Variations in Creatinine Generation Among Patients With Glomerular Disease: Evidence From the NEPTUNE and CureGN Studies

Rationale & Objective Estimation of glomerular filtration rate (GFR) assumes that creatinine generation (crG) is relatively stable. This study identified factors associated with crG variability and its impact on serum creatinine changes (Δ Scr) among patients with glomerular disease. Study Design An observational cohort study. Setting & Participants Nephrotic Syndrome Study Network and Cure Glomerulonephropathy adult and pediatric participants with at least one crG measurement. Predictors Potential predictors of crG levels included age, sex, disease diagnosis, weight status, estimated GFR (eGFR), urine protein, steroid use, and nonsteroid immunosuppressant use. crG change (Δ crG) was then used as an exposure to assess impacts on Δ Scr. Outcomes crG levels and Δ Scr. Analytical Approach The intraclass correlation coefficient illustrated crG variability within individuals. Multivariable linear mixed models identified factors associated with crG levels. Among those with 2+crG measurements, multivariable linear mixed models estimated the association between Δ crG and Δ Scr. Results Among 4,626 crG measurements from 1,081 participants, there was only moderate correlation between measurements within individuals (intraclass correlation coefficient=0.517, 95% CI, 0.482-0.548) overall. For pediatric participants, factors significantly associated with crG included age, sex, weight status, and urine protein. Among adults, significant factors were age, sex, disease diagnosis, weight status, eGFR, steroid use, and nonsteroid immunosuppressant use. Limitations The 24-hour urine collections may have collection error, measured GFR was unavailable, and edema status was unavailable. Conclusions crG was highly dynamic within individuals over time and varied with glomerular disease activity and treatments. The impact of Δ crG on Δ Scr —and subsequently on estimation of kidney function—is potentially large. Accounting for these changes or development of alternative kidney function measures are needed among glomerular disease patients. Plain Language Summary Creatinine generation is often assumed to be stable when using creatinine to estimate kidney function and track kidney function over time, but it can vary with chronic disease. This study showed high variability in creatinine generation within individuals with glomerular disease. Besides age, sex, and weight, important factors that can impact creatinine generation include urine protein in children and kidney function, disease diagnosis, steroid use, and nonsteroid immunosuppressant use in adults. Changes in creatinine generation may also impact changes in the serum creatinine, which would influence estimates of kidney function, but this relationship needs to be studied further. In the meantime, accounting for factors affecting creatinine generation or using alternative estimates of kidney function in patients with glomerular diseases is needed.