So there’s a new review of massage for fibromyalgia out.
“Massage therapy for fibromyalgia: a systematic review and meta-analysis of randomized controlled trials”
This is going to seem negative to a lot of folks (because, hey, it’s me), but let me be extremely clear right at the start here: I don’t think this paper shows anything one way or the other. I think it’s straightforwardly inconclusive. It has a positive sounding conclusion that isn’t really justified and has major caveats. But that doesn’t mean that I wouldn’t go get me some massage if I had fibromyalgia. I would! But that’s another story. This is about the paper, so here goes …
This paper epitomizes the “garbage in, garbage out” problem with meta-analysis. There was virtually no research on this topic worth analyzing to begin with. Trying to pool the results of several weak studies is basically meaningless. To the extent that the study results are generally inconclusive and ambiguous, the conclusions of any review are going to have more to do with the authors’ opinions than hard data.
This review is somewhat novel in that it includes some Chinese research, but it doesn’t really help. The introduction curiously boasts that “traditional Chinese massage is one of the most ancient massage therapies,” but there is not enough distinctive about Chinese massage that makes it worthy of any focus, and massage is ancient in every culture. This weird, prominently placed statement is a red flag: biased interpretation ahead! (And only one study of traditional Chinese massage made it past the selection criteria anyway.)
Unsurprisingly, the conclusions here are superficially positive: massage “significantly improved pain, anxiety, and depression in patients with FM.” But that’s statistical significance only, not a clinically significant degree of improvement: the size of the effect is trivial (much smaller than amplitude of the noise in the data). As usual, using the word “significantly” this way is technically correct and defensible, but otherwise misleading to all but the most alert readers.
Also, the conclusion is based in large part on the data about depression and anxiety, which are likely to improve with massage regardless of any effect on fibromyalgia. (This is the one truly evidence-based effect of massage.)
And more bad news! Despite the seemingly strong positive conclusion, the data is silent on longer term effects. Only two studies had any follow-up data at all. Without promising data about long-term effects, it would be hard to say massage “works” even if the short term data were much more clearly positive.
Both sides of this research question are highly problematic: fibromyalgia is hard to diagnose or define, and massage is hard to study. Even using official diagnostic criteria, which changed significantly in 2010, there’s a lot of wiggle room. (As Fred Wolfe has put it, “One doesn’t either have fibromyalgia or not have it. There is a gradual transition from the mild to the severe. The point at which we classify an individual as having fibromyalgia is arbitrary, but reasonable.”) The types of massage reviewed here were generally vague and all over the map, from the straightforward (Swedish massage) to trendy-but-meaningless “connective tissue massage” (the idea of isolating or even emphasizing connective tissue in massage is a biological absurdity, like trying to eat the gristle out of a steak without masticating anything else) to rank quackery like “therapeutic touch” (which is literally not massage at all and roughly on par with believing in magic). What a mess.
So here’s my conclusion: whoop-de-doo. There’s really nothing here, except maybe massage for fibromyalgia being damned by faint, ambiguous praise. •sigh•