Table 3—

Age-adjusted mean differences in IRS score per unit increase in personal income and education categories* stratified by neighborhood characteristics: the CARDIA study 1995–1996

White men (n = 782)
White women (n = 852)
Mean difference per unit increase in income categoryMean difference per unit increase in education categoryMean difference per unit increase in income categoryMean difference per unit increase in education category
Neighborhood quartiles I–II (score <4.99)−0.045 ± 0.044−0.131 ± 0.040−0.088 ± 0.039−0.198 ± 0.038
Neighborhood quartiles III–IV (score ≥4.99)−0.051 ± 0.041−0.051 ± 0.043−0.101 ± 0.038−0.099 ± 0.045
P value for interaction0.920.170.810.09
Black men (n = 599)
Black women (n = 860)
Mean difference per unit increase in income categoryMean difference per unit increase in education categoryMean difference per unit increase in income categoryMean difference per unit increase in education category
Neighborhood quartiles I–II (score <−0.71)0.135 ± 0.0580.116 ± 0.070−0.044 ± 0.048−0.028 ± 0.059
Neighborhood quartiles III–IV (score ≥−0.71)−0.031 ± 0.054−0.048 ± 0.056−0.118 ± 0.044−0.154 ± 0.048
P value for interaction0.040.070.250.10
  • Data are mans ± SE.

  • *

    * Categories correspond to those shown in Table 1. Mean difference per unit increase in income category corresponds to average change per unit difference in category obtained by including income and education categories in separate regression equations as ordinal covariates. P value for interaction corresponds to interaction between neighborhood score (in two categories as shown) and the ordinal income or education variable.