# Testosterone modifies U-Shaped association of eGFR with all-cause mortality in Chinese female centenarians: a prospective cohort study

## TL;DR

In a study of Chinese female centenarians, low kidney function is linked to higher mortality, but this risk is reduced by higher testosterone levels.

## Contribution

First report of a U-shaped eGFR-mortality relationship in female centenarians and testosterone's role as a modifier.

## Key findings

- eGFR below 45 mL/min/1.73 m² is associated with a 44.9% higher mortality risk.
- Testosterone levels significantly modify the eGFR-mortality relationship.
- A U-shaped curve with inflection points at 45 and 60 mL/min/1.73 m² was identified.

## Abstract

The relationship between kidney function and mortality in centenarians, particularly with respect to hormonal regulation, remains unclear. This study investigated the association between estimated glomerular filtration rate (eGFR) and all-cause mortality in female centenarians and explored the potential role of testosterone.

Within the China Hainan Centenarian Cohort Study, 701 female centenarians (median age: 102 years) were enrolled. eGFR was calculated using the CKD-EPI 2009 creatinine equation. Restricted cubic splines (RCSs) and multivariable Cox proportional hazards models were employed to assess nonlinear associations. Likelihood ratio tests were used to evaluate the interaction effect of testosterone.

During a median follow-up of 31 months, 643 participants (91.7%) died. RCS analysis revealed a significant nonlinear association (P-overall < 0.001; P-nonlinear = 0.004) between eGFR and all-cause mortality, characterized by a U-shaped curve, with an inflection point of 60.65 mL/min/1.73 m² identified in the adjusted RCS model. Multivariable-adjusted Cox regression analysis showed that each 10-mL/min/1.73 m² increase in eGFR was associated with a 6.8% reduction in all-cause mortality risk (hazard ratio = 0.932, 95% confidence interval: 0.882–0.985, P = 0.012). Compared with centenarians with eGFRs of 45–<60 mL/min/1.73 m², those with eGFRs of ≤ 45 mL/min/1.73 m² had a 44.9% increased risk of all-cause mortality (hazard ratio = 1.449, 95% confidence interval: 1.176–1.785, P < 0.001) and a significantly shorter median survival time (26 months vs. 34 months, P < 0.001). Higher testosterone levels attenuated the association between eGFR and all-cause mortality, indicating a protective effect (P for multivariable nonlinear interaction = 0.022).

In female centenarians, lower eGFR (< 45 mL/min/1.73 m²) is independently associated with all-cause mortality, and this association is modified by serum testosterone levels. These findings highlight testosterone as a potential modulator of the kidney function–mortality relationship in centenarians, although the causal relationship requires further investigation.

The online version contains supplementary material available at 10.1186/s12958-026-01529-w.

Novel U-shaped association: For the first time, we report a significant U-shaped association between estimated glomerular filtration rate (eGFR) and all-cause mortality in a large prospective cohort of female centenarians.

Critical renal thresholds: eGFR values of 45 and 60 mL/min/1.73 m² were identified as critical inflection points.

Testosterone as an effect modifier: Serum testosterone levels significantly modify the eGFR–mortality relationship, representing a novel finding in extreme aging.

Protective role of testosterone: Higher testosterone levels attenuate the increased mortality risk associated with low eGFR, suggesting a protective effect in this unique female population.

Sex-specific insights: This study challenges traditional paradigms by highlighting the potentially beneficial role of testosterone in renal function and overall survival among the oldest-old women.

The online version contains supplementary material available at 10.1186/s12958-026-01529-w.