Does collagen help with weight loss?

Possibly a little, and nobody has tested it the way the question needs. The strongest result is a meta-analysis of 12 placebo-controlled trials in which collagen peptides lowered fat mass by 2.7 lb (95% CI 0.64 to 4.7; metric: 1.21 kg, 0.29 to 2.13) and raised fat-free mass by 1.49 percent of body mass. Body weight was not pooled, and many of those trials paired collagen with resistance training. The one trial built around weight loss gave 64 adults with overweight or obesity two collagen bars a day for 12 weeks, had no placebo, and found 6.6 lb (3.0 kg) lost against 3.3 lb (1.5 kg) on diet advice alone.

Unsettled. The answer is open for a plain reason. No meta-analysis we found asks whether collagen makes people lose weight. The pooled trials measure body composition, mostly in people who were training. The single weight trial used a specially processed collagen in a bar and gave the control group nothing to eat, so the bar itself and the extra attention could explain the difference. Another trial, in women with overweight, found the opposite direction.

Kilograms lost, in the two results that point toward collagen

The first bar is fat mass against placebo, with its 95% confidence interval. Body weight was not pooled in that review. The second and third bars come from one open-label trial with no placebo, so the gap between them mixes collagen with extra attention and eating a bar before meals. That trial reports spreads rather than an interval, so no whiskers are drawn. Sources: cards ev-colw-01 and ev-colw-04 below.

What the trials found

Body composition, mostly with training

In the meta-analysis of 12 placebo-controlled trials, fat mass fell by 2.7 lb (1.21 kg) with collagen peptides, and the trials agreed with each other (I2 0%). Its authors note that many of the included trials combined collagen with resistance training, and the trials were short. We read only the abstract, so we could not count how many trials involved exercise.

A second meta-analysis looked only at adults who trained regularly. Across 8 trials in 418 people taking collagen peptides for at least 8 weeks alongside training, fat-free mass rose by a standardized difference of 0.48 (95% CI 0.22 to 0.74), rated moderate certainty. Six of those eight trials measured body composition with bioelectrical impedance. A third review, of 15 randomized trials of collagen with exercise, found only "certain improvements" in body composition and did not pool them. Twelve of its fifteen trials were in recreational athletes. None of these were people trying to lose weight.

Body weight, in people with overweight

In the open-label trial, 64 adults with overweight or obesity all got healthy diet advice. The collagen group also ate two 90 calories bars with bovine collagen, 45 minutes before lunch and dinner, 20 g of collagen a day. After 12 weeks they had lost 6.6 lb (3.0 kg) against 3.3 lb (1.5 kg), which is 3.36 against 1.80 percent of body weight (p = 0.045). There was no placebo bar.

Two trials point the other way. In a double-blind trial of 37 women with overweight, mean BMI 30.9, collagen was compared with whey protein for 8 weeks. BMI rose by 0.2 kg/m2 on collagen, and android fat, the fat around the middle, fell by 0.22 lb (0.1 kg) more on whey. The authors concluded that collagen "may not be an effective supplement" for women with overweight trying to change body composition. That trial had no placebo either. In a placebo-controlled trial of 56 adults with overweight or obesity, 10 g a day of porcine collagen for 4 weeks left weight and body measures unchanged.

Appetite, mixed

The appetite trials are tiny and single-meal. In 15 active women, 15 g a day of collagen peptides for 7 days cut the energy eaten at a meal after exercise by about 10 percent, roughly 41 calories, while hunger ratings did not change. The comparison drink had no calories, so some of that cut may simply be the body making up for the calories in the supplement. In 10 young women, a single 40 g dose of collagen changed neither hunger nor food eaten compared with whey. In 24 adults, a breakfast of gelatin with added tryptophan cut energy eaten at lunch by about 20 percent, 2.5 against 3.2 MJ, compared with casein, soy or whey with part of its protein removed (whey-GMP). A companion study found no difference in lunch intake.

Animals

A meta-analysis of 19 rodent studies, 339 animals on high-fat diets, found collagen peptides lowered body mass (standardized difference -1.87). Its authors rate the certainty very low and say the effect in humans "remains largely unexplored".

Who should be careful

Not for everyone. In the placebo-controlled trial of 56 adults with overweight or obesity, 10 g a day of porcine collagen for 4 weeks slightly raised fasting insulin and HOMA-IR, a measure of insulin resistance. The authors call the changes small, and they were seen in a short trial. If you already have insulin resistance or prediabetes, that is a reason to ask before adding collagen for weight.

Gelatin, a cooked form of collagen, raised urinary oxalate by 43 percent compared with whey in 10 healthy adults eating 1.1 oz (30 g) a day for 5 days. Stone risk itself was not measured, and these were people with no history of stones. The 20 g a day in the collagen bar trial is not far from that amount. If you have had kidney stones, this is the caution that applies to you.

We found no evidence on collagen in pregnancy, with medicines or in allergy. That is a gap in the research, not a sign that it is safe.

