Can a placebo actually change your body, not just your mood?
Yes, and the evidence shows it can affect concrete physiological measures. A 2025 meta-analysis of 15 randomized controlled trials involving 1,250 patients found that sham acupuncture (a placebo procedure) led to a statistically significant decrease in body mass index (BMI) in people with simple obesity [1]. The effect size was modest (standardized mean difference of 0.37), meaning the average person in the sham group lost some weight, but it wasn't a dramatic transformation. This shows that the ritual of treatment—needles, expectation, and the clinical setting—can trigger real metabolic changes.
However, the placebo effect isn't always beneficial. A 2023 meta-analysis on type 2 diabetes found that, on average, people in placebo groups actually saw a slight increase in their HbA1c (a key blood sugar marker) of 0.14% compared to baseline [3]. This means the placebo didn't improve their diabetes; in fact, it was associated with a small worsening. This highlights a crucial point: the placebo effect is context-dependent. It can produce measurable biological changes, but those changes aren't always therapeutic and can even be negative (the 'nocebo' effect).
How strong is the placebo effect, and what makes it stronger?
The strength of the placebo effect can be surprisingly large and is heavily influenced by context and belief. A key 2023 study showed that simply telling people a sham pain-relief machine was 'personalized' to their genetics and physiology made it significantly more effective [4]. In a double-blind experiment with 85 participants, those who believed the machine was tailored to them reported a greater reduction in pain intensity (a standardized effect of β = -0.20) compared to a control group who were told the machine was generally effective [4]. This demonstrates that the 'story' around a treatment—the idea that it's bespoke and high-tech—can amplify its biological impact.
This effect is not limited to elaborate machines. A 2023 study found that even 'imaginary pills'—where people simply imagined taking a pill—reduced test anxiety as effectively as real open-label placebos (pills known to be inactive) [2]. Both groups showed a large reduction in anxiety compared to a control group (Cohen's d = 0.71, a large effect) [2]. This suggests that the act of engaging with a treatment ritual, even a purely mental one, can trigger measurable psychological and physiological responses. The effect is also influenced by personality: the same study on personalized pain relief found that people with a high 'need for uniqueness' (a desire to feel distinct) benefited most from the personalized placebo [4].
What are the limits of the placebo effect?
While real, the placebo effect is not a magic bullet. Its magnitude varies greatly by condition and individual. The sham acupuncture study for obesity found that the effect was stronger with longer treatment durations and when combined with other interventions [1]. This means the placebo effect often works best as an add-on to real treatments, not as a replacement. Furthermore, the diabetes meta-analysis showed that the placebo effect on blood sugar was actually in the wrong direction (HbA1c increased), proving that placebos don't always produce positive biological changes [3].
It's also important to note that many placebo studies rely on subjective, self-reported outcomes like pain or anxiety, which are easier to influence than hard biological endpoints. While the studies above show effects on BMI and HbA1c, these are still relatively small. The effect on pain, while statistically significant, was also modest in absolute terms [4]. So, while the placebo effect has measurable biological mechanisms, it's typically not powerful enough to cure serious diseases on its own. Its power lies in enhancing the effects of active treatments and improving the patient's overall experience.
Sources used in this answer
Placebo response in sham acupuncture therapy trials for simple obesity: A systematic review and meta-analysis
Sham acupuncture led to a statistically significant decrease in BMI in 15 trials with 1,250 obesity patients (SMD 0.37), showing a measurable physiological placebo effect.
Imaginary pills and open-label placebos can reduce test anxiety by means of placebo mechanisms
Both open-label placebos and 'imaginary pills' significantly reduced test anxiety compared to a control group (Cohen's d = 0.71), demonstrating that even mental rituals can trigger placebo effects.
The power of the placebo effect in diabetes: A systematic review and meta-analysis
In type 2 diabetes, the placebo effect was associated with a slight increase in HbA1c (0.14%), indicating that placebo responses are not always beneficial and can be negative.
Author response: Presenting a sham treatment as personalised increases the placebo effect in a randomised controlled trial
Presenting a sham pain-relief machine as 'personalized' significantly increased its effectiveness (β = -0.20 for pain intensity), an effect moderated by the personality trait 'need for uniqueness'.
Decision letter: Presenting a sham treatment as personalised increases the placebo effect in a randomised controlled trial
This paper is the decision letter for paper 4, confirming the finding that framing a sham treatment as personalized increases its placebo effect.