Most Fibromyalgia Research is Worthless
I make it a point to try to be aware of all fibromyalgia research that is published by using PubMed’s free subscription service (http://www.ncbi.nlm.nih.gov/pubmed/). I get to read all of the abstracts this way and if I find articles that are of special interest to me I try to download them. PubMed reports 659 publications in the last 12 months relating to fibromyalgia. For those who are interested, there are 9,366 articles listed in the all the years that data are available. For 1990, the year the American College of Rheumatology 1990 fibromyalgia criteria were published, PubMed cites 95 articles. If you think that after all these years of research you and your patients are much better off, think again. A kind, conscientious physician treating a fibromyalgia patient in 1980 or 1990 will have done as well as the 2016 health workers with access to all of these new publications and expensive if not very efficacious medications.
Of all the 9,366 articles on fibromyalgia indexed in PubMed 1,268 have an attached tag of “clinical trial.” You can get a flavor for the research by looking at the two most recent articles. The first is titled, “The Effects of a Gluten-free Diet Versus a Hypocaloric Diet Among Patients With Fibromyalgia Experiencing Gluten Sensitivity-like Symptoms: A Pilot, Open-Label Randomized Clinical Trial.” That led me to ask myself. “Why would anyone want this information or want to do such a study?” The authors sav, “Patients with fibromyalgia frequently present with symptoms similar to those experienced by patients with gluten-related disorders, raising the possibility that a subgroup of these patients could be experiencing underlying gluten sensitivity. This study aimed to evaluate the effects of a gluten-free diet (GFD) compared with a hypocaloric diet (HCD) among patients with fibromyalgia? Well, the reason they did it was [they say] “…the possibility that a subgroup of these [fibromyalgia] patients could be experiencing underlying gluten sensitivity.” One reason for publications like this one is not quite the old “publish or perish.” Instead, think that the authors like a little of the glory that comes from being “scientists” and publishing “research.”
The second article is “Safety and efficacy of pregabalin in adolescents with fibromyalgia: a randomized, double-blind, placebo-controlled trial and a 6-month open-label extension study.” This is a special class of article,” one sponsored, funded, written and analyzed by Pfizer and led by physicians with a very long conflict of interest list. Pfizer has a long history of government fines, corrupt practices and ghost (or semi-ghost) written articles. The article concludes, “while this trial did not meet its primary efficacy outcome, improvements in secondary outcomes of pain and impression of change, together with a safety profile that was consistent with the known profile in adults with FM, suggest that patients might benefit from pharmacological treatments.” This isn’t the first time in Pfizer clinical trials that the authors point out the effectiveness of drugs that failed to meet their primary outcome goals. Few know that “Pfizer” is an old Klingon word that means beware.
Bad study design is a big problem
1) Normal controls. Most non-treatment studies of fibromyalgia make use of “normal” controls. This turns out to be a fatal flaw because persons with fibromyalgia virtually always have more severe abnormalities than healthy persons. They will suffer more, have greater costs, social disruption, pain, comorbidity, psychological distress. It doesn’t matter what you study if it is perceived to have some internal meaning. Studies against normals will tell you just what you always knew, fibromyalgia patients have more problems than those without fibromyalgia. Additional studies showing that patients with fibromyalgia have more of A, B and C are valueless. More than valueless, they tend to make clear that the authors don’t really understand the nature of fibromyalgia. Pfizer sponsors many studies that show how bad it is to have fibromyalgia.
If one wants to understand how those with fibromyalgia differ from those with pain who do not satisfy fibromyalgia, then controls have to come from pain groups. Because persons with fibromyalgia also usually have other comorbid illnesses and differences in psychosocial status, good studies of fibromyalgia need to seek similar (not normal) controls and have sufficient appropriate covariates. Many MRI studies, for example, that use normal controls may find differences that are not so much due to fibromyalgia as to other common comorbid symptoms. An example of this problem is that published studies began to report peripheral nerve abnormalities in fibromyalgia compared with normal controls. However, non-fibromyalgia pain controls also had such abnormalities. A number of early studies touted fibromyalgia vs. control data as examples of underlying abnormalities in fibromyalgia.
2) Biased studies and blinding. Convenience sample of volunteers with fibromyalgia as well as normal controls are usually biased, often in ways that are difficult to fully understand. Such biases are very common and truly invalidate studies, and statements by authors that there may be “limitations” never solve the problems. Fibromyalgia studies often have diagnosis and outcomes evaluated by persons who have an interest in the outcome. Studies that do this often get the results that the authors want.
3) Fibromyalgia studies are often underpowered, poorly analyzed and described. Only a small fraction follow guidelines for proportion publications.
4) Fibromyalgia studies often ask participants to remember events from many years in the past, something that always leads to unreliable data.
5 P-Hacking or taking multiple looks at the data before picking the analyses is common.
Why do “investigators do fibromyalgia studies? Because it is easy to find patients and there are always abnormalities. If you look hard enough almost any question can seem publishable. As I sit here with a cat at my side and sun streaming in the window, I wonder if cats or weather have something to do with fibromyalgia? Not a bad guess; more than 2,000 hits on pets and fibromyalgia in Google scholar. Weather and fibromyalgia? More than 5,000 hits. Chinese food? Not so good. Only 103 hits.