The fear behind this question is specific and reasonable: that you will do everything right, lose 20 pounds, and end up a smaller, softer, weaker version of yourself. It is reasonable because it does happen. It is also largely preventable, and the things that prevent it are not the things being advertised to you.
Four levers decide how much muscle survives a calorie deficit, and they are not equally weighted. Below is what the trials show for each, in the order they matter, with the supplement aisle arriving where it belongs: last, and short.
The short answer
- Lift, and everything else gets easier. Adding resistance training to a calorie deficit prevented about 93 percent of the muscle loss in obese older adults.
- Eat about 1.6 g of protein per kg of body weight, more if you are already lean. Above the RDA, extra protein saved about 0.36 kg of lean mass while dieting, and 0.65 kg in the two comparisons that combined dieting with lifting.
- Lose at 0.5 to 1 percent of body weight a week. The pace matters most for people who are already lean.
- Sleep is a body-composition variable. Cutting sleep to 5.5 hours on the same diet shifted the loss away from fat and toward lean tissue.
- Most of the early "muscle loss" on your scale is water and glycogen, not muscle.
- The supplement list is three items long, and two of them only help by making the first two levers easier.
Lean mass is not the same thing as muscle
Almost every panic about losing muscle starts with a number from a scale, a scan or a smart device, and almost every one of those numbers is reporting something broader than muscle. Fat-free mass, or lean mass, is everything in you that is not fat: muscle, yes, but also bone, organs, the contents of your gut, stored carbohydrate, and a great deal of water.
That matters enormously in the first weeks of a diet, because stored carbohydrate binds water with it. After about 5 percent weight loss, total body water had fallen by 3.2 kg on a 6-day fast, 1.2 kg on a very low calorie diet, and 0.3 kg on a moderate one. All of that lands in the lean-mass column. Researchers have demonstrated the effect deliberately: in 18 well-trained male cyclists, loading muscle glycogen and creatine changed estimated lean body mass by 2 to 3 percent over a few days, a window in which, as the authors put it, minimal change in muscle protein mass is likely. Even a single meal raised DXA lean-soft-tissue estimates by up to 1.7 percent on average, and more than 4.5 percent in some individuals.
Two further things drift during weight loss in ways that fool the machines. The water content of your fat-free mass itself rises, from about 74 percent to 77 percent in one study, which breaks the fixed-hydration assumption that bioimpedance devices depend on. And the fat-free mass you do lose is not all muscle: it includes the shrinking support tissue that came with the fat.
How much you will actually lose
The rule everyone repeats is that a quarter of the weight you lose will be fat-free mass. It is a useful rough anchor and a poor constant. The review that examined it in detail describes it as a widely cited rule of thumb rather than a measured law, and when 43 studies were pooled in which 2,379 people lost at least 5 percent of their body weight and then regained at least 2 percent of it, they had lost 10.9 percent of body weight over 13 weeks, of which 19.6 percent was fat-free mass. That sample was selected on regain, so read it as the figure for people who did not keep the weight off rather than as a population average.
The number moves with three things, and only the third is under your control on any given day.
How much fat you started with. There is an inverse curvilinear relationship between starting body fat and the share of weight lost as lean tissue: the leaner you are, the more of each pound comes from lean tissue, and the more careful you have to be. This is the single biggest reason generic advice fails. Someone at 35 percent body fat and someone at 12 percent are not running the same experiment.
How hard you cut. Across 26 cohorts, the degree of caloric restriction was positively associated with the percentage of weight lost as fat-free mass. In lean young men put on a 50 percent cut for three weeks, fat-free mass came off faster than fat did, at 159 grams a day against 114.
Whether you train and eat enough protein. Among middle-aged and older dieters, 81 percent of diet-only groups lost 15 percent or more of their body weight as fat-free mass. Among those who dieted and exercised, 39 percent did. That is the whole argument for the next two sections in a single comparison.
Lever 1: protein, and the number that is really supported
Protein is the most discussed lever and the most exaggerated. It works, the mechanism is sound, and the size of the effect is smaller than the internet implies.
