Dysfunctional HDL-cholesterol: a potential link between high levels of HDL-cholesterol and cardiovascular disease?

Abstract

Strategies which raise levels of high-density lipoprotein (HDL) are attracting increasing interest, based on the well-established inverse relationship between HDL levels and cardiovascular disease (CVD). However, recent evidence has shown that there may be considerable variation in the efficacy of HDL to protect from CVD. Therefore, the relationship between high levels of HDL and CVD needs to be investigated in the clinical setting, particularly in view of current interest in raising HDL levels.

Investigation of HDL levels in a lipid disorders clinic (n = 2,864) revealed HDL levels were higher in females and patients with increased age, but were inversely associated with triglyceride (TG) levels, obesity, insulin, HOMA-IR, glucose and glycated haemoglobin (HbA1c). HDL levels were lower in patients with diabetes, cigarette smokers and Maori ethnicity. Raised total cholesterol (TC) and low-density lipoprotein (LDL), rather than low HDL, explained most referrals to this specialist lipid service.

An incidental and unexpected finding was that Maori, Pacific Islanders and Asians were under-represented in the lipid disorders clinic despite their recognised high risk for CVD. Although females are generally perceived as having lower CVD risk than males, they are actually at higher risk when referred for lipid management, being older, having additional risk factors and higher Framingham risk score. The level of risk in women may be under-appreciated and may contribute to the higher percentage of deaths attributable to all forms of heart disease in New Zealand women.

Hypertension, diabetes, smoking, high-sensitivity C-reactive protein (hsCRP) and HOMA-IR were associated with the highest rates of CV events and mortality. Gender, TC, LDL and lipoprotein (a) were not associated with CV events and mortality. The Framingham risk score effectively indicated CV risk. Its precision may be increased by the addition of hsCRP, body mass index (BMI) and HOMA-IR to the model.

Clinic patients with HDL levels >95th percentile, had greater levels of CV events and CV mortality than expected, given their high levels of HDL and reduced levels of major risk factors including HOMA-IR, BMI, TG, hsCRP. A subset of 44 patients with high levels of HDL (males >1.69 mmol/L, n =23; females >2.30 mmol/L, n = 21) was recruited for more extensive testing. Evidence of sub-clinical CVD, defined by coronary artery calcification score (CACS), was found in 70% of the 44-patient cohort, with calcification levels described as zero (n =13), mildly elevated (n =16), moderate (n = 11), and high/extreme (n = 4). Although testing included inflammatory markers and a wide range of biochemical tests and clinical factors, the only elevated risk factor in patients with CACS >100 Agatston units was raised blood pressure. Determinants of HDL function, including cholesterol efflux and flow-mediated dilatation were lower in patients with higher CACS, or age- and gender-adjusted CACS percentiles, suggesting reduced HDL function in some patients with high HDL levels. Therefore, even in the absence of excessive risk factors, high HDL levels may not always provide protection from CVD. Targeting functionality, rather than raising the level, of HDL may be more important in reduction of CVD.