FitProven
Evidence Breakdown
Based on 5 studies
THE BIGGER QUESTION
Whether this one statement holds is settled above. What to actually do about it is a wider question, weighed across every claim that bears on it.
- I'm in a calorie deficit but not losing weight — why? — this claim is one of 2 weighed there
Evidence map
For & against, at a glance
Meta-Analysis
Pro
Murphy C & Koehler K · 2022Scandinavian Journal of Medicine & Science in SportsMeta-analysis and meta-regression of resistance-training studies performed in energy deficit, testing whether deficit magnitude predicts lean-mass outcomes. It found a dose-response relationship: each additional 100 kcal/day of deficit reduced the effect size for lean mass by ~0.031 units, and a deficit of roughly 500 kcal/day was the point at which gains in lean mass were prevented altogether. Strength gains, by contrast, were *not* impaired by energy deficiency — an important dissociation, since lifters often use maintained strength as evidence that muscle is being preserved. This is the strongest quantitative support for the claim, and it supplies the widely cited ~500 kcal/day ceiling.
0.75
Meta-analysis and meta-regression of resistance-training studies performed in energy deficit, testing whether deficit magnitude predicts lean-mass outcomes. It found a dose-response relationship: each additional 100 kcal/day of deficit reduced the effect size for lean mass by ~0.031 units, and a deficit of roughly 500 kcal/day was the point at which gains in lean mass were prevented altogether. Strength gains, by contrast, were *not* impaired by energy deficiency — an important dissociation, since lifters often use maintained strength as evidence that muscle is being preserved. This is the strongest quantitative support for the claim, and it supplies the widely cited ~500 kcal/day ceiling.
Design Meta-Analysis (1.0) × quality 0.75 = impact 0.75
View sourceSystematic Review
Pro
Chaston TB et al. · 2007International Journal of ObesitySystematic review pooling 55 cohorts (26 dietary/behavioural, 29 bariatric surgery) to quantify what fraction of weight lost is fat-free mass (%FFML). The degree of caloric restriction was positively associated with %FFML (r² = 0.31, p = 0.006) — the harder the restriction, the greater the share of the loss coming from lean tissue. Three RCTs within the review showed exercise reduced %FFML. The association is observational and across heterogeneous cohorts rather than a randomised comparison of deficit sizes, and severe-restriction cohorts also tended to differ in protein intake and training, which the correlation cannot disentangle.
0.45
Systematic review pooling 55 cohorts (26 dietary/behavioural, 29 bariatric surgery) to quantify what fraction of weight lost is fat-free mass (%FFML). The degree of caloric restriction was positively associated with %FFML (r² = 0.31, p = 0.006) — the harder the restriction, the greater the share of the loss coming from lean tissue. Three RCTs within the review showed exercise reduced %FFML. The association is observational and across heterogeneous cohorts rather than a randomised comparison of deficit sizes, and severe-restriction cohorts also tended to differ in protein intake and training, which the correlation cannot disentangle.
Design Systematic Review (0.9) × quality 0.50 = impact 0.45
View sourceRCT
Pro
Garthe I et al. · 2011International Journal of Sport Nutrition and Exercise Metabolism24 elite athletes were randomised to lose weight slowly (0.7% of body weight/week, ~8.5 weeks) or fast (1.4%/week, ~5.3 weeks), both while strength training and with matched protein intake. Both groups reached the same target weight, but body composition diverged: the slow group *gained* lean body mass (+2.1 ± 0.4%) while the fast group did not (−0.2 ± 0.7%, between-group p < .01), and the slow group lost more fat (−31 ± 3% vs −21 ± 4%). The aggressive deficit did not strip large amounts of muscle outright — it blocked the lean-mass gain that the moderate deficit allowed. Small sample of already-lean, well-trained athletes, so it may not transfer to untrained or higher-body-fat dieters.
0.44
24 elite athletes were randomised to lose weight slowly (0.7% of body weight/week, ~8.5 weeks) or fast (1.4%/week, ~5.3 weeks), both while strength training and with matched protein intake. Both groups reached the same target weight, but body composition diverged: the slow group *gained* lean body mass (+2.1 ± 0.4%) while the fast group did not (−0.2 ± 0.7%, between-group p < .01), and the slow group lost more fat (−31 ± 3% vs −21 ± 4%). The aggressive deficit did not strip large amounts of muscle outright — it blocked the lean-mass gain that the moderate deficit allowed. Small sample of already-lean, well-trained athletes, so it may not transfer to untrained or higher-body-fat dieters.
