Background: Type 2 diabetes (DM2) is associated with a greater risk of heart failure. The mechanisms underlying this association remain controversial and include diabetes-associated hypertension and obesity, impaired small and large artery function, and a distinct metabolic cardiomyopathy related to hyperglycaemia/
hyperinsulinaemia. The proximate causes of heart failure are left ventricular (LV) systolic dysfunction (SDF)
and diastolic dysfunction (DDF). We investigated, in a
population-based cohort (n=746), the association between glucose tolerance status and SDF and DDF .
Methods and results: The study population consisted of
274 individuals with normal glucose metabolism (NGM),
174 with impaired glucose metabolism (IGM) and 298 with DM2 (mean age 68.5 years). All participants underwent an LV echocardiogram. SDF was defined as ejection fraction <55%. DDF was determined by a sum score of peak A velocity (abnormal, ≥97 cm/s), the difference between A<sub>pv</sub> and A<sub>mv</sub> duration (≥41 ms), and left atrial volume (≥57 ml), where cut-off values were based upon the 90th percentile in NGM. In addition, we analysed the ratio of early to late diastolic filling (E/A ratio) on a continuous scale using linear regression analyses. The age- and sex-standardised prevalences in NGM, IGM and DM2 were 13, 14 and 30%
for SDF , and 26, 36 and 47% for DDF (P<sub>(trend)</sub> for both <0.001). After adjustment for sex, age, hypertension, body mass index, prior cardiovascular disease and (micro) albuminuria, DM2 was significantly associated with both SDF (odds ratio (95% CI) 2.04 (1.24 to 3.36)) and DDF (2.42 (1.63 to 3.60)) (90th percentile definition). This was also true for the analyses with the E/A ratio on a continuous scale (regression coefficient β (95% CI) -0.05 (-0.09 to -0.01). After adjustment for sex, age, hypertension, body mass index, prior cardiovascular disease and (micro) albuminuria IGM was not significantly associated with SDF (odds ratio (95% CI) 1.04 (0.58 to 1.88)) or DDF (1.33 (0.86 to 2.06)) using the definition based upon the 90th percentile. However, IGM was significantly associated with DDF if the E/A ratio was analysed on a continuous scale (regression coefficient β (95% CI) -0.05 (-0.10 to -0.01). Additional adjustment for brachial artery flow-mediated vasodilation or arterial stiffness, as measures of large artery function, did not materially alter the results. Hyperglycaemia and hyperinsulinaemia together explained ~30% of the association of DM2 with SDF and ~40% of that with DDF.
Conclusion: DM2 is independently associated with a 2.0-fold greater risk of SDF and a 2.4-fold greater risk of DDF. IGM was not associated with SDF , and the association with DDF was limited to the E/A ratio. These observations may therefore explain the increased risk of systolic and diastolic heart failure in elderly individuals with DM2.