The epidemiology of infections among older people with diabetes mellitus and chronic kidney disease

This thesis describes the epidemiology of community-acquired infections among older people with diabetes without a history of renal replacement therapy, according to markers of chronic kidney disease (CKD): proteinuria and reduced estimated glomerular filtration rate (eGFR). The thesis uses linked electronic health records from primary and secondary care, and mortality records. Among a cohort of 219,145 patients with diabetes aged ≥65 years there was a high burden of community-acquired infection: lower respiratory tract infections (LRTIs) having the highest crude rate (152.7/1,000 years) followed by urinary tract infections (male 51.4, female 147.9/1,000 years). All-cause 28-day mortality was 32.1% for pneumonia (as a subset of LRTI) (3,115/9,697) and 31.7% for sepsis (780/2,461). Reduced eGFR was associated with a strong and graded increased risk of community-acquired LRTI, pneumonia and sepsis incidence, after adjustment for co-morbidities, smoking status and characteristics of diabetes mellitus. The effect sizes were larger for sepsis than pneumonia, and for pneumonia than LRTI. Proteinuria was a marker of increased risk of infection incidence independently of eGFR, for LRTI (rate ratio 1.07: 95%CI 1.05–1.09), pneumonia (1.26:1.19–1.33), and sepsis (1.33:1.20–1.47), after adjustment for co-morbidities, smoking status and characteristics of diabetes. Advanced CKD (eGFR<30ml/min/1.73m2) was associated with 28-day mortality following community-acquired pneumonia (risk ratio=1.27:95%CI 1.10–1.47) and sepsis (RR=1.42:1.10–1.84) compared to eGFR≥60 ml/min/1.73m2), adjusted for age, sex, socioeconomic status, smoking status and co-morbidities. Lesser reductions in eGFR and proteinuria were not associated with mortality. The protective effects of pneumococcal vaccine against community-acquired pneumonia appeared to wane swiftly. There was scant evidence for any impact of influenza vaccination against the total burden of community-acquired LRTI. This study allows patients, clinicians and public health planners to quantify infection risks among older people with diabetes according to CKD status. Further research could explore mechanisms and prevention strategies, including enhanced vaccination schedules.