Traditional Cardiovascular Risk Factors and Coronary Microvascular Dysfunction in Women and Men: A Single-Center Study

Coronary microvascular dysfunction (CMD) is a common cause of effort intolerance and angina symptoms and is associated with adverse clinical outcomes, even in the absence of obstructive epicardial disease [1]. Previous studies have demonstrated that CMD is more common in women than in men [2, 3]. Additional sex-associated differences between women and men with CMD include lower coronary flow reserve (CFR) [4], higher prevalence of coronary vasospasm [5], and myocardial infarction with nonobstructive coronary arteries in women compared to men. The association between sex and traditional cardiovascular risk (TCR) factors in patients with CMD has not been extensively explored. Although the results of several studies suggest that TCR factors carry a role in the pathogenesis of CMD [6‒8], other studies have shown that these risk factors account only for a minority of patients with CMD [9]. A recent retrospective analysis of patients with CMD diagnosed by cardiac magnetic resonance showed that while diabetes and hyperlipidemia were associated with CMD in men, no association between traditional risk factors and CMD was found in women [10]. Although several noninvasive methods for the evaluation of CMD exist, during the last decade, catheter-based invasive assessment became the gold standard for the diagnosis of coronary microvascular dysfunction [11]. In the current study, we examined the association between TCR and CMD and sex-associated differences in TCR, in a prospective cohort of patients undergoing invasive microvascular function evaluation in a large tertiary medical center.

This is a retrospective analysis of prospectively collected data. Patients undergoing clinically indicated invasive evaluation of CMD were enrolled in a prospective registry at the Tel Aviv Medical Center. Clinical indication for coronary angiography could have been acute or chronic, and the decision to perform microvascular flow evaluation was according to operator’s clinical judgment. All patients had measurements of fractional flow reserve (FFR), CFR, index of microvascular resistance (IMR), and resistive reserve ratio (RRR) in the left anterior descending artery. Patients with clinically significant epicardial stenosis (FFR <0.8) and/or requiring revascularization in the index procedure were excluded from the current analysis (Fig. 1). Patients with previous significant epicardial stenosis were included in the registry. The study protocol was approved by the Tel-Aviv Sourasky Medical Center Review Board.

All measurements of microvascular function, including CFR, IMR, RRR, as well as FFR, were performed as previously described [11, 12]. In short, evaluation was performed using the bolus thermodilution technique to assess transit time with a specific intracoronary pressure wire (Aeris, Abbott Vascular, CA, USA) and the COROVENTIS system (Abbott Vascular, CA, USA). Pressure and temperature measurements were measured proximally at the level of the guiding catheter, and distally by the wire positioned at least 6 cm distally in the lumen of the coronary artery. An intravenous infusion of adenosine (140 μg/kg/min) was administered via a large peripheral vein to induce steady-state maximal hyperemia. Thermodilution was used to measure mean transit time (Tmn) and was performed by manual intracoronary bolus injections of 3 mL of normal saline (at room temperature) via the guiding catheter with the diagnostic guidewire in the left anterior descending coronary artery. Using these measurements, indices were derived to evaluate the coronary microvascular function. CFR was automatically calculated using resting Tmn divided by hyperemic Tmn. IMR was automatically calculated as the product of the mean distal coronary artery pressure and the Tmn during hyperemia. RRR was calculated as the product of CFR and the ratio between mean distal coronary artery pressure at rest and during hyperemia.

Microvascular dysfunction was defined as the presence of at least one abnormal parameter – CFR <2.5, IMR >25, or RRR <3.5 [13, 14]. As the cutoff of abnormal CFR varies between studies (2–2.5), we repeated the analysis also for the cutoff of CFR <2.0 [15]. Functional CMD was defined as reduced CFR with normal IMR [16]. Demographics and cardiovascular risk factors were obtained from patients’ medical records. Previous ischemic heart disease was defined as prior evidence of epicardial lesions with >70% stenosis or prior coronary revascularization.

Continuous variables with non-normal distribution (as assessed with the Kolmogorov-Smirnov test) are presented as median (inter-quartile range) and were compared using the Mann-Whitney U test. Categorical variables are presented as total (percentage) and were compared using χ² tests. Correlations were analyzed using the Pearson correlation coefficient. We performed multivariate linear regression analyses and included variables with significant correlations in the univariate analysis. The analyses were performed with SPSS software (version 28.0), and statistical significance was defined as p < 0.05.

