With protein and training held equal, a moderate energy deficit built the same fat-free mass as eating at maintenance and lost twice the fat
In people who started at BMI 26, which is the condition, not a footnote
Protocol
Thirty participants (23.0 ± 3.4 y, 174.3 ± 8.0 cm, 80.3 ± 16.0 kg, BMI 26.3 ± 4.5) randomised to three conditions for 10 weeks:
ISO (n=10) — isocaloric, high protein
DEF (n=10) — moderate energy deficit (size not given in the abstract; ~250 kcal/day per a secondary summary, not checked), the same protein as ISO (2.5 g/kg/day, per the abstract)
Control (n=10) — the same resistance-training programme, no dietary intervention and no supervision
ISO and DEF trained 4 days a week under supervision. Body composition by DXA at baseline and post.
Key Findings
| Fat mass | Fat-free mass | FFM adjusted for fat-free adipose tissue | |
|---|---|---|---|
| DEF (moderate deficit) | −2.94 kg (p < 0.001) | +1.04 kg (p = 0.009) | +1.50 kg (p < 0.001) |
| ISO (maintenance) | −1.41 kg (p = 0.051) | +0.97 kg (p < 0.001) | +1.22 kg (p = 0.002) |
| Control | no meaningful change | no meaningful change | all p ≥ 0.778 |
The muscle gain was the same in both diets. 1.04 kg against 0.97 kg over ten weeks — a moderate deficit did not cost fat-free mass when protein and training were held constant.
The Time × Group interaction was significant for fat-free mass (p = 0.034) and for FFM−FFAT (p = 0.006), and NOT significant for fat mass (p > 0.05). Read that carefully: the tested difference that reached significance is between the DIET groups and the unsupervised control, not between deficit and maintenance. The fat-loss gap of 2.94 vs 1.41 kg is a numerical difference the study was not powered to confirm.
The control group changed nothing at all despite following the same training programme — but it also had no nutritional supervision, so this comparison bundles diet with adherence support.
Practical Application
A moderate energy deficit was compatible with gaining fat-free mass here, with protein (2.5 g/kg/day) and training held constant. The deficit size is not in the abstract (a secondary summary reports ~250 kcal/day — not checked).
Do not read the fat-loss difference as established. Twice the fat lost sounds decisive and did not reach significance at n=10 per group. Say "the deficit group lost more fat, and the study could not confirm the difference".
Ask about body fat before recommending any of this. These participants had room to fund tissue growth from stored fat. Not from this paper (secondary literature, not checked): recomposition is said to be common up to a ~300 kcal deficit and rare beyond ~500, and leaner people to lose muscle instead.
Protein was the constant, not the variable. Both groups ate the same, high, amount. This study says nothing about whether protein could have been lower — see the tension below.
For an endurance athlete in a build block the answer is still usually no. Ten weeks of resistance training in recreationally trained young adults is a different physiological question from an athlete carrying 12-15 h/week of endurance load, where the energy-availability floor governs (the REDs entries in this section; REF-345 shows 5 days of deficit lowering resting muscle protein synthesis 27%).
Limitations
Ten per group. Small enough that the absence of a between-diet difference in fat loss is uninformative, and small enough that a 1 kg fat-free-mass change carries wide uncertainty.
The control was unsupervised. Diet and supervision are confounded, so "the diets worked and the control did nothing" cannot separate the nutrition from the attention.
Recreationally trained, ~1.5 years of experience, BMI 26.3. Both the training-status and the body-fat ceiling on generalisation apply, and the second is the one that matters clinically.
Ten weeks. Nothing here speaks to whether the same deficit remains non-catabolic over a season, which is the question an athlete actually asks.
DXA estimates fat-free mass; it does not measure muscle. FFM includes water and fat-free adipose tissue — the authors adjusted for the latter, which is a strength, but this is still not a direct measure of contractile tissue.
Full text not read — this entry is written from the abstract (the paper is not open access), so the dietary adherence data and the training programme detail are not verified here. The protein dose (2.5 g/kg/day) IS in the abstract; the ~250 kcal deficit and ~25% baseline body fat reported by a secondary summary are not, and are recorded as unverified.
How this entry was checked and changed
Content correction 2026-09-27
"~250 kcal deficit" stated as the study's → the abstract says only "moderate energy deficit"; the 250 kcal figure is labelled as an unverified secondary report.
Protein 2.5 g/kg/day was called unverified → it is stated in the abstract.
Claims that REF-345 "states a 250-300 kcal cap" removed → REF-345 is now the verified Areta 2014 study, which carries no such cap.