Haemoglobin A1c levels and subsequent cardiovascular disease in persons without diabetes: a meta-analysis of prospective cohorts.

AIMS/HYPOTHESIS: The aim of this meta-analysis was to determine the relationship between HbA(1c) levels and subsequent cardiovascular outcomes in individuals without diabetes. METHODS: We searched Medline, Embase and Scopus from initiation of the study until the end of 2009. One reviewer searched and another verified findings. Data were extracted by one reviewer and verified by another. We accepted prospective studies in any language reporting three or more quartiles for HbA(1c) levels. Within quartiles, authors must have presented both numbers of patient-years at risk and cardiovascular outcomes. Outcomes per person-time at risk were regressed on average HbA(1c) values using Poisson regression. We pooled ß coefficients using Cochran's semi-weighted (inverse variance) random-effects model. Study quality was assessed using the Downs-Black scale. RESULTS: We investigated 16 datasets (nine for total cardiovascular events and seven for death) from five papers with 44,158 patients (44% men) over 404,899 patient-years of follow-up. There were 1,366 cardiovascular deaths (3.1%; 3.37/1,000 person-years) and 2,142 cardiovascular events (4.9%; 5.29/1,000 person-years). The overall meta-analytic ß coefficients were 0.720 (95% CI 0.307-1.133) and 0.757 (95% CI 0.382-1.132) for cardiac death and events, respectively. Compared with the baseline value of 0.0427, an HbA(1c) level of 0.05 was associated with a relative risk for cardiovascular death of 1.13 (95% CI 1.05-1.21), a 0.06 value with 1.34 (95% CI 1.13-1.58), and a 0.07 HbA(1c) with relative risk 1.58 (95% CI 1.22-2.06). Results for total cardiovascular events were similar. The average study quality was 0.7 (70%). CONCLUSIONS/INTERPRETATION: We conclude that HbA(1c) was significantly associated with cardiovascular events and deaths in persons without diabetes.