The cleanest summary comes from a meta-analysis of 18 studies that compared protein above the RDA with protein at the RDA. Eating more preserved about 0.32 kg of lean mass overall and 0.36 kg during energy restriction specifically, rising to 0.65 kg in the two comparisons that combined a deficit with resistance training, while making no measurable difference at all in people who were neither dieting nor training. That last detail is the tell: protein is not a general tonic, it is insurance against a specific stress.
| Body weight | 1.6 g/kg most people | 2.2 g/kg already lean, or cutting hard | Per meal four meals at 1.6 g/kg |
|---|---|---|---|
| 150 lb (68 kg) | 109 g | 150 g | about 27 g |
| 180 lb (82 kg) | 131 g | 180 g | about 33 g |
| 200 lb (91 kg) | 145 g | 200 g | about 36 g |
| 220 lb (100 kg) | 160 g | 220 g | about 40 g |
| 250 lb (113 kg) | 181 g | 249 g | about 45 g |
Targets are per kilogram of body weight. If you have a lot of fat to lose, the per-kilogram figures overshoot: base them on your goal weight instead. Our protein guide covers the general targets in full.
The dramatic trials are real, but read their populations. Young men in a 40 percent deficit training six days a week gained 1.2 kg of lean mass on 2.4 g/kg against 0.1 kg on 1.2 g/kg, and lost more fat besides. Trained athletes cutting hard for two weeks lost 0.3 kg of lean mass on 2.3 g/kg against 1.6 kg on 1.0 g/kg. Both are studies of lean, hard-training people in aggressive deficits, which is the situation where protein matters most.
Push past that and the returns stop. In 39 adults held at a 40 percent deficit for three weeks, the step from 0.8 to 1.6 g/kg protected fat-free mass, while going on to 2.4 g/kg added nothing further, a null drawn from those 39 people split across the three intakes. In 21 recreational athletes lifting three times a week in a 25 percent deficit, 1.2, 1.6 and 2.2 g/kg produced the same result: every group gained fat-free mass and the protein dose made no difference. And in 61 overweight older adults on a 12-week 25 percent deficit, raising protein from 0.9 to 1.7 g/kg did not preserve lean mass, leg strength or physical performance.
Timing and distribution matter far less than the daily total. A meta-analysis of protein timing found that once total intake was accounted for, the timing effect disappeared, and in a randomized trial of women losing weight, spreading protein evenly across meals versus loading it at dinner produced no difference in body composition over 16 weeks. Even spacing is a reasonable default, not a requirement.
Lever 2: lifting, which does more than everything else combined
If you do one thing on this page, do this one. Resistance training during a calorie deficit is the only intervention here with large, repeatable effects, and it is the reason the protein numbers above look better when training is included.
In a meta-analysis of six randomized trials in obese older adults, adding resistance training three times a week to a calorie-restricted diet prevented 93.5 percent of the lean body mass that dieting alone would have cost. The diet-only groups lost 0.76 kg of lean mass; the groups that also lifted finished 0.05 kg up, essentially unchanged. Fat loss was the same in both. You are not trading fat loss for muscle retention, you are getting the second one free.
The type of exercise decides the outcome. In a 6-month trial of 160 obese older adults who all lost about 9 percent of their body weight, lean mass fell 5 percent with aerobic exercise, 3 percent with both, and 2 percent with resistance training, and strength rose 18 to 19 percent in the two lifting groups against 4 percent with aerobic exercise alone. A one-year trial in 107 obese adults aged 65 and older points the same way for training in general: dieters assigned to supervised sessions of aerobic work, resistance training and balance work three times a week lost 3 percent of lean mass against 5 percent for diet alone, and a third as much hip bone density. That trial did not compare one type of exercise with another, so it argues for training during a deficit rather than for lifting over cardio.
Two honest caveats. First, a deficit does blunt what lifting can build: pooled across trials, lean-mass gains were impaired in people training in an energy deficit, although strength gains were not. Second, the same analysis put a number on it: past a deficit of about 500 calories a day, lifting stopped adding lean mass, and the authors' own advice is that anyone training to preserve muscle during weight loss keep the deficit under that. That argues for the slower end of the rate table in the next section rather than the top of it.
