Table 2

Time-updated models for risk of cardiovascular events after worsening kidney function and hospitalization for heart failure

WKFCV death or HHF
Event (n=677),
n (%)
No event (n=117 985),
n (%)
Crude HR
(95% CI)
Adjusted HR
(95% CI)*
P value*
>30% eGFR drop (n=18 585)242 (35.7)18 343 (15.6)2.23 (1.88 to 2.64)1.97 (1.66 to 2.34)<0.001
>40% eGFR drop (n=8211)152 (22.5)8059 (6.8)2.75 (2.25 to 3.36)2.51 (2.06 to 3.07)<0.001
>50% eGFR drop (n=3246)83 (12.3)3163 (2.7)3.21 (2.48 to 4.16)3.03 (2.33 to 3.92)<0.001
WKFCV death
Event (n=331),
n (%)
No event (n=118 331),
n (%)
Crude HR
(95% CI)
Adjusted HR
(95% CI)*
P value*
>30% eGFR drop (n=18 585)96 (29.0)18 489 (15.6)2.09 (1.62 to 2.70)1.89 (1.47 to 2.45)<0.001
>40% eGFR drop (n=8211)56 (16.9)8155 (6.9)2.57 (1.87 to 3.51)2.41 (1.77 to 3.29)<0.001
>50% eGFR drop (n=3246)31 (9.4)3215 (2.7)3.35 (2.23 to 5.03)3.29 (2.19 to 4.92)<0.001
HHF (n=444)98 (29.6)346 (3.5)†8.58 (6.67 to 11.0)5.17 (3.91 to 6.83)<0.001
  • No interaction of WKF by baseline eGFR was observed; interaction p>0.1 in all models.

  • *Model adjusted for age (≤65 years vs >65 years), sex (female vs male), history of cardiovascular disease including hospitalization for heart failure (yes vs no), urinary albumin to creatinine ratio (≤30 mg/g vs >30 mg/g), waist circumference (≤88 cm for women and ≤102 cm for men vs >88 cm for women and >102 cm for men), glycated hemoglobin (≤8% vs >8%), and randomized treatment (intensive vs standard glucose-lowering treatment).

  • †Only the first HHF episode registered in the data set.

  • CV, cardiovascular; eGFR, estimated glomerular filtration rate; HHF, hospitalization for heart failure; WKF, worsening kidney function.