If you spent a year completely alone, you would almost certainly survive, but three things would drift. Your body clock would free-run on its own period of about 24.2 hours, your sense of how much time has passed would shrink, sometimes by about 40 percent (Siffre, 1962), and your sleep and activity would flatten out, as they did even for the six-person Mars-500 crew. Whether it ends as a retreat or as harm depends mostly on whether you chose it, how much daylight and structure you keep, and whether anyone knows you are there.
- 24.18 h: the average intrinsic period of the human body clock (Czeisler et al., 1999). Without light cues, a year drifts you about 3 to 17 days.
- Caver Michel Siffre spent about two months underground in 1962 and thought 25 days fewer had passed.
- Social isolation raises the odds of earlier death by 29%, loneliness by 26%, in a meta-analysis of studies from 1980 to 2014.
- About 5% of people on polar expeditions meet criteria for a psychiatric disorder; many others report growth.
- Robin Knox-Johnston sailed alone, non-stop, for 312 days in 1968–69 and was the only one of nine starters to finish.
"Completely alone" here means a full year with no conversation, no visitors and no messages, but with food, shelter and safety handled. That isolates the question to the mind. The evidence comes from four kinds of people who have done versions of it: cave volunteers, polar winter crews, solo sailors and simulated Mars crews.
What happens to your sense of time when you are alone?
It compresses. With no one to sync with and no daylight, the brain counts sleep-wake cycles as days, and those cycles stretch. The best-documented case is a French caver who did it deliberately.
On 16 July 1962, a 23-year-old French geologist named Michel Siffre went down into the Scarasson abyss in the Maritime Alps, about 130 metres underground, with no watch and no daylight. A team on the surface logged every time he called up to say he had woken, eaten or gone to sleep, but never told him the time. When they finally called on 14 September to say the experiment was over, Siffre believed it was 20 August. He thought he had about a month still to go. In a separate test he counted to 120 at what felt like one per second; it took him about five minutes. Ten years later he stayed 205 days in Midnight Cave, Texas, and at times fell into cycles of about 36 hours awake and 12 asleep without noticing anything unusual.
Cabinet Magazine: Caveman, an interview with Michel Siffre by Joshua Foer (2008)
Solving the drift
The governing idea is the free-running circadian rhythm: without cues called zeitgebers (light, meals, other people), the internal clock runs at its own period τ instead of exactly 24 hours.
- Best lab value. Czeisler's group measured τ = 24.18 h under dim, controlled light. Then \( N_{subj} = 8760 / 24.18 = 362.3 \), so you lose about 2.7 days in a year.
Calculated in this post from the intrinsic period of 24.18 h (Czeisler et al., 1999) and 365 days of 24 hours.
- Older free-run studies. Studies in which people chose their own light reported activity periods with a median of 25.2 h. Then \( N_{subj} = 8760 / 25.2 = 347.6 \): 17.4 days lost.
- Siffre 1962. Sixty calendar days (16 July to 14 September) felt like 35 (16 July to 20 August). If each of his "days" was one sleep-wake cycle, his average cycle was \( 60 \times 24 / 35 \approx 41 \) h. That is my inference from his dates, not a measured value.
- Scale Siffre's ratio to a year. \( 365 \times 35/60 \approx 213 \) subjective days: you would surface believing it was about August.
Sensitivity: the answer swings from 3 days to about 150 days depending on one variable, whether you control your own light. A lamp on a fixed timer collapses the error back toward the lab value. The model breaks down for long cycles like Siffre's, which mix the clock with very long wake periods the clock alone does not explain.
Calculated in this post from the 24.18 h intrinsic period and the 25.2 h median activity period (Czeisler et al., 1999) and from Siffre's 1962 dates (Cabinet Magazine interview with Siffre, 2008).
Earlier studies had reported human free-running periods anywhere from 13 to 65 hours, with activity rhythms at a median of 25.2 hours, but those subjects controlled their own lighting, which shifts the clock. Under carefully controlled dim light, melatonin, core body temperature and cortisol rhythms showed an intrinsic period averaging 24.18 hours in both young and older adults, with a tight spread. Limit: a lab protocol, not real isolation, so it gives the floor of the drift, not the typical case.
Read the abstractWhat did a 520-day isolation do to sleep?
It made people sleep and rest more and move less. The Mars-500 simulation in Moscow sealed six men in a mock spacecraft for 520 days from 2010 to 2011. They were not alone, which makes their results a conservative guide to solo isolation.
