The study by Henkin et al. sought to determine the effectiveness of transcranial magnetic stimulation (TMS) for treating chronic phantosmia and global oral phantogeusia. These chemosensory problems are extremely debilitating and not uncommon and often occur following upper respiratory infections, head trauma, and drug side effects . Although TMS has shown promise in treating depression, traumatic injury, and stroke , this study is the first to apply TMS to chemosensory disturbance.

These authors studied 17 patients with such problems on two occasions separated by 4 to 32 weeks. On each of these occasions, 2 “sham” TMS procedures were immediately followed by a “real” TMS procedure. The first sham consisted of presenting 20 stimuli at 1- to 5-second intervals sequentially over the right shoulder, then the left shoulder, and then to the back of the neck, all between 25% and 40% of maximum output (1.5 T). The second sham was a series of stimulations (10%–15% of maximum output) to the front, back, and sides of the skull. The “real” stimulation was directed to the same skull sights in the same order with the stimulator at 40%–55% of maximum output. Detailed measures of smell and taste function were taken at the beginning and end of each of the two 3-stimulation sessions. Brief assessments of the magnitude of the phantosmias “and/or” phantogeusias “and/or” responses to a single odorant presentation were assessed after each sham or real stimulation. If improvement was noted at one site, stimulation was continued until no further changes were noted.

This study, although pioneering, suffers from multiple methodological problems, including lack of basic controls, analytical rigor, and firm hypotheses. The study only provided anecdotal information on potential effects of the sites that were stimulated, and multiple variables and covariates were measured without consideration of interaction effects or correlations. Moreover, no correction was applied for multiple comparisons when examining 24 t tests for the different smell and taste measures.

A major problem with this study was the lack of double or even single blinding. The so-called “sham” procedures are particularly problematic. Despite the difficulties inherent in finding a sufficient TMS control condition or “sham” , efforts to lessen the likely placebo effect are absent from the study (eg, randomization, control sites over the cortex, and questioning of the patient as to which condition “felt” most effective). Both sham procedures were perceivably distinct from the “real” procedure, being either at a different location or employing a smaller magnitude stimulus. The large and distinctive physiological effect of the “treatment” and the fact that the experimenter was aware of the condition may well have induced a significant placebo effect. We suggest that the effect reported by Henkin et al. is, indeed, just that.

Although it is not unreasonable to assume that global magnetic stimulation might induce neurochemical or cortical electrical changes that modulate olfactory sensory function, this study—in light of its methodological and analytical problems—fails to adequately assess the efficacy of TMS on the chemosensory disturbances under study. Future studies with appropriate controls are needed to address this issue.