Overall, 302 patients were enrolled in our registry between November 2019 and March 2023. Of those, 57 (18.9%) had evidence of obstructive epicardial disease and therefore were excluded from the current analysis. The final study population included 245 patients (62.9% female, median age 68.0 years; interquartile range: 59, 75). Among those, 141 (57.5%) had CMD (at least one abnormal parameter) (Fig. 1). No difference was found between patients with and without CMD in age, sex, other baseline characteristics, and TCRs (Table 1).

Among 141 patients diagnosed with CMD (91/141 [64.5%] women, 50/141 [35.5%] men), women were older than men (71.0 [63, 77] vs. 65 [58, 72] years, p = 0.03) and had lower median CFR (1.9 [1.5, 2.5] vs. 2.3 [1.8, 2.9], p = 0.03) and RRR (2.3 [1.7, 2.9] vs. 2.7 [2.1, 3.3], p = 0.02) values (Fig. 2). Full data regarding TCR and angiographic indices for patients with CMD are presented in Table 2. In 41 patients diagnosed with functional microvascular dysfunction, there were more women (70.7% vs. 29%). The prevalence of TCRs was similar in men and women with functional microvascular dysfunction.

A multivariate analysis demonstrated that in women, older age and the presence of previous ischemic heart disease (IHD) were associated with lower CFR values (β = −0.29; p < 0.01 and β = −0.15; p = 0.05, respectively) and lower RRR values (β = −0.28; p < 0.01 and β = −0.17; p = 0.04, respectively). No associations between TCR and invasive indices of CMD were found in men in the univariate and multivariate analysis. A full description of the univariate and multivariate analysis is presented in Table 3. When using a threshold of <2.0 for CFR, similar results were found (online suppl. Tables S1 and S2; for all online suppl. material, see https://doi.org/10.1159/000539102).

In the current study, we examined the association between TCRs and CMD, and sex-associated differences in TCRs in a real-world prospective registry of patients with nonobstructive coronary artery disease (CAD) undergoing clinically indicated microvascular function evaluation. Our main findings are as follows: first, the prevalence of TCRs was similar in patients with and without CMD. Second, among patients with CMD, women have lower CFR and RRR than men while IMR is similar in women and men. Third, while in women, older age and previous IHD were associated with lower CFR and RRR values, no association between TCRs and indices of CMD was found in men.

Blood flows from the epicardial and intra-myocardial arteries (larger than 500 µm in diameter), via pre-arterioles (100–500 µm) into the arterioles (<100 µm). Each type of vessel has distinct vaso-regulatory properties that regulate blood flow to the myocardium [17]. Disruption of endothelial-dependent and -independent vasodilatory properties impairs the ability of the microcirculation to regulate coronary blood flow and cause CMD [18]. Previous studies evaluating the association between CMD and TCR provided conflicting data. Some cardiovascular risk factors have been associated with microvascular disease in other organs. A study of 2,535 patients with type 2 diabetes demonstrated that high triglyceride levels and low high-density lipoprotein cholesterol levels are associated with microvascular kidney disease [19]. High LDL-C and triglyceride levels have also been implicated in the pathogenesis of retinal microvascular disease [20].

Smoking has been implicated as a risk factor for CMD in some studies [21]. In healthy subjects, smokers were found to have lower CFR values compared to nonsmokers [6, 22, 23]. Notably, none of the participants in these trials had clinical symptoms of CMD. A study of patients with angina with nonobstructive CAD who underwent invasive assessment of CMD showed that the presence and severity of type 2 diabetes are associated in women, but not in men, with the presence and severity of CMD [24]. Increasing age was also associated with CMD [2, 25]. Hypertension is known to cause arterial remodeling and increased media thickness. Mechanistic studies have suggested that hypertension might contribute to the development of CMD [26]. Furthermore, a study of 20 patients with mild-moderate hypertension undergoing biopsies demonstrated a correlation between increasing media-to-lumen ratio of small arteries and decreased CFR [27].

In our study, age and previous IHD were found to be correlated with CMD in women but not in men. Coronary microvascular dysfunction can coexist with obstructive epicardial disease. In patients with IHD, CMD can be caused by prearteriolar and arteriolar constriction and abnormal subepicardial prearteriolar dilatation in the presence of increased myocardial oxygen demand. This was demonstrated in studies from the 1980s [28, 29], and updated evidence examining the association between existing epicardial IHD and CMD is lacking. In 2007, the WISE study [9] examined the association between TCRs and CMD in a large cohort of women undergoing invasive microvascular evaluation. This study found that increasing age was the only predictor of reduced CFR. None of the TCRs were found to be a predictor of reduced CFR in a multivariate analysis and TCRs accounted for <20% of the variability in adenosine-related coronary reactivity. The iPOWER study [30], which examined the association between CMD and TCRs in women, found that TCRs accounted for only a little of the variation in coronary flow velocity reserve, a parameter used to diagnose CMD. A recent study by Kwan et al. [10] that examined the association between TCRs and reduced myocardial perfusion reserve index (MPRI), a noninvasive index of CMD, demonstrated that while diabetes and hyperlipidemia were associated with reduced MPRI in men, no associations between TCRs and MPRI were found in women.

