How important is sleep?

Sleep is incredibly important, and can be considered crucial alongside diet and exercise. Proper sleep habits help sustain many biological processes, and bad sleep can cause these processes to be suboptimal or even malfunction.

Fat Mass and Obesity

Epidemiological (Survey) Research

Several studies have been conducted looking at the correlation (degree of association) between body fat and sleep. There appears to be an inverse correlation (less sleep nightly being associated with more body fat)[1][2] that is further associated with more fat mass gain over a period of 5 years[3]

When excluding possible confounding factors, this association extends beyond the genome[4] and appears to persist after controlling for demographic, lifestyle, work and health related factors.[5]

Although correlation research is not conclusive, there appears to be a persistent relationship between less sleep time and greater fat mass. This association persists after controlling for the most predictable potential confounding agents and is likely related to sleep time per se (Sleep time being what is most easily measured in epidemiological research)

Sleep Deprivation

During intentional caloric restriction (fat loss diets), it appears that a reduction of sleep by 3 hours (8.5 to 5.5) is associated with an unfavorable nutrient partitioning effect, making more weight loss come from lean mass rather than fat mass relative to a rested control.[6]

Sleep deprivation may adversely affect nutrient partitioning during weight loss

At least acutely, sleep deprivation appears to increase hunger[7][8] and may be more significant when the sleep deprivation coexists with a reduced caloric intake.[6] In otherwise healthy women, this has been quantified at around a 20% increase in voluntary energy intake (and a slight increase in body weight of 0.4kg over 4 days).[9]

Sleep deprivation has been noted to increase circulating leptin (29%).[9]

Sleep deprivation, over time, may lead to higher fat mass gains (possibly secondary to hunger, which is reliably increased) although even short term sleep deprivation appears to hinder fat loss attempts via reducing the percentage of weight loss that is fat mass

Cognitive Output

Restriction of sleep produces a neural sleep wave pattern that is sometimes observed in depression,[10][10] and well-being appears to be related to sleep as well.[11] A reduction in sleep reduces higher levels of cognition such as problem solving.[12]

Impaired sleep is associated with impaired cognitive function

Insulin sensitivity

Epidemiological Research

There is a correlation between abnormal sleep patterns and metabolic syndrome, with less sleep[13][14][15] and more irregular or disturbed sleep[16] being positively correlated with the occurrence of the comorbidities of metabolic syndrome (insulin resistance, hypertension, obesity). Extending from insulin resistance, an increased occurrence of diabetes is seen in persons with poor sleep patterns[16][17] although there appears to be an increased risk for both shortened sleep (5-6 hours; RR of 1.28 and 95% CI of 1.03-1.60[16]) and prolonged sleep (8-9 hours; RR of 1.48 and 95% CI of 1.13-1.96[16]) which has been noted in other trials and meta-analyses.[18][19]

Both shortened as well as excessive sleep are associated with insulin resistance and an increased risk of diabetes in survey research. Persons with a 7-8 hour sleep pattern seem to be at the lowest relative risk, with similar increases in both the 5-6 hour range and the 8-9 hour range


Acute sleep deprivation is able to impair insulin signalling in isolated adipocytes in otherwise healthy young adults[20] which is present after three weeks of 90 minute deprivation[21] or five days of more severe restriction (60% reduction in time).[22] This insulin resistance is associated with less Akt phosphorylation induced by insulin (a reduction of 20% after 4 hours sleep deprivation for a few days)[20] and has been quantified at around a 20+/-24% reduction in sensitivity (IVGTT) or 11+/-5.5% (euglycemic-hyperinsulinemic clamp) in otherwise healthy men.[23]

Studies administering significantly reduced sleep time for a single day (4 hours of sleep) also confirm insulin resistance in otherwise healthy persons.[24]

Interventions that reduce sleep time by as little as 2 hours daily can induce a state of insulin resistance in otherwise healthy persons within a week, and halving sleep time to 4 hours or less is able to induce insulin resistance after a single night


Sleep time appears to be an individual predicting factor for both total and free testosterone in the morning in older men,[25] and during the process of aging the decline in testosterone (associated with aging) is further associated with perturbed sleep patterns.[26] One study, however, has failed to find an associated with sleep time overall and serum testosterone.[27]

A study assessing chronotype of subjects (relationship of body functions to the time of day, such as a 'morning' or 'evening' person)[27] assessed via the Composite Scale of Morningness (CSM)[28][29] noted that chronotype was associated with testosterone rather than total sleep time, with evening-orientated persons being associated with a higher testosterone level.[27] It suggests that past studies finding a relationship between evening testosterone and sleep time (not deprivation studies, but associative studies) may be confounded with daily fluctuations of testosterone as chronotype is independent of sleep duration.[30]

In general, sleep appears to be somewhat associated with testosterone levels. The strength of the correlation is not remarkable, but studies have at least noted some form of relationship. One study suggests that this may be more indicative of chronotype than overall sleep time, however, with those two factors being correlated but independent

