Table 2.

Multivariate adjusted OR for myocardial infarction by level of GFR for each racial/ethnic categorya

ModelbWhite (n = 6,252)Black (n = 3,942)Mexican-American (n = 3,849)
OR (95% CI)POR (95% CI)POR (95% CI)P
Unadjusted GFR (ml/min)
    90 to 2001.00 (reference)1.00 (reference)1.00 (reference)
    60 to 892.44 (1.85 to 3.23)<0.00013.91 (2.68 to 5.72)<0.00014.16 (2.20 to 7.84)<0.0001
    <6010.42 (7.55 to 14.38)<0.000110.50 (6.03 to 18.28)<0.00019.74 (4.56 to 20.78)<0.0001
    c-statistic65.6%69.6%64.5%
Traditional GFR (ml/min)
    90 to 2001.00 (reference)1.00 (reference)1.00 (reference)
    60 to 891.06 (0.80 to 1.41)NS1.47 (1.01 to 2.12)<0.051.63 (0.93 to 2.85)NS
    <602.05 (1.48 to 2.86)<0.00011.64 (0.82 to 3.26)NS1.67 (1.02 to 2.73)<0.05
    c-statistic82.1%86.8%84.7%
Full model GFR (ml/min)
    90 to 2001.00 (reference)1.00 (reference)1.00 (reference)
    60 to 891.07 (0.80 to 1.44)NS1.60 (0.82 to 3.12)<0.05;1.68 (0.97 to 2.90)NS
    <601.97 (1.42 to 2.72)<0.00011.45 (1.01 to 2.08)NS1.39 (0.80 to 2.40)NS
    c-statistic82.4%86.8%85.6%
  • a The unadjusted model assessed the relationship of GFR with coronary artery disease in each racial/ethnic category. The traditional model adjusted for age, gender, diabetes, hypertension, family history of premature myocardial infarction, smoking status, obesity, physical inactivity during leisure time, and total cholesterol/HDL cholesterol ratio. Full model contains both traditional and nontraditional risk factors where the nontraditional factors are represented by C-reactive protein and white blood cell count. CI, confidence interval; OR, odds ratio.

  • b Three parallel models were built for each racial/ethnic group.\