A couple of comments (not a medical expert): It's likely that the placebo/nocebo effect is confounded by regression to the mean - a typical example are disease

A couple of comments (not a medical expert): It's likely that the placebo/nocebo effect is confounded by regression to the mean - a typical example are diseases with some sort of cyclical pattern, e.g. some forms of depression. But there is strong evidence that there is an effect beyond that: E.g. cancer patients die significantly sooner when they get a negative prognosis [1]. Acupuncture is so effective for treating certain forms of chronic (!) pain that in some cases it's covered by German health insurance since 2007 [2]. If you have been treated in a hospital for pain related issues, e.g. after a surgery, it's not unlikely you have received placebo instead of pain killers (depends a lot on jurisdiction). Regarding [3] specifically, the obvious elephant in the room is that 'absence of evidence is not evidence of absence'. Also see the refutations pointed out by user PaulDavisThe1st [4]. What I take away from that study is that yes, in some contexts, regression to the mean may account for a major portion of the placebo effect. [1] https://doi.org/10.1016/S0140-6736(73)91754-6 https://doi.org/10.1016/S0140-6736(73)91754-6 [2] https://www.verbraucherzentrale.de/wissen/gesundheit-pflege/aerzte-und-kliniken/akupunktur-wann-zahlt-die-kasse-12462 https://www.verbraucherzentrale.de/wissen/gesundheit-pflege/... [3] https://pubmed.ncbi.nlm.nih.gov/12535498/ https://pubmed.ncbi.nlm.nih.gov/12535498/ [4] https://news.ycombinator.com/item?id=29028133 https://news.ycombinator.com/item?id=29028133