The authors analyzed 7-day accelerometer data from 60,977 UK Biobank participants and found that the least regular sleepers had a 53% higher all-cause mortality risk, 57% higher cancer mortality, and 39% higher cardiometabolic mortality than the most regular sleepers. In head-to-head model comparisons, the Sleep Regularity Index outperformed sleep duration on nearly every mortality outcome, with effect sizes roughly double those of short or long sleep.
Surfaced the study on Hacker News where it drew 684 points and 360 comments, signaling that the developer community sees the finding as a meaningful reframing of conventional sleep advice. The submission's traction reflects interest in a metric — timing consistency — that is actionable and often overlooked relative to the '7-9 hours' rule.
The editorial argues that decades of sleep-mortality research relied on people accurately recalling bedtimes months later — an unreliable input — while this study used wrist-worn accelerometers that objectively capture timing. Isolating regularity as a variable was literally impossible in duration-only research, which is why the effect went undetected for so long.
The editorial notes that irregular sleep timing disrupts cortisol rhythms, insulin sensitivity, immune function, and DNA-repair cycles — mechanisms already implicated in shift-worker cancer risk. This mechanistic grounding strengthens the case that the observed mortality association reflects a genuine causal pathway rather than confounding.
A 2023 paper in the journal *Sleep* by Windred, Fripp, Vidafar, et al., titled *Sleep regularity is a stronger predictor of mortality risk than sleep duration*, has resurfaced on Hacker News with 684 points. The researchers pulled 7-day accelerometer data from 60,977 UK Biobank participants, computed each person's Sleep Regularity Index (SRI) — a score from 0 (random) to 100 (identical sleep-wake pattern every day) — and followed them for a median of 7.8 years, during which 1,859 deaths occurred.
The headline result: participants in the bottom fifth of sleep regularity had a 53% higher risk of all-cause mortality than the most regular sleepers, a 57% higher cancer mortality risk, and a 39% higher cardiometabolic mortality risk. The relationship was largely linear — more regularity, less risk — and it held after adjusting for age, sex, BMI, physical activity, smoking, alcohol, socioeconomic status, and, critically, average sleep duration.
When the authors ran a head-to-head model comparison, SRI outperformed sleep duration on nearly every mortality outcome. The effect size for irregularity was roughly double that of short or long sleep. This is not a subtle finding buried in a subgroup analysis. It reframes decades of sleep-hygiene advice that treated "7-9 hours" as the load-bearing metric.
Most of what practitioners know about sleep and health comes from self-reported duration studies — the classic U-shaped curve where both short and long sleepers die younger. Those studies rely on people accurately remembering when they went to bed six months ago, which is a stretch. This study used wrist-worn accelerometers, which don't lie, and it isolated a variable — timing consistency — that duration-only research literally could not see.
The biology is plausible. Circadian misalignment disrupts cortisol, insulin sensitivity, immune function, and DNA-repair cycles. Shift workers have known elevated cancer rates; the WHO classified night-shift work as a probable carcinogen in 2007. What's new is that you don't need to be a nurse working rotating 12-hour shifts to get the damage — a weekend social-jetlag pattern (staying up till 2am Friday and Saturday, then "catching up" Sunday) produces measurable SRI penalties.
Compare this to the standard advice ecosystem. Whoop, Oura, Eight Sleep, and Apple Health all surface sleep duration and stages prominently. Regularity metrics exist but are usually buried behind "consistency" scores that don't map cleanly to SRI. Community reaction on HN was blunt: multiple top comments pointed out that the entire wearables industry has been optimizing the wrong number for a decade. One commenter, a former ICU nurse, noted that hospital shift schedules — which the industry treats as an immutable operational constraint — may be quietly generating the mortality signal the paper describes.
The skeptic case deserves airtime. UK Biobank participants skew healthier and wealthier than the general population. Seven days of accelerometry is a snapshot, not a life history. And observational studies can't prove causation — it's plausible that irregular sleep is a marker of underlying illness or chaotic life circumstances that independently drive mortality. The authors adjusted for known confounders, but reverse causation remains the standard critique. Still, the effect size is large, the mechanism is biologically coherent, and the finding replicates smaller studies going back to the mid-2010s.
For developers — a population that skews toward late-night deep-work sessions, weekend hackathons, and on-call rotations — the operational read is uncomfortable. If you're sleeping seven hours a night but your bedtime varies by three or more hours across the week, the data says you're taking a bigger mortality hit than someone who consistently sleeps six. The industry's obsession with duration ('I got my 8') is measuring the wrong axis.
Practical implications, in rough order of leverage:
- On-call rotations are worse than they look. The standard defense of pager duty — "you can sleep in the next day" — assumes duration is what matters. If regularity dominates, a single 3am page followed by a compensatory late morning may cost more than a shorter but on-schedule night. Teams designing rotations should minimize schedule disruption, not just total sleep hours lost. - Weekend social jetlag is not free. The Friday/Saturday late nights that feel like recovery from a hard week are, per this data, a measurable health cost. The 90-minute rule of thumb from chronobiology (keep weekend wake times within 90 minutes of weekday) has empirical backing here. - Wearable dashboards need retuning. If your Oura or Whoop shows a green "sleep score" because you got 8 hours but your bedtime SD is 90 minutes, the score is lying to you by omission. Check whether your device exposes a regularity or consistency metric and weight it accordingly. - Async-first teams have a hidden benefit. Distributed teams that don't require synchronous 9am standups let people lock in a personal schedule and hold it. That's not just a lifestyle perk — it may be a longevity perk.
None of this requires new tooling. It requires treating bedtime as a hard constraint, the way you'd treat a deploy freeze. That's culturally harder than it sounds in an industry that romanticizes the 2am commit.
Expect the wearables industry to start marketing SRI-style metrics prominently within the next 12-18 months — Oura already has a rudimentary version, and the competitive pressure to differentiate on "health outcomes" rather than "data points collected" will force the others to follow. The bigger open question is whether interventions to improve regularity actually reduce mortality, which requires RCTs that haven't been run yet. Until then, the honest read of the evidence is: pick a bedtime, hold the line, and stop congratulating yourself for the 10-hour Sunday.
Top 10 dev stories every morning at 8am UTC. AI-curated. Retro terminal HTML email.