Independent prognostic value of left ventricular mass, diastolic function, and fasting plasma glucose: a population-based cohort stydy

OBJECTIVE: To explore the independent prognostic value of left ventricular (LV) mass, diastolic function, and fasting plasma glucose (FPG) for the prediction of incident cardiac events in a random population sample.

DESIGN AND METHOD: 415 women and 999 men aged 56-79 years, included between 2002-2006, underwent echocardiography based on groups defined by FPG, i.e. normal (NFG): FPG ≤ 6.0 mmol/L; impaired (IFG): FPG 6.1-6.9 mmol/L; and diabetes mellitus (DM): FPG ≥ 7.0 mmol/L, self-reported DM, and/or on anti-diabetic drugs. Additive prognostic value of FPG category and echocardiography (LV mass index (LVMI), LV hypertrophy (LVH), averaged E/é, and diastolic function graded as normal, grade 1, or grade 2 + 3 diastolic dysfunction) to a prediction model with traditional cardiovascular (CV) risk factors was assessed using Cox proportional hazards regression. Cardiac events were defined as myocardial infarction, coronary revascularization, or heart failure.

RESULTS: 37 % were classified as NFG, 26 % as IFG, and 37 % as DM. Median LVMI and E/é were 86 [74-102] g/m and 8 [6-10], respectively. Over a median follow-up time of 7.8 [7.2-8.7] years, 181 events occurred. The simple prediction model included age, gender, systolic blood pressure, heart rate, previous CV disease, and use of CV medication. Addition of averaged E/é (likelihood-ratio c 11.69, p < 0.001) or LVMI (likelihood-ratio c 4.52, p = 0.03) significantly improved the model, whereas FPG category did not (likelihood-ratio c 0.48, p = 0.79). Furthermore, we detected significant interactions between both FPG category and LVH (likelihood-ratio c 9.93, p = 0.007) and FPG category and diastolic function (likelihood-ratio c 11.65, p = 0.02) for prediction of cardiac events.

CONCLUSIONS: LVMI and E/é, but not FPG category provided additional adverse prognostic value on top of traditional CV risk factors. The combination of both glucometabolic and echocardiographic abnormalities was associated with a progressively greater risk of cardiac events.