One Week of Five-Hour Nights Cut Daytime Testosterone by 10 to 15 Percent in Ten Healthy Young Men, a Drop That Normal Aging Takes 10 to 15 Years to Produce. Here Is What Every Human Sleep and Testosterone Study Actually Found.
Every herb and mineral this series has examined shares one weakness: the best trials are small, short, often funded by the seller, and positive mainly in men who started deficient. Sleep is different in one respect and identical in another. It is different because nobody profits from telling men to sleep more, so the controlled experiments that exist were run by academic endocrinologists asking a physiological question rather than by a supplement maker asking a marketing one. It is identical because those experiments are still small, still short, and still mostly in men in their twenties. The headline result is real and reproducible in its direction: restrict a young man's sleep to four or five hours and his testosterone falls, by roughly a tenth to a seventh over the waking day in the best-designed study, with the loss concentrated in the afternoon and evening rather than in the early-morning sample most clinics draw. What happens to a 45-year-old who has slept six hours a night for a decade is a different question, and the answer rests on cross-sectional data in which short sleep travels with obesity, sleep apnea, shift work and alcohol, each of which lowers testosterone on its own. This article goes through the laboratory studies one at a time, then the population data, then the confounders, and finishes with what the evidence does and does not justify for a man over 40 trying to protect the hormone.
Key Takeaways: What the Sleep and Testosterone Evidence Actually Shows
One week of sleep restricted to five hours a night lowered daytime testosterone by 10 to 15 percent in ten healthy men with a mean age of 24, in the 2011 JAMA study by Rachel Leproult and Eve Van Cauter, and the fall was largest between 2 pm and 10 pm. Testosterone rises during sleep rather than at a fixed clock time: in the 2005 Axelsson study, seven young men who slept during the day still showed the rise, and in the 2001 Luboshitzky study, fragmenting sleep so that men did not reach REM blunted it. Partial restriction of two to five nights has produced inconsistent effects on the single morning sample: Reynolds 2012 found afternoon testosterone fell after five nights of four hours in 14 men while morning levels did not change significantly, and Schmid 2012 found no change in morning testosterone after two nights of 4.5 hours in 15 men. One full night without sleep lowers testosterone the next day, and the 2021 meta-analysis by Su and colleagues in Sleep Medicine found the pooled effect of total sleep deprivation was significant while the pooled effect of partial restriction was not. In 12 healthy men aged 64 to 74 in the 2007 Penev study, total sleep time measured by actigraphy independently predicted morning total and free testosterone, and men at the four-hour end had levels roughly half those of men sleeping eight hours. In 2,295 US men in NHANES 2011 to 2012, each hour of sleep lost was associated with about 5.85 ng/dL lower total testosterone after adjustment for age and body mass index, a small effect next to the effect of obesity itself. No randomized trial has tested whether extending sleep for months raises testosterone in men who habitually sleep short, so the long-term claim rests on physiology and observational data, not on an intervention study.
Six Studies, 64 Men, and Not One of Them Middle-Aged Except the Observational One
Men enrolled in each controlled human study of sleep and testosterone between 2001 and 2012. Violet marks the two timing and fragmentation experiments that established the mechanism: Luboshitzky 2001 fragmented the sleep of six young men and found the nocturnal testosterone rise disappeared in those who never reached REM, and Axelsson 2005 had seven men sleep by day or by night and found the rise followed sleep rather than the clock. Red marks Leproult and Van Cauter 2011, the eight-night, five-hour protocol in ten men aged about 24 with 24-hour blood sampling, in which daytime testosterone fell 10 to 15 percent. Amber marks the two partial-restriction studies with mixed results on the morning sample: Reynolds 2012, where five nights of four hours lowered afternoon but not morning testosterone in 14 men, and Schmid 2012, where two nights of 4.5 hours did not change morning testosterone in 15 men. Blue marks Penev 2007, the only study in men of the age that usually asks the question, 12 men aged 64 to 74 whose actigraphy-measured sleep predicted morning testosterone. Every experimental subject was young and lean; the effect in older, heavier men is inferred, not measured.
Why does sleep matter for testosterone at all?
