Two things happen, in order. A thumb leaning into a tight trapezius at moderate pressure disrupts the spasm-pain-spasm loop holding it tight; the muscle lets go and blood returns. Then the nervous system follows: sustained predictable pressure drops heart rate and cortisol and triggers the oxytocin release that makes touch feel like care ^{1 2}. Moderate is the load-bearing word: light-touch styles don't reliably produce the shift, and pressure hard enough to make you brace has gone past useful ¹.

The shiatsu-specific evidence is thin; the family it sits in is solid. The best shiatsu-only review found eleven small trials, "promising but limited," for back and neck pain, anxiety, and sleep ³. But moderate-pressure bodywork generally has a real base: consistent short-term relief for low-back pain ⁴, gains in function across pain populations ⁵, a moderate single-session anxiety effect that roughly doubles over a multi-week course ⁶, and improved pain, anxiety, and sleep even in fibromyalgia ⁷. Mid-range effect sizes for a specific cluster. Not a breakthrough.

One session tests the practitioner, not the treatment. Courses beat single sessions on every endpoint measured ⁶.

Three timescales. Same day: within twenty minutes the shift is felt, heart rate falling and shoulders giving, and the evening is often the calmest of the week, with softer sleep that night ⁸. Across the course: a couple of points off a pain score, a meaningful drop on anxiety, less of the braced look chronic stress writes into a face ⁶. Beyond that, six-month data shows gains hold at monthly sessions ⁹. The long-tail story is "kept the gain at lower frequency," not an effect that keeps growing.