Effectiveness of an AI-Enabled Mobile Intervention on Lifestyle Behaviors and Maternal-Infant Health

Summary

The goal of this pragmatic randomized controlled trial is to evaluate the effectiveness of an AI-enabled mobile health management application ("Better Pregnancy" app), grounded in the Theory of Planned Behavior, in preventing gestational diabetes mellitus (GDM) among pregnant women at high risk of GDM, and to assess its impact on maternal and infant health outcomes.

Study Population: Eligible participants are pregnant women aged 18-49 years, with a singleton pregnancy, gestational age \<12 weeks at enrollment, and at least one GDM risk factor (age ≥35 years, pre-pregnancy BMI ≥24 kg/m², family history of diabetes, previous history of GDM, prior delivery of a macrosomic infant \[birth weight ≥4000 g\], or polycystic ovary syndrome). Participants must have no heart, liver, or kidney diseases, use an Android smartphone, and provide written informed consent.

The main questions it aims to answer are:

Can the AI-driven mobile lifestyle intervention reduce the incidence of GDM in high-risk pregnant women?

Does the intervention improve lifestyle behaviors (diet, physical activity, sleep) and glycemic control (measured by continuous glucose monitoring) in participants?

What is the adherence, acceptability, and cost-effectiveness of this intervention in real-world clinical settings?

Comparison: Researchers will compare two groups of pregnant women: the intervention group (receiving the AI-enabled "Better Pregnancy" app plus routine antenatal care) and the control group (receiving routine antenatal care alone).

Participants will:

Enroll in the first trimester (\<12 weeks of gestation) and complete a baseline lifestyle and health status questionnaire

In the intervention group: watch 7 cognitive-attitudinal videos within the first week, then complete 12 weeks of AI-driven personalized intervention (daily check-ins, receiving tailored recommendations on diet, physical activity, sleep, etc.)

During the intervention period, participants in the intervention group will wear a continuous glucose monitoring device for 7 days to assess the impact of the intervention on glycemic fluctuations

Complete the first follow-up lifestyle and health status assessment after the intervention (approximately in the second trimester)

Undergo an oral glucose tolerance test during the second trimester (24-28 weeks)

Complete the second follow-up lifestyle and health status assessment in the third trimester (≥28 weeks)

Complete the third follow-up lifestyle and health status assessment at 42 days postpartum, along with collection of delivery outcomes and infant health information

A subset of participants may voluntarily provide blood and stool samples for mechanistic studies

Conditions

• Mobile Health Apps
• Lifestyle Intervention
• Prevention
• Multi-component Based Behavioral Intervention

Interventions

BEHAVIORAL

Multi-component lifestyle intervention

Participants receive a 12-week AI-powered intervention via the "PregSelfCare" app plus routine antenatal care. The intervention includes: (1) watching 7 theory-based videos and completing an attitude questionnaire in week 1; (2) daily check-ins (mood, water, fruit, bowel movements, sunlight, weight) and receiving AI-driven personalized recommendations (diet photo feedback, step goals, sleep advice, emotion-relief videos, supplement reminders) during weeks 1-12, with content dynamically adjusted based on baseline lifestyle assessment and adherence; (3) wearing a continuous glucose monitor for 7 days; (4) completing a satisfaction questionnaire after the intervention (week 12), with a subset invited for semi-structured interviews.

Sponsors & Collaborators

• Fangbiao Tao

  lead OTHER

Study Design

Allocation

RANDOMIZED

Purpose

PREVENTION

Masking

SINGLE

Model

PARALLEL

Eligibility

Min Age

18 Years

Sex

FEMALE

Healthy Volunteers

Yes

Timeline & Regulatory

Start

2026-08-10

Primary Completion

2027-12-31

Completion

2027-12-31

Countries

• China

Study Locations