Should Years of Schooling Be Used to Guide Treatment of Coronary Risk Factors?

I appreciate Dr Jaen's comments. The addition of educational level into PDA decision tools would help to insure that SES is routinely incorporated into cholesterol management in every day practice along with age and gender -two other non-modifiable sociodemographic factors that predict CHD risk.

I agree with Dr Diaz that changes in behavioral risk factors following baseline assessment likely contribute to educational disparities in CHD and that effective interventions designed to improve behavioral risk factors among members of this groups are needed. However, I disagree that we should wait for improvements in these risk factors among this population before we consider changing out treatment threshhold for the low education group for several reasons.

First, available evidence strongly suggest that psychosocial factors such as perceived control of one's life, the balance between job related effort and reward, and greater stress among others also contribute to this disparity (Michael Marmot, the renowned British epidemiologist, reviews this subject in his recent book "Status Syndrome"). Consequently, simply addressing behavioral risk factors is unlikely to eliminate this disparity.

Second, we currently lack evidence-based effective interventions for addressing most of these risk factors with the notable exception of smoking. Data from community-based interventions are even less encouraging. Given relatively weak behavioral interventions, it seems that we need complementary strategies for reducing these disparities in CHD.

Moreover, although definitive data are lacking, adherence to office- based counseling among persons with lower education is probably lower than among higher education due a combination of determinants including social norms, self efficacy, financial factors, and neighborhood facilitators (access to affordable, healthy food, walking/biking trails, fitness centers) and absence of neighborhood cues (bill board advertising and convenience stores). Thus, improvement in office-based counseling for behavioral risk factors will likely disproportionately benefit better educated patients possibley worsening rather than improving disparities in CHD.

Last, data from clinical trials show that relative risk reduction from statins is largely independent of baseline risk factors including baseline LDL. Because age and gender are currently included in cholesterol treatment guidelines, the threshhold for statin use is lower for older persons and men. Our data suggest a similar logic for educational level.

Competing interests:   None declared

This article describes the association between low educational level (LEL) and CHD mortality. This is not an unexpected finding based on previous studies, some of which are cited in this paper. First of all, I am struck by the differences in the groups to be compared in this study. The LEL group is older (they make up 64% of those 65-74 and only 19% of those 25-34 years old) and sicker (more uncontrolled blood pressure, diabetes and hypercholesterolemia). With only 10 years of follow up, looking at mortality in a group of predominantly 65+ versus another group that is predominantly much younger (25-44) seems problematic. However, I realize that the authors did control for both age and the higher prevalence of risk factors in the LEL group. Thus we should evaluate how to interpret these findings. I would hypothesize these findings are most likely due to environmental factors negatively affected by SES (such as diet, weight, use of preventive services and medications, access to care) that were not measured in this study. Considering the higher prevalence of uncontrolled disease in the LEL group, it seems developing interventions targeted to this group and defining barriers to controlling risk factors would be a helpful first step in overcoming this disparity. This step should be taken before we consider changing out treatment threshholds for the LEL group. Further identification of other factors that affect CHD mortality and are associated with LEL is also warranted to help overcome this disparity.

Competing interests:   None declared