechanism: two decades of daily blockade up-regulate adenosine receptors, and the dose required to clear the morning fog ratchets upward.

The 2025 edition read this gradient as a story about seniority and workplace pressure — "92% of senior professionals have a caffeine dependency." The cleaner reading is biological. The dose climbs because the receptor count climbs, and the receptor count climbs with years of exposure. The 50s band is small (n=222) and carries noise, but the direction is consistent across every adjacent band and matches the pharmacology.

Heavy use tracks stress more than it tracks anything demographic#

The single cleanest predictor of heavy caffeine use in the survey is not age, gender, city, or income. It is self-reported stress.

A respondent reporting high stress is 2.1× more likely to be a three-plus-cup drinker than one reporting low stress. The gap — 10.6 percentage points — is wider than the gap across any single demographic cut on this page. Caffeine intensity is a behavioural readout of strain — heavy use reads as a downstream symptom of the 40.4% high-stress burden the Stress page documents. The direction of causation runs both ways and the survey cannot separate them: stress drives people to the third cup, and caffeine at that volume worsens the sleep and elevates the cortisol that feed the stress back. The relationship is a loop, and the heavy-use band is where the loop tightens.

The heaviest users sit at the top of the income and seniority ladder#

The crisis-fuel intuition — that heavy caffeine use clusters among the lowest-paid and most precarious — does not hold in our sample. Heavy use is flat across the bottom and middle of the income range, then jumps at the top.

The bottom five income bands sit within half a point of each other, all hugging the 15% average. The top band — ₹40L and above — runs at 25.5%, two-thirds higher, though the sample there is thin (n=141). The same pattern shows in job function: General Management and Strategy roles report 28.2% heavy use, the highest of any function, while Engineering and Software Development sit lowest at 11.0%. Heavy caffeine use rises with seniority and pay, not with precarity. The people drinking the most are the people with the most decisions to make and the longest days to make them in.

Short sleep rises cup by cup#

The survey is not longitudinal and cannot prove that caffeine causes the lost sleep. But the dose-response is as clean as survey data gets.

Every additional cup is associated with a higher short-sleep rate, with no reversal across the five bands. The four-plus-cup drinkers sleep under six hours at 41.9% — well over half again the rate of the non-drinkers at 25.1%. The Sleep page found the same association from the other direction, with three-plus-cup drinkers running 1.38× more likely to be short sleepers; the full ladder here shows the relationship is graded, not a threshold effect. Caffeine and short sleep also share the stress driver above, which inflates the raw association. But the monotonic shape is exactly what the pharmacology predicts, and the gap between the ends of the ladder is too wide to be stress alone.

Two in three drink their last cup after 3 PM#

Timing is the most modifiable risk on this page, and the timing data is unflattering. Among the 3,339 respondents who answered the timing question:

65.4% take their last caffeinated drink after 3 PM, and 22.3% after 6 PM. For a substance with a five-hour half-life, an after-6 PM cup still has a quarter of its caffeine in the bloodstream around 2 AM. The short-sleep numbers follow the clock: 30.5% short sleep among those who stop before noon, 40.6% among those drinking after 6 PM — a ten-point spread.

The after-6 PM group is the clean read: the latest drinkers sleep worst. The 12–3 PM bar runs anomalously high (39.0%) on a small base of 344, and is best treated as noise rather than a real midday penalty. The actionable pattern is the late-evening tail.