Introduction

The traditional risk factors for atherosclerosis have been known for decades: blood lipid level alterations, cigarette smoking, hypertension, diabetes, lack of exercise, and obesity. Blood lipid screening has generally been used to predict the likely development of overt atherosclerosis and its complications.

Recent awareness of an infective component in atherosclerosis causation has led to the inclusion of other screening tests, such as C-reactive protein (CRP). In a previous study of apparently healthy middle-aged women, the addition of CRP determination to screening based on standard lipid levels improved identification of those subjects at increased risk for future myocardial infarction or stroke.

In order to determine the usefulness of different predictive tests, data from the Harvard Medical School's Physicians Health Study were analyzed and the findings published in JAMA.

Method

From 14,916 male physicians whose ages ranged from 40-84 and who were enrolled in the Physicians Health Study, some 140 without previous cardiovascular disease developed symptomatic peripheral arterial disease (PAD) during the 9-year follow-up period. The PAD endpoint was defined by the occurrence of intermittent claudication or hospitalization for peripheral arterial revascularization procedures. An equal number of control subjects (140) who had no PAD were selected at random from the remaining study subjects.

The study evaluated the predictive value for PAD of 11 possible risk factors for systemic atherosclerosis measured at baseline. These lipid and nonlipid risk factors were: total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), total cholesterol to HDL-C ratio, triglycerides, homocysteine, C-reactive protein (CRP), lipoprotein (a), fibrinogen, and apolipoproteins (apo) A-1 and B-100.

Conditional logistic regression analyses were used to compute relative risks of future PAD associated with increasing levels of each study parameter. By dividing the subjects into quartiles based on the distribution of control values, the authors were able to derive relative risks (RRs) associated with levels in the highest versus the lowest quartile for each of the 11 parameters. They also compared the potential additive value of the so-called novel parameters to standard lipid screens.

Results

The mean time to diagnosis of incident PAD for the study population was 5 years (range: 3 - 155 months). Traditional atherosclerotic risk factors such as diabetes, hypertension, and a family history of cardiovascular disease were more prevalent at baseline among those who subsequently developed symptomatic PAD compared with those who did not, but the difference was only statistically significant for diabetes. The mean age, smoking patterns and exercise habits were virtually identical in the case and control groups.

Among the lipid parameters, after adjustment for age, smoking status, hypertension, body mass index, family history, exercise level, and diabetes, TC, HDL-C, LDL-C, triglycerides, and apo B-100 were all significant predictors of risk, with the single strongest predictor being the TC/HDL-C ratio (RR, 3.9; 95% CI, 1.7-8.6).

Among the nonlipid parameters, the single strongest predictor was the plasma CRP after adjustment (RR, 2.8; 95% CI, 1.3-5.9). Plasma fibrinogen was a significant predictor, but homocysteine was not.

The additional analyses done showed that there was little benefit in assessing risk by adding results of other lipid parameters to the TC/HDL-C ratio, and by adding either lipoprotein (a) or homocysteine to standard lipid screening with TC alone or the TC/HDL-C ratio. On the other hand, adding either of the biomarkers for inflammation - CRP or fibrinogen - significantly improved risk assessment based on either TC or TC/HDL-C ratio; the effect of CRP was greater than that of fibrinogen.

The Message

For population screening for subclinical atherosclerosis, the authors recommend determining the TC/HDL-C ratio and the plasma CRP level to predict the development of symptomatic peripheral atherosclerosis, and thus as markers for systemic atherosclerosis. Plasma fibrinogen can be an effective substitute for CRP measurement. Although homocysteine levels were not found to be relevant in this study, other reports suggest that it is of value, especially in persons with renal impairment.

The authors recognize that the use of self-reported symptomatic PAD as their primary end point represents a possible weakness of study design, but they believe this approach is valid, as the study participants were physicians, who have previously delivered consistent subjective end-point information.

Assessment of PAD should include detection of distal pulses and the ankle-brachial index. More expensive objective tests include MRI, electron beam CT, and intima-media thickness measurement by ultrasound; such tests should be compared in future studies for cost-effectiveness with the biomarkers examined here.

This study provides yet another indication of the role of inflammation in the generation or progression of atherosclerosis. Further backing comes from a new publication by Dr Ridker ¹. In 5,742 subjects without known coronary artery disease, the risk of coronary events increased with increasing plasma levels of CRP. Treatment with the statin drug, lovastatin, reduced both CRP levels and the risk of coronary events, even when lipid levels were normal - possibly due to an anti-inflammatory effect of this drug class. These results may encourage the targeting of statin therapy to those persons who stand to benefit most from it. Even among persons with normal lipid levels, those who have elevated levels of CRP will probably benefit from statin therapy.