Design RCT (0.8) × quality 0.55 = impact 0.44
View sourceMeta-Analysis
Con
Ashtary-Larky D et al. · 2020British Journal of NutritionSystematic review and meta-analysis directly comparing gradual with rapid weight loss in people with overweight and obesity. Pooled across trials there was **no significant difference in fat-free mass** between gradual and rapid weight loss — the central prediction of the claim failed in the highest-weight evidence available. Gradual loss did produce greater reductions in fat mass (−1 kg; 95% CI −1.70, −0.29) and body-fat percentage (−0.83%; 95% CI −1.49, −0.17), and better preserved RMR (by ~407 kJ/day, ~97 kcal/day). So rate of loss affects fat loss and metabolic rate, but the specific muscle-sparing claim was not supported in this population.
0.65
Systematic review and meta-analysis directly comparing gradual with rapid weight loss in people with overweight and obesity. Pooled across trials there was **no significant difference in fat-free mass** between gradual and rapid weight loss — the central prediction of the claim failed in the highest-weight evidence available. Gradual loss did produce greater reductions in fat mass (−1 kg; 95% CI −1.70, −0.29) and body-fat percentage (−0.83%; 95% CI −1.49, −0.17), and better preserved RMR (by ~407 kJ/day, ~97 kcal/day). So rate of loss affects fat loss and metabolic rate, but the specific muscle-sparing claim was not supported in this population.
Design Meta-Analysis (1.0) × quality 0.65 = impact 0.65
View sourceRCT
Con
Longland TM et al. · 2016American Journal of Clinical Nutrition40 young men with overweight were randomised to a high-protein (2.4 g/kg/day) or lower-protein (1.2 g/kg/day) diet during a severe ~40% energy deficit, combined with 6 days/week of resistance and anaerobic training, for 4 weeks. Despite the aggressive deficit, the high-protein group *gained* lean mass (+1.2 ± 1.0 kg) and lost more fat (−4.8 ± 1.6 kg) than the lower-protein group (+0.1 ± 1.0 kg lean; −3.5 ± 1.4 kg fat). This is a direct counterexample to the claim as stated: a very large deficit did not cost muscle at all when protein and training were high, indicating deficit size is not independently decisive. Caveats: only 4 weeks, and both groups' protein was at or above typical recommendations.
0.52
40 young men with overweight were randomised to a high-protein (2.4 g/kg/day) or lower-protein (1.2 g/kg/day) diet during a severe ~40% energy deficit, combined with 6 days/week of resistance and anaerobic training, for 4 weeks. Despite the aggressive deficit, the high-protein group *gained* lean mass (+1.2 ± 1.0 kg) and lost more fat (−4.8 ± 1.6 kg) than the lower-protein group (+0.1 ± 1.0 kg lean; −3.5 ± 1.4 kg fat). This is a direct counterexample to the claim as stated: a very large deficit did not cost muscle at all when protein and training were high, indicating deficit size is not independently decisive. Caveats: only 4 weeks, and both groups' protein was at or above typical recommendations.
Design RCT (0.8) × quality 0.65 = impact 0.52
View sourceMeta-Analysis
Pro
Murphy C & Koehler K · 2022Scandinavian Journal of Medicine & Science in SportsMeta-analysis and meta-regression of resistance-training studies performed in energy deficit, testing whether deficit magnitude predicts lean-mass outcomes. It found a dose-response relationship: each additional 100 kcal/day of deficit reduced the effect size for lean mass by ~0.031 units, and a deficit of roughly 500 kcal/day was the point at which gains in lean mass were prevented altogether. Strength gains, by contrast, were *not* impaired by energy deficiency — an important dissociation, since lifters often use maintained strength as evidence that muscle is being preserved. This is the strongest quantitative support for the claim, and it supplies the widely cited ~500 kcal/day ceiling.
0.75
Meta-analysis and meta-regression of resistance-training studies performed in energy deficit, testing whether deficit magnitude predicts lean-mass outcomes. It found a dose-response relationship: each additional 100 kcal/day of deficit reduced the effect size for lean mass by ~0.031 units, and a deficit of roughly 500 kcal/day was the point at which gains in lean mass were prevented altogether. Strength gains, by contrast, were *not* impaired by energy deficiency — an important dissociation, since lifters often use maintained strength as evidence that muscle is being preserved. This is the strongest quantitative support for the claim, and it supplies the widely cited ~500 kcal/day ceiling.
