Effects of a Lifestyle Intervention to Prevent Deterioration in Glycemic Status Among South Asian Women With Recent Gestational Diabetes: A Randomized Clinical Trial

Abstract

Importance: Women with recent gestational diabetes (GDM) have increased risk of developing type 2 diabetes.

Objective: To investigate whether a resource-appropriate and context-appropriate lifestyle intervention could prevent glycemic deterioration among women with recent GDM in South Asia.

Design, setting, and participants: This randomized, participant-unblinded controlled trial investigated a 12-month lifestyle intervention vs usual care at 19 urban hospitals in India, Sri Lanka, and Bangladesh. Participants included women with recent diagnosis of GDM who did not have type 2 diabetes at an oral glucose tolerance test (OGTT) 3 to 18 months postpartum. They were enrolled from November 2017 to January 2020, and follow-up ended in January 2021. Data were analyzed from April to July 2021.

Interventions: A 12-month lifestyle intervention focused on diet and physical activity involving group and individual sessions, as well as remote engagement, adapted to local context and resources. This was compared with usual care.

Main outcomes and measures: The primary outcome was worsening category of glycemia based on OGTT using American Diabetes Association criteria: (1) normal glucose tolerance to prediabetes (ie, impaired fasting glucose or impaired glucose tolerance) or type 2 diabetes or (2) prediabetes to type 2 diabetes. The primary analysis consisted of a survival analysis of time to change in glycemic status at or prior to the final patient visit, which occurred at varying times after 12 months for each patient. Secondary outcomes included new-onset type 2 diabetes and change in body weight.

Results: A total of 1823 women (baseline mean [SD] age, 30.9 [4.9] years and mean [SD] body mass index, 26.6 [4.6]) underwent OGTT at a median (IQR) 6.5 (4.8-8.2) months postpartum. After excluding 160 women (8.8%) with type 2 diabetes, 2 women (0.1%) who met other exclusion criteria, and 49 women (2.7%) who did not consent or were uncontactable, 1612 women were randomized. Subsequently, 11 randomized participants were identified as ineligible and excluded from the primary analysis, leaving 1601 women randomized (800 women randomized to the intervention group and 801 women randomized to usual care). These included 600 women (37.5%) with prediabetes and 1001 women (62.5%) with normoglycemia. Among participants randomized to the intervention, 644 women (80.5%) received all program content, although COVID-19 lockdowns impacted the delivery model (ie, among 644 participants who engaged in all group sessions, 476 women [73.9%] received some or all content through individual engagement, and 315 women [48.9%] received some or all content remotely). After a median (IQR) 14.1 (11.4-20.1) months of follow-up, 1308 participants (81.2%) had primary outcome data. The intervention, compared with usual care, did not reduce worsening glycemic status (204 women [25.5%] vs 217 women [27.1%]; hazard ratio, 0.92; [95% CI, 0.76-1.12]; P = .42) or improve any secondary outcome.

Conclusions and relevance: This study found that a large proportion of women in South Asian urban settings developed dysglycemia soon after a GDM-affected pregnancy and that a lifestyle intervention, modified owing to the COVID-19 pandemic, did not prevent subsequent glycemic deterioration. These findings suggest that alternate or additional approaches are needed, especially among high-risk individuals.

Trial registration: Clinical Trials Registry of India Identifier: CTRI/2017/06/008744; Sri Lanka Clinical Trials Registry Identifier: SLCTR/2017/001; and ClinicalTrials.gov Identifier: NCT03305939.

Conflict of interest statement

Conflict of Interest Disclosures: Dr Zoungas reported participation in expert committees or educational meetings outside the submitted work on behalf of Monash University, which received payment from Boehringer-Ingelheim, Eli Lilly, Sanofi, AstraZeneca, Novo Nordisk, and Merck Sharp and Dohme. No other disclosures were reported.

Figures

Figure 1.. Planned Intervention

There were 84 text or voice messages and 9 phone calls from facilitators planned for each participant.

Figure 2.. Participant Flowchart

End-of-study visits were conducted at least 12 months postrandomization between July 1, 2020, and January 31, 2021. FPG indicates fasting plasma glucose; HbA1c, hemoglobin A1c; OGTT, oral glucose tolerance test.

Figure 3.. Kaplan-Meier Plot for Primary Outcome

The primary analysis consisted of a survival analysis of time to change in glycemic status at or prior to the final patient visit, which occurred at varying times after 12 months for each patient. Proportions shown in the plot are a function of the number of patients experiencing an event by a certain time in the subset of patients who were still at risk at that time (ie, after excluding patients who had already experienced an event or been censored). The curve in the usual care group decreases to 0% by 36 months, indicating that by 36 months, all patients had been censored or experienced an event (ie, no patient was still at risk).