Biological sex modifies the association between diabetes mellitus and cardiac remodelling: a UK Biobank cohort study

Introduction

Diabetes mellitus (DM) is associated with adverse cardiac remodelling (1,2), but this differs by sex remains poorly understood. DM is associated with substantially increased risk of incident heart failure (HF) with both reduced and preserved ejection fraction. (3) The HF incidence is even higher for females compared to males, being 5-fold vs 2.4-fold greater, respectively, even after adjustment for risk factors, such as age, coronary artery disease, and hypertension.(4) Cardiac Magnetic Resonance Imaging (CMR) is a valuable tool to characterize myocardial structure and function.(5)

Purpose

To compare CMR phenotypes associated with DM in men versus women.

Methods

We included all UK Biobank participants with baseline CMR data (n=80,889). They were stratified by sex and DM status. We defined DM as described in our published work.(6) Statistical analyses were conducted using R via the UK Biobank Research Analysis Platform. Linear regression models were used to define age-adjusted p values for interaction between DM and sex in their association with CMR and other baseline parameters.

Results

Amongst both sexes, DM was associated with higher BMI and HbA1c (table 1). These differences were more pronounced in females (p value for interaction <0.005). DM was associated with lower systolic blood pressure in men, with the opposite noted in women with DM (interaction p<0.005). No DM-sex interactions were observed for the cardiac biomarkers NT-proBNP or Troponin I, (table 1).

Significant sex–DM interactions were observed for multiple structural CMR parameters (table 1). In the left ventricle, DM was associated with reduced indexed end-diastolic and end-systolic volume in both sexes, but with a more pronounced phenotype in men (LVEDVi interaction p=0.012; LVESVi interaction p=0.015). DM was also associated with lower LV mass index, but only in men (interaction p<0.005). Despite structural differences, LV ejection fraction and myocardial strain (longitudinal, circumferential, radial) showed no significant sex–DM interaction. No interactions were observed for left atrial parameters (table 1).

Right ventricular indexed volumes were reduced in people with DM of both sexes, but to a greater extent in men (RVEDVi interaction p=0.010; RVESVi interaction p<0.005); right ventricular ejection fraction was reduced in women with DM, but not men (interaction p=0.009). Right atrial volumes showed a similar interaction pattern to right ventricular volumes, whereas right atrial ejection fraction was higher in women with DM, but did not differ in men with DM.

Conclusion

DM is associated with distinct CMR phenotypes in men versus women. Defining the mechanisms of these may help to define sex-specific interventions that address the differing risks of heart failure associated with DM in men and women.Table 1:Demographic & CMR results