Don’t eat before bed? Not so fast.
“Don’t eat before bed” is one of the most repeated pieces of nutrition advice out there. My patients come in convinced it’s true, and just as many come in convinced of the opposite. As with most things in nutrition, the real answer depends on who you are, what you ate, how much you move, and what you’re trying to achieve.
That’s why I sat down with Dr. Mike Ormsbee of Florida State University. Mike has likely done more research on pre-sleep nutrition than anyone in the field. He’s also a former college hockey player, a muscle physiologist, and one of the clearest science communicators I know. We covered pre-sleep protein, fat metabolism, resistance training for postmenopausal women, muscle loss during rapid weight loss, collagen, ketones, and much more.
Here’s what you need to know.
What “pre-sleep feeding” really means
Whenever someone says “don’t eat before bed,” Mike starts with a simple question: don’t eat what? How much? How active are you? Did you just train or compete? What’s your goal?
In his lab, pre-sleep feeding has a precise definition: a small, protein-centric snack eaten at least two hours after dinner and about 30 minutes before getting into bed. It is not a second dinner, and it is not a large mixed meal of carbohydrates, protein, and fat. Mike was clear that a Thanksgiving-sized plate right before bed is not what anyone should be aiming for. Some labs have reported that large mixed meals can upset heart rate variability and sleep.
Mike’s original studies found a small, statistically significant rise in next-morning resting metabolic rate after a pre-sleep snack. That was mostly in active young men, and carbohydrate raised it as much as protein did. Later studies in active women, in sedentary pre- and postmenopausal women, and in mildly overweight men found no change. Mike was refreshingly honest about what the bump meant even when it showed up: “It was significant, statistically significant. Meaningfully, I’m not sure. The number moved a little bit.”
The more important finding came from fat tissue itself. Mike’s team placed tiny microdialysis probes into abdominal fat (“almost like a belly piercing”) so they could watch what fat was doing overnight. In resistance-trained young women, taking 30 grams of casein 30 minutes before bed, instead of right after an evening workout, did not change overnight fat release or fat burning. The total protein was the same; only the timing moved. For active people, a small protein snack before bed doesn’t appear to blunt overnight fat metabolism or drive fat gain.
One notable exception: in young, sedentary women with obesity, a nightly snack raised next-morning insulin. It didn’t matter whether the snack was casein, whey, or carbohydrate. When Mike’s team put women from the same population through four weeks of exercise training while they kept up the nightly snack, that effect on insulin resistance was gone. That study didn’t include a snack-only group for comparison. Even so, to Mike it showed what a “blunt instrument” exercise is for controlling metabolism. Which brings us to the most powerful lever of all.
Casein or whey before bed? Pick the one you like
First, a quick word on the protein itself. For about a decade, casein was considered the gold-standard bedtime protein. It digests slowly during the day, so the logic was that it would drip-feed amino acids overnight.
Newer data say it doesn’t matter much. Work from Mike’s lab and from Luc van Loon’s group in Maastricht found no meaningful difference between whey and casein. That includes a trial led by Jorn Trommelen showing no difference in overnight myofibrillar or mitochondrial muscle protein synthesis. Why would casein lose its slow-release edge at night? Mike was upfront that nobody knows yet. In conversations with Luc, they’ve wondered whether lying flat, or small overnight changes in stomach pH, might be the reason. That’s a hypothesis, not a finding.
Form didn’t matter either. Mike’s team took cottage cheese off a grocery shelf, had it powdered into an exact replica, and compared the whole food with the shake and a placebo in active women. There were no differences in next-morning metabolic rate, fuel use, or hunger.
The practical takeaway: whey, casein, cottage cheese, Greek yogurt, or a protein bar. Choose what you enjoy and what keeps you consistent. Most of the studies used about 30 to 40 grams of protein, roughly 150 calories. Mike’s read is that roughly 200 to 300 calories of protein-dominant food looks fine, but beyond that, “we just don’t know.”
Exercise: your “metabolic apology”
Mike has a phrase I absolutely love: exercise is our metabolic apology for poor lifestyle choices. He pointed to two lines of evidence.
You can’t fully outrun a bad diet, but exercise helps. In a 2013 study from James Betts’s group, healthy young men were held to fewer than 4,000 steps a day and overfed by 50% for a week. One group also ran for 45 minutes a day at 70% of VO2 max, with extra calories added to cover what they burned. Both groups gained weight. But in the non-exercisers, the insulin response roughly doubled, while the exercisers largely kept insulin and other metabolic markers in check.
Exercise may protect you when sleep falls apart. Mike’s doctoral student Brett Cross cut participants’ sleep by 30% for three nights in a row, and within those three nights vascular markers such as the augmentation index worsened. The group meeting the ACSM guideline of 150 minutes of aerobic exercise per week was protected from those changes. This is new dissertation work, and it fits with other trials showing that exercise training limits the vascular damage from sleep loss. The protection isn’t universal, though: one trial in adults with obesity found daily exercise didn’t fully offset the metabolic effects of short sleep.
