Abstract 349: Comparison of Systolic, Diastolic, and Pulse Pressure, and Blood Pressure Variability, as Predictors of Kidney Function Loss in Persons With Chronic Kidney Disease in the Chronic Renal Insufficiency Cohort

Blood pressure (BP) predicts kidney function decline in CKD. It’’s unclear if systolic (SBP), diastolic (DBP), or pulse pressure (PP) is the strongest predictor, or if BP variability is a determinant. In a prospective study of CKD we examined the association of 3 brachial BP measures on a composite kidney function outcome: end-stage kidney disease (ESKD) or a halving in the estimated GFR compared to baseline. We also examined whether the amount of variability within the 3 BP measurements at baseline predicted the outcome. CRIC participants (∼half DM; ∼half M; eGFR 44.8(16.8) mL/min/1.73 m 2 ) had 3 seated brachial BP at study baseline. Mean BP [mmHg(SD)] and PP were 128(22)/72(13) and 57(19). Mean BP variability, as the SD of SBP (SBPSD) at baseline was 3.3(2.5). After 3.7 years, 766 participants had a halving in eGFR from baseline or ESKD. We modeled the time to the event with Cox proportional hazards regression. For comparability across different BP measures, results were expressed per 1 SD difference in SBP, DBP, and PP; adjusting for eGFR, proteinuria, age, gender, race, clinical site and diabetes(Y/N) in the 3379 participants with complete data. Death was a censoring event. Akaike Information Criterion (AIC; lower values better) evaluated overall model fit characteristics. The same regression parameters were used to model CKD associated with SBPSD, including the mean of the SBP at baseline as a covariate in this analysis. SBP was the strongest BP predictor of loss of kidney function. Variability of baseline SBP also predicted loss of kidney function, independent of level of SBP at baseline, but this association was considerably weaker than those observed for other BP measures.