Does Intermittent Fasting Improve Sleep?

Evidence on whether intermittent fasting improves sleep quality in adults with overweight or obesity.

Pooled across 16 randomised controlled trials of intermittent fasting in adults with overweight or obesity, there was no overall effect on sleep quality. The same pooled trials did produce real reductions in weight, fat mass and blood pressure, so this is not a set of studies where nothing happened. Something happened, and it was not sleep. KnowYourPrime's sleep desk rates this claim Unsupported. One qualification is load-bearing and travels with the finding: sleep was a secondary outcome in nearly all of these trials rather than the thing they were designed to detect, so this is an absence of evidence from studies that were not built to answer the question, not a demonstration that fasting worsens sleep or that it cannot help anyone.

What the pooled trials actually measured

The synthesis is a systematic review and meta-analysis of randomised controlled trials, published in Sleep Medicine Reviews in 2025, covering intermittent fasting diets in adults with overweight or obesity and reporting both sleep and cardiometabolic outcomes. Sixteen randomised trials contributed.

The cardiometabolic side of that pooling worked. Weight came down, fat mass came down, and blood pressure came down. Those are the outcomes intermittent fasting is usually studied for, and the trials detected them. The sleep-quality side produced no overall effect.

That internal contrast is what makes the result worth reporting rather than filing as a null. A set of trials that found nothing anywhere would raise the question of whether the intervention was delivered or the measurements were any good. These trials demonstrably detected change in the outcomes they were built around, and the sleep measure did not move with them.

Why an absence of evidence is not the same as evidence of absence

Sleep was a secondary outcome in nearly all 16 trials. A secondary outcome is one a study records but is not powered to detect, and it is measured with whatever instrument is cheap enough to add to a protocol built for something else, which for sleep usually means a self-reported questionnaire rather than polysomnography or even a wearable.

Two consequences follow. A real but modest effect could sit inside this pooled result without reaching significance, because the trials were sized for weight rather than for sleep. And a questionnaire asking how someone rates their sleep is a different measurement from how long they slept, how fragmented it was, or how much of it was spent in any particular stage.

So the honest reading of Unsupported here is specific. The claim that intermittent fasting improves sleep quality has been tested at scale and has not been supported. The question of whether a trial designed around sleep, powered for sleep and measuring sleep objectively would find an effect remains open, because that trial has not been published.

What this does and does not tell a reader who fasts

It says that an expectation of better sleep is not supported by the randomised evidence, which matters because better sleep is commonly listed among the reasons to try eating in a restricted window. The weight, fat mass and blood pressure reductions in the same pooled trials are supported.

It says nothing about any individual's experience. A pooled average of no overall effect is compatible with some people sleeping better, some sleeping worse, and the two cancelling in the mean. Trials of this design report the group, and the group is not a person.

It also does not address timing within the eating window, which is a separate question with its own evidence. How close a meal sits to bedtime, and how caffeine timing interacts with sleep onset, are not what these trials varied.

Common questions