I used to sell Bioflow magnetic devices for people with arthritis (many moons ago).
They came with a money back guarantee if they didnt work. 80% of people were satisfied and didnt claim their money back.
When people used to say its the placebo effect, I would point out that the magnetic dog collars they made for dogs with arthritis seemed to work 100% (and there were some really bad cases of dogs who had hip displasia,couldnt get the back legs to stand and had to drag themselves along).
They also used to make devices for horses with laminitis which wrapped around their legs with and again these seemed to have a very high success rate.
I dont know whether Bioflow are still going, but they had a unique patent which meant they had an alternating magnetic field, not like the cheap normal magnetic bracelet which usually has a weak static magnetic field.
The alternating field replicated the pulsing magnetic field equipment used by physiotherapists to treat injuries sustained by professional sports players.
The British Medical Journal, the main doctors magazine in the UK, did a study on the effect of Biolow magnets on hip and knee arthritis.
The study warned that low strength magnets are not effective. If your going to buy a wrist magnet for arthritis make sure it has a field strength at least 160 mTesla. If the supplier cant show this walk away.
Studies that have failed to show an effect on pain generally used weaker magnets (19 to 50 mTesla). Studies that have shown an effect used stronger magnets (47 to 180 mTesla), which were comparable with our standard strength magnets.
Together these findings suggest that field strength is important.