Do rub-on anti-inflammatory gels work for knee arthritis? For a minority, yes

The gels and creams you rub on a sore knee are not just placebo theatre. For knee osteoarthritis specifically, topical anti-inflammatory drugs have decent randomised evidence — they help a minority of people meaningfully, work about as well as the same drugs in pill form, and spare you most of the stomach and kidney risks that come with swallowing them. Guidelines have moved to reflect that.

What the trials show

The most careful summary is a 2016 Cochrane review of topical non-steroidal anti-inflammatory drugs (NSAIDs) for chronic musculoskeletal pain, drawing on 39 studies and 10,631 participants [s1]. A telling detail: every one of those studies was in osteoarthritis, and the efficacy results were almost entirely derived from people with knee osteoarthritis [s1]. So this is really a body of evidence about knee arthritis, whatever the label on the tube suggests.

In studies lasting 6 to 12 weeks, topical diclofenac and topical ketoprofen were significantly better than their carrier gel at reducing pain, with about 60% of participants reporting much reduced pain [s1]. The review measured benefit as "clinical success" — at least a 50% cut in pain — and expressed it as the number needed to treat (NNT), the number of people who must use the gel for one extra person to hit that mark. For topical diclofenac the NNT was 9.8 (95% CI 7.1 to 16) across six trials; for topical ketoprofen it was 6.9 (95% CI 5.4 to 9.3) across four trials [s1]. Both were graded moderate-quality evidence [s1]. Where trials compared a topical NSAID against an oral NSAID directly, the two showed similar efficacy, though that comparison rested on low-quality evidence [s1].

The catch: a big placebo effect

Rub-on treatments carry an unusually large placebo response. Clinical success occurred in around half of participants using the carrier gel alone [s1]. In fact, response rates with the carrier were about twice those seen with an oral dummy pill, and the reviewers noted emerging evidence that some of that benefit comes from the act of massaging a gel into the skin, with the NSAID adding on top [s1]. That is why the honest headline is "beyond carrier for a minority" rather than "works for everyone" — a lot of the felt relief is real but not specifically the drug.

The safety trade-off

The reason to prefer a gel is what you avoid. In the Cochrane data, systemic side effects such as gastrointestinal upset were no more common with topical NSAIDs than with carrier, and serious adverse events were infrequent and no different [s1]. The main downside was local: mild skin reactions were more common with topical diclofenac, though not with ketoprofen [s1].

A 2021 network meta-analysis in The BMJ put that trade-off in sharp relief [s2]. Pooling 192 trials and 102,829 participants across NSAIDs, opioids and paracetamol for knee and hip osteoarthritis, it found that 0% of the topical NSAID preparations carried an increased risk of dropouts due to side effects, and 0% carried an increased risk of any adverse event — against 18.5% and 29.8% respectively for oral NSAIDs [s2]. Topical diclofenac at 70 to 81 mg a day was judged effective, and the authors concluded it "should be considered as first line pharmacological treatment for knee osteoarthritis" because of its lower systemic exposure and dose [s2].

What this means for you

The practical reading is consistent across both analyses. If your problem is knee osteoarthritis, a topical NSAID is a sensible first thing to try: a real chance of meaningful relief, efficacy close to oral NSAIDs, and far less systemic risk [s1][s2]. Give it a genuine trial of several weeks, since the benefit was measured over 6 to 12 weeks [s1]. If it does nothing after that, you may be among the many for whom it does not clear the bar — and the large carrier effect means some of any relief is the rubbing, not the drug [s1].

Two honest limits: the good evidence is for knee (and to a lesser extent hip) osteoarthritis, not for back pain or other chronic pain, where the review found no support [s1]; and "topical" is not "risk-free" for everyone — people already taking oral NSAIDs, or with kidney or ulcer risk, should still check with a clinician. Persistent, worsening or hot, swollen joints deserve a proper diagnosis rather than a standing habit of self-treating with gel.

This article describes what the studies found and is not medical advice.

Sources

• Topical NSAIDs for chronic musculoskeletal pain in adults — Cochrane Database of Systematic Reviews, 2016-04-22
• Effectiveness and safety of non-steroidal anti-inflammatory drugs and opioid treatment for knee and hip osteoarthritis: network meta-analysis — BMJ, 2021-10-12