-Nico Dragano, Pablo Emilio Verdea, Susanne Moebus, Andreas Stang,
Axel Schmermund, Ulla Roggenbuck, Stefan Möhlenkamp, Richard Peter,
Karl-Heinz Jöckel, Raimund Erbel, Johannes Siegrist and for the
Heinz Nixdorf Recall Study
Background Social inequalities of manifest coronary heart diseases are well documented in modern societies. Less evidence is available on subclinical atherosclerotic disease despite the opportunity to investigate processes underlying this association. Therefore, we examined the relationship between coronary artery calcification as a sign of subclinical coronary atherosclerosis, socio-economic status and established cardiovascular risk factors in a healthy population.
Baseline data on 2264 women and 2037 men are displayed in Table 1. There were marked gender differences in the distribution of the two indicators of social status. Mean
income was higher for men, and men reported more years of formal education. As expected, coronary risk factors were more pronounced in men than in women. The median CAC score for the whole sample was 12.6 with an interquartile range of 117.6 (25%=0; 75%=117.6). On average values were lower in women.
A stratification of the CAC score by education and income is presented in Table 2. The median values of the score varied considerably between the social groups in men and women: calcification decreased with increasing education or income. This trend was, however, not consistent across all age groups. It was most pronounced in younger participants aged 45–54 years, but less consistent in the middle and older age groups.
Table 3 displays the results of multivariate analyses, using education as an indicator of SES. For men and women, the crude model confirmed the bivariate findings. As can be seen consecutive adjustment for risk factors diminished substantially the strength of the association between education and CAC. On the basis of the model-fit, cholesterol (men only), diabetes (women only), triglycerides, systolic blood pressure, body mass index and smoking contributed most to the model-fit. Importantly, in both genders the association between education and CAC was no longer statistically significant in the most parsimonious model.
Analogous findings for income are given in Table 4. The weak relationship between income and CAC for the male study participants was largely explained by age. Further adjustment ruled out any variation of the dependent variable in relation to income. In women associations were more pronounced. In the age-adjusted model (2) women in the lowest quartile had a 50% higher CAC than women in the upper quartile of the income distribution. Again, adjustment for risk factors resulted in a reduction of effects. (weiter…)