Saturday, June 11, 2011

Placebo Effects Foster Headache Therapy Success

Headache Pain Writing in the June edition of the Journal of Manipulative and Physiological Therapeutics, a research team from The Netherlands examine trials of therapies for tension-type headache and migraine to assess benefits among patients in placebo and “no treatment” control groups. Surprisingly, on average, more than a third of patients in control groups — not receiving any therapy expected to be helpful — experienced recovery from their headaches.

After examining 8 systematic reviews of randomized controlled trials published prior to 2005, the authors identified 119 trials encompassing 7,119 participants for analysis [de Groot et al. 2011]. In total, 85 trials were on tension-type headache (TTH) studying adults and children (n=5,913), and 34 were on migraine studying only children and adolescents (n=1,206). The number of participants per group ranged from 4 to 447, with a mean of about 54; a little more than a third of the studies (35.3%, mostly pharmacologic) were considered to be of high quality with low risk of bias.

All trials had either “no treatment” or placebo interventions in control groups that were classified into 6 subgroups.

  • Pharmacologic therapy trials were classified as either (1) acute, or (2) prophylactic [ie, preventative therapies].

  • Nonpharmacologic trials were designated as either (3) wait-list controls, (4) behavioral [eg, pseudo-EMG biofeedback, pseudo-relaxation, and pseudo-acupuncture], (5) general discussion/information groups, or (6) “others.”

Overall, results indicated a relatively high headache recovery rate in “no treatment” and placebo control groups (35.7%). Recovery was variously defined as a reduction of headache frequency, duration, and intensity, or being pain free 2 hours after medication administration, or merely global ratings of being pain free. More specifically, the following outcomes were reported…

  1. Control groups in pharmacological trials showed a higher response rate than nonpharmacologic trials (38.5% vs 15.0%).

  2. Children had a higher recovery rate than adults in control groups of pharmacological trials (45.1% vs 36.5%). Conversely, adults were more likely to recover in control groups of nonpharmacologic studies.

  3. Control group recovery rates in acute medication trials were higher compared with prophylactic trials (39.6 vs 32.8).

  4. Participants in waiting list and discussion/information control groups recovered more than those in the other types of nonpharmacologic/behavioral control groups.

  5. With all studies combined, participants in control groups of migraine trials recovered more than TTH trial patients (40.7% vs 34.1%). However, the migraine studies in this analysis did not include adults.

The authors conclude that recovery rates in control groups, largely due to placebo effects, were substantial. However, recovery rates varied between type of intervention and patients, and there was much diversity in the design of “no treatment” control groups.

COMMENTARY: The intention of control interventions — whether placebo therapies or “no treatment” interventions as described in this study — is to be neutral or ineffective. However, placebo effects are a complex, potent, and prevalent influence in most pain research studies [as previously discussed in other Pain-Topics UPDATES here].

Among the interesting findings in this present study was that placebo responses are higher in pharmacological trials than in behavioral or nonpharmacologic interventions. Along with that, it seems unexpected that merely being on a waiting list or attending an unrelated information session would exert greater placebo effects than sham acupuncture or biofeedback, or other pseudo-behavioral control interventions.

It does make sense that there might be higher responses with placebo interventions during acute versus prophylactic therapy trials, since results would be expected more immediately in the former. And, children responding more favorably to a pharmacologic placebo is understandable since behavioral approaches might require a higher level of cognitive understanding; hence, adults responded better than children to such control interventions as pseudo-relaxation.

Knowing that a substantial portion of patients improve without active treatment is important when considering the benefits and risks of daily headache therapies. The authors believe their study might support the possible benefits of a wait and see approach, comparable to a “no treatment” placebo, especially in the TTH population. At the same time, however, one might wonder what, if anything, does not exert a placebo effect to a greater or lesser extent.

Clearly, there is much still to learn about placebo effects and how they should be accounted for in pain research outcomes and/or harnessed for bolstering the success of clinical pain therapies. Unfortunately, the quality of evidence to date is disappointing; out of 119 trials identified for inclusion in this study by de Groot and colleagues only about a third were of good quality and many had low numbers of participants that probably biased outcomes.

SIDE NOTE: During a recent presentation, David W. Dodick, MD, President of the American Headache Society, stressed that an urgent need to conquer migraine is an infusion of more public funding to spur advances in treatment. The U.S. National Institutes of Health allocates less than $13 million a year to migraine research when, given the burden of this disease on individuals and society, that figure should be more like $260 million, Dodick asserted. About 36 million Americans suffer from migraine, more than have asthma or diabetes combined. The disease costs the U.S. more than $20 billion each year in direct medical expenses (e.g. doctor visits, medications) and indirect expenses (eg, missed work, lost productivity). [See Medical News Today report here.]

REFERENCE: de Groot FM, Voogt-Bode A, Passchier J, et al. Headache: The Placebo Effects in the Control Groups in Randomized Clinical Trials; An Analysis of Systematic Reviews. J Manipulative Physiol Ther. 2011(Jun);34(5):297-305 [abstract here].