Late calories do not count double#
No hour of the day converts food into more energy than it contains. A 400-calorie meal at 10pm and a 400-calorie meal at 6pm deliver the same 400 calories, and the studies that hold total intake fixed do not find a large penalty for putting them late. If your evening eating is displacing earlier eating rather than adding to it, the clock is close to a non-issue.
But the clock is not entirely innocent either, and the interesting part is that two bodies of research disagree about how guilty it is. Controlled feeding experiments that fix the calories tend to find a measurable late-eating cost. Free-living studies that track when people eat tend to find the association disappears once you account for how much they ate. That disagreement has a specific cause — the two literatures are not measuring the same thing by "late" — and it is worth unpacking, because it decides whether the advice "don't eat after 8pm" means anything for you personally.
The mechanism behind the myth fails a direct test#
The folk version has a mechanism attached: food eaten before bed is stored as fat because you're not moving, and it stops your body burning fat overnight. That claim is specific enough to test, and it has been tested.
Twelve men with obesity (BMI 36.1 ± 1.9) took either 30 g of casein protein (120 kcal) or a non-nutritive placebo about 30 minutes before sleep in a randomized, double-blind crossover. Overnight lipolysis — measured as interstitial glycerol in subcutaneous abdominal fat, a direct readout of fat being broken down — did not differ between conditions (177.4 ± 26.7 vs 183.8 ± 20.2 μmol/L, p = 0.83), and neither did next-morning values or resting energy expenditure2. Eating shortly before sleep did not switch off overnight fat breakdown.
That is one small crossover with one food at one dose, so it doesn't license eating anything in any quantity at midnight. But it does dispose of the specific claim that pre-sleep calories are metabolically quarantined into fat. Whatever late eating costs you, it isn't that.
Hold the calories fixed and a real, modest cost appears#
The other direction has support too. A randomized crossover trial with tight control of nutrient intake, activity, sleep and light exposure ran the same meals early versus late in adults with overweight and obesity. Late eating increased hunger (p < 0.0001), raised the ghrelin-to-leptin ratio (p = 0.006 over 24 hours), lowered waking energy expenditure (p = 0.002), and shifted adipose-tissue gene expression toward lipid storage3. The size of that effect once it reaches the scale is covered in whether meal timing affects fat loss, which owns the trial in detail.
So we have controlled trials finding a timing effect and observational work often finding none. The standard reconciliation is "the lab effect is real but too small to see in the noise of real life." That's partly true. It is not the whole answer.
The sharper split: whose clock is "late" measured against?#
Almost every study in the free-living literature defines late eating by the wall clock — calories after 8pm, or after 9pm. That definition contains a hidden assumption: that 8pm means the same physiological thing to everyone. It doesn't. Body clocks run on different schedules, and the gap between an early and a late chronotype can be several hours. For one person a 10pm meal arrives well inside biological night; for another it lands in mid-evening, hours before their system starts shutting down for sleep.
One study measured this directly rather than assuming it away, anchoring meal timing to each participant's own melatonin onset instead of to the clock. Timing relative to that internal marker tracked body fat and BMI; the clock hour of eating did not1. The distinction is discussed further in whether when you eat matters more than what you eat.
If "after 8pm" describes biological night for some people and ordinary evening for others, a study built on that cutoff is averaging two different exposures — and averaging shrinks whatever effect is there.
That is a mechanical reason the free-living literature would report weaker timing effects than the controlled trials, quite apart from the effect genuinely being small. The two research designs are not disagreeing about an answer so much as asking slightly different questions: the lab asks what happens when identical calories move later against a fixed internal clock, and the survey asks what happens to people whose clocks are all in different places. Neither is wrong. Only one of them is measuring the exposure the mechanism actually cares about.