Think about new mothers, shift workers, and anyone in a season of poor sleep. Movement may be one of your best defenses.
Resistance training is a fat-loss tool, not just a muscle tool
Mike’s dissertation, more than 20 years ago, was the first to ask how resistance training affects fat metabolism specifically, rather than muscle growth. His reasoning was simple: the leanest people he knew lifted weights.
First, an important distinction. Lipolysis is fat being released from the fat cell into the blood. Fat oxidation is that fat actually being burned for fuel. A cup of coffee can raise lipolysis without changing body composition. Real fat loss needs both. And where does the fat go? Ultimately, you breathe most of it out as carbon dioxide.
The case that lifting supports this process has been building for decades:
- Metabolism stays up after a hard session. Studies going back to the late ’90s show metabolic rate stays elevated after a hard lifting session. In some studies it’s still measurably higher about a day later. The extra calories are modest, so think of this as a small bonus rather than a fat-loss engine.
- Lifting produces a bigger hormonal response. Resistance exercise produces a larger catecholamine and growth hormone response than walking. Growth hormone is both anabolic for muscle and lipolytic for fat, which we often forget.
- Programs that include lifting work well. In Mike’s early work with Paul Arciero, a program combining three days of resistance training, three days of interval running, and a higher-protein diet beat a general activity plan for fat loss, including visceral fat, over 12 to 16 weeks. Mike was candid that the interval training and the extra protein make it impossible to credit the lifting alone. In broader comparisons, vigorous aerobic exercise and interval training reduce visceral fat the most, while resistance training on its own produces smaller visceral fat reductions. Lifting’s unique strengths are building strength and preserving muscle. In practice, the combination of the two wins.
- Muscle talks to fat. Signals released from working muscle communicate directly with adipose tissue.
One caveat: people with obesity often show a blunted fat-release response to both resistance and endurance exercise. Training doesn’t always normalize that response, even when body composition and metabolic health improve. That doesn’t mean exercise isn’t working. It means progress may take longer. As Mike put it, it took years to get here, so expect it to take time to get out. Give yourself grace and stay the course.
Rapid weight loss without resistance training costs you muscle
This was the finding that stopped me in my tracks, and it’s more relevant than ever in the era of GLP-1 medications.
Around 2016, Mike’s team worked with patients considering bariatric surgery who had been placed on a medically prescribed 800-calorie diet. Mike immediately noticed that their protein intake was dangerously low, around 0.3 to 0.4 g/kg. With approval from the ethics board, his team increased protein for everyone, which brought intake to roughly 1,100 to 1,200 calories.
One group received standard of care: light calisthenics. The other did structured resistance training in the gym.
Share of weight lost from lean tissue, by group (DEXA):
- Standard of care (calisthenics): ~25%
- Resistance training: ~4%
Both groups lost a lot of weight. The difference was what they lost. The study was very small (fewer than 14 people) and DEXA has its limits, as Mike pointed out himself. Still, the signal fits the broader research. Typically a fifth to two-fifths of diet-induced weight loss comes from lean tissue, and resistance training substantially reduces that loss.
If you’re using a GLP-1 or any aggressive weight-loss approach, prioritize adequate protein and resistance training. Most of the evidence comes from calorie-restriction studies rather than GLP-1 trials specifically, but the logic carries over. Protect your muscle.
And you don’t need a fancy gym to start. Most successful studies used two to three full-body sessions per week, often on machines. During COVID, one of Mike’s students ran a study in frail patients after heart-valve replacement, who trained at home over Zoom with water jugs and grocery bags. They still got significantly stronger.
Postmenopausal women: lift, and maybe sleep better too
Mike and his colleague Dr. Bob Hickner are wrapping up a five-year NIH-funded study in postmenopausal women with obesity and prediabetes. Their team screened nearly 2,000 people to enroll about 60 women, a population that has historically been told to “just go walk.”
The women were randomized to 12 weeks of either treadmill training or circuit-style resistance training, matched for calories burned. Matching was harder than it sounds: burning 200 calories took 30 to 50 minutes on the treadmill but 75 to 105 minutes of circuit lifting. Before and after, each woman completed an eight-hour testing day with insulin clamps, isotope tracers, microdialysis, and muscle biopsies.
The full analysis is still coming. Early results presented at conferences show strength improving in both groups, with greater gains in the resistance group. Mike expects lifting to come out equal or superior on fat metabolism outcomes, but that’s his prediction until the data are analyzed. The women didn’t need special “menopause-specific” movements, either. A standard, well-coached program worked.