Wrist actigraphy and light sensors recorded 4.396 million minutes of data from the six-man crew, plus 888 weekly computer-based tests. Over the mission the crew became more sedentary (hypokinesis), slept and rested more, and were exposed to less light. Most crew members had at least one disturbance of sleep quality, vigilance lapses or sleep-wake timing, which pointed to weak circadian entrainment. Limit: six men, one simulation, and a crew rather than a single person.
Read the paper (free)The lesson for a solo year is direct: light, meals and exercise at fixed times are the replacement for other people. The Mars-500 authors concluded that future crews will need habitats that recreate Earth's daily signals.
Is a year alone bad for your health?
Long-term isolation is clearly linked to earlier death; one year of it is a smaller, less-studied dose. In 2015 Julianne Holt-Lunstad's team pooled studies from 1980 to 2014 and compared three related exposures.
See the data as a table
| Odds ratio (odds ratio) | |
|---|---|
| Living alone | 1.32 |
| Social isolation | 1.29 |
| Loneliness | 1.26 |
Source: Holt-Lunstad et al., Perspectives on Psychological Science 2015 · Weighted averages from studies that controlled for confounds. 1.0 means no added risk.
A meta-analysis of studies published from January 1980 to February 2014 with mortality outcomes. After controlling for confounds, social isolation carried an odds ratio of 1.29, loneliness 1.26 and living alone 1.32, meaning 26 to 32% higher odds of death over follow-up. Objective isolation and the feeling of loneliness mattered about equally, and the effect was larger in samples younger than 65. Limit: observational data over years, so causation and the effect of a single year cannot be read off directly.
Read the abstractTurning an odds ratio into a one-year risk
Odds ratios are not risk multipliers, so convert them properly. If p₀ is your baseline chance of dying in a year, the isolated risk p₁ is:
- Pick an illustrative baseline of 1 in 500 a year, \( p_0 = 0.002 \). This is a round number for a young adult, not a statistic.
- Apply OR = 1.29. \( p_1 = 0.00258 / (0.998 + 0.00258) = 0.00258 \).
- Excess risk is about 0.06 percentage points, roughly 1 extra death per 1,700 such people.
Calculated in this post from the odds ratio of 1.29 (Holt-Lunstad et al., 2015) and the example baseline of 1 in 500 above, with these example inputs.
- Sensitivity. At a baseline of 2% a year (much older), the excess is about 0.6 points: the same odds ratio matters ten times more.
Calculated in this post from the odds ratio of 1.29 (Holt-Lunstad et al., 2015) and an example baseline of 2% a year, with these example inputs.
The honest caveat: those odds ratios come from people isolated for years, often with illness, poverty or depression in the mix. A healthy volunteer choosing a year alone is not the same exposure, so treat the numbers as an upper-end guide.
Your one-year risk if isolation acts like the meta-analysis says
What do polar crews and solo sailors show?
They show that isolation produces a predictable set of symptoms, and that most people come through them. A 2008 Lancet review by Lawrence Palinkas and Peter Suedfeld summarised decades of polar expeditions.
- Common symptoms: disturbed sleep, slower thinking, low mood and irritability.
- Named syndromes: the "winter-over syndrome", the "polar T3 syndrome" (a thyroid-linked pattern) and mild seasonal affective disorder.
- Clinical cases: about 5% of expedition members meet formal criteria for a psychiatric disorder.
- The other side: many report salutogenic effects: more self-reliance, better health and personal growth.

| Who | When | How long | Alone? | What drifted |
|---|---|---|---|---|
| Michel Siffre, Scarasson abyss | 1962 | about 2 months | Yes | Time sense lost 25 days |
| Michel Siffre, Midnight Cave | 1972 | 205 days | Yes | 36 h awake / 12 h asleep cycles at times |
| Robin Knox-Johnston, Suhaili | 1968–69 | 312 days | Yes, at sea | Only finisher of nine starters |
| Mars-500 crew | 2010–11 | 520 days | No, six men | Less movement, more sleep, weak entrainment |
| Polar winter-over crews | ongoing | a polar winter | No, small teams | Sleep, mood, about 5% clinical |
The 1968 Golden Globe Race is the starkest contrast. Knox-Johnston finished on 22 April 1969 in a 32-foot boat. Bernard Moitessier chose to abandon the race and kept sailing to Tahiti. Donald Crowhurst faked his position reports and died at sea. Same isolation, three very different outcomes.