Lever 3: the pace of loss
Rate of loss is usually given in pounds, which makes it useless: two pounds a week is cautious at 300 pounds and reckless at 130. Use a percentage of body weight instead.
| Body weight | 0.5% a week conservative | 0.75% a week the common middle | 1% a week the upper end |
|---|---|---|---|
| 150 lb (68 kg) | 0.8 lb | 1.1 lb | 1.5 lb |
| 180 lb (82 kg) | 0.9 lb | 1.4 lb | 1.8 lb |
| 200 lb (91 kg) | 1.0 lb | 1.5 lb | 2.0 lb |
| 220 lb (100 kg) | 1.1 lb | 1.6 lb | 2.2 lb |
| 250 lb (113 kg) | 1.2 lb | 1.9 lb | 2.5 lb |
Most people do well between 0.5 and 1 percent a week. Lower is not automatically better: a deficit so small that you abandon it is worse than one you keep. If picking a weekly number is already making you anxious, or you have any history of disordered eating, read the box at the top of this guide again before you go further.
In the one randomized trial to put two rates head to head, 24 elite athletes who all lifted four times a week were assigned to lose weight slowly or fast, targeting 0.7 or 1.4 percent of body weight a week. In practice the slow group averaged 0.7 percent a week and gained 2.1 percent in lean mass, while the fast group averaged 1.0 percent a week and finished with lean mass unchanged. Both lost the same total weight, about 5.5 percent of body weight; the slower group simply took longer and kept more. A meta-analysis of trials in people with overweight or obesity, comparing gradual with rapid weight loss at matched totals, found gradual dieting produced about 1 kg more fat loss and better preserved resting metabolic rate, though fat-free mass itself did not differ significantly between the two paces. And when people with overweight lost the same weight on a very low calorie diet versus a moderate one, fat-free mass was 8.8 percent of the loss in the crash group against 1.3 percent in the moderate one.
The counterweight, and the reason this is a lever rather than a law: in 35 people with obesity randomized to lose the same weight rapidly over 4 weeks or gradually over 8, changes in body weight and composition were similar in both groups. The pace penalty is real and sharpest in lean, trained people. If you have a substantial amount of fat to lose, it is a smaller effect than the internet suggests, and adherence matters more.
One structural option worth knowing: 51 men with obesity were randomized to 16 weeks of dieting either straight through or split into eight 2-week blocks separated by 2-week breaks at maintenance, which stretched that arm to 30 weeks on the calendar. Among the men who finished per protocol, the intermittent group lost 14.1 kg against 9.1 kg for continuous dieting, with similar fat-free mass loss. More weight came off, but it took nearly twice as long. That is a single trial in one population, so treat diet breaks as a legitimate option rather than a requirement.
Lever 4: sleep, the one nobody sells you
This is the lever missing from nearly every article on this subject, and the evidence for it is better than the evidence for most of the supplement shelf.
Ten overweight adults ate the same reduced-calorie diet twice, once with 8.5 hours in bed and once with 5.5. They lost the same total weight both times. What changed was where it came from: fat loss fell by 55 percent (1.4 kg against 0.6 kg) and fat-free mass loss rose by 60 percent (1.5 kg against 2.4 kg). Same diet, same weight on the scale, opposite body composition. That is one 14-day crossover in ten people. A larger randomized trial, 36 adults with overweight or obesity dieting for 8 weeks, cut sleep more gently, by about an hour on five nights a week, and found the two groups lost similar amounts of weight, fat mass and lean mass, with only the proportion of the loss coming from fat favoring the normal-sleep group. Read sleep as a lever worth pulling because it is free and it moves appetite, not because the 1.5 against 2.4 kg figure has been reproduced.
The mechanisms are documented separately. Five nights of four hours in bed lowered the rate at which young men built new muscle protein, from 1.53 to 1.24 percent a day across three groups of eight, although a third group on the same four hours who did three high-intensity interval sessions held that rate at normal-sleep levels. A single night without sleep reduced muscle protein synthesis by 18 percent while raising cortisol. Short sleep also makes the diet harder to keep: pooled across intervention studies, partial sleep deprivation increased next-day energy intake by 385 calories without increasing energy expenditure, and two nights of short sleep were enough to drop leptin 18 percent and raise ghrelin. It degrades the training half too, with exercise performance falling about 7.6 percent after sleep loss.
It runs the other way as well. Coaching habitually short sleepers to extend their time in bed added about 1.2 hours a night and reduced their energy intake without any diet instruction at all.
Does cardio burn muscle?
Not by itself, and the myth does real damage by talking people out of the activity that makes their deficit sustainable. What the evidence actually shows is that cardio is the weaker tool for protecting muscle, not a destroyer of it.
