Type 2 Diabetes Subgroups, Risk for Complications, and Differential Effects Due to an Intensive Lifestyle Intervention

Abstract

Objective: We reevaluated the Action for Health in Diabetes (Look AHEAD) intervention, incorporating diabetes subgroups, to identify whether intensive lifestyle intervention (ILI) is associated with differential risk for cardiovascular disease (CVD) by diabetes subgroup.

Research design and methods: In the Look AHEAD trial, 5,145 participants, aged 45-76 years, with type 2 diabetes (T2D) and overweight or obesity were randomly assigned to 10 years of ILI or a control condition of diabetes support and education. The ILI focused on weight loss through decreased caloric intake and increased physical activity. To characterize diabetes subgroups, we applied k-means clustering to data on age of diabetes diagnosis, BMI, waist circumference, and glycated hemoglobin. We examined whether relative intervention effects on the trial's prespecified CVD outcomes varied among diabetes subgroups.

Results: We characterized four subgroups related to older age at diabetes onset (42% of sample), poor glucose control (14%), severe obesity (24%), and younger age at diabetes onset (20%). We observed interactions (all P < 0.05) between intervention and diabetes subgroups for three separate composite cardiovascular outcomes. Randomization to ILI was associated with increased risk for each cardiovascular outcome only among the poor-glucose-control subgroup (hazard ratio >1.32). Among the three other diabetes subgroups, ILI was not associated with increased risk for CVD.

Conclusions: Among overweight and obese adults with T2D, a lifestyle intervention was associated with differential risk for CVD that was dependent on diabetes subgroup. Diabetes subgroups may be important to identify the patients who would achieve benefit and avoid harm from an ILI.

Trial registration: ClinicalTrials.gov NCT00017953.

Figure 1

Forest plot of HRs and 95% CIs for the ILI in the Look AHEAD trial primary and secondary outcomes stratified by diabetes subgroup. Older onset, older age at diabetes onset; younger onset, younger age at diabetes onset.

Figure 2

Cumulative hazard curves for the primary cardiovascular composite end point according to intervention arm and diabetes subgroup (inset includes expanded y-axis). Gluc. cont., glucose control; older onset, older age at diabetes onset; younger onset, younger age at diabetes onset.