Abstract

Although several reports have described early changes of cardiac structure and function in diabetic patients, controversy persists regarding the existence of a clinically distinct diabetic cardiomyopathy. To this end, sex-specific linear regression analyses were used to examine the contribution of diabetes mellitus and glucose intolerance to age-adjusted echocardiographic parameters in 1,986 men (mean age 48 years) and 2,529 women (mean age 50 years) from the original Framingham Study cohort and the Framingham Offspring Study. Subjects with evidence of cardiovascular disease at the time of echocardiogram were excluded. Diabetics had higher heart rates than nondiabetics (67.9 vs 64.0 beats/min (p = 0.002) in men, and 73.1 vs 68.3 beats/min (p = 0.004) in women). Diabetic women had increased left ventricular (LV) wall thickness (18.7 vs 17.1 mm, p < 0.001), relative wall thickness (0.403 vs 0.377, p = 0.008), LV end-diastolic dimension (46.9 vs 45.7 mm, p = 0.03) and LV mass corrected for height (100.4 vs 82.2 g/m, p < 0.001). Women with glucose intolerance showed similar, less significant trends (p = 0.007 for wall thickness, p < 0.01 for LV mass). In diabetic men, fractional shortening was slightly reduced (0.355 vs 0.360, p < 0.05). In a multivariate model that included potentially confounding factors, diabetes remained an independent contributor to LV mass (p = 0.004) and wall thickness (p = 0.008) in women. In a separate linear regression model, which assessed the association of age with LV mass, the age-coefficient for diabetic women was much higher than that for nondiabetics (13.6 vs 6.6 g/m per 10-year increment in age). In conclusion, many echocardiographic differences in diabetics are explained by the contributions of other concomitant factors such as obesity, hypertension and smoking, but an independent association of diabetes with LV mass and LV wall thickness is evident in women.