Table 3. Association of APOL1 High-Risk Group, AKI, AKI Stages, and Death in Veterans With African Ancestry Hospitalized With COVID-19^a.

Variable	No.	Primary outcome, acute kidney injury		Secondary outcomes
				AKI severity stages		Death
		Odds ratio (95% CI)	P value	Odds ratio (95% CI)	P value	Odds ratio (95% CI)	P value
All patients
Minimally adjusted
2 Copies of APOL1 RVs	125	1.80 (1.21-2.69)	.004	1.88 (1.30-2.71)	.001	1.92 (1.13-3.17)	.01
1 Or 0 copies of RVs	865	1 [Reference]		1 [Reference]		1 [Reference]
Fully adjusted model
2 Copies of APOL1 RVs	121	1.95 (1.27-3.02)	.002	2.03 (1.37-2.99)	<.001	2.15 (1.22– 3.72)	.007
1 Or 0 copies of RVs	812	1 [Reference]		1 [Reference]		1 [Reference]
Subgroup GFR ≥60 mL/min
Minimally adjusted
2 Copies of APOL1 RVs	92	1.88 (1.18-2.99)	.008	1.98 (1.28-3.06)	.002	2.54(1.32-4.72)	.004
1 Or 0 copies of RVs	609	1 [Reference]		1 [Reference]		1 [Reference]
Fully adjusted
2 Copies of APOL1 RVs	88	1.93 (1.15-3.26)	.01	2.11 (1.31-3.39)	.002	2.51(1.21-5.05)	.01
1 Or 0 copies of RVs	569	1 [Reference]		1 [Reference]		1 [Reference]

Abbreviations: AKI, acute kidney injury; APOL1, apolipoprotein L1; BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); eGFR, estimated glomerular filtration rate; RV, risk variants.

^a A recessive model of inheritance was used for these analyses. Logistic regression was used to evaluate the association of APOL1 high-risk group and AKI as a binary outcome. Partial proportional odds logistic regression was used to evaluate the association of APOL1 high-risk groups and AKI stages (controls, stage 1, stage 2 and stage 3). Model 1 or minimally adjusted: adjusted for age, sex, baseline eGFR and 10 PCs of ancestry. Model 2 or fully adjusted = model 1 + BMI, diabetes, hypertension, COPD, CHF, liver disease, smoking, systolic blood pressure, diastolic blood pressure, ACE inhibitor /ARBs outpatient (180 days prior to admission). Inpatient vancomycin, inpatient NSAID, ACE inhibitor/ARB inpatient, Remdesivir, dexamethasone (definitions included in eTable 3 in Supplement 1). All drug administration exposures were accounted for if they occurred prior to development peak creatinine.