Towards standardisation of pre- and post-analytical variables in the assessment of lipid risk factors for ischaemic heart disease.

A wide disparity exists between statistical and optimal reference ranges for total cholesterol (TC) levels in South Africa. The haphazard downward adjustments by some laboratories over the years have resulted in highly variable and confusing 'normal cholesterol' values. Using essentially the same enzymatic method and equipment the TC upper limits varied by almost 50% (5.7-8.5 mmol/l) in 24 private laboratories, while it ranged from 5.7 mmol/l to 6.7 mmol/l in 6 academic hospital laboratories. The greatest disparity occurred in the lower limits of 'normal' for high-density lipoprotein cholesterol (HDLC), which ranged from 0.45 mmol/l to 1.43 mmol/l. Although triglycerides may not be an independent risk factor for ischaemic heart disease, the upper limit ranged from 1.69 mmol/l to 2.5 mmol/l despite comparable methods and equipment used. Lipid data obtained on approximately 5,000 healthy adults from the Transvaal, Natal and the Orange Free State revealed remarkable similarities to those obtained in the CORIS study. It therefore confirms the applicability of the CORIS epidemiological findings to other westernized groups in different regions of South Africa. This congruency in blood lipid values and methodologies used in different laboratories should also enable all state and private laboratories to use the same age-specific limits for desirable TC and HDLC levels.