Can Just One Week of Poor Sleep Affect Your Blood Sugar Control?
Yes. In this 12-day inpatient study, cutting healthy men from 10 hours a night in bed to 5 hours a night in bed for one week reduced their insulin sensitivity by 20 ± 24% on intravenous glucose tolerance testing and by 11 ± 5.5% on the euglycemic-hyperinsulinemic clamp, the reference method for measuring insulin action. These were healthy men aged 20 to 35 with a BMI between 20 and 30, eating a controlled diet, with activity controlled too. One week was enough to move the needle.
Dr. Kumar’s Take
This is a small, tightly controlled inpatient study, and that is exactly why I take it seriously. Diet and activity were held constant, so the change in insulin sensitivity has nowhere else to come from. Both measurement methods pointed the same direction, and the clamp result is the one I weigh most heavily because it measures insulin action directly rather than inferring it from a model.
The size of the drop on the IVGTT (20 ± 24%) carries a wide spread, so I would not tell a patient to expect a specific number for themselves. What I take from this is directional and still clinically meaningful: seven nights of 5 hours in bed measurably degraded glucose handling in young, healthy men who had no metabolic disease to begin with. If sleep restriction can do that to the healthiest possible group, I am not comfortable assuming it does nothing in patients who already carry metabolic risk.
One more thing worth knowing: the researchers also tested modafinil, a wake-promoting drug, during the sleep-restricted week. It did not protect against any of the metabolic changes. Feeling more alert is not the same as being metabolically intact.
Key Findings
Twenty healthy men participated. They spent 10 hours per night in bed for at least eight nights, including three inpatient nights, in the sleep-replete condition, then 5 hours per night in bed for seven nights in the sleep-restricted condition. On the last two days of each condition, the investigators measured glucose metabolism with an intravenous glucose tolerance test and a euglycemic-hyperinsulinemic clamp.
IVGTT-derived insulin sensitivity was reduced by 20 ± 24% after sleep restriction (p = 0.001), with no significant change in the insulin secretory response. Clamp-derived insulin sensitivity was reduced by 11 ± 5.5% (p < 0.04). Glucose tolerance and the disposition index were both reduced by sleep restriction.
Salivary cortisol rose by 51 ± 8% with sleep restriction (p < 0.02). But the changes in insulin sensitivity did not correlate with the changes in cortisol, with 24-hour urinary catecholamines, or with slow wave sleep. So the mechanism behind the drop in insulin sensitivity is not settled by this data.
Subjects received either 300 mg/day modafinil or placebo during the sleep restriction week. None of the metabolic outcomes were affected by modafinil treatment.
Brief Summary
This 12-day inpatient general clinical research center study examined the effects of sleep restriction on glucose metabolism in 20 healthy men aged 20 to 35 years with a BMI of 20 to 30 kg/m². Subjects spent 10 hours per night in bed for at least eight nights (sleep-replete), followed by 5 hours per night in bed for seven nights (sleep-restricted). Diet and activity were controlled. Glucose metabolism was assessed by intravenous glucose tolerance test and euglycemic-hyperinsulinemic clamp on the last two days of each condition. Salivary cortisol, 24-hour urinary catecholamines, and neurobehavioral performance were also measured.
Study Design
This was a double-blind, placebo-controlled, randomized clinical study conducted at an inpatient general clinical research center. The two sleep conditions ran in fixed sequence, sleep-replete first and then sleep-restricted, so this was not a randomized crossover. The randomization was to drug: subjects received 300 mg/day modafinil or placebo during the sleep restriction week.
Healthy male subjects were recruited through newspaper ads, flyers, and website postings, and screened for sleep patterns and medical and psychological history. Procedures were approved by the human research committee of Brigham and Women’s Hospital and conducted according to the principles of the Declaration of Helsinki. All subjects gave written informed consent. Because the study was inpatient, diet and activity were controlled rather than self-reported.
Results You Can Use
After seven nights of 5 hours in bed, insulin sensitivity fell by 20 ± 24% on the IVGTT (p = 0.001) and by 11 ± 5.5% on the clamp (p < 0.04). Glucose tolerance and the disposition index were both reduced. The insulin secretory response did not change significantly, which means the problem was on the insulin sensitivity side rather than the insulin production side.
Salivary cortisol went up 51 ± 8%, but that increase did not track with the drop in insulin sensitivity. Neither did urinary catecholamines or slow wave sleep. Modafinil, which is used to counter the alertness and performance costs of sleep loss, did not change any of these metabolic outcomes.
