Table 3.

Adjusted associations between CKD and collateral supply to culprit and nonculprit vessels

OutcomeOdds Ratio95% Confidence IntervalP Value
Patient-level analysis (binary measures)
 any collaterals to infarct vessela1.240.55–2.790.60
 adequate collateral supply to infarct vessela1.260.52–3.050.62
 any collaterals to noninfarct vesselb1.640.61–4.390.33
 adequate collateral supply to noninfarct vesselb1.560.56–4.340.39
Vessel analysis
 any collaterals to nonculprit vesselsb2.42(0.74, 7.92)0.14
 adequate collaterals to nonculprit vesselsb2.33(0.70, 7.69)0.17
 collateral score in infarct arteryc0.91(0.40, 2.11)0.83
 collateral score in noninfarct arteriesd1.97(0.71, 5.51)0.19
  • Patient-level and vessel-level adjusted analyses of the association between CKD and collateral formation. For patient-level analysis, the individual patient with or without CKD served as the unit of analysis and only one outcome per patient was allowed in the logistic regression models. For vessel-level analyses, the individual vessels served as the unit of analysis in logistic models with the Rentrop score (0–3) serving as the dependent variable in a multivariable ordered logistic regression models. Robust SEMs were used in the vessel analysis of noninfarct collateral score to account for correlations between multiple vessels within an individual patient. All models were adjusted for age, sex, race, diabetes, MI type, hypertension, smoking, and infarct location in addition to the covariates listed below in the footnotes.

  • a None.

  • b Number of diseased vessels.

  • c Hyperlipidemia.

  • d Hyperlipidemia and percent diameter stenosis.