One study (measuring testosterone for one day during waking hours) in young male subjects sleeping 8 hours routinely that cut sleep by 3 hours for a period of 5 days reduced testosterone by an average 10.4% relative to rested control,[31] suggesting that acute sleep deprivation is able to influence testosterone levels. Another (similarly small) controlled study noted that despite weird fluctuations in FSH between persons, that sleep deprivation in young men noted a 30.4% decrease in testosterone that was accompanied by a decrease in DHT (26.4%) and Androstenedione (32.6%).[32] One study using 60% sleep deprivation (10 hours routinely was then reduced to 4) for a period of 5 days noted a trend for reduced testosterone but failed to reach statistical significance despite an increase in SHBG.[22]

This decrease in androgens occurs during 24 hours of sleep deprivation as well, but doesn't increase further beyond this time point.[33]

One study has been conducted on outright sleep deprivation for one night (33 hours straight without sleep) has found an acute reduction in testosterone levels coupled with less reactive aggressiveness.[34]

A single night without sleep is enough to decrease androgen production, and moderate daily sleep deprivation reliably reduces androgen levels (Testosterone levels being the most frequently measured) by some 10–30%.


Cortisol is a hormone that mediates the process of waking up, and shows a predictable circadian rhythm of being high in the morning while lower at night prior to sleep.

While some studies have admittedly found no significant effect or mild sleep deprivation over a few days on cortisol,[31][32] numerous studies have noted increases seen with a single night[35] or 5 days[22] which can reach 51+/-8%.[23]

The studies that find increases in cortisol tend to measure whole-day cortisol secretion, and may be showing an increase due to cortisol being increased in the evening following sleep deprivation.[35] Morning readings of cortisol following sleep deprivation are actually reduced,[36][37][38] and as such sleep deprivation appears to dysregulate and normalize the normally circadian rhythm of cortisol although overall exposure to cortisol goes upwards.

Cortisol normally follows a pulsative pattern and is higher in the morning and lower at night. Sleep deprivation dysregulates this, and causes a normalization of sorts of this pulsatile pattern (reducing morning cortisol, increasing serum cortisol) while whole-day exposure to cortisol goes up slightly

Thyroid Hormones

Sleep deprivation in otherwise healthy females has been noted to increase thyroid hormones T3 (19%) and T4 (10%)[9] although other trials have failed to find an alteration in thyroid hormones with 33 hours acute sleep deprivation[39] or chronic sleep deprivation.[40] One study that measured thyroid stimulating hormone (TSH) noted that it was elevated during acute sleep deprivation[41] but this has not been observed with chronic sleep deprivation.[40]

Practical sleep deprivation either absolutely for one day or a reduction over a few days does not have consistent evidence for its effects on thyroid hormone levels

Diet induced thermogenesis may be slightly suppressed with sleep deprivation[42] and cold-induced thermogenesis does not appear to be affected.[39] At least in rats, sleep deprivation actually increases thermogenesis (resulting in weight loss despite increased food intake[43]) in accordance with symptoms of sleep deprivation in rats (hyperphagia, weight loss, elevated energy expenditure, increased plasma catecholamines, hypothyroidism, reduction in core temperature, deterioration in physical appearance).[44]

Studies that measure metabolic rate or total energy expenditure fail to find significant differences between normal sleep patterns and deprivation[9] or possibly an overall increase in metabolic rate that is due to more spontaneous physical activity (ie. movement).[42]

Metabolic rate is not reduced with sleep deprivation, and some evidence suggests that it is actually increased with sleep deprivation (either inherently as seen in rats, or secondary to an increase in physical activity)

Growth Hormone

A major pulse of growth hormone occurs shortly after falling asleep in relation to slow wave sleep[45][46] and delta waves (0.5–3.5Hz),[47] this spike accounts for approximately 50% of the daily AUC (Area-under-curve; a measure of overall exposure) of growth hormone in otherwise healthy young men.[48] The association between the GH pulse and slow wave sleep is not seen at all times[49][50] and some authors suspect that slow wave sleep is not an inducer of growth hormone but a coordinator of pulses (thus forcing a correlation).[51]

Overall growth hormone secretion appears to be greater in youth and in women relative to older individuals and men, respectively; the increase seen in women is due to higher daytime levels being positively influenced by estrogen.[52]

Sleep mediates the largest daily spike of growth hormone, which in young persons accounts for approximately half of daily exposure

Studies using sleep deprived persons that note a decline in growth hormone (due to lack of sleep and thus missing the pulse) note that daily GH production increases, but only enough to approximately compensate.[53][51] This may be due to chronic sleep deprivation (6 nights of 4 hours sleep) causing a predictable biphasic GH pulse pattern[54] or one study noting that night workers who had a small (16.8+/-3.3%) sleep pulse had sporadic pulses throughout the day to normalize the AUC.[51]

It appears that when the pulse of GH seen with sleep is disrupted, that the body compensates during the day and overall daily exposure to GH is left not significantly different

For more information, see ten tips for better sleep.