Risk factors for CKD amongst HIV-positive persons have been well established, but insights into the prognosis after CKD including the role of modifiable risk factors for serious clinical outcomes (SCO) are limited.
D:A:D participants developing CKD (confirmed, >3 months apart, eGFR<60mL/min/1.73 m[sup]2[/sup] or 25% eGFR decrease when eGFR<60mL/min/1.73m[sup]2[/sup]) after 2004 were followed from date of CKD date incident SCO (end stage renal (ESRD) and liver disease (ESLD), cardiovascular disease (CVD), AIDS- and non-AIDS defining malignancies (ADM and NADM), other AIDS events or death), 6 months after last visit or Feb 1st 2016. SCO rates in persons with CKD were compared to rates in persons without CKD followed from eGFR> 60mL/min/1.73 m[sup]2[/sup] to CKD, 6 months after last visit or Feb 1st 2016. Poisson regression models considered associations between individual SCO and modifiable risk factors.
2467 persons with and 33427 persons without CKD were included. During 2.7 (IQR 1.1-5.1) years median follow-up after CKD 595 persons with CKD (24.1%) developed a SCO (IR 68.9/1000PYFU [95%CI 63.4-74.4]) with 7.9% [6.9-9.0] estimated to have a SCO at 1 year. In persons without CKD the SCO IR was 23.0/1000PYFU [22.4-23.6] with 2.8% [2.6-3.0] estimated to have a SCO at 1 year. In persons with CKD, death was the most common SCO (12.7%), followed by NADM (5.8%), CVD (5.6%), other AIDS (5.0%), ESRD (2.9%), ESLD (1.0%) and ADM (0.8%). In adjusted models poor HIV control (2.72 [2.01-3.69]), low BMI (1.68 [1.14-2.48]), diabetes (1.60 [1.19-2.15]), smoking (1.48 [1.06-2.07]) and higher eGFR (0.74 [0.68-0.80]) were strongly associated with death; poor HIV control (3.05 [1.87-4.95]), low BMI (1.96 [1.11-3.47]) and smoking (1.75 [1.02-3.00]) with other AIDS; smoking (1.78 [1.07-2.99]) and diabetes (1.65 [1.05-2.57]) with NADM; dyslipidaemia (2.22 [1.40-3.52]), smoking (1.98 [1.22-3.19]), diabetes (1.81 [1.16-2.81]) and higher eGFR (0.81 [0.72-0.92]) with CVD (figure).
In an era where many HIV-positive persons require less monitoring due to efficient antiretroviral treatment, persons with CKD have a high SCO burden requiring close monitoring. Our data suggest modifiable risk factors including smoking, diabetes, BMI, HIV-control and dyslipidaemia play a central role for post-CKD morbidity and mortality.