Where does solitude turn from useful to harmful?
Solitude helps when it is chosen, bounded and quiet; it harms when it is imposed, open-ended and unpredictable. The difference shows up in the research as the gap between being alone and feeling lonely, which Holt-Lunstad found carry similar risks once they persist.
There is also a simple arithmetic loss. In a 2007 Science study, Matthias Mehl's team recorded 396 people and estimated that both men and women speak about 16,000 words a day. A year alone removes roughly \( 16{,}000 \times 365 \approx 5.8 \) million spoken words, and with them the constant small corrections other people give to your memory and mood.
Shorter, deliberate solitude is a different and more promising thing; the evidence is in what silence and solitude do to thinking, and a structured version of it is the 10-day Vipassana course. If isolation is not chosen, finding affordable therapy in India is a practical next step.
A year completely alone is survivable: Knox-Johnston managed 312 days at sea and Siffre 205 days in a cave. Without changes, expect your clock to free-run (losing 3 to 17 days of count, or far more if your cycles stretch like Siffre's), activity to fall and mood to sag, with isolation's long-term odds ratio of about 1.29 for earlier death as the background risk. With a timed light, fixed routines, exercise, a purpose and a check-in, the year looks more like a long retreat than a sentence.
What's new
What if next?
- What if you could upload your mind? The loneliest version of solitude: a copy of you running in a server with nobody else in it.
- What if humans didn't need sleep? Alone in a cave, your sleep drifted. What if there were none to drift?
- What would the Gita say to an AI? The Gita's forest-dwellers chose solitude; its advice on acting without attachment has a surprising machine version.
Questions people ask
Can a person go crazy from being alone for too long?
Prolonged, forced isolation is linked to anxiety, depression, disturbed sleep and in severe cases perceptual distortions. Chosen, structured solitude, as in Siffre's caves or solo sailing, is usually tolerated. About 5% of polar expedition members meet criteria for a psychiatric disorder.
What happens to your brain when you are isolated?
The clearest measured changes are in sleep timing, activity and vigilance, as in the Mars-500 crew. Without light and social cues, the body clock free-runs at its own period of about 24.2 hours, and the sense of elapsed time can shrink sharply.
Is being alone the same as being lonely?
No. Being alone is a situation; loneliness is the feeling of lacking connection. In the 2015 meta-analysis both carried similar mortality risk once they persisted, which is why chosen, short solitude and chronic loneliness are treated differently.
How long did Michel Siffre stay in the cave?
About two months in 1962 in the Scarasson abyss in the Alps, and 205 days in Midnight Cave, Texas, in 1972.
The opposite problem is the one most of India has, never getting an hour alone: What if you were never alone for a year?
Sources
- Foer J.: Caveman, an interview with Michel Siffre, Cabinet Magazine (2008) (source for the 1962 and 1972 cave details)
- Wikipedia: Michel Siffre (accessed 2026) (source for the cave name and depth)
- Czeisler C.A. et al.: Stability, precision, and near-24-hour period of the human circadian pacemaker, Science (1999) (source for τ = 24.18 h and the 25.2 h median)
- Basner M. et al.: Mars 520-d mission simulation reveals protracted crew hypokinesis, PNAS (2013)
- Holt-Lunstad J. et al.: Loneliness and social isolation as risk factors for mortality, Perspectives on Psychological Science (2015) (source for the odds ratios)
- Holt-Lunstad J., Smith T.B., Layton J.B.: Social relationships and mortality risk, PLoS Medicine (2010) (background: 148 studies, 308,849 people)
- Palinkas L.A., Suedfeld P.: Psychological effects of polar expeditions, The Lancet (2008) (source for the 5% figure and syndromes)
- Mehl M.R. et al.: Are women really more talkative than men?, Science (2007) (source for 16,000 words a day)
- Wikipedia: Robin Knox-Johnston (accessed 2026) (source for the race dates, Suhaili and the other entrants)
- WHO: Social connection linked to improved health and reduced risk of early death (2025)
- Lokdam N.T. et al.: Mental health disorders and risk of solitary confinement, BMJ Mental Health (2026)
- Ministry of Health and Family Welfare: Tele MANAS
More on the mind on the Manthan world page.