The head-to-head data are in the training section above: with the same weight loss, lean mass fell 5 percent with aerobic exercise against 2 percent with lifting. A smaller 12-week trial points the same way: 20 adults, 17 of them women, ate an 800 calorie liquid diet and either lifted three days a week or did an hour of cardio four times a week. Lean body weight fell in the cardio group, from 51 kg to 47 kg, and did not fall in the lifting group. Twenty people is a thin basis for the exact numbers, but the direction matches the larger trials above. Cardio during a deficit is not neutral, but the problem is the absence of a lifting stimulus rather than the presence of cardio.
The practical answer is to keep the walking and the cycling, which help your deficit and your heart, and to stop treating them as a substitute for resistance training.
How to tell whether you are losing muscle
This is where most people go wrong, because the tools they reach for cannot resolve the change they are worried about.
| Method | Smallest change it can honestly detect | What that means |
|---|---|---|
| Your gym log | A sustained drop in the weight or reps of the same top set | The most sensitive tool you own, and free. Strength holds up better than muscle does in a deficit, so a genuine multi-week decline is a real signal. |
| DXA, two scans the same day | A lean-mass change of about 0.6 kg | The best case, and it still cannot see a month of careful dieting. |
| DXA, scans on different days | A lean-mass change of about 2.1 kg | Which is the situation you are actually in. The error more than triples once the scans are days apart. |
| Bioimpedance scale or InBody | A change of roughly 4 kg in either direction | For tracking change during weight loss the limits of agreement against DXA ran to plus or minus 3.7 to 4.6 kg. That is larger than all the muscle most people will lose in a year. |
| Skinfold calipers | About 0.6 percentage points of body fat between repeat measurements by the same tester | In the precision study this comes from, 49 adults at a mean BMI of 23.5, calipers were the one method of three graded as having poor absolute reliability, and they underestimated body fat by about 4.5 points against air displacement plethysmography. Use it for direction, not for a level. |
| The mirror and the waistband | Slow, unquantified change | Poor for a month, honest over six. |
Detection limits are from precision studies of each method. A change smaller than the figure in the middle column is indistinguishable from measurement noise.
Read that table again with a realistic number in mind. If you are losing half a kilogram of lean mass a month and doing well, a bioimpedance scale with limits of agreement of several kilograms will tell you nothing, and two DXA scans taken weeks apart can disagree by 2 kg on lean mass alone.
So use the gym instead. Log the load and reps of two or three fixed lifts and watch the trend over weeks. Strength is preserved better than muscle in a deficit, which is exactly what makes a sustained decline meaningful: in pooled trials, an energy deficit impaired lean-mass gains from training while strength gains were comparable to training without a deficit. A bad session means you slept badly. Four weeks of declining top sets, while you are eating enough protein and sleeping, means the deficit is too aggressive.
The supplement layer, in order of evidence
Here is the part the rest of the internet puts first. It goes last because that is where it belongs, and because the honest version is short.
| What | What the best evidence shows | Amount studied | Verdict |
|---|---|---|---|
| Resistance training | Not a supplement, and the largest single effect here. Adding it to a calorie deficit prevented about 93 percent of the muscle loss in obese older adults, and beat cardio head to head in a 6-month trial. | 2 to 3 sessions a week | Do this first |
| Enough protein | Eating above the RDA while dieting preserved about 0.32 kg of lean mass overall, 0.36 kg during energy restriction, and 0.77 kg when resistance training was added. Real, and smaller than the internet suggests. | 1.6 g/kg a day, more if lean | Do this second |
| Sleep | Free, and the lever nobody sells. Cutting sleep to 5.5 hours on the same diet shifted the loss away from fat and toward lean tissue. | 7 to 9 hours | Do this third |
| Protein powder | A convenient way to hit the number above, not a separate intervention. In a meta-analysis in women, whey raised lean mass by 0.90 kg across the six comparisons with energy restriction, but made no significant difference across the seven comparisons that included resistance training. | 20 to 40 g per serving | Worth it if you struggle to hit your target |
| Creatine monohydrate | The best-supported supplement in this area, though most trials were not run in a deficit: added to resistance training it increased lean body mass by 1.14 kg. Part of the early gain is water, not muscle. | 3 to 5 g a day | Worth trying, with training |
| Caffeine | Helps you train hard when calories are low. The effect on strength is real but small (SMD 0.20), and it does nothing directly for muscle retention. | 3 to 6 mg/kg before training | Optional, for training quality |
| Psyllium fiber | Nothing for muscle, and nothing cited in this review. It is here because some people find a soluble fiber dose taken with water blunts hunger enough to keep a moderate deficit going. Treat it as a cheap thing to try rather than a proven effect. | 5 to 10 g in a full glass of fluid | Optional, and uncited |
| HMB, BCAAs, EAAs | Sold hardest as anti-catabolic muscle savers. HMB produced no gain in fat-free mass, no fat loss and no strength benefit in younger adults, and BCAAs cannot build muscle without the other essential amino acids. | Not established | Skip |
| Fat burners, CLA, carnitine, garcinia | Of 315 randomized trials of 14 weight-loss supplements, only 52 were low risk of bias and sufficient to judge efficacy, and just 16 of those 52 showed a significant difference in weight against a comparator. | Not applicable | Skip |
Ordered by the strength of the evidence, not by how heavily each is marketed.
