Bidirectional associations between kidney function decline and carotid plaque progression: a longitudinal cohort...

Atherosclerosis and chronic kidney disease are major contributors to cardiovascular disease (CVD) and premature mortality worldwide. However, how kidney function decline and carotid plaque (CP) progression influence each other over time remains unclear. In the context of predictive, preventive, and personalised medicine (PPPM/3PM), we investigated the bidirectional associations between kidney function decline and CP progression by leveraging both baseline and repeated measurements of estimated glomerular filtration rate (eGFR) and total plaque area (TPA). Understanding these relationships may facilitate early risk stratification at the subclinical stage and guide targeted preventive and personalised interventions for high-risk individuals, ultimately improving long-term cardiorenal outcomes.

We derived three sub-cohorts from the Beijing Health Management Cohort. Sub-cohort 1 included 11,657 participants who underwent at least two examinations between 2010 and 2018; cross-lagged panel analyses were conducted to evaluate the bidirectional associations between eGFR and TPA. Sub-cohort 2 comprised 4173 participants free of CP at baseline; Cox proportional hazards models were used to assess associations of eGFR slope and cumulative eGFR with incident CP. Sub-cohort 3 consisted of 7601 participants with baseline eGFR ≥ 60 mL/min/1.73 m²; Cox models were applied to examine associations between TPA slope, cumulative TPA, and kidney function decline. Hazard ratios (HRs) and 95% confidence intervals (CIs) were calculated. C-statistics, integrated discrimination improvement, and the net reclassification index were used to estimate the incremental predictive value.

In sub-cohort 1, cross-lagged panel analyses demonstrated a significant bidirectional association between eGFR and TPA after adjusting for confounders. Higher baseline eGFR predicted lower subsequent TPA (β = -0.029, P < 0.001), whereas higher baseline TPA predicted lower subsequent eGFR (β = -0.070, P < 0.001). In sub-cohort 2, during a median follow-up of 3.98 years, 922 participants developed incident CP. The eGFR slope (HR: 0.804, 95%CI: 0.750-0.862) and cumulative eGFR (HR: 0.805, 95%CI: 0.751-0.863) were negatively associated with incident CP. In sub-cohort 3, over a median follow-up of 4.86 years, kidney function decline occurred in 239 participants. The TPA slope (HR: 1.222, 95%CI: 1.133-1.317) and cumulative TPA (HR: 1.244, 95%CI: 1.136-1.362) were positively associated with kidney function decline. Finally, incorporating eGFR and TPA measures, particularly their slopes and cumulative levels, yielded greater incremental improvements in predicting incident CP and kidney function decline, respectively.

These findings demonstrate a bidirectional association between kidney function decline and CP progression, supported by baseline levels, slopes, and cumulative exposure metrics. This reciprocal relationship underscores the potential of integrated, dynamic monitoring of eGFR and TPA as predictive diagnostic tools for identifying subclinical cardiorenal risk, thereby guiding personalised preventive strategies within the PPPM/3PM paradigm.

The online version contains supplementary material available at 10.1007/s13167-025-00425-4.