Background: Heart failure is common in diabetes and ischaemic heart disease is the most likely link. Still, it has been suggested that the relation extends beyond such disease.
Methods: 7060 subjects with two or more visits in the Reykjavík Study were followed--during 30 years from 1967. All underwent oral glucose tolerance tests. Disease status was defined according to the glycaemic level and presence of heart failure. The incidence and predictive factors for these diseases were determined.
Findings: Age and sex standardized incidence of heart failure was 5.3/1000/year, of diabetes 4.6/1000/year and abnormal glucose regulation 12.6/1000/year. Body mass index (BMI) and fasting glucose predicted the development of these conditions (p<0.001). Increasing fasting glucose by 1 mmol/l increased the risk for heart failure by 14% (p=0.04) after adjusting for IHD, BMI and other risk factors for CVD. There was a strong association between diabetes and heart failure, OR 3.0 (2.3-4.0), and abnormal glucose regulation and heart failure, OR 1.8 (1.5-2.3). Diabetes and heart failure were, however, not independent predictors of each other.
Interpretation: There was an independent relationship between increases in fasting glucose and development of heart failure. BMI was a strong predictor of heart failure. Although fasting glucose and BMI were significant risk factors for glucose disturbances and heart failure the conditions themselves did not independently predict each other.