A double-blind study of guaifenesin vs. placebo was completed at the University of Oregon in June 1995 by Dr. Robert Bennett. In this study, patients in the placebo group compared to patients in the guaifenesin group appeared to improve equally (not by much) over the course of this one-year study, a finding which suggested a placebo effect. It is our belief that this study was flawed, due to our collective lack of knowledge and possibly some bad luck. We take full responsibility for the errors in the study design which includes both of these issues.

In the past, while using gout medications, our patients had been routinely cautioned to avoid aspirin, because we knew that aspirin blocked the uricosuric effect of the medications. (This is well documented in medical research.)  We had also discovered that aspirin rendered the medications ineffective when treating fibromyalgia. One of our very first patients on guaifenesin agreed to be a guinea pig and took aspirin too.  We quickly found that guaifenesin did not reverse fibromyalgia when taken alongside salicylates.

Dr. Bennett's patients were duly warned to avoid aspirin and aspirin-containing compounds such as Alka Seltzer and Excedrin.

However, three months into the study, we began to encounter patients who, while using Myoflex, Ben Gay, or sunscreens containing salicylates, had reversed their initial progress and were worse. In the seventh month, we saw the same effect from an oral herbal medication.

Through search of the literature, we learned that plants make salicylates (this had just been published in Science Magazine in 1994,printed AFTER the Oregon study had begun) as a defense mechanism. These salicylates are easily absorbed by humans through the skin.  Although we warned Dr. Bennett of these problems as we encountered them, it was not until six weeks after the study was completed that we came to understand the enormity of the problem due to the widespread and increased presence of plant extracts or salicylates in cosmetics and lotions. (This was 1995).

It is also clear from reviewing the results of the patient wellness questionnaires that the screening process had failed to cull out the hypoglycemics, something we had spoken to Dr. Bennett about before the study began.  Without dietary correction, the probable 30 plus percent of patients with overlapping symptoms would not have felt better.

Just as important, we were constrained by the study design to give the patients a single dosage of guaifenesin.  That is, we could not raise the dose in patients who did not respond as we do in practice.  We knew that a dose of 1200 mg. a day would prove effective for some 80%, in theory so we selected it.  We felt that if we picked a higher dose patients who were low dose responders would experience too much pain and drop out of the study. Since we had only 20 patients taking guaifenesin that would have been a disaster. It is possible that a certain number of the 16 patients who completed the study were not taking enough guaifenesin to be effective.  (It could have been all of them or none of them --we do not know.  But in theory, 20 percent of this 16 would not have been taking enough).

So, it's important, as well, to point out the size of the Bennett study. Each group contained less than 18 patients at the end.  With such a small number of enrollees only a few needed to block or be at the wrong dose for the study to statically fail.

Dr. Bennett makes a point that no increase excretion of phosphate was seen in patients.  He also did not detect urate excretion (which is a documented finding in medical literature, not St. Amand speculation).  Since other researchers found an increase in urate excretion to the point where it is mentioned as fact in the guaifenesin monograph we do not know why Dr. Bennett did not see this. However, if the patients were blocked or at an insufficient dosage we would not expect any changes whatsoever to be seen.

As consultant, and the only one with experience in this treatment, I accept full responsibility for the errors. The reader may take this as a lame excuse for failure or accept our disclaimer. In our hands and in the hands of other physicians nationwide, results from guaifenesin have been too obvious to dispute.

Our sole purpose in writing this paper is to promote among patients and physicians an effective treatment for fibromyalgia. We seek to replace the dismal set of ineffective medications currently in use with a simple, nontoxic one that works at some very fundamental level. We are pleased to help any physician who is interested in using this approach.  We have a moral obligation to pursue dissemination of this information and experience.

R. Paul St. Amand, M.D.

Points of rebuttal to the conclusion of the Oregon Study:
Following the Oregon study, Dr. Bennett suggested that my charisma and warmth toward patients are the mechanisms of a placebo effect that leads to patient improvement. We appreciate the compliment, but we must decline and reject it as an explanation. Below are our seven points of rebuttal to Dr. Bennett's belief that we merely saw a placebo effect in this study.

1. Most patients come to us while taking many medications, including all varieties of herbs, antidepressants, tranquilizers, sleeping medications, anti-inflamatories, and muscle relaxants. No placebo effect ensued despite charismatic physicians along the way.

2. We have a list of  physicians worldwide who have had the same experience as we when they treat with guaifenesin.

3. Were a placebo effect the beneficial mechanism, it would seem strange that 80 percent of our patients fail to improve when we begin with only 300 mg of guaifenesin twice daily. When we reach 600 mg twice daily, 30 percent still fail. Some continue to fail at double that amount, and we increase to 3600 mg. per day. We would assume our level of care is equal at each dosage, and if charisma were responsible for patients improving, the percentage who improve would not be dosage-dependent.

4. Sometimes initiating guaifenesin at 600 mg. twice daily for a stoic patient results in a violent cycle that forces me to reduce the dosage. Improvement proceeds at the lowered amount as before. When raising the dosage on an improving patient to speed the process, even a small increment of 300 mg. daily will again precipitate acute symptoms. This sometimes forces us to withdraw to the lower, previous effective level and accept a less greedy road to recovery. No patient can anticipate these events.

5. Many drugs we tried in past years, such as allopurinol, failed, though none of us could predict results. Over the past 36 years, only the uricosuric agents, probenecid, Robinul, sulfinpyrazone, Flexin (no longer marketed) and guaifenesin (the weakest uricosuric of all) have worked, one better than the others, in the reverse order listed. The placebo effect should have been similar with each, since patients had no insight into comparative strengths of varying milligram dosages. For example, 1000 mg. of probenecid equaled 200 mg. of sulfinpyrazone that equaled 300 mg. of guaifenesin.

6. Only lesions we can feel are drawn on my maps, not areas of patient tenderness or presenting pain. Improvement on repeated mappings would have to extend the placebo effect to my hands as well.

7. Numerous patients reverse their improvement subjectively and objectively, shown on our mapping, when they inadvertently add something that would block guaifenesin. When told of this change, patients usually find offending substances at home. Neither of us has any way to predetermine this.

Conclusion:
Though highly theoretical, there are facts in this paper. Guaifenesin has proven our most effective medication to date. Any source of salicylate will block guaifenesin's benefit at a renal, tubular level, as it does in gout. All plants make varying amounts of salicylate. Salicylates are readily absorbed through the skin. Even small amounts in cosmetics and other topicals will negate or slow the effects of all agents we have used. We have many maps made during treatment that illustrate lesions becoming static or worsening in previously improving patients who, unwittingly, began using such preparations. Once warned, some patients find that over one-half of their usual skin preparations could have been deleterious. Susceptibility to blockage seems genetically determined and highly variable. Some patients are blocked by tiny amounts of offending agents, yet others improve despite moderate usage. Many patients are carbohydrate intolerant and must also be treated dietarily. These statements are factual and must be respected, or there will be no improvement. We contend that the outcome of the Oregon study of guaifenesin treatment of fibromyalgia patients was due to faults such as blockade by salicylates and the failure to exclude hypoglycemics from the study as well as the possibility of an inadequate dosage due to the small size of the cohort.

Clinical Bulletin of Myofascial Therapy, Vol. 2(4) 1997. c 1997 by The Haworth Press. Inc. All rights reserved.

Return to Research