RESUMO
AIM: LDL cholesterol (LDL-C) is considered an important cardiovascular disease (CVD) risk factor. Less is known in Type 1 diabetes. We assessed LDL-C and total cholesterol to HDL cholesterol ratio (TC/HDL-C) as predictors of CVD in Type 1 diabetes. METHODS: The study monitored 30 778 people with Type 1 diabetes, baseline 2003-2006, to 31 December 2011. Cox regression analyses were performed with LDL-C and TC/HDL-C as predictors of fatal/non-fatal CVD. Models were adjusted for traditional CVD risk factors. RESULTS: Hazard ratios (HR) (with 95% CI) per 1 mmol/l increase in LDL-C for CVD were 1.09 (1.01-1.18) in people without lipid-lowering medication and 1.02 (0.95-1.09) in people with lipid-lowering medication (P = 0.02 and 0.65). In people aged 40 years or older having a CVD risk factor, and in people with a history of CVD, HR was 1.07 (0.99-1.16) and 1.02 (0.92-1.13) (P = 0.07 and 0.66). HR per 1 unit increase in TC/HDL-C was 1.12 (1.05-1.20) in people without lipid-lowering medication and 1.08 (1.02-1.15) in people with lipid-lowering medication (P < 0.001 and 0.01). For people aged 40 or older and people with a history of CVD, HR was 1.16 (1.09-1.24) and 1.04 (0.95-1.14) (P < 0.001 and 0.43). Broken down into octiles, LDL-C was not a significant predictor of CVD in any group. Higher octiles of TC/HDL-C were significant predictors for CVD in people without lipid-lowering medication and in those aged 40 years or older. CONCLUSION: In this study of people with Type 1 diabetes in clinical practice, LDL-C was not a good predictor of CVD. We found no support for an LDL-C cut-off point < 2.6 mmol/l. TC/HDL-C seems more reliable as a marker for CVD risk when considering primary prevention.