REF-123 · Heileson et al., 2023 · Journal of the International Society of Sports Nutrition, · Paper

The Effect of Fish Oil Supplementation on Resistance Training-Induced Adaptations

Blimited evidenceMeasured on recreational to trained athletesWritten before October 2026, being re-checked

Heileson et al. tested whether fish oil supplementation (FOS; concentrated EPA + DHA) augments the adaptations to a 10-week progressive resistance-training programme (3 days/week) in young, recreationally trained men and women, against placebo (PL); n=28 were randomized and 21 completed (FOS 10 after 4 withdrawals; PL 11 after 3).

Blood EPA+DHA rose 109.7% in FOS (p < .001) and was unchanged in PL (+1.3%). Body-composition changes were similar between groups (LBM +3.4% vs +2.4%, p = .457; fat mass −5.2% vs 0.0%, p = .092 but a large between-group effect, d = 0.84; %BF −5.9% vs −2.5%, p = .136).

Bench-press 1RM improved more with FOS (absolute +17.7% vs +9.7%, p = .047; relative +17.6% vs +7.3%, p = .011); absolute squat 1RM was similar (+28.8% vs +20.5%, p = .191) while relative squat 1RM favoured FOS (+29.3% vs +17.9%, p = .045). No soreness or recovery outcomes are reported in the abstract.

Participants

n=28 randomized young, recreationally trained men and women; 21 completed (FOS n=10, PL n=11).

Protocol

10-week progressive resistance-training programme, 3 days/week.
FOS (concentrated EPA + DHA) vs placebo; compliance by self-report, bottle collection every two weeks and dried-blood-spot fatty acids at PRE and POST. Dose per the paper: 2.275 g EPA + 1.575 g DHA daily (full text, checked 2026-09-27).
Outcomes: DXA body composition (LBM, fat mass, %BF); 1RM back squat and bench press; PRE (week 0) and POST (week 10).

Key Results

Blood EPA+DHA: FOS +109.7% (p < .001); PL +1.3% (p = .938).
Bench press 1RM absolute: FOS +17.7% vs PL +9.7% (p = .047); relative: +17.6% vs +7.3% (p = .011).
Back squat 1RM absolute: +28.8% vs +20.5% (p = .191, ns); relative: +29.3% vs +17.9% (p = .045).
LBM +3.4% vs +2.4% (p = .457); fat mass −5.2% vs 0.0% (p = .092, d = 0.84); %BF −5.9% vs −2.5% (p = .136).

Practical Application

In 21 young, recreationally trained adults doing 10 weeks of resistance training, fish oil (EPA + DHA) roughly doubled the bench-press 1RM gain over placebo (+17.7% vs +9.7% absolute; +17.6% vs +7.3% relative to body mass) and improved relative squat 1RM more (+29.3% vs +17.9%), with similar lean-mass changes — a small trial, but a real strength signal.
Inference, not tested: the mechanism (inflammation resolution, protein-synthesis signalling) is not measured here; the trial shows an outcome, not why.
The trial did not measure soreness (DOMS) or recovery speed; do not quote it for those.
Not from this paper; the base's own practice: when an athlete wants to try fish oil around a strength block we suggest a dose in the range this trial used and a full block (about 10 weeks) before judging the effect; oily fish remains the food-first default. For inflammation mechanisms see REF-122.

Limitations

Small (n=21 completers, 7 withdrawals), recreationally trained young adults, upper-body effect clearer than lower-body; no cycling or endurance outcomes; 10 weeks only.
How this entry was checked and changed
Why this level

regraded 2026-09-27 from B+ (previously marked unverified): parallel RCT, 21 completers of 28 randomized (10 vs 11)

Content correction 2026-09-27
"Bench press 1RM +3.2%", "DOMS 30-40% faster resolution", "2-3 days faster return to baseline", "effect after 4-6 weeks", "+1.5-2% strength acceleration" -> the abstract reports bench 1RM +17.7% vs +9.7% (relative +17.6% vs +7.3%), relative squat +29.3% vs +17.9%, and measured no soreness or recovery outcome; all invented figures deleted.
"Athletes" -> young, recreationally trained men and women; n=28 randomized, 21 completers added; Tier 2-3 -> Tier 1-2; Sport "cycling, sprint" -> strength.
DOI line added from the PMID already in the entry; Level B+ [pending] -> B. Dose kept only as a labelled full-text figure.
Lines labelled "full text; not checked against the abstract" were checked against the open-access full text (PMC9970203): every number in them appears there; relabelled "full text, checked 2026-09-27".