Design Meta-Analysis (1.0) × quality 0.75 = impact 0.75
View sourceSystematic Review
Pro
Chaston TB et al. · 2007International Journal of ObesitySystematic review pooling 55 cohorts (26 dietary/behavioural, 29 bariatric surgery) to quantify what fraction of weight lost is fat-free mass (%FFML). The degree of caloric restriction was positively associated with %FFML (r² = 0.31, p = 0.006) — the harder the restriction, the greater the share of the loss coming from lean tissue. Three RCTs within the review showed exercise reduced %FFML. The association is observational and across heterogeneous cohorts rather than a randomised comparison of deficit sizes, and severe-restriction cohorts also tended to differ in protein intake and training, which the correlation cannot disentangle.
0.45
Systematic review pooling 55 cohorts (26 dietary/behavioural, 29 bariatric surgery) to quantify what fraction of weight lost is fat-free mass (%FFML). The degree of caloric restriction was positively associated with %FFML (r² = 0.31, p = 0.006) — the harder the restriction, the greater the share of the loss coming from lean tissue. Three RCTs within the review showed exercise reduced %FFML. The association is observational and across heterogeneous cohorts rather than a randomised comparison of deficit sizes, and severe-restriction cohorts also tended to differ in protein intake and training, which the correlation cannot disentangle.
Design Systematic Review (0.9) × quality 0.50 = impact 0.45
View sourceRCT
Pro
Garthe I et al. · 2011International Journal of Sport Nutrition and Exercise Metabolism24 elite athletes were randomised to lose weight slowly (0.7% of body weight/week, ~8.5 weeks) or fast (1.4%/week, ~5.3 weeks), both while strength training and with matched protein intake. Both groups reached the same target weight, but body composition diverged: the slow group *gained* lean body mass (+2.1 ± 0.4%) while the fast group did not (−0.2 ± 0.7%, between-group p < .01), and the slow group lost more fat (−31 ± 3% vs −21 ± 4%). The aggressive deficit did not strip large amounts of muscle outright — it blocked the lean-mass gain that the moderate deficit allowed. Small sample of already-lean, well-trained athletes, so it may not transfer to untrained or higher-body-fat dieters.
0.44
24 elite athletes were randomised to lose weight slowly (0.7% of body weight/week, ~8.5 weeks) or fast (1.4%/week, ~5.3 weeks), both while strength training and with matched protein intake. Both groups reached the same target weight, but body composition diverged: the slow group *gained* lean body mass (+2.1 ± 0.4%) while the fast group did not (−0.2 ± 0.7%, between-group p < .01), and the slow group lost more fat (−31 ± 3% vs −21 ± 4%). The aggressive deficit did not strip large amounts of muscle outright — it blocked the lean-mass gain that the moderate deficit allowed. Small sample of already-lean, well-trained athletes, so it may not transfer to untrained or higher-body-fat dieters.
Design RCT (0.8) × quality 0.55 = impact 0.44
View sourceMeta-Analysis
Con
Ashtary-Larky D et al. · 2020British Journal of NutritionSystematic review and meta-analysis directly comparing gradual with rapid weight loss in people with overweight and obesity. Pooled across trials there was **no significant difference in fat-free mass** between gradual and rapid weight loss — the central prediction of the claim failed in the highest-weight evidence available. Gradual loss did produce greater reductions in fat mass (−1 kg; 95% CI −1.70, −0.29) and body-fat percentage (−0.83%; 95% CI −1.49, −0.17), and better preserved RMR (by ~407 kJ/day, ~97 kcal/day). So rate of loss affects fat loss and metabolic rate, but the specific muscle-sparing claim was not supported in this population.
0.65
Systematic review and meta-analysis directly comparing gradual with rapid weight loss in people with overweight and obesity. Pooled across trials there was **no significant difference in fat-free mass** between gradual and rapid weight loss — the central prediction of the claim failed in the highest-weight evidence available. Gradual loss did produce greater reductions in fat mass (−1 kg; 95% CI −1.70, −0.29) and body-fat percentage (−0.83%; 95% CI −1.49, −0.17), and better preserved RMR (by ~407 kJ/day, ~97 kcal/day). So rate of loss affects fat loss and metabolic rate, but the specific muscle-sparing claim was not supported in this population.