Testosterone matters for sleep because most of the daily rise in the hormone happens while a man is asleep, and the rise is triggered by sleep itself rather than by the time of night. In young men serum testosterone is roughly 20 to 30 percent higher in the early morning than in the late afternoon, which is why the Endocrine Society tells clinicians to draw the diagnostic sample before 10 am. The overnight climb begins within an hour or so of falling asleep, tracks the first three hours of consolidated sleep and the first REM period, and depends on a rise in luteinizing hormone pulses from the pituitary during the night. Two small experiments established that the driver is sleep and not the clock. Axelsson and colleagues in the Journal of Clinical Endocrinology and Metabolism in 2005 had seven healthy young men sleep either at night or during the day and sampled blood around the clock: testosterone rose during sleep whenever sleep happened, and stayed low during wakefulness at night. Luboshitzky and colleagues in the same journal in 2001 fragmented the sleep of six healthy young men with repeated awakenings and found that the men who never reached REM sleep did not show the normal nocturnal rise, while those who did reach REM despite the fragmentation kept it. The nocturnal rise is also part of what disappears with age. A second Luboshitzky study in 2003 compared eight middle-aged men with eight young men and found the overnight increase in testosterone was blunted in the older group, and Bremner and colleagues had shown in 1983 that the diurnal rhythm itself flattens in men over 60. The mechanism is not fully mapped, but the picture is that sleep opens a window during which the hypothalamus and pituitary drive the testis harder, and that shortening or shattering the window shrinks the output.
Six Studies, 64 Men, and Not One of Them Middle-Aged Except the Observational One
Men enrolled in each controlled human study of sleep and testosterone between 2001 and 2012. Violet marks the two timing and fragmentation experiments that established the mechanism: Luboshitzky 2001 fragmented the sleep of six young men and found the nocturnal testosterone rise disappeared in those who never reached REM, and Axelsson 2005 had seven men sleep by day or by night and found the rise followed sleep rather than the clock. Red marks Leproult and Van Cauter 2011, the eight-night, five-hour protocol in ten men aged about 24 with 24-hour blood sampling, in which daytime testosterone fell 10 to 15 percent. Amber marks the two partial-restriction studies with mixed results on the morning sample: Reynolds 2012, where five nights of four hours lowered afternoon but not morning testosterone in 14 men, and Schmid 2012, where two nights of 4.5 hours did not change morning testosterone in 15 men. Blue marks Penev 2007, the only study in men of the age that usually asks the question, 12 men aged 64 to 74 whose actigraphy-measured sleep predicted morning testosterone. Every experimental subject was young and lean; the effect in older, heavier men is inferred, not measured.
What happened when healthy young men slept five hours a night for a week?
One week of five-hour nights lowered daytime testosterone by 10 to 15 percent in the Leproult and Van Cauter study published in JAMA in June 2011, the single best-controlled experiment on sleep and testosterone in men. Ten healthy men with a mean age of 24.3 years and a mean body mass index of 23.5 lived in the University of Chicago sleep laboratory for the whole protocol. They first spent three nights with ten hours in bed to clear any pre-existing sleep debt, then eight nights with only five hours in bed, from 12:30 am to 5:30 am, with their sleep recorded by polysomnography. On the last day of each condition blood was drawn every 15 to 30 minutes for 24 hours, so the investigators had a complete hormone profile in the rested state and a complete profile after a week of restriction, in the same men, under the same lighting, meals and activity. Daytime testosterone levels were 10 to 15 percent lower after the restricted week, with the biggest gap between 2 pm and 10 pm, and the men's self-rated vigor on a standard mood inventory fell steadily over the eight nights in step with the hormone. Twenty-four-hour cortisol did not change, which matters because it rules out the simplest alternative explanation, that stress hormones were suppressing the testis. The authors made the comparison that has been quoted ever since: since testosterone in men declines by about 1 to 2 percent a year from the thirties onward, a 10 to 15 percent drop is the amount of loss normal aging takes 10 to 15 years to produce, and it appeared in one week. The study's limits are the usual ones. Ten men is a small sample, all were young and lean, and the protocol lasted eight nights, so it says nothing directly about a middle-aged man who has slept six hours a night for years. What it does show, cleanly, is that the connection is causal and the effect size is not trivial.