Creatine monohydrate has the strongest case. Added to resistance training it increased lean body mass by 1.14 kg compared with training alone, and reduced body fat percentage slightly. Two honest caveats: most of those trials were not run in a calorie deficit, and creatine increases total body water, so part of the early gain on a scan is water rather than muscle. In a 42-day trial in female collegiate dancers who kept dancing but did no resistance training, the creatine group showed increases in both total body water and DXA lean mass, and the authors attributed the DXA change to the extra fluid, since DXA counts body water inside lean mass. It is a small study, so read it as an illustration of what a scan cannot separate rather than a measurement of how much of a gain is water. The closest thing to a test in a real deficit is a four-day study of male resistance trainers eating 18 calories per kilogram a day. Fat-free mass fell less on creatine, 1.4 percent against 2.4 percent, but urinary nitrogen losses were similar in both groups and the authors concluded creatine did not affect body fat or protein loss. Over four days, a smaller drop in fat-free mass is hard to separate from the water creatine adds. Our creatine buying guide covers forms and testing.
Protein powder is a delivery method, not an intervention. It earns its place only if you are missing your target with food. In a meta-analysis of 13 randomized trials in women, whey supplementation raised lean mass by 0.37 kg overall, and by 0.90 kg across the six comparisons that included energy restriction. The same analysis cuts the other way too: across the seven comparisons in which the women also did resistance training, whey made no significant difference to lean mass (0.23 kg, 95 percent confidence interval minus 0.17 to 0.63). Read together, that is an argument for using a powder to close a protein gap rather than as something to add on top of training. Whether it comes from whey isolate, casein or pea protein matters much less than the daily total.
Caffeine does nothing for muscle retention directly. It helps you train hard when calories are low, which is not nothing. The effect on strength is real but small, with a pooled standardized mean difference of 0.20 across ten one-rep-max studies, and in that analysis it reached significance for upper body lifts (0.21) but not for lower body ones (0.15, 95 percent confidence interval minus 0.05 to 0.34). Being a habitual coffee drinker does not blunt it, so there is no need to cycle off. Doses in trials run 3 to 6 mg/kg about an hour before training, which for a 200 pound adult is roughly 270 to 545 mg. More is not better: in the analysis of habitual intake, doses above 6 mg/kg produced no significant benefit. Caffeine also costs some people their sleep, which is the lever above this one, so keep it away from evening sessions. If you are pregnant or breastfeeding, or you have a heart, blood pressure or anxiety condition, clear added caffeine with your doctor first.
Psyllium fiber is on this list for adherence, not muscle. It does nothing for lean mass and is not pretending to. It is also the one item here we have not backed with a citation. If hunger is the thing that ends your diets, a cheap soluble fiber is a more useful purchase than anything in the next section. Two rules come with it, both straight off the label: at least a full glass of liquid per dose, because dry psyllium can swell and choke, which is why the FDA-mandated notice tells anyone with difficulty swallowing not to take it at all, and a gap of a couple of hours from any other medicine you take by mouth.
The links below are affiliate links. NutraSmarts may earn a commission if you buy through them, at no extra cost to you.
Only if you are missing your protein target with food. A powder is a convenience, not a separate lever.
- Optimum Nutrition Gold Standard 100% Whey24 g per scoop at 120 calories, Informed Choice tested, and the best balance of protein, quality and price in our protein powder guide.