Design Meta-Analysis (1.0) × quality 0.65 = impact 0.65
View sourceRCT
Con
Longland TM et al. · 2016American Journal of Clinical Nutrition40 young men with overweight were randomised to a high-protein (2.4 g/kg/day) or lower-protein (1.2 g/kg/day) diet during a severe ~40% energy deficit, combined with 6 days/week of resistance and anaerobic training, for 4 weeks. Despite the aggressive deficit, the high-protein group *gained* lean mass (+1.2 ± 1.0 kg) and lost more fat (−4.8 ± 1.6 kg) than the lower-protein group (+0.1 ± 1.0 kg lean; −3.5 ± 1.4 kg fat). This is a direct counterexample to the claim as stated: a very large deficit did not cost muscle at all when protein and training were high, indicating deficit size is not independently decisive. Caveats: only 4 weeks, and both groups' protein was at or above typical recommendations.
0.52
40 young men with overweight were randomised to a high-protein (2.4 g/kg/day) or lower-protein (1.2 g/kg/day) diet during a severe ~40% energy deficit, combined with 6 days/week of resistance and anaerobic training, for 4 weeks. Despite the aggressive deficit, the high-protein group *gained* lean mass (+1.2 ± 1.0 kg) and lost more fat (−4.8 ± 1.6 kg) than the lower-protein group (+0.1 ± 1.0 kg lean; −3.5 ± 1.4 kg fat). This is a direct counterexample to the claim as stated: a very large deficit did not cost muscle at all when protein and training were high, indicating deficit size is not independently decisive. Caveats: only 4 weeks, and both groups' protein was at or above typical recommendations.
Design RCT (0.8) × quality 0.65 = impact 0.52
View sourceTap any node to expand its detail.
Evidence
PRO (3)
PRO Meta-Analysis0.75 Murphy C, Koehler K (2022)
Lean-mass gains impaired in energy deficit (ES -0.57, p = 0.02) while strength gains were not (ES -0.31, p = 0.28); each additional 100 kcal/day of deficit cost ~0.031 effect-size units, and ~500 kcal/day prevented lean-mass gain entirely
Meta-analysis and meta-regression of resistance-training studies performed in energy deficit, testing whether deficit magnitude predicts lean-mass outcomes. It found a dose-response relationship: each additional 100 kcal/day of deficit reduced the effect size for lean mass by ~0.031 units, and a deficit of roughly 500 kcal/day was the point at which gains in lean mass were prevented altogether. Strength gains, by contrast, were not impaired by energy deficiency — an important dissociation, since lifters often use maintained strength as evidence that muscle is being preserved. This is the strongest quantitative support for the claim, and it supplies the widely cited ~500 kcal/day ceiling.
Weighted 0.75 — The pooled participant total is not reported in the accessible record, and the underlying RCT literature on resistance training performed in a deficit is small, so the widely quoted ~500 kcal/day ceiling is an extrapolation from a meta-regression rather than a directly tested dose. Design and analysis are otherwise sound and the lean-mass/strength dissociation replicates across both of the authors' analyses. Funding could not be verified (paywalled).
Scandinavian Journal of Medicine & Science in Sports
PRO RCTn=240.55 Garthe I, Raastad T et al. (2011)
Slow loss (0.7%/wk) gained lean mass +2.1%; fast loss (1.4%/wk) did not (-0.2%, p<0.01); fat loss -31% vs -21%
24 elite athletes were randomised to lose weight slowly (0.7% of body weight/week, ~8.5 weeks) or fast (1.4%/week, ~5.3 weeks), both while strength training and with matched protein intake. Both groups reached the same target weight, but body composition diverged: the slow group gained lean body mass (+2.1 ± 0.4%) while the fast group did not (−0.2 ± 0.7%, between-group p < .01), and the slow group lost more fat (−31 ± 3% vs −21 ± 4%). The aggressive deficit did not strip large amounts of muscle outright — it blocked the lean-mass gain that the moderate deficit allowed. Small sample of already-lean, well-trained athletes, so it may not transfer to untrained or higher-body-fat dieters.
Weighted 0.55 — 36 athletes were recruited, 30 completed and only 24 were analysed after a subgroup of 6 female athletes was excluded post hoc - a small final sample and a departure from intention-to-treat. Participants were elite, already-lean athletes, so the result transfers poorly to untrained or higher-body-fat dieters. Independently funded.