One Week of Five-Hour Nights Did What Aging Takes a Decade to Do
Percent change in daytime serum testosterone after eight nights of five hours in bed in Leproult and Van Cauter, JAMA 2011, shown as the midpoint of the reported 10 to 15 percent fall with the range as error bars, next to the pace of normal age-related decline. Ten healthy men with a mean age of 24.3 and a mean body mass index of 23.5 spent three ten-hour nights in the laboratory to reach a rested baseline, then eight nights with time in bed restricted to 12:30 am to 5:30 am, with blood drawn every 15 to 30 minutes for 24 hours at the end of each condition. Daytime testosterone was 10 to 15 percent lower after restriction, with the deepest trough between 2 pm and 10 pm, and 24-hour cortisol did not change. Testosterone in men declines by roughly 1 to 2 percent a year from the thirties onward, so the one-week loss matched the loss of 10 to 15 years of aging, which is the comparison the authors drew. The aging bars are population averages, not data from the same men, and the study was eight nights in young lean subjects, so it demonstrates causation and effect size in the short term rather than what chronic six-hour sleep does to a 50-year-old.
Does the loss show up in the morning blood test, or only later in the day?
Short-term sleep restriction lowers testosterone most clearly in the afternoon and evening, and a single early-morning sample can miss it entirely. Reynolds and colleagues published the most direct test in PLoS ONE in 2012: 14 healthy men with a mean age of about 27 slept four hours a night for five nights in a laboratory, and blood was drawn across the day at baseline and after restriction. Afternoon testosterone fell after the five restricted nights, cortisol in the afternoon rose, and glucose handling worsened, but morning testosterone did not change significantly. Schmid and colleagues in Clinical Endocrinology in 2012 restricted 15 healthy men to 4.5 hours for two nights, in either the first half or the second half of the night, and found that two nights was not enough to lower the morning testosterone sample regardless of timing, though the 24-hour profile shifted with when the men slept. The 2021 systematic review and meta-analysis by Su and colleagues in Sleep Medicine pooled the available experiments and reached the conclusion the individual studies point to: total sleep deprivation, meaning a full night or more without sleep, produced a significant fall in serum testosterone, while partial sleep restriction produced a pooled effect that did not reach statistical significance, largely because most partial-restriction studies relied on morning samples taken after a few hours of sleep that had already triggered part of the nocturnal rise. This is the detail most articles on the topic skip. The morning testosterone reading is the one clinics use and the one men compare against reference ranges, and it is the reading least sensitive to a moderately short night. The Leproult result was so clear because the investigators sampled for 24 hours; a man whose only data point is an 8 am draw after a six-hour night may read normal and still be spending his afternoons at a level 10 percent below where his sleep-replete self would sit.
Zero Hours Always Lowered Testosterone, Five Hours for a Week Did, and Two Nights of 4.5 Hours Did Not Move the Morning Sample
Hours of sleep allowed per night in each experimental protocol, with the number of nights in the label and the bar coloured by what happened to testosterone. Red marks protocols in which testosterone fell across the day: the one-night total deprivation studies by Cote 2013 and Arnal 2016, and the eight-night five-hour protocol of Leproult 2011 in which daytime testosterone fell 10 to 15 percent. Amber marks Reynolds 2012, five nights of four hours in 14 men, where afternoon testosterone fell but the morning sample did not change significantly. Grey marks Schmid 2012, two nights of 4.5 hours in 15 men, where morning testosterone did not change whichever half of the night the men slept in. Green marks the three ten-hour nights Leproult used to bring men to a rested baseline before restricting them. The pattern across the protocols is that the effect depends on how much sleep is lost and for how many nights, and on when in the day the blood is drawn: a morning sample after a few hours of sleep has already captured part of the nocturnal rise, which is why the 2021 Su meta-analysis found a significant pooled effect for total deprivation and a non-significant one for partial restriction.
- Testosterone fell across the day
- Afternoon fell, morning unchanged
- Morning sample unchanged
- Rested baseline
What does one night without any sleep do?
