Abstracts from the 6th Joint Meeting of ABCD & UKKA 2023

Objectives: Globally, 40% of people receiving peritoneal dialysis have a diagnosis of diabetes.High-quality data on the potential impact of improving glycaemic control are needed to inform the KDIGO recommendation to individualise HbA1c targets.Methods: The association between first HbA1c and all-cause mortality in people on peritoneal dialysis (PD) for kidney failure recruited into PDOPPS1(2014-2017) and PDOPPS2(2018-2022) identified as diabetic was estimated using Cox proportional hazards models adjusted for age, sex, race, country, albumin, haemoglobin and co-morbidities.To inform HbA1c individualisation, subgroup analyses drawn from these adjustment variables were performed.Results: From 24,259 individuals recruited into PDOPPS, 13,646 were identified as diabetic.Of these, 9,722 had HbA1c performed after a mean of 11.2 months' PD therapy, mean follow-up 17.2 months.Mean HbA1c was 6.9%.Mean HbA1c ranged from 6.4% in Japan to 7.3% in Canada.In people with type 2 diabetes (T2DM), relative to HbA1c 6.0-7.0%,there was weak evidence for increased all-cause mortality for HbA1c >9% (HR 1.18, p=0.067), becoming more robust in those aged <65 years (HR 1.4, p=0.01).In individuals with albumin >3.0g/dL, or those with no previous coronary artery disease (CAD), the threshold for significantly increased mortality dropped to >8% (HR ~1.2 ).The hazard ratio for mortality for the >8% threshold climbed to 1.92 for those aged <65 years, with both serum albumin >3.0g/dL and no previous CAD.Conclusions: In diabetic adults receiving PD for kidney failure, the associations seen between HbA1c and mortality argue for tighter individualised targets for particular patient subgroups (younger, non-inflamed and without established CAD) than clinical practice guidelines have previously suggested.