Depression, immune-metabolic alterations and mortality in haemodialysis: a prospective cohort study

Introduction: Depression is common among people receiving haemodialysis and is associated with adverse outcomes, but the biological pathways underlying this relationship remain uncertain. We examined associations between depressive symptoms, mortality, immune-metabolic markers and stress-related transcriptional profiles, including potential sex differences in routine biomarkers. Methods: We studied 297 adults receiving maintenance haemodialysis across North and East London and Essex. Moderate/severe depressive symptoms were defined as a 17-item Hamilton Depression Rating Scale score >18. Cox regression examined all-cause mortality from enrolment until death or administrative censoring on 24 March 2026. Linear regression assessed associations between depressive-symptom severity and routinely measured immune-metabolic markers, overall and by sex. In a laboratory subset, inflammatory proteins and transcriptional profiles were examined overall only. Results: Moderate/severe depressive symptoms were associated with higher mortality in the adjusted model, although the association was attenuated after full adjustment (HR= 1.49; 95% CI 1.00-2.23; p=0.055). Higher white-cell count was associated with greater depressive-symptom severity before, but not after, adjustment for body mass index. Diastolic blood pressure was the only routine marker showing evidence that its association with depressive symptoms differed by sex (interaction p=0.022). Exploratory analyses did not provide convincing evidence that the measured biomarkers accounted for the depression-mortality association. In the laboratory subset, no inflammatory protein or transcriptional measure was significantly associated with depressive symptoms in fully adjusted analyses. Conclusion: Depression may identify a clinically relevant risk state among people receiving haemodialysis. The exploratory biological findings require validation in larger, prospectively sampled cohorts.