HDL Cholesterol and Residual Risk After Potent Statin Therapy: A JUPITER Analysis

This prespecified JUPITER trial analysis tested whether HDL-C and ApoA-I retained predictive value for cardiovascular events in patients treated with rosuvastatin who achieved very low LDL-C levels.

In the placebo arm, HDL-C and ApoA-I were significantly inversely associated with cardiovascular events — the expected epidemiological finding. In the rosuvastatin arm — where LDL-C was reduced to a median of approximately 55 mg/dL — HDL-C and ApoA-I were no longer predictive of events. The disappearance of HDL-C as a risk predictor in the context of very low LDL-C suggests that the protective signal attributed to HDL-C in epidemiological studies may reflect its inverse relationship with atherogenic particle number rather than an independent protective effect of HDL itself.

This analysis provided important causal insight into the HDL-C paradox. Multiple RCTs of HDL-raising therapies — niacin, CETP inhibitors — had already failed to reduce events. The JUPITER data offered a mechanistic explanation: when LDL-C is brought very low, the remaining residual risk is no longer driven by the LDL-versus-HDL balance that HDL-C measures capture. It is driven instead by inflammation and other risk factors that HDL-C cannot capture or predict in that context.

The practical implication: in patients on potent statin therapy at LDL-C targets, HDL-C monitoring adds little prognostic value, and treating low HDL-C pharmacologically is not indicated.

We rate the evidence strong. A rigorous JUPITER prespecified analysis resolving the HDL-C prediction paradox in statin-treated patients — supporting the abandonment of HDL-C as a residual risk biomarker and treatment target in treated patients.