One full night without sleep lowers testosterone the following day, and this is the most consistent finding in the experimental literature. Cortes-Gallegos and colleagues reported in the early 1980s that 48 hours of sleep deprivation reduced circulating androgens in healthy men, and Gonzalez-Santos and colleagues in 1989 found the same after 48 hours awake. Cote and colleagues in the journal Sleep in 2013 kept healthy young men awake for one night and measured salivary testosterone alongside a laboratory aggression task; testosterone fell after the sleepless night and reactive aggression fell with it, an observation that runs against the popular idea that tired men are more hostile. Arnal and colleagues in the Journal of Clinical Endocrinology and Metabolism in 2016 asked whether banking sleep in advance protects the hormone. Healthy young men spent a week either at their habitual sleep length or with extended time in bed, then stayed awake for one full night, then had a recovery night. Testosterone fell after the sleepless night in both conditions, the prior week of extended sleep did not prevent the fall, and levels moved back toward baseline after recovery sleep. The Su 2021 meta-analysis found this class of study, total sleep deprivation, was the one with a statistically robust pooled effect. The practical reading is that the acute response is fast in both directions: a lost night costs testosterone the next day, and a recovered night gives most of it back. Whether repeated cycles of loss and partial recovery, which is what most shift workers and new fathers actually experience, leave a residual deficit is not something a two-night laboratory protocol can answer.
Zero Hours Always Lowered Testosterone, Five Hours for a Week Did, and Two Nights of 4.5 Hours Did Not Move the Morning Sample
Hours of sleep allowed per night in each experimental protocol, with the number of nights in the label and the bar coloured by what happened to testosterone. Red marks protocols in which testosterone fell across the day: the one-night total deprivation studies by Cote 2013 and Arnal 2016, and the eight-night five-hour protocol of Leproult 2011 in which daytime testosterone fell 10 to 15 percent. Amber marks Reynolds 2012, five nights of four hours in 14 men, where afternoon testosterone fell but the morning sample did not change significantly. Grey marks Schmid 2012, two nights of 4.5 hours in 15 men, where morning testosterone did not change whichever half of the night the men slept in. Green marks the three ten-hour nights Leproult used to bring men to a rested baseline before restricting them. The pattern across the protocols is that the effect depends on how much sleep is lost and for how many nights, and on when in the day the blood is drawn: a morning sample after a few hours of sleep has already captured part of the nocturnal rise, which is why the 2021 Su meta-analysis found a significant pooled effect for total deprivation and a non-significant one for partial restriction.
- Testosterone fell across the day
- Afternoon fell, morning unchanged
- Morning sample unchanged
- Rested baseline
How much testosterone do short sleepers lose in real-world data?
In observational data, men who sleep less have lower testosterone, and the size of the gap ranges from dramatic in one small study of older men to modest in the largest national survey. Plamen Penev published the small study in the journal Sleep in 2007: 12 healthy men aged 64 to 74 wore wrist actigraphs at home for about a week, spent a night under polysomnography, and had morning blood drawn. Total sleep time was an independent predictor of both morning total and free testosterone after accounting for age and body composition, and in Penev's own summary the men who slept around four hours had morning testosterone in the range of 200 to 300 ng/dL while the men who slept around eight hours were in the range of 500 to 700 ng/dL, roughly a two-to-one gradient across a group of otherwise similar retirees. Twelve men is too few to build a dose-response curve on, but the direction matched the laboratory experiments and the population was for once the right age. The largest analysis is Patel and colleagues in the World Journal of Urology in 2019, using the 2011 to 2012 cycle of the US National Health and Nutrition Examination Survey, in which 2,295 adult men had both a measured serum testosterone and self-reported sleep duration. After adjustment for age, body mass index and other health conditions, each hour of sleep lost was associated with a decrease of about 5.85 ng/dL in total testosterone. On that model a man sleeping four hours instead of eight would sit about 23 ng/dL lower, which is a real but small shift against a normal range that runs from roughly 300 to 1,000 ng/dL, and body mass index in the same model was a much stronger predictor than sleep. A Hong Kong cohort of 1,274 men aged 65 and over, reported by Auyeung and colleagues in 2015, found that the relationship was not a straight line: both disturbed sleep and very long sleep were associated with lower testosterone and with lower muscle mass and strength, with the highest hormone levels around a conventional eight hours. The honest summary of the population data is that sleep duration tracks testosterone in every dataset that has looked, that the effect per hour is modest once weight is accounted for, and that the small older-male study showing a two-fold gradient has never been repeated at scale.
