Inflammation 04: Anti-inflammatory drugs

Short answer. Non-steroidal anti-inflammatory drugs (NSAIDs) and acetaminophen suppress the prostaglandin pathway. Eleven controlled trials have given one during a resistance-training program and measured muscle growth: two found more growth with the drug, one found less, and eight found no difference. The trials differ in drug, dose, comparator, population and measurement, and none was designed to say which of those decides the result. This entry sets out what each one found and leaves the question open.

What it is

Non-steroidal anti-inflammatory drugs taken around training. Ibuprofen, diclofenac, aspirin, naproxen. Widely used by recreational exercisers, usually for soreness or a nagging joint, sometimes prophylactically before a hard session. (→ Inflammation 02: Delayed-onset muscle soreness)

Acetaminophen (paracetamol) is not classed as an NSAID, but it also inhibits cyclooxygenase, and two of the trials below tested it, so it is included.

This entry is the reason the rest of this branch exists. It is also the one where the literature refuses to settle.

The proposed mechanism

NSAIDs inhibit cyclooxygenase enzymes, which produce prostaglandins. Prostaglandins are involved in a wide range of processes including muscle protein metabolism and regeneration.

This matters because it is the only intervention in this branch that unambiguously does what it is advertised to do. Cold does not measurably reduce inflammation in human muscle. (→ Inflammation 01: Cold water immersion) High-dose antioxidants change signaling without moving outcomes. (→ Inflammation 03: Antioxidant supplements) NSAIDs do suppress the inflammatory pathway. (→ Inflammation sub-hub)

That makes this the cleanest available test of the premise underneath the whole branch: if post-exercise inflammation is part of the adaptive signal, then a drug that does suppress it should cost you something.

The test has been run eleven times. It did not come out cleanly.

What has actually been tested

These are every controlled trial we could find that gave one of these drugs during a resistance-training program of weeks or months and measured muscle size, counted once per group of participants. Papers reporting other outcomes from the same participants are noted under the trial they came from. Single-session studies of protein synthesis or soreness are left out, because they measure a response to one workout rather than adaptation to a program.

More muscle with the drug: two trials.

• Older adults, 12 weeks of knee-extensor training. Acetaminophen 4 g a day or ibuprofen 1.2 g a day, against placebo. Quadriceps volume rose 12.5% and 10.9% with the drugs against 8.6% with placebo (P < 0.05, adjusted for starting size and strength), and strength rose more with both drugs (Trappe et al., 2011).
• Seventeen trained young men, 12 weeks. Diclofenac 75 mg a day, against placebo. Quadriceps cross-sectional area rose 8.6% against 3.9% at 12 weeks (P = 0.038). There was no difference at 4 weeks, and none in strength at any point (Mallinson et al., 2025). Authors from the laboratory behind the next trial have published a letter questioning its methods and interpretation, and the authors have replied (Mandić et al., 2026; Mallinson et al., 2026). Neither letter has an abstract and we have not read them.

Less muscle with the drug: one trial.

• Thirty-one young men and women, 8 weeks of knee-extensor training. Ibuprofen 1.2 g a day, against low-dose aspirin at 75 mg a day, which is itself a cyclooxygenase inhibitor rather than a placebo. Quadriceps volume rose 3.7% against 7.5% (P = 0.029). Strength gains were smaller in the leg trained on a flywheel and the same in the leg trained on a weight stack (Lilja et al., 2018). Later papers on the same participants found that muscle fiber size did not change with training in either group (Lilja et al., 2023), and that mitochondrial function fell by a similar amount in both (Cardinale et al., 2017).

No difference: eight trials.

