In recent years, research using the principles of evidence-based medicine (EBM) has been required for Kampo medicine. In the present study, we focus on Yokukansan, for which clinical trials have been accumulated, and study it from a methodological perspective. Research on the effects of Yokukansan on behavioral and psychological symptoms of dementia (BPSD) began with an observational study by Hara on 48 older patients who showed emotional disturbance. He found marked improvement in 32 patients (67%), improvement in 11 patients (23%), slight improvement in three patients, and no improvement in two patients, and reported that it was particularly effective for treatment of insomnia, irritability, excitement, and delirium. Randomized trials began in 2005 when Iwasaki, et al. reported an observer-blind, single-blind study. Fifty-two dementia patients, including those with Alzheimer's disease (AD) and dementia with Lewy bodies (DLB), were randomized to receive 7.5 g of Yokukansan, and the effects were evaluated by the Neuropsychiatric Inventory (NPI) after 4 weeks. Subsequently, Mizukami, et al. conducted an 8-week crossover study in collaboration with 20 institutions in the Kanto area of 108 patients with dementia, including AD and DLB, who were randomly divided into two groups: Group A took 7.5 g of Yokukansan for the first 4 weeks and Group B took it for the second 4 weeks. Matsuda, et al. conducted a meta-analysis of four reports that included these two reports, and found that Yokukansan had effects on hallucinations, delusions, and agitation. These reports led to the inclusion of Yokukansan as a treatment option for BPSD in treatment guidelines; however, as there was a lack of placebo-controlled trials, the scientific evidence was considered insufficient. Furukawa, et al. conducted the first multicenter, placebo-controlled trial of Yokukansan. A Yokukansan group (n=75) and a placebo group (n=70) were evaluated for changes in BPSD after 4 weeks. Although there was no significant difference in NPI scores between the two groups, there was an improvement in symptoms in patients with agitation and hallucinations. Furthermore, Matsunaga, et al. conducted a meta-analysis using five reports, including that of Furukawa et al. and reported effects on hallucinations, delusions, and excitement. However, a meta-analysis of three reports for AD only found no significant effect.
 It is necessary to continue to study Kampo medicine to accumulate a high level of evidence. As it is known that the efficacy rate is high in cases with rectus abdominis muscle tension, it may be useful to incorporate the perspective of therapy in accordance with oriental medical treatment into clinical trials of Yokukansan.
 Author's abstract