Plasma galectin 3 and heart failure risk in the P hysicians' H ealth S tudy

Aims We sought to test the hypothesis that plasma galectin 3 (Gal‐3) is positively associated with the risk of heart failure (HF) in male subjects. Methods and results While Gal‐3 has been reported as prognostic factor in HF patients, limited data are available on the role of Gal‐3 in the development of HF. We used a prospective nested‐case control study (n = 462 cases and 462 controls) within the Physicians' Health Study for current analyses. For each case of HF, we randomly selected one control among subjects that were alive and free of HF at the time of index case occurrence and matched on age, race, and time of blood collection. Gal‐3 was measured using ELISA and we used conditional logistic regression to compute adjusted odds ratios. Mean age at baseline was 58.3 y and median log‐Gal‐3 was 1.50 (IQR: 1.20–1.73) ng/ml. Cubic splines suggested a non‐linear relation between Gal‐3 and HF. Odds ratios (95% CI) for HF were 1.0 (ref), 0.89 (0.58–1.38), 1.08 (0.71–1.67), and 1.57 (1.03–2.39) across consecutive quartiles of Gal‐3 after adjustment for body mass index, diabetes, atrial fibrillation, hypertension, C‐reactive protein, alcohol, smoking, and exercise. The Gal‐3‐HF relation was seen for HF with and without antecedent coronary heart disease. Conclusions Our data are consistent with a positive non‐linear association between Gal‐3 and HF risk in male subjects.