W LIM1,2, D JOHNSON3,4,5, C HAWLEY3,4,5, C LOK6, K POLKINGHORNE7,8,9, M ROBERTS10,11, N BOUDVILLE1,2, G WONG12,13,14
1Sir Charles Gairdner Hospital, Perth, Australia, 2University of Western Australia, Perth, Australia, 3Princess Alexandra Hospital, Brisbane, Australia, 4University of Queensland, Brisbane, Australia, 5Translational Research Institute, Brisbane, Australia, 6University of Toronto, Toronto, Canada, 7Monash Medical Centre, Melbourne, Australia, 8Monash University – Department of Medicine, Melbourne, Australia, 9Department of Epidemiology and Preventive Medicine, Monash University, Melbourne, Australia, 10Eastern Health Clinical School, Monash University, Melbourne, Australia, 11Eastern Health Integrated Renal Service, Melbourne, Australia, 12University of Sydney, Sydney, Australia, 13Centre for Kidney Research, The Children’s Hospital at Westmead, Sydney, Australia, 14Centre for Transplant and Renal Research, Westmead Hospital, Sydney, Australia
Aim: To examine the association between T2DM with and without diabetic nephropathy and CVD mortality in incident dialysis patients and to determine whether this association was modified by age.
Background: The overall survival of dialysis-dependent end-stage kidney disease (ESKD) patients with T2DM may be different for those with diabetic nephropathy than those with ESKD attributed to another cause.
Methods: Using data from the ANZDATA Registry, all incident ESKD patients commencing dialysis in Australia and New Zealand between 1980-2014 were included. The association between diabetes status (i.e. no diabetes, T2DM with ESKD attributed to diabetic nephropathy or T2DM with ESKD attributed to non-diabetic nephropathy) and CVD mortality was examined using adjusted Cox regression and competing risk analyses.
Results: Of 56,552 patients followed for a median of 2.5 years (193,549 person-years), 15,829 (30.0%) and 4993 (8.8%) had T2DM with diabetic nephropathy and non-diabetic nephropathy, respectively. Compared to patients without diabetes, the adjusted hazard ratio (HR) for CVD mortality was 1.63 (95%CI 1.55-1.71) for patients with T2DM and diabetic nephropathy, and 1.31 (1.22-1.40) for patients with T2DM and non-diabetic nephropathy. There was an interaction between age and CVD mortality (pinteraction<0.001), with the excess risk of CVD mortality being greater in younger patients with T2DM and diabetic nephropathy or T2DM and non-diabetic nephropathy (age ≤30 years: 3.25 [1.87, 5.63] and 1.64 [0.71, 3.80]; >70 years: 1.24 [1.13, 1.35] and 1.21 [1.09, 1.34], respectively). Estimates from the competing risk models were similar.
Conclusions: Prevalent T2DM is associated with a survival disadvantage in incident dialysis patients, particularly younger patients, with this disadvantage more pronounced when ESKD was attributed to diabetic nephropathy as opposed to another cause.