Our study (1) demonstrates that an LDL-related lipid factor in coronary patients with type 2 diabetes is 1) not elevated, 2) not associated with significant coronary stenoses, and 3) not significantly predictive for future events. Whereas the first two findings did not even show a trend for the LDL-related factor, our prospective data need to be confirmed by a longer follow-up with more end points. We have amply discussed this in our report, and Schulze and Hoffmann (2) also raise this point.
Since currently ∼58% of coronary patients are on statins (3) and lipid-lowering medication very frequently is required in diabetes to achieve the stringent LDL cholesterol goals that have been put forth for this “CAD risk equivalent,” exclusion of diabetic patients on lipid-lowering therapy in our investigation would have excluded diabetic coronary patients receiving up-to-date medical care and thus would have rendered the study population utterly unrepresentative of today’s clinical setting.
Factor analysis is the appropriate technique to extract an integral factor representing diabetic dyslipidemia from the individual lipid parameters measured. The power of this factor to predict vascular events can be subsequently evaluated in Cox regression models. This approach is novel in the study of lipid risk factors in patients with diabetes but has been previously applied to investigate the metabolic syndrome (4).
We included apolipoprotein B levels in our factorial analysis. Non-HDL cholesterol represents the total cholesterol carried by VLDL, IDL (intermediate-density lipoprotein), and LDL, of which each particle contains one molecule of apolipoprotein B. Serum levels of apolipoprotein B are thus closely related to non-HDL cholesterol.
Even though cardiovascular risk is significantly reduced in diabetic patients by statins, it remains at a very high level in these patients (5). The consistent association of the HDL-related factor with glycemia, with significant stenoses and with the future incidence of vascular events, provides compelling and conclusive evidence that the main lipid risk factor in our patients with diabetes is the triad of low HDL cholesterol, small LDL particles, and high triglycerides.
