Refining insights on type 2 diabetes and aortic valve replacement—reply to Liu et al.

Liu et al. raise points relevant to the relationship between type 2 diabetes and incident aortic valve replacement (AVR), as was assessed in our recent publication.1 These comprise possible further analyses of associations between incident AVR and diabetes duration, blood glucose-lowering medication classes, and possible sarcopenia-related distal symmetrical polyneuropathy (DSPN), in the cohort of participants with type 2 diabetes. They also suggest further granular analyses of post-operative complications and deaths, presumably in people with versus without type 2 diabetes, including whether the surgery was urgent versus elective. As we stated in our paper,1 multivariable analyses were not performed in the participants with type 2 diabetes, given the limited number of incident AVR hospitalisations in this group (n = 24 or 1.7% of the total cohort). A larger study would be needed to assess significant independent predictors of AVR. Depending on the range of variables available, such a study could accommodate the suggested stratification of diabetes duration and blood glucose-lowering medication classes, as well as an analysis of a possible interaction between sarcopenia (key indices of which including dual-energy X-ray absorptiometry for lean muscle mass were not measured in The Fremantle Diabetes Study Phase II (FDS2)) and DSPN. In the context of the similar post-operative outcomes in people with versus without type 2 diabetes in the present study,1 a comparative analysis of prognostic determinants, including diabetes-related complications and whether AVR was urgent or planned, would also only be feasible using larger, well-curated cohorts. The FDS2 provides very detailed patient-level longitudinal data2 while the much larger Spanish national database study of AVR which generated similar relative risks3 had only limited broad disease-based cross-sectional variables for inclusion in analyses. The trade-off between sample size and data coverage is well recognised in epidemiological studies,4, 5 but is being addressed through the use of comprehensive national health databases such as the Swedish National Diabetes Register.6 Open access publishing facilitated by The University of Western Australia, as part of the Wiley - The University of Western Australia agreement via the Council of Australasian University Librarians.