Background: No information is available regarding the association between low plasma glucose levels and cause-specific and all-cause mortality in patients with coronary artery disease (CAD). We aimed to investigate the relationship between hypoglycaemia and all-cause, cardiovascular and cancer mortality in a large population of patients with CAD.
Design: Patients were recruited from the BIP (Bezafibrate Infarction Prevention) registry, a secondary prevention prospective multicentre randomized, placebo-controlled, double-blind trial aimed to assess the efficacy of bezafibrate in reduction of coronary events.
Methods: The study included 14,670 CAD patients aged 45-74, divided into six groups: (1) hypoglycaemic (up to 69 mg/dl); (2) low normal (70-79 mg/dl); (3) euglycaemic (80-109 mg/dl); (4) impaired fasting glucose (IFG) (110-125 mg/dl); (5) borderline diabetics (126-139 mg/dl); (6) diabetics (> or 140 mg/dl).
Results: Patients comprised 131 with hypoglycaemia (0.9%), 731 with low normal glucose (5%), 9308 euglycaemic (63.4%), 1577 with IFG (10.7%), 617 borderline diabetics (4.2%) and 2306 diabetics (15.7%). Over a mean 8-year follow-up, crude all-cause mortality was higher in both diabetic (31.8%) and hypoglycaemic groups (25.2%) as compared with euglycaemics (14.9%; P<0.0001); CAD mortality was higher in diabetic and borderline groups (17.8 and 13.3%, respectively, versus 7.9% in euglycaemics; P<0.0001). The highest prevalence of cancer mortality was documented in the hypoglycaemic group (6.1 versus 2.9% in euglycaemics; P<0.02). Actuarial survival curves showed the lowest mortality in euglycaemic and low normal groups; the highest was seen in diabetic and hypoglycaemic patients. Intermediate values were found in borderline and IFG patients. After adjustment for variables, a significantly higher mortality rate was seen in hypoglycaemics when compared with euglycaemics (P<0.0001). Hypoglycaemia was identified as a predictor of increased all-cause and cancer mortality with a hazard ratio (HR) of 1.84 [95% confidence interval (CI) 1.29-2.61] and 2.26 (95% CI 1.12-4.57), respectively, but not of increased CAD mortality, with HR 1.30 (95% CI 0.73-2.29).
Conclusions: Over a mean 8-year follow-up, hypoglycaemia emerges as a marker for substantially increased all-cause and cancer mortality among patients with CAD presenting with low fasting glucose levels.