The practical upshot is that "don't eat after 8pm" is advice keyed to the wrong reference frame. It is strict for early risers, lenient for night owls, and calibrated to neither.
| Claim | What the evidence shows |
|---|---|
| Calories after 8pm count double | No — no metabolic basis found |
| Pre-sleep food blocks overnight fat breakdown | No — lipolysis unchanged, p = 0.83 |
| Late eating costs something at matched calories | Yes — modest, in controlled trials |
| Wall-clock time of eating predicts body fat | No — not in the melatonin-anchored study |
| Timing relative to your own body clock predicts body fat | Yes — 1.1 h closer, p = 0.009, cross-sectional |
In real life, late eating is usually extra eating#
All of the above concerns timing at matched calories, which is a laboratory condition and not a description of most evenings. Outside the lab, a long evening is mostly a longer opportunity. The calories that arrive after 10pm are disproportionately discretionary — snacks rather than meals — and they are additions rather than substitutions, because the earlier meals already happened.
There is direct evidence for this being the operative variable. In 52 adults wearing actigraphy and keeping food logs for a week, calories consumed after 8:00 pm predicted BMI even after controlling for sleep timing and sleep duration — while sleep timing itself did not predict BMI once duration was accounted for4. The amount landing in the late window carried the association; the schedule around it did not.
The composition skews too. Late sleepers in that study ate more fast food and full-calorie soda and fewer fruits and vegetables. What gets eaten late is not usually a second dinner of vegetables; a tired brain reaches for energy-dense options, a bias documented in what sleep loss does to cravings. And a late-eating pattern often runs alongside a short-sleep pattern, so the two effects arrive together and get credited to the wrong one — the sleep and appetite hormone pillar untangles which parts of that cascade actually replicate.
This is why the practical answer converges even though the science splits. Whether the mechanism is circadian or purely arithmetic, the intervention is the same: the problem to solve is the extra intake in the late window, not the hour itself. Someone who eats dinner at 9:30 and stops has a very different situation from someone who eats dinner at 7 and grazes until midnight, even though a clock-based rule condemns the first and misses the second. That distinction is also why a log that records when alongside what is more informative here than either alone — the pattern only becomes visible across a week.
If you want a rule that survives the evidence, it isn't a cutoff time. It's this: close the kitchen at a point that reliably ends intake rather than at a point that feels virtuous, and shift it earlier only if the late calories are additional. Advice for actually doing that is in how to stop night eating.
FAQ#
Is there a cutoff time after which I shouldn't eat?#
No cutoff has survived testing as a clock hour. In the study that measured people's internal clocks directly, the clock hour of eating showed no association with body fat, while timing relative to individual melatonin onset did1. A fixed hour applies one rule to people whose biology is running hours apart.
Does food eaten before bed get stored as fat overnight?#
Not on the evidence available. Thirty grams of casein 30 minutes before sleep left overnight fat breakdown statistically unchanged versus placebo (p = 0.83), with no difference in resting energy expenditure2. Late calories still count toward your total; they are not diverted to fat by the hour.
Does a bedtime snack ruin a calorie deficit?#
Only if it puts you over your total. A snack that fits inside the day's intake is a scheduling choice. The reason bedtime snacks have a bad reputation is that they are usually added on top of a completed day of eating rather than planned into it.
Sources#
- McHill AW, Phillips AJK, Czeisler CA, et al. Later circadian timing of food intake is associated with increased body fat. American Journal of Clinical Nutrition. 2017;106(5):1213-1219.
- Kinsey AW, Cappadona SR, Panton LB, et al. The effect of casein protein prior to sleep on fat metabolism in obese men. Nutrients. 2016;8(8):452.
- Vujović N, Piron MJ, Qian J, et al. Late isocaloric eating increases hunger, decreases energy expenditure, and modifies metabolic pathways in adults with overweight and obesity. Cell Metabolism. 2022;34(10):1486-1498.
- Baron KG, Reid KJ, Kern AS, Zee PC. Role of sleep timing in caloric intake and BMI. Obesity. 2011;19(7):1374-81.
