RESUMO
AIMS: Although hypertensive patients with low baseline HDL cholesterol levels have a higher incidence of diabetes mellitus, whether changing levels of HDL over time are more strongly related to the risk of new diabetes in hypertensive patients has not been examined. METHODS: Incident diabetes mellitus was examined in relation to baseline and in-treatment HDL levels in 7485 hypertensive patients with no history of diabetes randomly assigned to losartan- or atenolol-based treatment. RESULTS: During 4.7 ± 1.2 years follow-up, 520 patients (6.9%) developed new diabetes. In univariate Cox analyses, compared with the highest quartile of HDL levels (> 1.78 mmol/l), baseline and in-treatment HDL in the lowest quartile (< 1.21 mmol/l) identified patients with > 5-fold and > 9 fold higher risks of new diabetes, respectively; patients with baseline or in-treatment HDL in the 2nd and 3rd quartiles had intermediate risk of diabetes. In multivariable Cox analyses, adjusting for randomized treatment, age, sex, race, prior anti-hypertensive therapy, baseline uric acid, serum creatinine and glucose entered as standard covariates, and in-treatment non-HDL cholesterol, Cornell product left ventricular hypertrophy, diastolic and systolic pressure, BMI, hydrochlorothiazide and statin use as time-varying covariates, the lowest quartile of in-treatment HDL remained associated with a nearly 9-fold increased risk of new diabetes (hazard ratio 8.7, 95% CI 5.0-15.2), whereas the risk of new diabetes was significantly attenuated for baseline HDL < 1.21 mmol/l (hazard ratio 3.9, 95% CI 2.8-5.4). CONCLUSIONS: Lower in-treatment HDL is more strongly associated with increased risk of new diabetes than baseline HDL level.