- Nutricost Whey Protein Concentrate25 g per scoop at roughly two thirds the price, if cost is the thing stopping you.
- Garden of Life Sport Organic Plant Protein30 g per serving and NSF Certified for Sport, for anyone avoiding dairy.
Plain monohydrate. The fancier forms cost more and are studied less.
- Thorne CreatineNSF Certified for Sport, which matters if you are drug tested. Creapure monohydrate.
- Optimum Nutrition Micronized CreatineThe value pick: the same 5 g of monohydrate for about 20 cents a serving.
- Nutricost Creatine MonohydrateThird-party tested, unflavored, and about as cheap as creatine gets.
For appetite and regularity while you are eating less, not for muscle. Mix each dose into at least a full glass of water. The label warning is not decorative: taken without enough fluid psyllium can swell and block the throat, so do not take it if you have any difficulty swallowing, and keep it a couple of hours away from other medicines you take by mouth.
- Konsyl Original 100% Psyllium Fiber6 g of psyllium husk per dose with no sweeteners or additives, which our fiber guide puts at about 3 g of soluble fiber.
- Metamucil 4-in-1 Psyllium FiberAbout 3 g of fiber per dose, close to what a Konsyl dose delivers once you account for husk versus fiber, in a flavored powder that is easy to find.
How we make money: the links above are affiliate links. If you buy through one, NutraSmarts may earn a small commission at no extra cost to you. That does not change what we recommend: every pick here is one our own buying guides already reviewed and ranked on published specs, dose, third-party certification and price per serving. We do not lab test products ourselves, and we do not accept payment from brands to feature them. The list is short because the evidence is short. See our affiliate disclosure.
What to skip, and what can actually hurt you
Start with the category-level finding, because it is more useful than any individual verdict. A systematic review screened more than 20,000 citations and found 315 randomized trials covering 14 weight-loss supplements and therapies. Only 52 of those trials, 16.5 percent, were low risk of bias and sufficient to support efficacy. Of those 52, only 16 found a significant difference in weight against a comparator, and the differences ranged from 0.3 to 4.93 kg.
HMB and BCAAs. These are marketed precisely as anti-catabolic muscle savers for dieters, which is why they need naming. A meta-analysis in younger adults found HMB produced no improvement in fat-free mass, fat mass or strength alongside resistance training. BCAAs fail for a reason that is arithmetic rather than opinion: they supply three of the nine essential amino acids, so the muscle they can build is capped by the six they do not provide. We cover that in full in do BCAAs work.
CLA and L-carnitine. Both produce statistically significant, practically trivial effects: about 0.35 kg of body mass for CLA and 1.21 kg for carnitine, with carnitine's effect shrinking the longer people take it. Neither earns a place for muscle, but for different reasons. The carnitine meta-analysis reported body weight, BMI, fat mass, waist circumference and body fat percentage, and never measured fat-free mass at all, so there is no muscle result either way. The CLA meta-analysis did measure it and found a small increase of 0.27 kg (95 percent confidence interval 0.09 to 0.45), alongside the 0.35 kg drop in body mass, while its higher-quality trials showed no fat loss at all. CLA has a further problem: a trial using the purified c9,t11 isomer found it decreased insulin sensitivity by 15 percent in obese men.
Green tea extract and EGCG. The Cochrane review found a mean difference in weight loss of 0.04 kg in studies conducted outside Japan. That is not a small effect, it is no effect. It also carries the clearest safety signal in this section: concentrated extracts have been linked to liver injury, in published case reports, at EGCG intakes from 140 to about 1000 mg a day. The US Pharmacopeia has added a cautionary labeling requirement to its powdered decaffeinated green tea extract monograph. And in a secondary analysis of a 12-month trial in postmenopausal women taking 843 mg of EGCG a day, alanine aminotransferase was up 78 percent from baseline at six months in one genotype subgroup, against 28 percent in the reference genotype.
Garcinia, ketone esters and thermogenic blends. Garcinia cambogia was tested properly, in a real deficit with fat mass measured, and produced no difference against placebo. The human case for ketone esters preserving muscle rests on a single-dose appetite study in 15 people that never measured body composition. And stimulant blends carry a real adulteration problem: analyses have repeatedly found banned or never-tested stimulants in them, and across the FDA's tainted-products database, weight-loss products are one of the two most commonly adulterated categories, most often with sibutramine, a drug withdrawn from the US market for cardiovascular harm.
