Sleep

Of 7,569 urban working Indians who answered the sleep-hours question in WHI 2026, only 30.7% report sleeping the clinically recommended seven hours or more. Another 30.7% — by coincidence the same share — sleep under six. The modal band, 6–7 hours, captures 38.6% of the working population: short of the US National Sleep Foundation recommendation of seven-to-nine hours for working-age adults, but invisible to any standard medical workup, and the experience most Indian working professionals would call "fine".

Roughly two-thirds of the working population falls below the recommendation. The technical term for the modal experience here is sleep-truncation — short of clinical adequacy but adapted to. It is the default of the Indian salaried class.

Sleep is moving the wrong way#

The 2026 reading is on a sample 2.6× the size of 2025. The direction is unambiguous — urban working Indians slept less in the year just ended than in the year before it.

The honest framing isn't a "sleep crisis", because the word crisis belongs to acute events. There is no acute event here. There's the slow year-on-year erosion of a basic biological input, in a population that's also reporting more stress, more caffeine, more sedentary work, and worse glucose. Each move alone is small. Together they describe a working life the human body wasn't designed for, conducted at the speed for which it was.

Sleep deteriorates after 40, not in the 30s#

The sleep curve isn't the glucose curve. Glucose dysfunction doubles between the 20s and the 30s; sleep is essentially flat across the first two decades of adult working life, with the 30s actually marginally better than the 20s. The deterioration sets in after 40, with a meaningful jump in the 50s.

The 50s sample is smaller (n=229) and introduces some noise, but the direction is consistent with the international literature. Older adults experience more fragmented sleep, earlier waking, and reduced slow-wave duration — biology that compounds the social pressures of late-career responsibility.

Women sleep eight points less than men#

Eight percentage points isn't a rounding artefact. The gap replicates across age bands, cities, and industries — which is what you'd expect from a finding rooted partly in domestic-load asymmetry and partly in hormonal architecture, neither of which is going away just because we count the data more carefully.

The international sleep literature documents two contributing mechanisms. The first is social: in dual-earner households where housework and care are shared unequally, the woman's sleep window compresses from both ends — she goes to bed later because the last household task is hers and wakes earlier because the first one is hers also. The second is biological: progesterone fluctuations across the menstrual cycle alter sleep-onset latency and REM density; perimenopause adds thermoregulatory disturbance (Joffe et al., Menopause, 2010). Neither mechanism alone explains an eight-point gap. Both contribute.

Mumbai again sleeps least#

Mumbai was also the worst-sleeping major Indian city in WHI 2025, on a smaller sample. The composition is well-documented and unsurprising: long daily commutes, dense housing, late-evening eating, hot summer nights — and, as the Mumbai page sets out, the city that sleeps worst also carries the metros' worst female anemia and glucose dysfunction. Pune at the other end sits ten points below — smaller scale, easier commute, cooler climate. Across two editions of the report and very different sample sizes, the metropolitan ordering on sleep is stable.

IT/SaaS sleeps best; EdTech sleeps worst#

Among industries with at least 200 respondents:

The IT/SaaS reading is the one worth pausing on. The industry most associated in public conversation with poor sleep is, in our sample, the best-sleeping major industry by a clear margin. Two contributing factors fit: the IT cohort skews younger (and the 20s/30s are the best-sleeping age bands), and the sector reports the highest rates of hybrid and work-from-home arrangements among the major industries — both of which protect evening time and morning recovery.

EdTech / Education at the other end — 38.8%, eight points above IT/SaaS — fits the inverse pattern. Early-morning teaching schedules, longer commutes, and a workforce that skews into the older sleep-deteriorating bands stack up.

One in four wakes up at night#

Of the 7,561 respondents who answered the sleep-disorders question:

One in four wakes up at night. One in five takes longer than they should to fall asleep. One in five can't get out of bed without effort. The clinical category for this pattern is sub-syndromal insomnia — not severe enough to bring the patient into a clinic, severe enough to shape the next working day, and the day after that, and the year of working life of which those days are the building blocks.

3.6% are using alcohol as a sleep aid#

Of the 3,304 respondents who answered the sleep-aids question:

The 3.6% who report reaching for alcohol to fall asleep is small in share, large in clinical implication. In a population the size of urban professional India, that share translates to several hundred thousand working adults reaching for a drink each night to switch off. The Alcohol page resolves the behaviour by drinking intensity — among weekly drinkers it runs to 22.3%, the heaviest tier feeding a self-defeating loop.

High stress is the cleanest predictor of short sleep#

Within-survey associations. Not causal, but consistent.

The stress association is the strongest by some margin — a respondent reporting high stress is 1.73× more likely than a respondent who isn't to be sleeping under six hours. The Stress page reads the same loop from its own end, where short sleep is the cleanest readout of high stress.

The WFO finding cuts against the lazy prior. The intuition is that the commute is the sleep thief and removing it gives the worker the sleep back. In our sample, WFO and WFH come out within three points of each other; the hybrid arrangement (two or three days in the office) does best. The likely explanation isn't that the office is restful but that the absence of an external structure to the day, in pure WFH, gives the late-night creep of work and screen time room to grow. Hybrid preserves enough rhythm to protect the sleep.

A note on the I'm a six-hours person claim#

The phrase is common in Indian offices, dinners, and profile interviews. It is presented as a fact about personality.

The clinical evidence is unambiguous. A single short-sleep gene (DEC2, identified in a Californian family in 2009) accounts for roughly one carrier in a million. The proportion of adults who genuinely function well on under six hours, by the standard biomarkers — error rate, reaction time, memory consolidation, mood regulation, glucose control — is a fraction of 1%. Statistical near-zero.

The rest of the "six-hours people" are adapted to the experience of six hours. They've stopped noticing the deficit. The biomarkers, when measured, still notice it.