• Eighteen young adults, 6 weeks of arm training. Ibuprofen 400 mg after training one arm, placebo after training the other (Krentz et al., 2008).
• Twenty-three recreationally trained young men, 6 weeks of upper-body training. Naproxen against placebo (Brewer et al., 2015).
• Thirty-six knee osteoarthritis patients aged 50 to 70, 12 weeks. Ibuprofen against placebo. No difference in muscle size, but larger strength gains with the drug (Petersen et al., 2011). The patients had painful knees, and the trial cannot separate pain relief during testing from adaptation.
• Twenty-eight postmenopausal women, 9 weeks. Ibuprofen 400 mg after each session (Candow et al., 2013).
• Ninety women averaging about 65, 9 months. Ibuprofen 400 mg after each session (Duff et al., 2016).
• Men aged 50 and over, 16 weeks. Acetaminophen 1,000 mg before each session (Jankowski et al., 2013).
• 189 adults aged 60 to 75, 36 weeks. Ibuprofen 400 mg before or after each session (Jankowski et al., 2015).
• Nineteen men aged 60 to 80, two weeks with one leg in a cast and six weeks of retraining with a protein supplement. Ibuprofen 1.2 g a day (Dideriksen et al., 2016).

No meta-analysis pools these trials. The one meta-analysis we found covers performance in the 48 hours after a session, not adaptation across a program (de Oliveira et al., 2024).

Our reading

These are the facts the trials share or do not share, and what each one does and does not allow.

The count is two, one and eight. Most trials found no difference. A count is not a verdict: it weighs a trial of 17 the same as a trial of 189, and the three trials that found a difference had 8 to 16 people per group.

Every trial that found a difference, in either direction, gave a high dose every day. Ibuprofen 1.2 g, acetaminophen 4 g, or diclofenac 75 mg. All five trials that gave 400 mg of ibuprofen or 1 g of acetaminophen once per session found no difference. One trial at 1.2 g a day also found none, in a cast-then-retraining design. That makes dose one candidate. It does not explain why a high daily dose went up in two trials and down in one.

Age does not separate the directions. More growth came in an older group and in a young trained group; less growth came in a young group. Among healthy adults over 50, one trial found more and five found no difference.

The three trials that found a difference measured whole-muscle volume or area by MRI. Most of the trials that found none used lean mass by DXA or thickness by ultrasound. A small effect may be invisible to a coarser instrument, or the MRI results may be noise. The data do not say which.

The one trial that found less growth used an active comparator. Low-dose aspirin inhibits the same enzymes, so that result compares two drugs, not a drug and nothing.

What the facts leave open. Whether these drugs cost growth, add to it, or do nothing at the doses people take, and whether dose, drug, age, training status, sex or the measuring instrument decides it. No trial has yet tested the same drug at the same dose in the same kind of lifter twice.

OMNID programs do not reference medication and will not. Dosing and timing are outside what we will publish on.

Where the goals diverge

Managing real pain. A clinical question with a clinician attached. Non-pharmacological soreness management is covered separately. (→ Cold and heat exposure sub-hub)

Taking it before sessions. Five trials gave a dose once per session and found no difference in growth. The trials that found a difference gave a dose every day.

Starting out. The two trials in young adults whose training status was not stated split: one found less growth at a high daily dose against aspirin, one found no difference at a low dose against placebo.

Trained for years. Two trials in trained young men: one found more growth with diclofenac, one found no difference with naproxen.

The case against

Every trial here is small. Only two had more than 40 participants, and both of those gave a low dose once per session.

Counting trials is a weak way to read them. A tally treats a null result from a coarse instrument the same as a null result from a precise one, and a trial against aspirin the same as a trial against placebo.

The branch's premise is not settled here either way. If suppressing inflammation reliably cost you adaptation, this is where it would show, and in most trials it does not show at all.

What is not known

Whether dose, drug, comparator, age, training status, sex or the measuring instrument accounts for the split.

Whether occasional use behaves like daily use. The trials gave either a daily dose or a dose with every session.

What the 2026 letters argue. Neither has an abstract and we have not read them.

Whether a drug selective for one cyclooxygenase isoform, such as celecoxib, behaves like the others. We found no long-term training trial of one, and none of aspirin against placebo.

Safety

We do not cover drug dosing or timing. If you take an NSAID for pain, that is a decision for you and a doctor or pharmacist who knows your history, and nothing here should influence it.