There are large lessons to learn from the Cardiovascular Risk Study in Young Finns, published in this issue of the Journal (1). Two reiterate what we already knew but have largely ignoredfirst, whether we are paying attention or not, atherosclerosis is well underway by the third and fourth decades of our lives and, second, this disease is directly traceable to the atherogenic imbalance in the plasma lipoproteins that is identifiable by puberty. As important as these lessons are, Juonala et al. (1), break much fresh ground.
See page 293They found that the plasma apolipoprotein (apo) B and the apoB/apoA-I ratio at age 12 to 18 years were directly related to carotid artery intima-media thickness (IMT) at age 24 to 39 years, whereas apoA-I was indirectly related to carotid IMT. Moreover, these relations were not altered when adult levels of the apolipoproteins and non-lipid risk factors were taken into account. In addition, the same relations were noted, although in samples from an even earlier age, between the levels of the apoB, apoB/apoA-I ratio and impaired flow-mediated vasodilatation (FMD). Also of interest was the lack of any significant gender interaction between the association of the apolipoproteins and either carotid IMT or FMD.That the plasma lipoproteins matter in atherogenesis and that atherogenesis starts early should be easy for all to accept. The challenge lies in what follows. Juonala et al. (1) demonstrate convincingly that apoB is superior to both low-density lipoprotein (LDL) cholesterol and non-highdensity lipoprotein (HDL) cholesterol as an estimate of the atherogenic lipoproteins, that apoA-I is superior to HDL cholesterol as a marker of the antiatherogenic lipoproteins, and that the apoB/apoA-I ratio was significantly better than either the LDL/HDL cholesterol or the non-HDL/HDL cholesterol ratio as overall estimates of the lipoprotein-related vascular disease (1). Not only were the standardized beta coefficients of the apolipoproteins substantially greater than their cholesterol counterparts, but significant differences in favor of the apoB/apoA-I ratio were also evident by the c-statistic.Perhaps no other issue in lipidology has been as contentious as the debate as to whether cholesterol or apolipoproteins are better markers of risk. All major guideline groups had already embraced cholesterol, and prior commitment and the perceived need for continuity are potent arguments for the status quo. However as the number of studies comparing apoB with LDL cholesterol has mounted, the comparison has become one-sided. Except in the oldest subjects where LDL by any measure is not predictive (2), in all other groups, whether those with symptomatic disease or those without, in men or women, those receiving therapy or those not treated, apoB has come out on top against LDL cholesterol (Online Appendix, supplementary references 1 to 23).Understandably, it is not easy for most clinicians to fully appreciate just how clear the difference in predictive power is. Hazard ratios, c-statistics, and p values do n...