Body Composition, Coronary Microvascular Dysfunction, and Future Risk of Cardiovascular Events Including Heart Failure

Abstract

Background: Body mass index (BMI) is a controversial marker of cardiovascular prognosis, especially in women. Coronary microvascular dysfunction (CMD) is prevalent in obese patients and a better discriminator of risk than BMI, but its association with body composition is unknown.

Objectives: The authors used a deep learning model for body composition analysis to investigate the relationship between CMD, skeletal muscle (SM), subcutaneous adipose tissue (SAT), and visceral adipose tissue (VAT), and their contribution to adverse outcomes in patients referred for evaluation of coronary artery disease.

Methods: Consecutive patients (n = 400) with normal perfusion and preserved left ventricular ejection fraction on cardiac stress positron emission tomography were followed (median, 6.0 years) for major adverse events, including death and hospitalization for myocardial infarction or heart failure. Coronary flow reserve (CFR) was quantified as stress/rest myocardial blood flow from positron emission tomography. SM, SAT, and VAT cross-sectional areas were extracted from abdominal computed tomography at the third lumbar vertebra using a validated automated algorithm.

Results: Median age was 63, 71% were female, 50% non-White, and 50% obese. Compared with the nonobese, patients with obesity (BMI: 30.0-68.4 kg/m²) had higher SAT, VAT, and SM, and lower CFR (all P < 0.001). In adjusted analyses, decreased SM but not increased SAT or VAT was significantly associated with CMD (CFR <2; OR: 1.38; 95% CI: 1.08-1.75 per -10 cm²/m² SM index; P < 0.01). Both lower CFR and SM, but not higher SAT or VAT, were independently associated with adverse events (HR: 1.83; 95% CI: 1.25-2.68 per -1 U CFR and HR: 1.53; 95% CI: 1.20-1.96 per -10 cm²/m² SM index, respectively; P < 0.002 for both), especially heart failure hospitalization (HR: 2.36; 95% CI: 1.31-4.24 per -1 U CFR and HR: 1.87; 95% CI: 1.30-2.69 per -10 cm²/m² SM index; P < 0.004 for both). There was a significant interaction between CFR and SM (adjusted P = 0.026), such that patients with CMD and sarcopenia demonstrated the highest rate of adverse events, especially among young, female, and obese patients (all P < 0.005).

Conclusions: In a predominantly female cohort of patients without flow-limiting coronary artery disease, deficient muscularity, not excess adiposity, was independently associated with CMD and future adverse outcomes, especially heart failure. In patients with suspected ischemia and no obstructive coronary artery disease, characterization of lean body mass and coronary microvascular function may help to distinguish obese phenotypes at risk for cardiovascular events.

Keywords: body composition; cardiometabolic disease; coronary microvascular dysfunction; heart failure; ischemia and no obstructive coronary artery disease; lean body mass; obesity.

FIGURE 1. Abdominal Body Composition Compartments and Relationship to BMI

(A) Representative computed tomography (CT) slice selection at the third lumbar vertebra (L3) level and segmentation of subcutaneous adipose tissue (SAT) (green), skeletal muscle (SM) (red), and visceral adipose tissue (VAT) (yellow). (B) Regression output generated by a DenseNet architecture model after slice processing and assignment of an estimated z-axis (craniocaudal) offset from L3. After sigmoid curve fitting, the zero-crossing is selected as the L3 level (solid vertical red line). Relationship between body mass index (BMI) (C) and indices of SAT (D), VAT (E), and SM (F).

FIGURE 2. Freedom From Major Adverse Events by Coronary Microvascular Dysfunction and Sarcopenia

Kaplan-Meier (unadjusted) (A) and adjusted (B) freedom from major adverse events by coronary microvascular dysfunction (CMD) and sarcopenia. Major adverse events include death and hospitalization for nonfatal myocardial infarction or heart failure. *Adjusted for pretest clinical score, race, estimated glomerular filtration rate, left ventricular ejection fraction, and total adipose fat index. †P value refers to interaction between coronary flow reserve and skeletal muscle index as continuous variables.

FIGURE 3. Adjusted Annualized Rate of Major Adverse Events by CMD and Sarcopenia

Major adverse events include death and hospitalization for nonfatal myocardial infarction or heart failure. Poisson regression was adjusted for pretest clinical score, race, estimated glomerular filtration rate, left ventricular ejection fraction and total adipose fat index. Abbreviation as in Figure 2.

FIGURE 4. Major Adverse Events by CMD and Sarcopenia in Younger, Female, and Obese Patients

Kaplan-Meier (unadjusted) freedom from major adverse events by CMD and sarcopenia stratified by younger age (A), female sex (B), and obesity (C). Major adverse events include death and hospitalization for nonfatal myocardial infarction or heart failure. Abbreviations as in Figures 1 and 2.

CENTRAL ILLUSTRATION. A Sarcopenic CMD Phenotype is Associated With Highest Risk of Adverse Cardiovascular Events, Including Heart Failure, in INOCA Patients With Normal or Elevated BMI

Assessments of BMI are insufficient to stratify cardiovascular disease risk, especially in women. Characterization of body composition metrics in conjunction with evaluation for coronary microvascular dysfunction (CMD) may help to elucidate hidden at-risk phenotypes among patients with cardiometabolic risk and clarify mechanistic pathways in adverse cardiovascular outcomes including heart failure. BMI = body mass index; Cre = creatinine; LVEF = left ventricular ejection fraction; INOCA = ischemia and no obstructive coronary artery disease.