The evidence for Shiatsu: a systematic review of Shiatsu and acupressure

BACKGROUND: Shiatsu, similar to acupressure, uses finger pressure, manipulations and stretches, along Traditional Chinese Medicine meridians. Shiatsu is popular in Europe, but lacks reviews on its evidence-base. METHODS: Acupressure and Shiatsu clinical trials were identified using the MeSH term 'acupressure' in: EBM reviews; AMED; BNI; CINAHL; EMBASE; MEDLINE; PsycARTICLES; Science Direct; Blackwell Synergy; Ingenta Select; Wiley Interscience; Index to Theses and ZETOC. References of articles were checked. Inclusion criteria were Shiatsu or acupressure administered manually/bodily, published after January 1990. Two reviewers performed independent study selection and evaluation of study design and reporting, using standardised checklists (CONSORT, TREND, CASP and STRICTA). RESULTS: Searches identified 1714 publications. Final inclusions were 9 Shiatsu and 71 acupressure studies. A quarter were graded A (highest quality). Shiatsu studies comprised 1 RCT, three controlled non-randomised, one within-subjects, one observational and 3 uncontrolled studies investigating mental and physical health issues. Evidence was of insufficient quantity and quality. Acupressure studies included 2 meta-analyses, 6 systematic reviews and 39 RCTs. Strongest evidence was for pain (particularly dysmenorrhoea, lower back and labour), post-operative nausea and vomiting. Additionally quality evidence found improvements in sleep in institutionalised elderly. Variable/poor quality evidence existed for renal disease symptoms, dementia, stress, anxiety and respiratory conditions. Appraisal tools may be inappropriate for some study designs. Potential biases included focus on UK/USA databases, limited grey literature, and exclusion of qualitative and pre-1989 studies. CONCLUSIONS: Evidence is improving in quantity, quality and reporting, but more research is needed, particularly for Shiatsu, where evidence is poor. Acupressure may be beneficial for pain, nausea and vomiting and sleep.