Accuracy non-fasting vs fasting lipid measures for CVD prediction clarified

19 August 2008

MedWire News: Researchers from the Women's Health Study have clarified which lipid measures are best measured fasting and which non-fasting for the most accurate prediction of cardiovascular disease (CVD) risk.

Current recommendations from the National Cholesterol Education Program Adult Treatment Panel III suggest that initial screening of patients for levels of total cholesterol, low-density lipoprotein (LDL) cholesterol, high-density lipoprotein (HDL) cholesterol, and triglycerides should be carried out when patients are in a fasting state.

The guidelines also state that total cholesterol, HDL cholesterol, and non-HDL cholesterol may also be measured non-fasting.

However, recent studies have suggested that these recommendations may not provide the most accurate measures for CVD risk prediction, particularly for triglycerides.

Samia Mora and colleagues from the University of Harvard in Boston, USA investigated further by collating baseline lipids, measured in a fasting (n=19,983) and non-fasting (n=6347) state, with incident CVD in a cohort of 26,330 women from the Women's Health Study aged 54.4 years on average.

Incident CVD was recorded over an 11.4-year follow-up period and was defined as non-fatal myocardial infarction, percutaneous coronary intervention, coronary artery bypass grafting, nonfatal stroke, or cardiovascular death.

Over the study period, the investigators observed 961 cases of incident CVD which occurred in 754 fasting and 207 non-fasting patients.

Writing in the journal Circulation, Mora et al noted that stronger associations with CVD were found when comparing fasting with non-fasting baseline measurements for total cholesterol (hazard ratio [HR]=1.22 vs 1.07 per 1 standard deviation [SD] increment), LDL cholesterol (HR=1.21 vs 1.00), apolipoprotein (apo)B (HR=1.36 vs 1.20), non-HDL cholesterol (HR=1.29 vs 1.15), and the apoB-to-apoA-I ratio (HR=1.39 vs 1.18).

In contrast, the fasting and non-fasting measurements for HDL cholesterol, apoA-I, and the total-to-HDL cholesterol ratio were very similar.

Fasting and non-fasting levels of triglycerides were both related to CVD risk, but further adjustment for total and HDL cholesterol produced a stronger association with non-fasting triglycerides.

The team concludes: "These observations suggest that non-fasting blood draws may be highly effective and practical when limited to HDL cholesterol, total-to-HDL cholesterol ratio, triglycerides, and apolipoprotein A-1.

"However, these data also suggest that a fasting sample is preferred if risk assessment is based on total cholesterol, LDL cholesterol, non-HDL cholesterol, apolipoprotein B-100, and apolipoprotein B-100/A-1 ratio."

Circulation 2008; Advance online publication