Abstract
BACKGROUND: To determine temporal changes in the prevalence and associates of lower extremity amputation (LEA) complicating type 2 diabetes. METHODS: Baseline data from the longitudinal observational Fremantle Diabetes Study (FDS) relating to LEA and its risk factors collected from 1296 patients recruited to FDS Phase 1 (FDS1) from 1993 to 1996 and from 1509 patients recruited to FDS Phase 2 (FDS2) from 2008 to 2011 were analysed. Multiple logistic regression was used to determine associates of prevalent LEA in individual and pooled phases. Generalised linear modelling was used to examine whether diabetes related LEA prevalence and its associates had changed between Phases. RESULTS: There were 15 diabetes-related LEAs at baseline in FDS1 (1.2 %) and 15 in FDS2 (1.0 %; P = 0.22 after age, sex and race/ethnicity adjustment). In multivariable analysis, independent associates of a baseline LEA in FDS1 were a history of vascular bypass surgery or revascularisation, urinary albumin:creatinine ratio, peripheral sensory neuropathy and cerebrovascular disease (P ≤ 0.035). In FDS2, prevalent LEA was independently associated with a history of vascular bypass surgery or revascularisation, past hospitalisation for/current foot ulcer and fasting serum glucose (P ≤ 0.001). In pooled analyses, a history of vascular bypass or revascularisation, past hospitalisation for/current foot ulcer at baseline, urinary albumin:creatinine ratio (P < 0.001), as well as FDS Phase as a binary variable [odds ratio (95 % confidence interval): 0.28 (0.09-0.84) for FDS2 vs FDS1, P = 0.023] were associated with a lower risk of LEA at study entry. CONCLUSIONS: The risk of prevalent LEA in two cohorts of patients with type 2 diabetes from the same Australian community fell by 72 % over a 15-year period after adjustment for important between-group differences in diabetes-related and other variables. This improvement reflects primary care foot health-related initiatives introduced between Phases, and should have important individual and societal benefits against a background of a progressively increasing diabetes burden.