Sex-Specific Associations With Abnormal Myocardial Flow Reserve in Non-Obstructive Coronary Artery Disease: Insights From a Real-World Cadmium-Zinc-Telluride SPECT Study.

BACKGROUND

Coronary microvascular dysfunction (CMVD) is prevalent in non-obstructive coronary artery disease (NOCAD), and potential sex-related determinants remain incompletely characterized. This study explored sex-specific associations with abnormal myocardial flow reserve (MFR) using cadmium-zinc-telluride single-photon emission computed tomography (CZT-SPECT) in a real-world cohort.

METHODS

We retrospectively analyzed 393 NOCAD patients (56.7% female) undergoing CZT-SPECT. Abnormal MFR was defined as < 2.5. Candidate variables were identified through group comparisons (p < 0.10) and univariate logistic regression, followed by multicollinearity assessment. A pooled multivariable logistic regression model incorporating main effects and sex interaction terms was constructed. Receiver operating characteristic (ROC) analysis evaluated discriminatory performance.

RESULTS

Abnormal MFR was observed in 49.1% of patients. In the multivariable model, LDL-C was associated with abnormal MFR (OR: 1.471, 95% CI: 1.072-2.018; p = 0.017). A significant sex × MDRD-eGFR interaction was identified (OR: 1.024, 95% CI: 1.003-1.046; p = 0.026), with a borderline sex × LDL-C interaction (OR: 0.650; p = 0.089). The overall model demonstrated moderate discrimination (AUC: 0.680). In females, LDL-C showed modest discriminatory ability (AUC: 0.586). Among patients with abnormal MFR, females exhibited higher LCX and RCA territory resting myocardial blood flow and higher E/e', and RCA-rMBF was inversely correlated with E/e'.

CONCLUSIONS

Sex-modified associations of renal and lipid parameters with abnormal MFR were observed in this cohort. These interaction patterns and female-specific phenotypes provide preliminary insights into cardiorenal-microvascular relationships in CMVD, but should be interpreted as exploratory and hypothesis-generating and warrant further validation.