Some of these women have kept training in an alumni group. One recently won a gold medal in the deadlift at the Senior Games for her age group. That’s what’s possible.
What about sleep? PhD student Laney Thompson is now studying whether pre-sleep protein can improve sleep in postmenopausal women, a group that commonly struggles with it. The focus is alpha-lactalbumin, a milk protein naturally rich in tryptophan. Across the published studies the results are mixed. The most consistent signal is falling asleep faster, and several trials in athletes found no benefit. Mike’s own trial in elite soccer players found no benefit either, but the product turned out to be only about 44% alpha-lactalbumin. The positive studies used roughly 93%, and Mike thinks the purity difference may explain the result. The next study will use the high-purity version. Stay tuned.
Protein, aging, and a myth that won’t die
I recently heard physicians at a scientific meeting claim that excess protein is primarily stored as fat. Mike and I agree: the evidence does not support this. Jose Antonio’s studies in resistance-trained adults, feeding about 3.3 to 4.4 g/kg of protein per day, found no increase in fat mass.
For older adults, a bedtime protein snack is an opportunity. Luc van Loon’s group showed that older adults can increase muscle protein synthesis from protein consumed before and during sleep. In older men, a larger dose (about 40 grams) did this, while 20 grams did not reliably do so. Long-term gains in muscle mass from bedtime protein in older adults haven’t been shown yet. Even so, Mike would be very hesitant to tell an aging adult to skip a protein-only snack before bed. In his experience, most older adults simply don’t eat enough protein, and evenings at home are an easy window to add some.
What about how protein is spread across the day? Here the science is genuinely mixed. A French group (Arnal and colleagues) actually found the opposite of what you might expect. In older women, eating most of the day’s protein in one large midday meal (“pulse feeding”) produced better protein retention than spreading it over four meals. In a large Canadian cohort, older adults who ate protein more evenly across meals had more lean mass, but they didn’t lose muscle any more slowly over two years.
Mike’s practical point is about real life. Appetite declines with age, and it’s hard to meet protein needs in one sitting. His own father was getting “all his protein” in one meal, and Mike nudged him to split it up. For most older adults, the bigger win is simply eating enough protein in total, at whatever meals make that possible. As Mike likes to say: just because you can do something doesn’t mean you should.
The one caveat: you are an n of one. If you know you sleep poorly after eating, find another time to get that protein in.
Collagen, creatine, and ketones: what Mike takes
Collagen changed his mind. Mike once considered collagen “the worst type of protein I’ve ever seen,” and for muscle it’s incomplete. But his lab ran what he believes is the longest collagen trial to date: up to nine months, comparing 0, 10, and 20 grams per day in active middle-aged adults, many with joint pain. Surprisingly, 10 grams outperformed 20 on most outcomes, improving daily function and joint pain. The pain benefit showed up mainly in people exercising more than about 180 minutes per week. The trial was small, and broader guidelines still consider the evidence on collagen for joint health limited. Mike now takes 10 grams daily.
Creatine is another staple. Mike takes 10 grams a day, partly with brain health in mind, though he’s quick to add, “I don’t know if it’s working.”
Ketones? Not yet. Exogenous ketones haven’t consistently improved performance. Mike’s lab is testing whether they help decision-making in SWAT officers after exhausting exercise in full gear, and whether they influence erythropoietin in elite runners. Until the data are in, he’s not taking them.
He skips fish oil because he eats plenty of fish, and gets his vitamin D from the Florida sun. A short, evidence-based list.
The bottom line
- A small, protein-dominant snack 30 minutes before bed appears to do no harm and may help if you’re active, especially if you train in the evening, are an athlete, or are an older adult who under-eats protein. If you’re sedentary and have obesity, pair it with exercise. A large mixed meal right before bed is a different story.
- Whey or casein, shake or cottage cheese: choose what you enjoy.
- Exercise is your metabolic apology. It buffers much of the damage from overeating and poor sleep.
- Resistance training belongs in every fat-loss plan. It builds strength and protects muscle during rapid weight loss, including on GLP-1s. Pair it with aerobic or interval work for the best results on visceral fat.
- Eat enough protein as you age, at the meals that work for you, and don’t fear higher intakes.
When I asked Mike for his single best piece of advice for a postmenopausal woman looking to change her body composition, he didn’t hesitate: resistance training. In his words, endurance training alone tends to make you “a smaller version of that same shape,” while lifting “can completely change the entire shape of that body.” Aerobic work still pulls its weight, especially for visceral fat, but if you’re only going to start one thing, Mike would start with the weights.
You can follow Dr. Mike Ormsbee on Instagram at @mikeormsbee for behind-the-scenes looks at his research. His Great Courses series, Changing Body Composition Through Diet and Exercise, is available on Amazon and is written for everyday people, not scientists.
Listen to the full episode for the complete conversation.