What expert bodies say

Nothing, as far as we could find. The US Office of Dietary Supplements has no fact sheet on collagen. The EU register of health claims holds claims for collagen and none of them is about weight or appetite. The US food labeling rules have no collagen claim either.

How we searched

Searched: a local copy of the PubMed 2026 baseline, three queries for collagen, collagen peptides, collagen hydrolysate or gelatin with weight, body composition, obesity, fat mass, lean mass, satiety, appetite or energy intake. The two broad queries returned 345 and 197 records and we screened the top 40 and top 30. A narrow query on gelatin and satiety against other proteins returned 4. We also searched ClinicalTrials.gov, the EU and US health claim registers, a drug interaction table and the open web for newer reviews and official positions. Search run on 7 October 2026, and every included paper was checked against Retraction Watch.

Included: two meta-analyses and one systematic review of randomized trials, six randomized trials, one controlled feeding study for safety and one meta-analysis of animal studies.

Excluded: trials about skin or muscle protein synthesis without a weight outcome, multi-ingredient supplements where collagen could not be separated, a capsule study where gelatin was only the shell, and news and blog pages.

What we read: full text for the meta-analysis of trials with training and for the collagen bar trial, the paper itself for the gelatin and oxalate study, and abstracts for the rest.

What we could not get: the full text of the 12-trial meta-analysis, so its list of trials and how many involved exercise are unchecked.

What would change this answer

A placebo-controlled trial with weight loss as its main outcome, using the ordinary collagen powder people buy. We found none, and none is running in the ClinicalTrials.gov records we searched.
A trial of the collagen bar against an identical bar without collagen. That would separate the collagen from the act of eating something before meals.
Longer trials in people with insulin resistance, to see whether the small rise in insulin holds up or fades.
The trial list of the 12-trial meta-analysis, to know how much of the fat mass result is training rather than collagen.

The food behind this question

Collagen supplements: the skin benefit fades when the sponsor leaves — numbers, evidence and safety

The same question for other foods (weight loss)

