Introduction:
Increasing evidence demonstrates a role of cardiovascular fat in the pathogenesis of CAD. It is unknown whether heart and vascular fat depots are related to atherosclerotic burden in women transitioning through menopause, a time of increasing CAD risk.
Hypothesis:
We hypothesize that volumes of epicardial (EAT), pericardial (PAT), total heart (TAT=EAT+PAT) and peri-aortic (PVAT) adipose tissues are associated with presence and severity of coronary artery calcification (CAC) in a sample of white and black midlife women.
Methods:
CAC and cardiovascular fat depots were quantified by electron beam CT. Outcomes were presence of CAC (none vs. any: Agatston score >0), and severity of CAC (CAC Agatston score). Logistic and tobit regression were used. Final models were adjusted for age, race, study site, menopausal status, obesity (BMI ≥30 Kg/m
2
), systolic blood pressure, lipids, homeostasis model assessment insulin resistance index, current smoking, physical activity, comorbidity (history of hypertension, stroke, angina, heart attack or diabetes) and medication use (cholesterol lowering, antihypertensive or antidiabetic medications).
Results:
The study included 509 women (37.9% black; 58.4% pre-/early perimenopausal, 41.6% late peri-/postmenopausal) aged 46-59 years with data on any of the 4 fat depots. CAC was found in 47.4% of the participants. Odds ratios (95% CI) from final logistic regression models showed that higher volumes of EAT (2.43 (1.22, 4.86), PAT (1.57 (1.04, 2.37), and TAT (2.43 (1.22, 4.87), were significantly associated with higher odds of presence of CAC. Similarly, tobit regression models showed that higher volumes per 1 log-unit increase of EAT (β (SE): 28.0 (10.7)), PAT (16.5 (6.6)), and TAT (30.0 (10.7)), were significantly associated with greater severity of CAC in final adjusted models. PVAT was not associated with either CAC presence or severity in final models.
Conclusions:
Heart fat (EAT, PAT and TAT), but not peri-aortic fat were independently associated with greater presence and severity of CAC in women at midlife, suggesting that local cardiovascular fat depots may contribute to CAD in midlife women. Future work is warranted to understand the underlying mechanistic pathways.
Empirical L2-distance lack-of-fit tests for Tobit regression models