Funding: Norwegian Olympic Sports Center and Norwegian School of Sport Sciences
International Journal of Sport Nutrition and Exercise Metabolism
PRO Systematic Review0.50 Chaston TB, Dixon JB et al. (2007)
The degree of caloric restriction was positively associated with the share of weight lost as fat-free mass (r2 = 0.31, p = 0.006); 3 RCTs within the review showed exercise reduces %FFML
Systematic review pooling 55 cohorts (26 dietary/behavioural, 29 bariatric surgery) to quantify what fraction of weight lost is fat-free mass (%FFML). The degree of caloric restriction was positively associated with %FFML (r² = 0.31, p = 0.006) — the harder the restriction, the greater the share of the loss coming from lean tissue. Three RCTs within the review showed exercise reduced %FFML. The association is observational and across heterogeneous cohorts rather than a randomised comparison of deficit sizes, and severe-restriction cohorts also tended to differ in protein intake and training, which the correlation cannot disentangle.
Weighted 0.50 — k=55 cohorts (26 dietary/behavioural, 29 surgical); the review reports participants per arm rather than one author-stated pooled total, so the field is omitted. The core finding is a cross-cohort CORRELATION, not a randomised comparison of deficit sizes, and severe-restriction cohorts also differed in protein intake and training — confounds the correlation cannot separate. No funding or conflict-of-interest statement could be retrieved from the paywalled paper. Heavily discounted despite the systematic-review base weight.
International Journal of Obesity
AGAINST (2)
AGAINST Meta-Analysisn=3610.65 Ashtary-Larky D, Bagheri R et al. (2020)
No significant difference in fat-free mass between gradual and rapid weight loss; gradual loss produced greater fat-mass loss (-1 kg; 95% CI -1.70, -0.29), greater body-fat reduction (-0.83%; 95% CI -1.49, -0.17) and better preserved RMR (about 407 kJ/day, roughly 97 kcal/day)
Systematic review and meta-analysis directly comparing gradual with rapid weight loss in people with overweight and obesity. Pooled across trials there was no significant difference in fat-free mass between gradual and rapid weight loss — the central prediction of the claim failed in the highest-weight evidence available. Gradual loss did produce greater reductions in fat mass (−1 kg; 95% CI −1.70, −0.29) and body-fat percentage (−0.83%; 95% CI −1.49, −0.17), and better preserved RMR (by ~407 kJ/day, ~97 kcal/day). So rate of loss affects fat loss and metabolic rate, but the specific muscle-sparing claim was not supported in this population.
Weighted 0.65 — k=7 trials, 361 participants pooled (167 gradual, 194 rapid). Explicitly reports that no financial support was provided and no conflicts of interest — an independent synthesis directly on the claim. Weight held at 0.65 because seven small, heterogeneous trials is a thin evidence base for a null on fat-free mass; the non-difference could reflect low power as much as no effect.
Funding: none declared
British Journal of Nutrition
AGAINST RCTn=400.65 Longland TM, Oikawa SY et al. (2016)
high-protein group gained lean mass (+1.2 kg) and lost more fat (-4.8 kg) vs +0.1 kg lean and -3.5 kg fat on lower protein, under a ~40% energy deficit
40 young men with overweight were randomised to a high-protein (2.4 g/kg/day) or lower-protein (1.2 g/kg/day) diet during a severe ~40% energy deficit, combined with 6 days/week of resistance and anaerobic training, for 4 weeks. Despite the aggressive deficit, the high-protein group gained lean mass (+1.2 ± 1.0 kg) and lost more fat (−4.8 ± 1.6 kg) than the lower-protein group (+0.1 ± 1.0 kg lean; −3.5 ± 1.4 kg fat). This is a direct counterexample to the claim as stated: a very large deficit did not cost muscle at all when protein and training were high, indicating deficit size is not independently decisive. Caveats: only 4 weeks, and both groups' protein was at or above typical recommendations.
Weighted 0.65 — 40 men analysed (20 per group, no dropouts reported), but only 4 weeks long and single-blind by the authors' own admission. Carbohydrate was held constant, so fat intake also differed between arms — the authors concede it cannot be stated conclusively that protein was responsible for the effects, and call it a proof-of-principle trial. Publicly funded, though the senior author discloses dairy-industry support and the intervention drinks were dairy-based.
Funding: Natural Sciences and Engineering Research Council of Canada
American Journal of Clinical Nutrition