Sources

ev-colw-01 · Meta-analysis or systematic review · systematic review and meta-analysis of randomized placebo-controlled trials · n = 12 trials
Pooled data revealed that CPS significantly decreased fat mass (-1·21 kg; 95 % CI: -2·13, -0·29; I2 = 0·0 %; P = 0·010) and increased fat-free mass, based on body mass percentage (1·49 %; 95 % CI: 0·57, 2·42; I2 = 0·0 %; P = 0·002).
Who: adults in randomized placebo-controlled trials of collagen peptide supplementation
Effect: fat mass -2.7 lb (95% CI -4.7 to -0.64; I2 0%; metric: -1.21 kg, -2.13 to -0.29); fat-free mass +1.49% of body mass (95% CI 0.57 to 2.42)
Certainty: Body weight itself not pooled; many included trials combined collagen with resistance training; short trials.
Br J Nutr, 2023 · checked 2026-10-07 · we read the abstract
ev-colw-02 · Meta-analysis or systematic review · systematic review and meta-analysis of RCTs · n = 418
Eight studies including 418 subjects investigating fat-free mass changes following long-term CP supplementation (Fig. 2) revealed a significant pooled effect size (ES) of 0.48 [p < 0.01, confidence interval (CI) 0.22–0.74].
Who: healthy normal to overweight adults over 17 doing regular training, collagen peptides daily for at least 8 weeks
Effect: fat-free mass SMD 0.48 (95% CI 0.22 to 0.74); GRADE moderate certainty for body composition
Certainty: Fat-free mass, not weight loss; effect seen alongside training; six of eight trials used BIA.
Sports Med, 2024 · checked 2026-10-07 · we read the fulltext
ev-colw-03 · Meta-analysis or systematic review · systematic review of randomized controlled trials, no meta-analysis · n = 15 trials
Certain improvements in body composition, strength and muscle recovery were present.
Who: 12 trials in recreational athletes, 2 in older adults, 1 in untrained pre-menopausal women
Effect: certain improvements in body composition; strongest effect on joint pain; no pooled estimate
Certainty: Narrative synthesis; weight loss not a primary outcome; populations were not people trying to lose weight.
Amino Acids, 2021 · checked 2026-10-07 · we read the abstract
ev-colw-04 · Randomized controlled trial(s) · randomized parallel trial, control got no bars or placebo · n = 64
After three months (V3), a reduction in body weight of −3.0 ± 2.0 kg was observed in the group who consumed collagen compared to a reduction of −1.5 ± 1.3 kg in the control group (3.36% vs. 1.80% of weight loss, p = 0.045).
Who: adults with overweight or obesity; healthy diet advice in both arms, plus two 90 calories bars with bovine collagen 45 min before lunch and dinner in the collagen arm
Effect: weight -6.6 lb (-3.0 kg) vs -3.3 lb (-1.5 kg) (3.36% vs 1.80% of body weight, p = 0.045) at 12 weeks
Certainty: No placebo, so extra attention and pre-meal eating are confounded; specially processed high-swelling collagen, not ordinary collagen powder; single trial.
Nutrients, 2024 · checked 2026-10-07 · we read the fulltext
ev-colw-05 · Randomized controlled trial(s) · randomized double-blind trial, collagen vs whey, no placebo · n = 37
Additionally, our findings suggest that collagen may not be an effective supplement for overweight women who are attempting to alter body composition.
Who: women with overweight, mean age 40.6, mean BMI 30.9; 38 g/day hydrolyzed collagen or 40 g/day whey concentrate
Effect: BMI +0.2 kg/m2 with collagen (p = 0.044); android fat -0.22 lb (-0.1 kg) with whey vs collagen (p = 0.031)
Certainty: Active comparator, not placebo; small; 8 weeks.
Nutrients, 2019 · checked 2026-10-07 · we read the abstract
ev-colw-06 · Randomized controlled trial(s) · randomized double-blind crossover trial · n = 15
Ad libitum energy intake was ∼10 % (∼41 kcal) lower in the CP trial (P = 0·037).
Who: healthy active females aged about 23, 15 g/day collagen peptides or taste-matched non-energy control for 7 days
Effect: ad libitum energy intake about 10% (41 calories) lower; GLP-1 higher; subjective appetite unchanged
Certainty: One meal, small sample; control had no energy, so part of the cut may be compensation for the supplement calories.
Br J Nutr, 2025 · checked 2026-10-07 · we read the abstract
ev-colw-07 · Randomized controlled trial(s) · randomized double-blind crossover pilot study · n = 10
There was no difference between treatment groups in the perception of hunger (P = 0·983), desire to eat (P = 0·326), full stomach feeling (P = 0·567) or food consumption (P = 0·168).
Who: healthy adult women, mean age 22.4, single 40 g dose of hydrolyzed collagen or whey protein
Effect: no difference in hunger (p = 0.983), desire to eat (p = 0.326), fullness (p = 0.567) or food intake (p = 0.168)
Certainty: Pilot, single dose, small.
Br J Nutr, 2022 · checked 2026-10-07 · we read the abstract
ev-colw-08 · Randomized controlled trial(s) · randomized single-blind crossover trial of protein types · n = 24
Both at the level of 10 and 25 En% from protein, EI at lunch was approximately 20% lower after an alpha-lactalbumin or gelatin (+TRP) breakfast (2.5+/-0.2 MJ) compared with after a casein, soy, or whey-GMP breakfast (3.2+/-0.3 MJ, p<0.05).
Who: healthy adults, mean BMI 24.8, mean age 25; custard breakfasts with different protein sources, including gelatin with added tryptophan (+TRP)
Effect: lunch energy intake 2.5 vs 3.2 MJ (about 20% lower); appetite ratings about 40% different at 180 min
Certainty: Single meals; gelatin is a cooked form of collagen; a companion study (PMID 19017422) found no lunch intake difference.
Clin Nutr, 2009 · checked 2026-10-07 · we read the abstract
ev-colw-09 · Randomized controlled trial(s) · randomized placebo-controlled parallel trial (NCT05282641) · n = 56
Other parameters, anthropometrics, physical activity, and energy/nutrient intakes were not significantly changed.
Who: men and women with overweight or obesity, 10 g/day porcine collagen hydrolysate or erythritol placebo for 4 weeks
Effect: anthropometrics unchanged; fasting insulin (p = 0.005), HOMA-IR (p = 0.008) and post-meal glucose AUC rose; authors call changes small
Certainty: Short; small changes of limited clinical relevance per the authors; relevant to people with insulin resistance.
Clin Nutr, 2025 · checked 2026-10-07 · we read the abstract
ev-colw-10 · Observational data · controlled crossover feeding study, randomization not stated · n = 10
Urinary glycolate excretion was 5.3-fold higher on the gelatin diet compared to the whey diet and urinary oxalate excretion was 43% higher.
Who: healthy subjects with no history of nephrolithiasis, two 5-day controlled diets with 30 g gelatin or whey
Effect: urinary oxalate +43%, glycolate 5.3-fold vs whey
Certainty: Small, short; healthy people; stone risk itself not measured. The 20 g/day used for weight in ev-colw-04 is near this load.
Kidney International, 2006, Hydroxyproline ingestion and urinary oxalate and glycolate excretion · checked 2026-10-07 · we read the abstract
ev-colw-11 · Laboratory or animal data · systematic review and meta-analysis of animal trials · n = 339 animals
However, CP's anti-obesity effects in humans remain largely unexplored, necessitating further research.
Who: rodents on high-fat or high-calorie diets given collagen peptides for at least 3 weeks
Effect: body mass SMD -1.87; food intake SMD -1.43; very low certainty of evidence
Certainty: Animal data; moderate to high risk of bias.
Int J Obes (Lond), 2026 · checked 2026-10-07 · we read the abstract

How this page was made. Evidence was extracted from primary sources into cards, every number was re-checked against the source, and the text was written from those cards. Bioma Learn has no human medical reviewer at this time, and we say so rather than invent one. Methodology. This is information, not medical advice.