On the relationship between beta-blockers,migraines and post-traumatic stress disorder: Response to Peterlin et al.

Dear Sir,

We appreciate the intriguing comments of Dr Peterlin and colleagues about our study (1) and their astute understanding of the complex relationship between post-traumatic stress disorder (PTSD) and migraine headache. As the authors point out, compared to our previous studies evaluating back pain (28% co-prevalence rate of psychiatric illness), neck pain (26% prevalence rate), and non-cardiac chest pain (25% rate) in evacuated service members, the incidence of coexisting psychiatric disease in general (40%) and PTSD in particular (15%) is considerably higher in headache evacuees (2–4). However, when placed in context of the much higher rate of battle-related injuries in the headache patients (31%) compared to spinal pain (<5%) and non-cardiac chest pain (18%), these findings are not surprising. When comparing the rate of PTSD in various headache cohorts, it was actually lower in service members diagnosed with migraines (5.9%) than it was in those evacuated for tension headaches (8.1%). The odds ratio for having PTSD in someone with migraines compared to other headaches was 0.25 (95% CI, 0.15 to 0.41), which favorably compared to 0.49 for tension headache (95% CI, 0.26 to 0.92).

The hypothesis that the comorbid conditions for which beta-blockers are utilized, or conceivably even the medications themselves, may predispose patients to PTSD is an intriguing prospect with significant clinical implications. This is especially true because beta-blockers are sometimes used to alleviate the symptoms of PTSD, although the evidence for this indication is weak (5,6). Whereas this paradox may seem incompatible, the list of medications used to treat a condition that can cause or worsen the same condition is substantial, and includes antidepressants and suicide (7), chemotherapeutics and neuropathy (8), anti-arrhythmics and arrhythmias (9), and analgesics and headaches (10).

Inherent in any retrospective study, our methodology is incapable of establishing a cause-and-effect relationship between demographic and clinical variables, treatment parameters and outcomes. Other methodological factors that could have influenced any of these variables are the lack of standardization for diagnosis and frank misdiagnosis (e.g. tension headaches misdiagnosed as migraines), treatment, and disparate criteria for returning a service member to duty. Although the relationship between beta-blockers and PTSD clearly warrants further investigation, considering our uncertainties and the over-arching fact that the evidence in favor of beta-blockers for refractory migraines is better than the indirect evidence that they may predispose patients to migraines, we would recommend exercising caution with respect to withholding beta-blocker therapy in combat-wounded individuals who may benefit.