In Twelve Men Aged 64 to 74, Eight-Hour Sleepers Had Roughly Twice the Morning Testosterone of Four-Hour Sleepers
Morning total testosterone in ng/dL by habitual sleep duration in Penev, Sleep 2007, plotted as the midpoint of the ranges the author reported, with the ranges as error bars. Twelve healthy men aged 64 to 74 wore wrist actigraphs at home for about a week, spent a night under polysomnography, and had morning blood drawn. Total sleep time was an independent predictor of both total and free testosterone after accounting for age and body composition. In the author's summary, men sleeping around four hours had morning testosterone of 200 to 300 ng/dL while men sleeping around eight hours had 500 to 700 ng/dL. The dashed line marks 300 ng/dL, the cut-off most guidelines use for biochemical hypogonadism, and the short sleepers sat at or below it. The study is a cross-section of 12 men, so it cannot show that making the short sleepers sleep longer would raise their hormone, and the gradient has not been reproduced at scale; the far larger NHANES analysis found a much smaller per-hour effect once body mass index was accounted for. Its value is that it is the one dataset in men of the age that usually asks the question, and it points the same way as the laboratory experiments in the young.
Does shift work and irregular sleep timing matter?
Shift work lowers testosterone mainly when it produces genuine sleep loss and misalignment, and men who work shifts but sleep well look much like men who do not work shifts. The physiology predicts this. Because the Axelsson experiment showed that testosterone rises during sleep whenever sleep occurs, a night worker who gets a full consolidated block of daytime sleep should still produce the nocturnal rise, just displaced. The problem is that most rotating and night shift workers do not get that block: daytime sleep is shorter, lighter and more fragmented, and the Luboshitzky fragmentation data suggest that fragmented sleep that fails to reach REM is precisely what blunts the rise. Balasubramanian and colleagues examined this in a men's health clinic population of 766 men and separated shift workers who met criteria for shift work sleep disorder, meaning insomnia or excessive sleepiness attributable to the schedule, from shift workers who did not. Men with shift work sleep disorder had lower testosterone and worse scores on a hypogonadal symptom questionnaire than both non-shift workers and shift workers without the disorder, while the shift workers who slept adequately did not differ meaningfully from day workers. The Schmid 2012 study adds a laboratory footnote: shifting 4.5 hours of sleep to the early or the late half of the night changed the shape of the 24-hour testosterone profile even when it did not change the morning value. For a man who works nights, the evidence points at the quantity and continuity of sleep rather than at the clock time as the variable that matters for the hormone, which is a more tractable problem than the schedule itself.
In 2,295 US Men, Each Hour of Sleep Lost Cost About 5.85 ng/dL, a Real but Modest Effect Once Weight Is Accounted For
Modelled decrease in total testosterone by hours of sleep lost from Patel and colleagues, World Journal of Urology 2019, using the 2011 to 2012 cycle of the US National Health and Nutrition Examination Survey. The analysis included 2,295 adult men with a measured serum testosterone and self-reported habitual sleep duration. After adjustment for age, body mass index and other health conditions, each hour of sleep lost was associated with a decrease of about 5.85 ng/dL in total testosterone, so the bars are that coefficient multiplied out: a man sleeping four hours rather than eight would be modelled at about 23 ng/dL lower. Body mass index was a considerably stronger predictor in the same model. Against a normal range of roughly 300 to 1,000 ng/dL, a 23 ng/dL shift is small, and it is far less than the 10 to 15 percent fall in the Leproult laboratory study or the two-fold gradient in Penev's older men. Part of the gap is that NHANES testosterone was drawn in the morning, the sample least sensitive to sleep loss, part is that self-reported sleep is imprecise, and part is that the large restriction effects were measured in young lean men under laboratory control. The population estimate is the conservative anchor; the laboratory estimate is the upper bound for a man whose sleep is genuinely cut to five hours.
How much of the association is really obesity and sleep apnea?
A large share of the observational link between short sleep and low testosterone runs through body weight and obstructive sleep apnea, both of which lower testosterone independently and both of which shorten and fragment sleep. In the NHANES analysis by Patel and colleagues, body mass index was the strongest predictor of testosterone in the model, and the sleep effect of about 5.85 ng/dL per hour was what remained after adjusting for it. Obstructive sleep apnea is the more insidious confounder because it is common, underdiagnosed and closely tied to central weight. Men with untreated sleep apnea have lower testosterone than weight-matched men without it, the nocturnal rise in testosterone is blunted in proportion to how fragmented the night is, and the cross-sectional studies of sleep duration rarely screened for it. Whether treating apnea with positive airway pressure alone raises testosterone is disputed; the trials are small and most found little change, which suggests that the weight driving the apnea is doing much of the hormonal damage rather than the apnea itself. Weight loss, by contrast, raises testosterone reliably in men with obesity, and the recent tirzepatide trials in obstructive sleep apnea, which this site has covered separately, reduced apnea severity substantially through weight loss. Alcohol sits in the same tangle: it fragments the second half of the night, suppresses REM, and lowers testosterone directly. The laboratory restriction studies matter precisely because they cut through all of this. Leproult's ten men were lean, young, screened for sleep disorders and drank nothing, and their testosterone still fell 10 to 15 percent when their sleep was cut. So the sleep effect is real on its own, but in the population at large the man who sleeps five hours is usually also the man who weighs more, snores, and drinks in the evening, and the hormone reflects all four.
