Sleep and weight loss, what four studies found

Key takeaways

• In small, short experiments, sleep changed what people lost on a diet, how hungry they felt and how many calories they took in. In Nedeltcheva's 2010 study, 10 adults on the same reduced-calorie diet lost 1.4 kilograms of fat over 14 days with 8.5 hours in bed a night and 0.6 kilograms with 5.5 hours.¹²³
• In Tasali's 2022 trial, 80 adults who usually slept less than 6.5 hours a night stayed at home with no set diet. The group counseled toward 8.5 hours in bed gained about 1.2 hours of sleep a night over the control group, and its change in calorie intake across 2 weeks was about 270 calories a day lower than the control group's.³
• In Spiegel's 2004 study, 12 healthy young men went through 2 days of restricted sleep and 2 days of extended sleep. With restricted sleep, leptin was 18 percent lower, ghrelin was 28 percent higher and hunger ratings were 24 percent higher.²
• In Benedict's 2011 study, 14 normal-weight men stayed awake for 1 full night. The next morning their energy use at rest was about 5 percent lower than after 8 hours of sleep, and no difference was detected in how much they ate at a buffet that afternoon.⁴
• No sleep condition in any of the four studies lasted longer than 2 weeks. Three of the studies had 14 people or fewer, two included men only, and none followed body weight for months.¹²³⁴

Does sleep affect weight loss?

One of the four studies tested sleep during a diet. In it, sleep changed what the lost weight was made of, not how much weight came off. Nedeltcheva's team had 10 overweight adults, 3 women and 7 men with an average age of 41, stay in a sleep laboratory for two 14-day periods at least 3 months apart. Both times they ate the same diet, with calories set at 90 percent of each person's resting metabolic rate. One period allowed 8.5 hours in bed a night and the other 5.5, in random order.¹

Weight loss was about the same both times, roughly 3 kilograms, or about 6.6 pounds. With 8.5 hours in bed, 1.4 kilograms of the loss was fat and 1.5 kilograms was fat-free mass. With 5.5 hours in bed, 0.6 kilograms was fat and 2.4 kilograms was fat-free mass. The abstract reports that short sleep cut the proportion of weight lost as fat by 55 percent and raised the loss of fat-free mass by 60 percent.¹

Fat-free mass in that study is scale weight minus the body fat measured by a DXA scan, so it counts everything in the body that isn't fat. The paper points out that lean tissue is mostly water.¹

The gap didn't come from eating more on short sleep. The group averaged 1,447 calories a day in the longer sleep periods and 1,450 in the shorter ones.¹

Of 12 people enrolled, 10 finished. The abstract says the nature of the study limited its duration and sample size.¹

What did each sleep study measure?

The four studies put different questions to different people. In three, the researchers set the sleep schedule. In the fourth, people slept at home and half of them had a counseling session.¹²³⁴

Spiegel's abstract names two limitations, that the study included only 12 young men and that it didn't measure energy expenditure.²

Tasali's 2022 trial is the largest of the four and the only one run in daily life. People wore a wrist monitor that tracked their sleep and got no diet or exercise instructions. After 2 baseline weeks, half had one counseling session aimed at 8.5 hours in bed, with a follow-up visit a week later.³

Does lack of sleep cause weight gain?

None of the four experiments ran long enough to show weight building up over months of short sleep. For that link, the papers point to population studies. Tasali's paper says prospective studies suggest short sleep is a risk factor for weight gain. Benedict's abstract says population evidence links long-term short sleep and body weight, and it calls the mechanisms behind that link largely unknown.³⁴¹²

Benedict's team tested one possible mechanism, energy use. The morning after a night awake, the men's energy use at rest was about 5 percent lower, and their energy use after eating 20 percent lower, than after a night of sleep.⁴

The closest thing to a weight gain result in the four papers is Tasali's control group. Those 40 people kept their usual short sleep, and over the 2 weeks their average intake rose by about 115 calories a day and their average weight by 0.39 kilograms. The authors list that rise as a limitation, because it may have added to the difference between the groups.³

Does less sleep make you hungrier?