Why This Matters For Health And Performance
Average sleep duration in the U.S. has fallen below 7 hours per night, a drop of roughly 2 hours per night over the last century and more than 1 hour per night over the last 40 years. Cross-sectional and longitudinal studies have linked short sleep duration or poor sleep quality to increased risk of obesity, diabetes, hypertension, cardiovascular disease, the metabolic syndrome, and early mortality.
Those are associations. This study is different: it is an experiment, in a controlled inpatient setting, where sleep was the variable that changed. That is what gives it weight. The authors’ own conclusion is that sleep restriction of 5 hours per night for one week significantly reduces insulin sensitivity, raising concerns about the effects of chronic insufficient sleep on disease processes associated with insulin resistance.
Insulin resistance has also been reported in two other models of disrupted sleep, sleep apnea and experimental disruption of slow wave sleep. In the slow wave sleep model, the extent of disruption predicted the reduction in insulin sensitivity. In this study, though, the change in insulin sensitivity did not correlate with slow wave sleep, so short sleep and fragmented sleep may not be acting through the same route.
How to Apply These Findings in Daily Life
- Protect the sleep opportunity, not just the sleep: This study manipulated hours in bed. Going from 10 hours in bed to 5 hours in bed for a week was enough to reduce insulin sensitivity.
- Treat a week of short nights as metabolically real: Seven consecutive nights was the exposure here, and it was enough to change glucose handling in healthy young men.
- Do not count on stimulants to cover the gap: Modafinil at 300 mg/day did not protect insulin sensitivity, glucose tolerance, or the disposition index in this study.
- Bring sleep into the metabolic conversation: If you are being evaluated for diabetes risk, your sleep schedule is relevant information for your clinician.
- Be careful generalizing to yourself: The participants were healthy men aged 20 to 35 with a BMI of 20 to 30. If you sit outside that description, this study does not directly describe you.
Limitations To Keep In Mind
The study enrolled 20 healthy men aged 20 to 35 with a BMI of 20 to 30, so it says nothing about women, older adults, people with higher BMI, or people with existing metabolic disease. The sleep conditions ran in fixed sequence rather than randomized order, which means order effects cannot be separated from the effect of sleep restriction. The standard deviation on the IVGTT result (20 ± 24%) is larger than the mean effect itself, so individual responses varied a great deal. The exposure was one week, so this does not tell you what months or years of short sleep do. And the inpatient setting that makes the result clean, with diet and activity controlled, is also what makes it unlike ordinary life.
Related Studies And Internal Links
- Sleep Duration and Type 2 Diabetes Risk Meta-Analysis
- Short Sleep and Poor Diet: Double Hit for Diabetes Risk
- Prevalence and Geographic Patterns of Self-Reported Short Sleep Duration Among US Adults
- Glycine Ingestion Improves Subjective Sleep Quality in Human Volunteers
- How to Sleep Better: Science Daily Playbook
FAQs
How much sleep restriction did this study actually test?
Five hours per night in bed for seven consecutive nights, compared against 10 hours per night in bed. That is the exposure, and it produced a 20 ± 24% reduction in IVGTT-derived insulin sensitivity and an 11 ± 5.5% reduction in clamp-derived insulin sensitivity.
Was the problem that the pancreas stopped making insulin?
No. The insulin secretory response did not change significantly. Insulin sensitivity dropped, meaning the body responded less well to the insulin it was making. The disposition index, which accounts for both, was also reduced.
Does a wake-promoting drug protect against this?
Not in this study. Subjects took either 300 mg/day modafinil or placebo during the sleep-restricted week, and the metabolic outcomes were not affected by modafinil.
Is elevated cortisol the explanation?
Salivary cortisol did rise by 51 ± 8% with sleep restriction, but the changes in insulin sensitivity did not correlate with the changes in cortisol, urinary catecholamines, or slow wave sleep. The mechanism is not established here.
Conclusion
Sleep restriction to 5 hours per night for one week significantly reduced insulin sensitivity in healthy men, by 20 ± 24% on intravenous glucose tolerance testing and 11 ± 5.5% on the euglycemic-hyperinsulinemic clamp, with glucose tolerance and the disposition index reduced as well. The authors’ conclusion is that this raises concerns about the effects of chronic insufficient sleep on disease processes associated with insulin resistance. I read it the same way.
