Covariates | Geometric mean ratio of CAC scores (95% CI)* | χ^{2} | P value |
---|---|---|---|

Sex (male versus female) | 3.4 (1.4–8.3) | 6.8 | 0.0091 |

Scanning site (n = 19) | — | 28.1 | 0.0611 |

DCCT baseline | |||

Age (year) | 1.3 (1.2–1.4) | 81.7 | <0.0001 |

Diabetes duration (year) | 1.3 (1.1–1.6) | 4.1 | 0.0441 |

Albumin excretion rate (mg/24 h)† | 1.3 (1.1–1.5) | 7.3 | 0.0069 |

Cohort (primary versus secondary) | 1.7 (0.7–4.4) | 1.2 | 0.2651 |

EDIC year 7–9 (prior to CT scan) | |||

Smoking (yes versus no) | 7.1 (3.0–16.9) | 19.7 | <0.0001 |

Waist-to-hip ratio (%)† | 2.6 (1.5–4.5) | 11.5 | 0.0007 |

Hypercholesterolemia (yes versus no) | 2.8 (1.4–5.7) | 8.6 | 0.0033 |

Hypertension (yes versus no) | 2.8 (1.4–5.7) | 8.6 | 0.0034 |

DCCT mean A1C† | 1.4 (1.1–1.7) | 9.3 | 0.0022 |

Analysis was performed using Tobit regression: Y = log CT − log (lowest detectable CT score).

- *
* Geometric mean ratio of CAC scores is the ratio of predicted CAC scores for a 1-unit increase in quantitative variables or change in status for dichotomous variables if without notation;

- †
† geometric mean ratio of CAC scores is based on 10 mg/24 h increase in albumin excretion rate, 10% increase in waist-to-hip ratio, and 10% increase in DCCT mean A1C.