Urinary Prostaglandin E2 Excretion and the Risk of Cardiovascular and Kidney Disease

Abstract

Background: Inhibition of prostaglandin synthesis by nonsteroidal anti-inflammatory drugs is associated with cardiovascular mortality and kidney disease. This study hypothesizes that urinary prostaglandin E2 (PGE2) and PGE2 metabolite (PGEM) excretions are markers of cardiovascular and kidney health, because they reflect both systemic and kidney-derived PGE2 production.

Methods and results: PGE2 and PGEM were measured in spot urine samples from 2291 participants (≥55 years old) of the population-based Rotterdam Study. Urinary PGE2 and PGEM excretions were analyzed using linear regression analyses to identify cross-sectional associations with cardiovascular risk factors and baseline estimated glomerular filtration rate (eGFR). Longitudinal associations with cardiovascular mortality and kidney outcomes (eGFR <60 or <45 mL/min per 1.73 m² and the composite outcome 40% eGFR loss or kidney failure) were assessed with Cox regression. Urinary PGE2 and PGEM excretions were higher with increasing age, lower eGFR, smoking, diabetes, and albuminuria. A 2-fold higher urinary PGE2 and PGEM excretion was associated with a higher risk of cardiovascular mortality (28 825 patient-years; 160 events; PGE2 hazard ratio [HR], 1.27, [95% CI, 1.06-1.54]; PGEM HR, 1.36 [95% CI, 1.10-1.67]). Higher PGE2 excretions were also associated with a higher risk of incident eGFR <60 mL/min per 1.73 m² (31 530 person-years; 691 events; HR, 1.13 [95% CI, 1.02-1.25]) with similar HRs for the other kidney outcomes.

Conclusions: Urinary PGE2 and PGEM excretions are novel markers for the presence and progression of cardiovascular and kidney disease. Future studies should address whether these associations are causal and can be targeted to improve cardiovascular and kidney outcomes.

Keywords: cardiovascular disease; chronic kidney disease; prostaglandin E2.

Figure 1. Determinants of urinary PGE2 and PGEM excretions.

Effect plots showing the nonlinear associations between urinary prostaglandin E2 (PGE2) (top row) and prostaglandin E2 metabolite (PGEM) (bottom row) excretion and age, body mass index (BMI) (A) and creatinine‐based estimated glomerular filtration rate (eGFR) and cystatin C‐based eGFR (B). These plots are prediction plots from multivariable models corrected for age, sex, BMI, creatinine‐based eGFR, smoking status, presence of diabetes, albuminuria, and hypercholesterolemia; for the plot with cystatin C‐based eGFR, creatinine‐based eGFR is not used as a covariate. All covariates are set at either their median (age 61.7 years, BMI 26.6 kg/m², eGFR 82.1 mL/min per 1.73 m²) or most common value (female sex, never smoker, no diabetes, no albuminuria, no hypercholesterolemia). P values are from Wald tests for the overall effect of adding the nonlinear term. On the x axis a rug plot depicting data point density is shown.

Figure 2. Cardiovascular mortality.

Cumulative incidence curves for cardiovascular and noncardiovascular mortality for tertiles of urinary prostaglandin E2 (PGE2) (A) and prostaglandin E2 metabolite (PGEM), (B) excretions, respectively. P values are from log‐rank tests.

Figure 3. Survival outcomes for incident estimated glomerular filtration rate <60 mL/min per 1.73 m².

A, Kaplan–Meier survival curves and risk tables for incident estimated glomerular filtration rate (eGFR) <60 mL/min per 1.73 m² stratified to tertiles of prostaglandin E2 (PGE2, n=2009; events, n=677) and prostaglandin E metabolite (PGEM, n=2010; events, n=678). P value from log‐rank test. B, Hazard plots from the fully corrected Cox proportional hazard model (model 3) using a restricted cubic spline with 4 knots, showing nonlinearity for urinary PGEM excretion (P=0.014) but not for urinary PGE2 excretion (P=0.7). On the x axis a rug plot with the density of PGE2 and PGEM measurements is shown.

Figure 4. Forest plot of the associations between urinary PGE2 and PGEM excretions and kidney outcomes.

Hazard ratios (HRs) for the association of urinary prostaglandin E2 (PGE2, blue) and prostaglandin E2 metabolite (PGEM, red) and the 4 different kidney outcomes. The HRs are corrected for the confounders age, sex, body mass index, smoking status, and baseline estimated glomerular filtration rate (eGFR).