Importance of very low proteinuria levels in immunoglobulin A nephropathy: evidence of a non-linear association with long-term kidney survival

Abstract Background Proteinuria is a major predictor of disease progression in immunoglobulin A nephropathy (IgAN), and lower levels are associated with more favorable kidney outcomes. Prior studies have largely evaluated proteinuria using predefined categories (e.g., < 0.5 g/d; ≥0.5 - <1.0 g/d; ≥1.0 g/d etc.), limiting the continuity of the dose-response and potential non-linear associations between proteinuria and long-term kidney survival. This study sought to characterize the association between proteinuria as a continuous variable and long-term kidney survival. Methods The relationship between proteinuria as a continuous variable and kidney survival was examined to describe the functional form using two complementary approaches based on different data sources. Kidney survival was defined as the time free from kidney failure. The first approach was based on the best-fitting model from a dose-response meta-analysis (informed by two recent systematic literature reviews). The second approach used the International IgAN Prediction Tool developed by Barbour et al. (the prediction tool recommended by the KDIGO 2025 guideline), which is a validated Cox proportional hazards model incorporating patient risk factors such as proteinuria, eGFR, and histologic features. In both approaches, proteinuria was varied while other inputs were held constant, while kidney survival was plotted across proteinuria levels to characterize the shape of the association. Results Both approaches produced similar results demonstrating a non-linear relationship between proteinuria and kidney failure. In the first approach, a dose-response meta-analysis of 11 studies including 6,955 patients showed that model fit statistics supported selection of a quadratic model. In the quadratic model, both the main term (HR 4.10; 95% CI 2.70–6.22) and the quadratic term (HR 0.82; 95% CI 0.76–0.88) were statistically significant, indicating a non-linear association. When median kidney survival was estimated across proteinuria levels, equivalent absolute reductions produced larger survival gains at lower proteinuria levels: a reduction in proteinuria from 0.5 to 0.3 g/day corresponded to an approximately 3-year increase in median kidney survival, versus approximately 2 months for the same absolute reduction from 3.0 to 2.8 g/day. The second approach, the International IgAN Prediction Tool, showed a similar non-linear pattern. Conclusions The findings from this study demonstrate the non-linear relationship between proteinuria and long-term kidney outcomes. Specifically, at low proteinuria levels, absolute reductions in proteinuria were associated with large gains in long-term kidney survival, while the same absolute reductions at higher proteinuria levels were associated with smaller gains in kidney survival. The findings suggest that reductions of proteinuria to very low levels (< 0.3 g/day) are beneficial even when proteinuria levels are already low.