In two of the four studies, people rated their hunger higher on less sleep. Spiegel's team controlled the calorie intake and physical activity of its 12 healthy young men. With restricted sleep, leptin, a hormone that lowers appetite, averaged 18 percent lower, and ghrelin, which raises it, averaged 28 percent higher. Hunger ratings were 24 percent higher, appetite ratings 23 percent higher, and appetite for calorie-dense foods high in carbohydrate 33 to 45 percent higher.²

Nedeltcheva's study also recorded more hunger during the 5.5-hour periods, and higher 24-hour levels of acylated ghrelin at the end of them. Leptin didn't match Spiegel's result. It fell as the weight came off in both conditions, and sleep loss had no significant effect by itself.¹²

Benedict's study found higher ghrelin on the morning after a night awake, and still detected no difference in buffet intake that afternoon.⁴

Ratings and hormone levels aren't the same as food eaten. Spiegel's and Nedeltcheva's studies both fixed how many calories people got, so neither one shows whether the hungrier condition would have led to more eating.²¹

What happened to calorie intake when people slept longer?

Tasali's trial set out to lengthen sleep in people's own homes. Its 80 participants were 41 men and 39 women aged 21 to 40 who usually slept less than 6.5 hours a night. By the wrist monitors, sleep in the extension group rose by about 1.2 hours a night compared with the control group.³

The change in calorie intake from baseline was about 270 calories a day lower in the extension group than in the control group. The paper gives a 95 percent confidence interval of 147 to 393 calories for that gap. Total energy use didn't change significantly.³

Weight change in the extension group was 0.87 kilograms lower than in the control group over the 2 weeks. Most of that gap was fat-free mass, 0.62 kilograms, and 0.25 kilograms was fat. The authors describe the weight change as modest.³

Where do the four sleep studies differ?

Lined up next to each other, the four papers answer some of the same questions differently. Their designs differ too. They ran from 1 night to 4 weeks, studied men only or both men and women, and either fixed what people ate or left it up to them.¹²³⁴

What don't the four sleep studies show?

They don't show what happens over months. The longest sleep condition in any of them was 2 weeks. Nedeltcheva's paper says more studies are needed on longer-term effects, and Tasali's says the trial gives no information on how long the changed sleep habits could be kept up.¹²³⁴

They don't cover many people. Three of the studies had 10, 12 and 14 participants, and Spiegel's and Benedict's included men only. Tasali's authors write that their selective eligibility criteria may limit how far the findings apply to more diverse groups.¹²³⁴

Only Nedeltcheva's 10-person study combined short sleep with a reduced-calorie diet. Tasali's participants had no prescribed diet, and one person found to be following a weight loss regimen was ruled ineligible, so the 270 calorie figure comes from people who weren't on one.¹³

Snoring, daytime sleepiness or trouble sleeping is something to bring to a doctor.

Worked example

Picture two made-up people who sit exactly on the group averages in Tasali's trial. The first is in the control group and takes in about 2,665 calories a day in the 2 baseline weeks and about 2,780 in the next 2, a rise of 114.9. The second is in the sleep extension group and goes from about 2,977 to about 2,821, a fall of 155.5. The trial's 270 is the gap between those two changes, 114.9 plus 155.5, or 270.4. In the last 2 weeks the second person still takes in more than the first, because that group's average started higher. The paper notes that comparing groups, not each group against its own baseline, is the recommended way to read a randomized trial. These are group averages from one 4-week trial, not a prediction for any one person.³

A closer look

Tasali's 270 calorie figure isn't a count of food. The team measured each person's total energy use with doubly labeled water, took daily home weights and body composition scans, and calculated intake as energy used plus the change in the body's energy stores. The paper says the doubly labeled water method has a precision of 5 percent, which may put some uncertainty into the intake numbers.³

Questions and answers

Related guides

• How big a calorie deficit is safe
• Not losing weight in a calorie deficit
• What is TDEE?
• Metabolic adaptation, measured
• Water weight vs fat loss
• Building muscle in a calorie deficit
• Calorie deficit calculator

Sources

1. Annals of Internal Medicine. Nedeltcheva, Kilkus, Imperial, Schoeller, Penev. 2010.
2. Annals of Internal Medicine. Spiegel, Tasali, Penev, Van Cauter. 2004.
3. JAMA Internal Medicine. Tasali, Wroblewski, Kahn, Kilkus, Schoeller. 2022.
4. American Journal of Clinical Nutrition. Benedict, Hallschmid, Lassen, Mahnke, Schultes, Schiöth, Born, Lange. 2011.