If you want the longer version of that argument, we wrote it up in do fat burners work.
If you are over 60, postmenopausal, or on a GLP-1
Over 60. The protein floor moves up before you even start dieting: expert consensus recommends at least 1.0 to 1.2 g/kg a day for healthy older adults, and 1.2 to 1.5 g/kg for most older adults dealing with an acute or chronic illness. The same consensus names one exception, and it matters at this age: older people with severe kidney disease, meaning an estimated glomerular filtration rate below 30, who are not on dialysis may need to limit protein rather than raise it. If you do not know your kidney numbers, get them from your doctor before you raise your protein intake. Weight loss itself is also not a blanket recommendation at this age, and the joint position of nutrition and obesity societies reserves it for older people whose obesity is causing functional impairment or medical complications. The resistance-training case is strongest here, which we cover in supplements for sarcopenia.
Postmenopausal women. One retrospective analysis of a 20-week diet and exercise trial in postmenopausal women found they ate an average of 0.62 g/kg of protein and lost 32 percent of their total weight as lean mass, well above the 19.6 percent average across weight-loss studies. Within that group, the women who ate more protein lost less lean mass. That is a correlation inside a trial that was not designed to test protein intake, so read it as a signal rather than proof. In this group protein supplements moved lean mass and strength only when paired with resistance training; without training the pooled effect was not beneficial. Lifting is not optional here. Our protein guide for women has the general targets.
On a GLP-1 medication. Pooled across 22 trials, lean mass fell 0.86 kg, about a quarter of total weight lost, and a 2026 review of incretin trials put the median muscle share at 28.3 percent. That is above the roughly 19.6 percent seen across ordinary weight-loss studies, and the one analysis cited here that compares modalities at matched weight loss of 10 percent or more found fat-free mass made up 33.3 percent of the weight lost on incretin therapies against 14.9 percent for diet with or without exercise, with pooled fat-free mass losses of 4.8 kg and 1.8 kg. So the lean-tissue share does look larger on these medications. That is an argument for training and protein alongside the drug, not an argument for stopping one your prescriber put you on: any change to the dose or the medication is their call. It is the same four levers, with protein made harder by reduced appetite. We cover that specific situation in supplements to take with GLP-1 medications and creatine on a GLP-1.
The plan, in order
- Set the pace first. Pick 0.5 to 1 percent of body weight a week from the table above and build the deficit to match. If you are already lean, take the lower end.
- Lift twice a week, minimum. Full body, close to failure, progressing the load. This is the lever, not the accessory.
- Hit the protein number from the table, from food where you can. Spread it across meals if that is easy; do not lose sleep over the distribution.
- Protect the sleep. Seven hours is a body-composition intervention, and it is free.
- Track strength, not scans. Two or three fixed lifts, logged. Re-measure body weight weekly and judge it on the four-week trend.
- Then, and only then, consider the three supplements above, and only the ones that fix a problem you actually have.
Frequently asked questions
Can you lose fat without losing any muscle?
Often yes, and sometimes you can gain a little. Across 43 studies of people who lost at least 5 percent of their weight and then regained at least 2 percent of it, about 19.6 percent of what came off was fat-free mass. Those studies were selected because the weight came back, so it is not a general average. In trials where people lifted weights and ate enough protein, lean mass held steady or rose: obese older adults who added resistance training to a calorie deficit avoided about 93 percent of the muscle loss, and young men in a severe deficit who ate 2.4 grams of protein per kilogram while training hard gained 1.2 kg of lean mass in four weeks. The honest summary is that some lean-tissue loss is normal, most of it is avoidable, and almost none of it is decided by a supplement.
How much protein do I need to keep muscle in a calorie deficit?
About 1.6 grams per kilogram of body weight a day is a sensible target for most people, rising toward 2.2 if you are already lean or cutting aggressively. For a 200 pound adult that is roughly 145 to 200 grams a day. Be careful with the very high numbers you see quoted: the widely repeated 2.3 to 3.1 g/kg figure comes from a review of lean, resistance-trained athletes and is expressed per kilogram of fat-free mass, not body weight, so it does not translate to someone with a lot of fat to lose. If that describes you, base the number on your goal weight rather than your current weight, and ask your doctor for a target if you have kidney disease.
How fast can I lose weight without losing muscle?
