Most effective pain reduction for arthritic thumbs?

The short version: for thumb arthritis pain, a custom neoprene splint worn at night is the only intervention tested against a real control group, and it cut pain by about half.

The best evidence here is a randomized trial of 109 people that compared a custom neoprene splint plus usual care against usual care alone [5]. At 12 months, the splint group dropped from 46 to 24 on a 100-point pain scale, while the control group went from 48 to 39. That is a between-group difference of 22 points, p=0.006. No adverse events were reported, and 86% of people still wore the splint at least five nights a week at one year. This is the only trial in this set that used a true control group and measured a hard clinical endpoint (pain). It was funded by French hospitals, not by a manufacturer.

Home-based hand exercises also have decent evidence. A randomized trial of 80 women found a 1.1-point pain reduction on a 0-10 scale compared to information alone, plus meaningful gains in grip strength and thumb web space [11]. The effect was moderate but real, and adherence was good. A smaller pilot trial found that general hand exercises actually outperformed specific "dynamic stability" exercises for pain and function at 3 and 6 months [15], which suggests that the specific type of exercise matters less than doing something.

Steroid injections are common but the evidence here is thin. One study of 43 people found that if pain persists one week after an ultrasound-guided steroid injection, the odds of needing surgery within a year are about 3 times higher [7]. That is useful as a prognostic test, not as proof that injections reduce long-term pain. No placebo-controlled trial of injections appears in this evidence.

Surgery works for severe cases. Trapeziectomy with ligament reconstruction produces large and sustained pain reductions: PRWHE pain scores dropped from 35 at baseline to 8 at 5 years in one study [12]. But surgery carries real risks, and the evidence here compares surgical techniques against each other, not against continued conservative care. A meta-analysis found that joint replacement improves function slightly more than trapeziectomy but carries double the complication rate and five times the revision surgery rate [14].

My call: start with a custom neoprene splint worn at night and daily hand exercises. Those are the only interventions tested against a real control group, and both show meaningful pain reduction with essentially no risk. Steroid injections may help some people short-term but the evidence does not prove it. Surgery is reserved for when conservative care fails and pain is severe enough to accept the risks. Confidence: moderate for splints and exercises, low for injections, moderate for surgery in severe cases.