# Study protocol for a randomized controlled trial to evaluate the effectiveness of an artificial intelligence-based health education accurately linking system based on traditional Chinese medicine body constitution in patients with chronic disease multimorbidity

**Authors:** Qingqing Liu, Yuqi Yang, Yi Hu, Xiaohua Yang, Suxiang Liu, Rantong Bao, Yibo Wu

PMC · DOI: 10.3389/fpubh.2026.1773974 · 2026-02-27

## TL;DR

This study tests an AI-based health education system that uses Traditional Chinese Medicine to personalize lifestyle advice for people with multiple chronic diseases.

## Contribution

The study introduces an AI system combining TCM constitution theory with modern technology for personalized chronic disease management.

## Key findings

- The AI-HEALS system will be tested for its feasibility and preliminary efficacy in a randomized controlled trial.
- The study will compare generalized and TCM-based AI interventions against standard care for chronic disease outcomes.
- Changes in TCM constitution classification will be explored as an additional outcome.

## Abstract

The prevalence of chronic disease multimorbidity among Chinese adults is high and presents a serious public health challenge. Personalized and accessible lifestyle interventions are key to improving long-term outcomes. Existing mobile health technologies often deliver generalized content, offering high accessibility but overlooking the potential of individualized interventions informed by Traditional Chinese Medicine (TCM) body constitution. TCM emphasizes identification of constitution types to develop more targeted and comprehensive personalized lifestyle guidance. Therefore, we developed the Artificial Intelligence–Health Education and Lifestyle Support system based on TCM body constitution (AI-HEALS) to explore its feasibility and assess its preliminary efficacy in improving clinical outcomes, enhancing self-management behaviors among patients with chronic disease multimorbidity.

This study adopts a three-arm parallel randomized controlled trial design. A total of 207 patients with hypertension, type 2 diabetes, and/or hyperlipidemia will be recruited and randomized in a 1:1:1 ratio into a routine chronic disease management group, a generalized AI-HEALS intervention group, or a TCM-based AI-HEALS intervention group. The routine management group will receive standard chronic disease care only. AI-HEALS is deployed via a WeChat Mini Program and features automated health-education content delivery, digital diaries for data logging, and intelligent Q&A functions. The backend houses a knowledge base covering chronic disease multimorbidity and TCM constitution-based wellness. The generalized AI-HEALS group will receive health-education content targeting chronic disease multimorbidity. The TCM-based AI-HEALS group will first complete a constitution assessment through the Mini Program, after which the system will deliver personalized health-management content tailored to their constitution type, drawing from the combined knowledge base of chronic disease multimorbidity and TCM constitution wellness. The intervention will last 3 months with follow-ups to month 6. The primary outcome is the change in systolic blood pressure from baseline to the end of intervention. Secondary outcomes include blood glucose and lipid levels, chronic-disease self-management behaviors, among others. Changes in TCM constitution classification will be treated as an exploratory outcome.

This study integrates TCM constitution theory with the AI-HEALS system to provide an innovative, personalized, and accessible management strategy for chronic disease multimorbidity, and seeks to explore the value of combining traditional medical theory with modern information technology. This protocol describes a randomized controlled trial designed to generate preliminary evidence on effectiveness.

https://www.chictr.org.cn/index.html, identifier ChiCTR2600118267

## Linked entities

- **Diseases:** type 2 diabetes (MONDO:0005148), hyperlipidemia (MONDO:0021187)

## Full-text entities

- **Diseases:** type 2 diabetes (MESH:D003924), HEALS (MESH:C563468), hypertension (MESH:D006973), hyperlipidemia (MESH:D006949), chronic disease (MESH:D002908)
- **Chemicals:** lipid (MESH:D008055), blood glucose (MESH:D001786)
- **Species:** Homo sapiens (human, species) [taxon 9606]

## Figures

1 figure with captions in the complete paper: https://tomesphere.com/paper/PMC12982435/full.md

---
Source: https://tomesphere.com/paper/PMC12982435