Abstract
The Oxford (split) and Japanese Histological Grade (JHG [lumped]) classifications of immunoglobulin A nephropathy (IgAN) are widely used in Japan. We compared their prognostic ability and explored the benefits of combining them. Data were from a prospective IgAN cohort (2005–2015). The primary endpoint was a ≥ 50% decline in estimated glomerular filtration rate (eGFR) or end-stage renal disease. Cox regression with Uno C assessed prognostic discrimination. The integrated discrimination improvement index evaluated the added value of histologic to clinical variables (eGFR, urinary protein, and mean arterial pressure) at 1, 5, and 10 years. Kaplan–Meier analysis stratified patients into JHG 1 and 2–4 to assess the prognostic value of mesangial hypercellularity (M) and interstitial fibrosis/tubular atrophy (T). Among 938 patients (median follow-up: 66 months), 58 (6.2%) reached the endpoint. M and T (Oxford) and JHG were significantly associated with outcomes. Both systems showed good discrimination (Uno C: 0.83 and 0.86). Adding Oxford to clinical variables improved prediction at 5 and 10 years; adding JHG improved prediction at 1, 5, and 10 years. M and T were predictive in JHG 2–4 but not in JHG 1. Both classifications demonstrated comparable discrimination. Their combined use can help enhance predictive precision. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1038/s41598-026-37412-2.