LAUSR

have read the paper entitled ‘Efficacy and feasibility of antidepressants for the prevention of migraine in adults: a meta-analysis’ [1] written by Xu et al. which was published in the European Journal of Neurology in May 2017 [1]. The authors reported that antidepressants had a significant advantage over placebo in reducing the migraine frequency or index of adults with a standardized mean difference (SMD) of −0.79 (−1.13, −0.45) and they concluded that antidepressants are effective in the prophylaxis of migraine in adults. Although the data reported by Xu et al. are interesting, it is important to emphasize that, although an SMD of −0.79 is statistically significant, clinically these differences may be unimportant. The reader should distinguish the difference between statistical significance and clinical importance. A larger sample size, higher standard deviation and bigger means difference can increase the chance of finding a statistically significant SMD [2, 3]. In meta-analyses, the larger sample size arising from pooling individual studies can easily lead to a significant P value. Therefore, these conclusions should be interpreted with caution because clinical judgements have been overlooked.