Association between BMI measured within a year after diagnosis of type 2 diabetes and mortality.

Logue, J and Walker, JJ and Leese, G and Lindsay, R and McKnight, J and Morris, A and Philip, S and Wild, S and Sattar, N and Group, Scottish Diabetes Research Network Epidemiology (2013) Association between BMI measured within a year after diagnosis of type 2 diabetes and mortality. Diabetes Care, 36 (4). pp. 887-893. ISSN 0149-5992

Abstract

OBJECTIVE To describe the association of BMI with mortality in patients diagnosed with type 2 diabetes. RESEARCH DESIGN AND METHODS Using records of 106,640 patients in Scotland, we investigated the association between BMI recorded around the diagnosis of type 2 diabetes mellitus (T2DM) and mortality using Cox proportional hazards regression adjusted for age and smoking status, with BMI 25 to <30 kg/m2 as a referent group. Deaths within 2 years of BMI determination were excluded. Mean follow-up to death or the end of 2007 was 4.7 years. RESULTS A total of 9,631 deaths occurred between 2001 and 2007. Compared with the reference group, mortality risk was higher in patients with BMI 20 to <25 kg/m2 (hazard ratio 1.22 [95% CI 1.13–1.32] in men, 1.32 [1.22–1.44] in women) and patients with BMI ≥35 kg/m2 (for example, 1.70 [1.24–2.34] in men and 1.81 [1.46–2.24] in women for BMI 45 to <50 kg/m2). Vascular mortality was higher for each 5-kg/m2 increase in BMI >30 kg/m2 by 24% (15–35%) in men and 23% (14–32%) in women, but was lower below this threshold. The results were similar after further adjustment for HbA1c, year of diagnosis, lipids, blood pressure, and socioeconomic status. CONCLUSIONS Patients categorized as normal weight or obese with T2DM within a year of diagnosis of T2DM exhibit variably higher mortality outcomes compared with the overweight group, confirming a U-shaped association of BMI with mortality. Whether weight loss interventions reduce mortality in all T2DM patients requires study. Compared with normal weight, obesity is associated with increased all-cause mortality in the general population, as well as increased cardiovascular and cancer mortality (excluding smoking-related cancers) (1–3). However, the findings in patients with type 2 diabetes mellitus (T2DM) have been variable, with some studies reporting an increase in all-cause mortality with increasing body mass (4–6), but many studies finding no relationship (7–10) or a decrease in mortality with higher BMI (11). This may be due to limited power, as those studies describing no association all included <1,000 individuals (7–10) and limitations of the analytical approaches. In the general population, the relationship between BMI and mortality is U-shaped (1) with excess mortality seen at BMI <22.5 kg/m2 compared with 22.5–25 kg/m2. Consequently, if the lowest BMI group is used as the reference group, excess mortality in the obese range compared with the BMI group with the lowest mortality (i.e., the low normal range) is not always apparent. There is also the potential for reverse causality as a consequence of another medical condition causing both weight loss and death and confounding by smoking, which can lead to smokers having a lower BMI than nonsmokers yet a high risk of smoking-related disease. Higher BMI is associated with increased risk of coronary heart disease and cardiovascular mortality among people with T2DM (6,12), in keeping with results from the general population. As these studies (4–11) represent cross-sections of populations with established T2DM, there is the potential that diabetes control, treatment, and duration will have influenced BMI. These complexities and conflicting results have in some way deflected attention from the potential importance of weight reduction in patients with diabetes. The aim of the present work, based on data from the Scottish Care Information Diabetes Collaboration (SCI-DC) database, was to examine the relationship between BMI within a year after diagnosis of T2DM and all-cause, cardiovascular, cancer, and respiratory mortality in the medium term. The use of a large contemporary database has allowed closer examination of any graded association between obesity and mortality than among smaller studies. The availability of BMI recorded within a year after diagnosis limits the potential effects that diabetes treatment and glycemic control can have on BMI.