Abstract Background Reduced serum levels of high density lipoprotein cholesterol (HDL-C) is a hallmark of the metabolic syndrome. Epidemiologic studies have reported an inverse correlation between HDL- C and cardiovascular disease (CVD) risk. However, recent works suggest that the association is strongest in healthy individuals. Changes in particle number and functional properties of HDL may more closely reflect CVD prognosis in patients with pre-existent disease. The majority of such patients receive statin treatment, which affects both HDL-C levels, and particle composition. Hence, serum apoA1 and the HDL-C:apoA1 ratio have been proposed as more sensitive indicators of cardio-metabolic prognosis. Purpose We studied the associations of serum HDL-C, apoA1 and the HDL-C: apoA1 ratio to long term risk of CVD mortality and incident type 2 diabetes (T2D) patients with suspected stable angina pectoris (SAP). Methods A total of 41064 patients underwent elective coronary angiography in 2000–2004 and were followed-up for CVD mortality throughout 2016. In a subgroup of 2519 participants without verified or possible diabetes at baseline, the associations to incident type 2 diabetes (T2D) were evaluated throughout 2014. Information on clinical endpoints was obtained through national health registries. Risk estimates are reported per 1 SD increment of (log transformed) biomarkers and were calculated by cox or logistic regression. We explored risk classification by calculating the continuous net reclassification improvement (NRI). Results At inclusion, median (25th-75th percentiles) age was 62 (55–70) years, 28% were women 76% had obstructive coronary artery disease and 80% received statins. During median (25th-75th percentiles) 13.9 (12.0–15.3) years of follow-up, 14.1% of the participants died from CVD. After multivariate adjustment (age, gender, body mass index, HbA1c, triglycerides, statin treatment, fasting status) HDL and apoA1, but not the HDL: apoA1 ratio, significantly predicted CVD mortality. The hazard ratio (HR) and 95% confidence interval (CI) was: 0.86 (0.78–0.94), 0.88 (0.80–0.98) and 0.96 (0.86–1.03) for HDL-C, apoA1 and the HDL-C:apoA1 ratio, respectively. HDL-C was the only of the evaluated biomarkers providing a significant NRI (95% CI) of 0.14 (0.04–0.19). In the subset evaluated for incdent T2D, HDL-C provided multivariate adjusted odds ratios (OR; 95% CI) and NRI (95% CI) of 0.69 (0.58–0.82) and 0.34 (0.21–0.47) for new onset TSD. The corresponding OR (95% CI) and NRI (95% CI) for apoA1 were: 0.85 (0.73–0.99) and 0.20 (0.06–0.33), respectively. The HDL:apo A1 ratio provided an OR (95% CI) of 0.66 (0.55–0.80) and NRI (95% CI) of 0.24 (0.11–0.37) for T2D. No significant effect modifications according to statin treatment were found (P≥0.22). Conclusion Among patients with suspected SAP, of which the majority received statins, HDL-C was non-inferior to apoA1 and the HDL:apoA1 ratio in predicting long term risk of CVD mortality and T2D. Funding Acknowledgement Type of funding source: Public hospital(s). Main funding source(s): Department of Heart Disease, Haukeland University Hospital; Department of Clinical Science, University of Bergen
Read full abstract