Nonlinearity of the inverse relationship between high-density lipoprotein (HDL) cholesterol and incident cardiovascular risk: Is it time to revisit the "HDL hypothesis"?

BACKGROUND: Low levels of high-density lipoprotein cholesterol (HDL-C) are clearly associated with atherosclerotic cardiovascular disease (ASCVD), but the risk curve is not well defined, especially at very high and low HDL-C levels. Current proportional hazards prediction models assume inverse linearity of effect, which may not accurately represent risk at these levels. SOURCES OF MATERIAL: Clinical inattention to risk associated with low HDL-C may derive from randomized controlled trials (RCTs) aimed at raising HDL-C, though most failed to reduce ASCVD events when combined with statin-based therapy. However, these prior trials enrolled patients with HDL-C levels largely in the 35-45 mg/dL range. ABSTRACT OF FINDINGS: Mounting post hoc evidence from RCTs as well as new genetic and observational data suggest that very low HDL-C (less than 30 or 35 mg/dL) may signal a further increase in incident cardiovascular events. Moreover, when HDL-C exceeds 90 mg/dL, monotonic reduction of ASCVD risk appears to reverse. Because a pervasively agnostic view of the importance of both very low and high levels of HDL-C now exists, consideration should be given to incorporating nonlinear effects of HDL-C into future risk prediction models such that very low HDL-C and/or very high HDL-C levels could be considered as new risk-enhancing factors to promote more optimal risk stratification. CONCLUSION: When revision of the U.S. Cholesterol Guideline recommences, consideration should be directed to whether HDL-associated risk matches the assumptions of current statistical models. Thus, it may be both timely and opportune to revisit the "HDL hypothesis" based on evolving scientific evidence.

Duke Scholars

Author John Richard Guyton Medicine, Endocrinology, Metabolism, and Nutrition