Does Weight at Diabetes Diagnosis Affect Mortality?

Gregory A. Nichols, PhD


October 29, 2012

Association of Weight Status With Mortality in Adults With Incident Diabetes

Carnethon MR, De Chavez PJD, Biggs ML, et al
JAMA. 2012;308:581-590

Study Summary

This pooled analysis of 5 well-known longitudinal cohort studies (Atherosclerosis Risk in Communities, Cardiovascular Health Study, Coronary Artery Risk Development in Young Adults, Framingham Offspring Study, and Multi-Ethnic Study of Atherosclerosis) studied 2625 individuals with diabetes to determine whether weight at diagnosis was associated with mortality.

Participants who were free of diabetes at baseline but subsequently had fasting glucose ≥126 mg/dL or self-reported use of anti-hyperglycemic medications were followed from when diabetes was newly identified until they died, reached the end of their particular cohort study, or were lost to follow-up.

The primary outcome was all-cause mortality, but the investigators also separately looked at cardiovascular and noncardiovascular mortality. The main analysis variable of interest was weight status defined in body mass index (BMI) categories of normal weight (18.5-24.9 kg/m2), overweight (25-29.9), and obese (≥ 30). Cox regression analyses initially adjusted for age, sex, race, and education, then further adjusted for waist circumference, total cholesterol, systolic blood pressure, HDL cholesterol, and smoking status.

Mean age ranged from 41 years in the Coronary Artery Risk Development in Young Adults study to 76 years in the Cardiovascular Health Study. Half of the participants across all cohorts were women. Across cohorts, 293 participants (11.2%) were of normal weight when diagnosed with diabetes.

Total mortality was higher in normal-weight participants (284.8 per 10,000 person-years) with diabetes compared with rates among overweight or obese participants (152.1 per 10,000 person-years). Following adjustment for all covariates, participants with normal-weight diabetes experienced a significantly elevated total mortality (HR, 2.08; 95% CI, 1.52-2.85) and noncardiovascular mortality (2.32; 1.55-3.48). The hazard for cardiovascular mortality was elevated but not statistically significant (1.52; 0.89-2.58). Participants with normal-weight diabetes had higher mortality from all causes than overweight/obese participants across strata of sex, age, race, and smoking.

When weight at the time of diabetes was stratified into 3 levels, total mortality was higher in normal-weight as compared with overweight participants, whereas mortality hazards did not differ between obese and overweight.


We tend to associate diabetes with obesity and blame the explosion of diabetes prevalence largely on the obesity epidemic. Indeed, recommended criteria for testing for diabetes in asymptomatic adults under age 45 key in on the individual being considered overweight.[1]

However, nearly a quarter of normal-weight adults in the United States have at least 2 cardiometabolic risk factors.[2] In the current study and depending on the cohort, between 9% and 21% of the participants with incident diabetes were of normal weight.

Thus, it is apparent that risk for diabetes, while clearly higher among overweight/obese patients, should not be ignored among those of normal weight. Furthermore, the present analysis demonstrates that diabetes in normal-weight individuals is particularly lethal.

Unfortunately, Carnethon and colleagues could not account for glucose levels. Following the ACCORD results which suggested increased mortality among those in the intensive glycemic control arm, as well as a subsequent observational analysis associating low A1c levels with higher mortality risk,[3,4] one wonders how glycemia might differentially affect the relationship between mortality and weight. That is an important topic for future study.

In the meantime, it is critical that patients not misinterpret these findings and decide that gaining weight following diagnosis of diabetes might be beneficial. Weight reduction for most diabetic patients remains the cornerstone of diabetes care.