Think in percentages, not pounds: roughly 0.5 to 1 percent of body weight a week. In 24 elite athletes who all lifted four times a week, those averaging about 0.7 percent of body weight a week gained 2.1 percent in lean mass, while those averaging about 1.0 percent a week finished with lean mass unchanged. In people carrying more body fat the penalty for going faster is smaller, and one trial that matched total weight loss found no difference in body composition between a 4-week and an 8-week version of the same diet. The leaner you already are, the more the pace matters.
Am I losing muscle or just water and glycogen?
Early in a diet, mostly the latter. Each gram of stored carbohydrate holds roughly three grams of water with it, so the first week's drop is largely fluid: after about 5 percent weight loss, a 6-day fast had shed 3.2 kg of body water, a very low calorie diet 1.2 kg, and a moderate diet only 0.3 kg. That water sits inside what a scan calls lean mass, which is why body composition devices can show muscle disappearing that was never muscle. Manipulating muscle glycogen and creatine alone moved DXA lean mass estimates by 2 to 3 percent in 18 trained male cyclists, over a few days in which actual muscle protein is unlikely to have changed.
Does cardio burn muscle?
Not on its own, but it is the weaker tool for this job. In a 6-month trial of dieting older adults, lean mass fell 5 percent in the group doing aerobic exercise, 3 percent in the group doing both, and 2 percent in the group lifting, and only the lifting groups improved strength meaningfully. In a 12-week trial of 20 adults on an extreme 800 calorie diet, lean body weight fell about 4 kg in the group doing cardio and did not fall in the group lifting. The practical reading is not to avoid cardio, which is good for your heart and your deficit, but not to let it replace resistance training.
What supplements actually help you keep muscle while dieting?
A short list. Protein powder helps only if it gets you to a protein target you were missing. In a meta-analysis of 13 trials in women, whey raised lean mass by 0.90 kg across the six comparisons with energy restriction, and made no significant difference across the seven comparisons in which the women also did resistance training. Creatine monohydrate has the best evidence of any supplement here, adding about 1.14 kg of lean body mass when combined with resistance training, though most of those trials were not run in a deficit and some of the early gain is water rather than muscle. Caffeine helps you train hard on low calories. Everything else sold for this purpose, including HMB, BCAAs, CLA and fat burners, is either unsupported or trivially small: across 315 randomized trials of 14 weight-loss supplements, only 52 were good enough to judge efficacy, and just 16 of those showed a significant difference in weight.
Will I lose muscle on Ozempic or another GLP-1 medication?
Some of what you lose will be lean tissue, as with any large weight loss. Pooled across 22 trials, lean mass fell by 0.86 kg on GLP-1 medications, roughly a quarter of the total weight lost, and a 2026 review of incretin trials put the median muscle share at 28.3 percent. That is above the roughly 19.6 percent average across ordinary weight-loss studies, and the one cited analysis comparing modalities at matched weight loss found fat-free mass was 33.3 percent of the loss on incretin therapies against 14.9 percent for diet with or without exercise, so the lean-tissue share does look larger on these medications. The response is the same as in this guide: resistance training, enough protein, and a conversation with your prescriber rather than stopping a medication on your own.
Should I do a diet break or just keep going?
Both work, and the evidence for breaks is better than it used to be. In one trial, 51 men with obesity dieted for 16 weeks either straight through or in eight 2-week blocks separated by 2-week breaks at maintenance, which stretched that arm to 30 weeks. Among the men who finished per protocol, those who took breaks lost 14.1 kg against 9.1 kg for continuous dieting, with similar fat-free mass loss in both groups, but they took nearly twice as long to do it. That is one trial in one population, so treat it as a reasonable option rather than a requirement. If a break is what stops you abandoning the diet entirely, it is doing its job.
The bottom line
Losing fat without losing muscle is mostly a training and protein problem, partly a pacing and sleep problem, and barely a supplement problem. The single largest number on this page is that resistance training prevented about 93 percent of diet-induced muscle loss. The largest supplement number is roughly a kilogram of lean body mass from creatine alongside training, and part of that is water.
If you lift twice a week, eat around 1.6 grams of protein per kilogram, lose at half a percent to a percent of your body weight weekly and sleep seven hours, you have already captured nearly all of the available benefit. The shopping list exists to make those four things easier, not to replace them, and anything sold to you as a way around them is selling the one thing that is not for sale.