In Twelve Men Aged 64 to 74, Eight-Hour Sleepers Had Roughly Twice the Morning Testosterone of Four-Hour Sleepers
Morning total testosterone in ng/dL by habitual sleep duration in Penev, Sleep 2007, plotted as the midpoint of the ranges the author reported, with the ranges as error bars. Twelve healthy men aged 64 to 74 wore wrist actigraphs at home for about a week, spent a night under polysomnography, and had morning blood drawn. Total sleep time was an independent predictor of both total and free testosterone after accounting for age and body composition. In the author's summary, men sleeping around four hours had morning testosterone of 200 to 300 ng/dL while men sleeping around eight hours had 500 to 700 ng/dL. The dashed line marks 300 ng/dL, the cut-off most guidelines use for biochemical hypogonadism, and the short sleepers sat at or below it. The study is a cross-section of 12 men, so it cannot show that making the short sleepers sleep longer would raise their hormone, and the gradient has not been reproduced at scale; the far larger NHANES analysis found a much smaller per-hour effect once body mass index was accounted for. Its value is that it is the one dataset in men of the age that usually asks the question, and it points the same way as the laboratory experiments in the young.
Does sleeping more raise testosterone back?
Recovering lost sleep restores testosterone quickly in the short term, but no randomized trial has tested whether extending sleep for weeks or months raises testosterone in men who habitually sleep short. The acute evidence is clear. In the Arnal 2016 experiment, testosterone that fell after a sleepless night returned toward baseline after one recovery night, and the Leproult protocol began with three ten-hour nights specifically to bring the men to a rested baseline before restriction, which implies the investigators expected and observed that extra sleep normalizes the profile. The long-term question is open. Sleep extension studies exist in athletes and in short-sleeping adults, and they have reported improvements in reaction time, mood, insulin sensitivity and appetite hormones over one to six weeks, but none of the well-known ones measured testosterone as a primary outcome in men. The Penev data in older men suggest the potential effect could be large, since men sleeping eight hours had roughly double the morning testosterone of men sleeping four, but that was a cross-section of 12 men, not a trial in which short sleepers were made to sleep longer. Two cautions apply. The Auyeung cohort found very long sleep was also associated with lower testosterone, which is more likely to reflect illness, depression or undiagnosed sleep disorders in the long sleepers than a harm from sleep itself, but it means the target is a normal night, not the longest one possible. And sleep that is longer but still fragmented, whether by apnea, alcohol, pain or a child, may not deliver the consolidated three-hour block and REM period that the Luboshitzky work suggests the nocturnal rise depends on. The defensible statement is that a man who moves from five hours to seven and a half of consolidated sleep should expect his daytime testosterone to rise toward his own rested level, and that nobody has yet measured how far that is over a year.
One Week of Five-Hour Nights Did What Aging Takes a Decade to Do
Percent change in daytime serum testosterone after eight nights of five hours in bed in Leproult and Van Cauter, JAMA 2011, shown as the midpoint of the reported 10 to 15 percent fall with the range as error bars, next to the pace of normal age-related decline. Ten healthy men with a mean age of 24.3 and a mean body mass index of 23.5 spent three ten-hour nights in the laboratory to reach a rested baseline, then eight nights with time in bed restricted to 12:30 am to 5:30 am, with blood drawn every 15 to 30 minutes for 24 hours at the end of each condition. Daytime testosterone was 10 to 15 percent lower after restriction, with the deepest trough between 2 pm and 10 pm, and 24-hour cortisol did not change. Testosterone in men declines by roughly 1 to 2 percent a year from the thirties onward, so the one-week loss matched the loss of 10 to 15 years of aging, which is the comparison the authors drew. The aging bars are population averages, not data from the same men, and the study was eight nights in young lean subjects, so it demonstrates causation and effect size in the short term rather than what chronic six-hour sleep does to a 50-year-old.
