Consequences of kidney dysfunction in the community : health care-based epidemiological studies from the Stockholm creatinine measurements (scream) project

Abstract

Chronic kidney disease (CKD) is affecting an increasing share of the world’s population, and presents a global health problem. The prevalence of CKD in society and its implication for various health outcomes have been difficult to assess due to unawareness and underuse of clinical diagnoses for CKD in clinical practice. The present work aims to inform the reader of the epidemic of CKD in Sweden through a newly created repository of health care data.

Paper I describes the creation of the Stockholm CREAtinine Measurements (SCREAM) project, a health care extraction containing all creatinine values collected between 2006 and 2011 in the Stockholm region, together with a host of associated laboratory data, dispensed prescriptions, and diagnostic and demographic information. SCREAM has a good representation of the general population of Stockholm county, particularly of the elderly, with a coverage exceeding 90% among persons aged 65 or above. The coverage of people with diabetes was 98%, and that of ischemic heart disease was 97%. SCREAM captured 89% of deaths occurring in the region those years.

Paper II investigates how causes of death correlate with estimated glomerular filtration rate (eGFR) in the last year of life. We observed a higher proportion of death from cardiovascular disease and infections with lower eGFR. Among cardiovascular causes, heart failure and arrhythmias became more common with lower eGFR. Diabetes complications were also more common in lower kidney function strata.

Paper III evaluates the association between kidney function and the risk of incident fractures among people with CKD stages 3-5 (not on dialysis), and whether developing a fracture predicts subsequent risk of death or major cardiovascular adverse events (MACE). We found a gradual increase in the risk of fractures with lower eGFR in persons with CKD 3b-5 (HR 2.47 (1.94- 3.15) for hip fractures and 1.50 (1.25- 1.80) for non-hip fractures, in CKD 5 vs CKD 3a). In the aftermath of a fracture, mortality and MACE incidence were considerably increased, in both short (<90 days) and long term (>= 90 days) follow-up, as compared to non-fracture periods.

Paper IV investigates the risk and severity of hypoglycaemia in relation to kidney function, among people with diabetes followed in outpatient care. We observed a significant, gradual rise in the multivariable-adjusted risk of hypoglycaemia with lower eGFR. Several other risk factors for hypoglycaemia were identified, among them type 1 diabetes, presence of diabetic complications, and liver diseases. The risk of fatal hypoglycaemia was also higher with lower eGFR.