Summary: In a randomised trial in type 2 diabetes with cognitive impairment, moxibustion showed a temperature dose-effect: the highest range (44–46°C) improved cognitive scores and lowered HbA1c more than lower temperatures, although fasting glucose did not change.
PICO Summary
| Element | Detail |
|---|---|
| Population | 66 patients with type 2 diabetes and cognitive impairment; China. |
| Intervention | Moxibustion at GV20, GV14, GV24 at 44–46°C (high), 20 min/session, 3×/week for 3 months, added to glycaemic treatment. |
| Comparison | Moxibustion at 41–43°C (medium) or 38–40°C (low). |
| Outcome | The high-temperature group had higher MoCA and MMSE scores and better memory and short-term-memory measures than medium and low groups, and lower HbA1c than the low group. Total effective rates were 75% (high), 50% (medium), 15% (low). Fasting plasma glucose did not differ within or between groups. |
Moxibustion temperature and cognition in type 2 diabetes
RCT · type 2 diabetes · 3 months
Higher-temperature moxibustion (44-46°C) produced a markedly higher total effective rate for cognition than low temperature (75% vs 15%), with a clear temperature dose-effect. Small, single-centre evidence; treat as adjunctive only.
Expert Commentary
This is a methodologically interesting trial because, rather than asking the usual yes-or-no question about an alternative therapy, it tests a dose-response by varying moxibustion temperature, which is a more sophisticated design and yields a cleaner internal signal: cognitive scores and HbA1c improved progressively with higher temperature, and the effective rate climbed from 15% to 75%. A genuine dose-effect relationship is harder to explain away than a simple active-versus-nothing difference, so the internal consistency is a point in its favour. I would still interpret cautiously. The sample is small at 66 across three arms, the trial is single-centre, and an inherent difficulty with moxibustion is blinding, since patients and practitioners can perceive heat, so expectation effects are hard to exclude. It is also notable that fasting glucose did not move even though HbA1c did, which warrants a degree of circumspection about the glycaemic claim. Can I use this with my patients? Cautiously and as an adjunct at most. Cognitive impairment in diabetes has few good treatments, so I would not dismiss a patient’s interest in moxibustion, while being honest that this is preliminary, single-centre evidence and that glycaemic optimisation and vascular risk management remain the priorities.
References
Wei Y, Qu Y, Yuan A, et al. [Moxibustion at different temperatures for cognitive impairment in type 2 diabetes mellitus: a randomized controlled trial]. Zhongguo Zhen Jiu. 2025;45(9):1233–1240. doi:10.13703/j.0255-2930.20240903-k0006