How should a man over 40 use this evidence?
A man over 40 should treat sleep as the one testosterone lever with clean causal human evidence, aim for seven or more hours of consolidated sleep on a regular schedule, and get screened for sleep apnea before spending money on anything sold as a booster. The controlled experiments in young men establish that cutting sleep to four or five hours lowers testosterone within days and that the loss is concentrated in the afternoon and evening, when energy, mood and libido are most noticed. The observational data in older men point the same way and suggest the gradient may be steeper with age, though the evidence is thin. Against that, the effect size for a moderately short night, six hours rather than seven and a half, is probably in the range of a few percent rather than the 10 to 15 percent seen at five hours, and it is smaller than the effect of carrying 15 kilograms of excess weight. The American Academy of Sleep Medicine recommends at least seven hours a night for adults, and that figure is the sensible target. Regular wake times matter because the nocturnal rise is sleep-dependent and a consolidated block is what the experiments protected. Snoring, morning headaches, a neck circumference over 43 centimetres, or a partner who reports pauses in breathing are each reasons to ask for a sleep study, because untreated apnea will blunt the overnight rise no matter how many hours are spent in bed, and because losing weight, whether by diet or with medical help, is the intervention that improves both the apnea and the testosterone. Alcohol in the evening should be understood as a sleep-fragmenting drug with a direct testicular effect. Finally, a man whose morning testosterone reads in the low 300s after a run of short nights should recognize that the 8 am sample is the one least affected by sleep loss, that the 2 pm to 10 pm window is where the deficit lives, and that a month of proper sleep before repeat testing is a reasonable and free first step. None of this is a treatment for any condition; it is a description of what the human experiments found and what they leave unanswered.
In 2,295 US Men, Each Hour of Sleep Lost Cost About 5.85 ng/dL, a Real but Modest Effect Once Weight Is Accounted For
Modelled decrease in total testosterone by hours of sleep lost from Patel and colleagues, World Journal of Urology 2019, using the 2011 to 2012 cycle of the US National Health and Nutrition Examination Survey. The analysis included 2,295 adult men with a measured serum testosterone and self-reported habitual sleep duration. After adjustment for age, body mass index and other health conditions, each hour of sleep lost was associated with a decrease of about 5.85 ng/dL in total testosterone, so the bars are that coefficient multiplied out: a man sleeping four hours rather than eight would be modelled at about 23 ng/dL lower. Body mass index was a considerably stronger predictor in the same model. Against a normal range of roughly 300 to 1,000 ng/dL, a 23 ng/dL shift is small, and it is far less than the 10 to 15 percent fall in the Leproult laboratory study or the two-fold gradient in Penev's older men. Part of the gap is that NHANES testosterone was drawn in the morning, the sample least sensitive to sleep loss, part is that self-reported sleep is imprecise, and part is that the large restriction effects were measured in young lean men under laboratory control. The population estimate is the conservative anchor; the laboratory estimate is the upper bound for a man whose sleep is genuinely cut to five hours.
Frequently Asked Questions
Have more questions? Contact us or explore our full peptide intelligence platform.
Related Articles
Is NAD+ a Peptide? No. It Is a Coenzyme Built From Two Nucleotides, and the Peptide Label Comes From the Clinic, Not the Chemistry
Is NAD+ a peptide? No. NAD+ is a coenzyme built from two nucleotides. What NAD+ is used for, what IV NAD+ studies found, and how it differs from SS-31.
Epitalon vs Pinealon: One Lab Built Both. Epitalon Has an FDA Review and Three Small Human Studies. Pinealon's Case Rests Mostly on Cell Dishes and Rats.
Epitalon vs pinealon compared: structure, targets, sleep and melatonin data, brain studies, human evidence, FDA status, and whether stacking them is tested.
Epithalon vs Epitalon: Same Peptide, Two Spellings. The Name That Means Something Different Is Epithalamin, and Most of the Human Data Belong to It.
Epithalon vs epitalon: one AEDG tetrapeptide, two spellings. Why epithalamin is a different substance and what FDA found in its 2026 epitalon review.