Weight change in diabetes and glycemic and blood pressure control

Abstract

Objective: Weight loss in type 2 diabetes is undisputedly important, and data from community settings are limited. We evaluated weight change and resulting glycemic and blood pressure control in type 2 diabetic patients at an HMO.

Research design and methods: Using electronic medical records, this retrospective cohort study identified 2,574 patients aged 21-75 years who received a new diagnosis of type 2 diabetes between 1997 and 2002. We estimated 3-year weight trajectories using growth curve analyses, grouped similar trajectories into four categories using cluster analysis, compared category characteristics, and predicted year-4 above-goal A1C and blood pressure by group.

Results: The weight-trajectory groups were defined as higher stable weight (n = 418; 16.2%), lower stable weight (n = 1,542; 59.9%), weight gain (n = 300; 11.7%), and weight loss (n = 314; 12.2%). The latter had a mean weight loss of 10.7 kg (-9.8%; P < 0.001) by 18 months, with near-complete regain by 36 months. After adjusting for age, sex, baseline control, and related medication use, those with higher stable weight, lower stable weight, or weight-gain patterns were more likely than those who lost weight to have above-goal A1C (odds ratio [OR] 1.66 [95% CI 1.12-2.47], 1.52 [1.08-2.14], and 1.77 [1.15-2.72], respectively). Those with higher stable weight or weight-gain patterns were more likely than those who lost weight to have above-goal blood pressure (1.83 [1.31-2.57] and 1.47 [1.03-2.10], respectively).

Conclusions: A weight-loss pattern after new diagnosis of type 2 diabetes predicted improved glycemic and blood pressure control despite weight regain. The initial period postdiagnosis may be a critical time to apply weight-loss treatments to improve risk factor control.

Figure 1

The four dominant weight-change trajectory patterns in the three years after diagnosis of type 2 diabetes. •, group with higher stable weight (n = 418; 16.2%); ▵, weight-loss group (n = 314; 12.2%); ▴, weight-gain group (n = 300; 11.7%); ▪, group with lower stable weight (n = 1,542; 59.9%); and %▵, percentage weight change from baseline. **Estimated baseline weights differ from mean baseline weights in Table 1. *Percentage weight changes (month 0–month 18 and month 0–month 36) are significantly different from zero (P < 0.001 based on Wilcoxon's signed-rank test).