Longitudinal evaluation of glomerular filtration rate across adolescence and early adulthood in youth with type 1 diabetes: associations with clinical and social risk factors.

BACKGROUND: The continuous evaluation of GFR among youth with T1D has traditionally been difficult during the transition to adult care given distinct pediatric & adult estimated GFR (eGFR) formulas recommended by clinical practice guidelines. We aimed to assess longitudinal trends in kidney function in type 1 diabetes from adolescence to early adulthood using updated eGFR formulae and to examine associations between eGFR and clinical as well as social factors over time in this population. METHODS: Clinical factors including age, sex, diabetes duration, insulin therapy, BMI and HbA1c in addition to social factors including ethnicity & material deprivation (MD) were collected alongside creatinine (Cr) and cystatin c (CysC) from Canadian participants in the AdDIT & CanSOLVE CKD-SPOR studies (2009-2023). Estimated glomerular filtration rates (eGFR) derived using updated Chronic Kidney Disease in Children under 25 (CKiD U25) Cr, CysC & combined Cr-CysC formulas were evaluated longitudinally using linear mixed effects regression, as were urine albumin-creatinine ratios (uACR). RESULTS: Youth with type 1 diabetes ages 10 to <25 years were evaluated (N = 137) at a median 5 timepoints (IQR: 4-6) over 9.8 ± 1.3 years. As age increased, eGFRCr increased mildly (β:+1.0 ml/min/1.73m2/year; P<0.0001), whereas eGFRCysC decreased until the age of 18 (β:-3.3 ml/min/1.73m2/year; P<0.0001) and stabilized thereafter (β: +0.1 ml/min/1.73m2/year; P = 0.433). These trends nullified one another across the age range using the combined eGFRCr-CysC, which remained stable (β:-0.1 ml/min/1.73m2/year) from ages 10-25 years (P = 0.452). Sex, BMI, insulin therapy, HbA1c and MD were associated with eGFRCr-CysC over time with a large difference of +6.5 ml/min/1.73m2 seen in high relative to low MD groups (P = 0.0025). Log-transformed uACR was negatively associated with male sex (P = 0.017) and baseline diabetes duration (P = 0.049) yet positively associated with HbA1c (P = 0.021). CONCLUSIONS: Different eGFR trends were observed using Cr- vs. CysC-derived estimates in youth with diabetes using updated eGFR formulae. Elevations in eGFR, alongside alterations in markers of albuminuria, were associated with increased glycemia and marginalization, providing insights into key clinical and social risk factors for kidney disease monitoring in this population. (NCT01581476).