Two endotypes of CMD have been identified: functional and structural. Functional CMD is characterized by reduced CFR and normal IMR and is partly caused by increased coronary flow in resting conditions. In structural CMD, which is characterized by reduced CFR and increased IMR, changes in the microvasculature anatomy cause impaired microcirculation during maximal vasodilation. A recent study comparing 5-year outcomes of patients with functional versus structural CMD demonstrated similar prevalence of TCRs, other than hypertension (which was more prevalent in patients with structural CMD), for both endotypes [16]. In our cohort, neither hypertension nor any other TCRs or clinical characteristics differed between patients with and without CMD, regardless of whether CMD was functional or structural.

Although the association between TCRs and CMD has generated much interest, the existing data presented above are conflicting and partly based on older, mechanistic studies or studies in healthy subjects. To the best of our knowledge, our study is the first to examine the association between sex and TCRs in a cohort of “real-world” patients undergoing invasive assessment of CMD. The difference in findings in our study compared to the study by Kwan et al. [10] could be partly explained by the older age of our cohort (68 vs. 57 years) and the fact that CMD in our study was assessed invasively. Our study also differs from the WISE study [9] as the WISE study included only women, and the mean age of the participants was younger than our cohort (54.8 vs. 68 years). In addition, 20% of the participants in the WISE study had obstructive CAD, while these patients were excluded from our study.

Similar to other studies, we have found that CFR in women with CMD is lower compared to men [31]. The lower CFR in women is attributed to increased myocardial blood flow at rest in women compared to men and is hypothesized to be due to lower vascular tone at rest. This has been demonstrated in angiographic studies and noninvasive positron emission tomography studies. In addition, women have smaller size coronary vessels compared to men, which can also contribute to higher resting coronary flow [32]. Unlike sex-related differences in CFR, we found no sex-related difference in IMR in patients with CMD. This finding has already been demonstrated in previous studies [4, 32]. Importantly, IMR has been shown to be better reproducible than CFR when measured by bolus thermodilution and is not affected by resting blood flow or epicardial stenosis [33].

This finding raises the question of whether in women, IMR rather than CFR may better represent true microvascular dysfunction and should have a more substantial role in the diagnosis of CMD. Identifying risk factors for CMD is vital for early detection, prevention, and treatment of this disorder, especially since recent studies have suggested that CMD confers an increased risk for mortality and heart failure [34]. It is very likely that there are other, currently unknown risk factors for CMD as well as unique risk factors for each of its endotypes, beyond TCRs. It is also unclear which parameter of microvascular dysfunction is the most accurate, and whether different parameters should be used in women and men. To further elucidate the epidemiology and pathophysiology of CMD, there is still a need for generational, prospective large-scale multicenter studies such as the “Farmingham Heart” study and the “Nurses Health” study, that will follow up with healthy subjects for several decades.

Our study has several limitations: first, while our data were obtained from a prospectively collected database, we conducted a retrospective and non-pre-specified analysis, and therefore, our results are subject to the effects of possible confounders and may be biased by the nature of this design. Second, it is a relatively small, single-center study, and generalization of our results is limited. Third, the lack of consensus regarding the threshold of pathologic CFR possesses an obstacle for assessing CMD. To account for this, we have examined associations between TCRs and CMD using both the threshold of <2.5 and <2.0, and the results were similar. Last, we did not assess the relation between alcohol consumption, a possible risk factor for cardiovascular disease, and CMD.

The prevalence of TCRs was similar in patients with and without CMD. In women, older age and previous IHD were associated with lower CFR and RRR values while no association between TCRs and indices of CMD was found in men.

The study was approved by the Tel-Aviv Sourasky Medical Center Review Board (Approval number: 0680-19-TLV). All patients signed an informed consent form prior to the procedure according to the study protocol.

The authors have no conflict of interest to declare.

The authors received no financial support for the research, authorship, and/or publication of this article.

Lior Zornitzki and Aviel Shetrit: analysis and interpretation of data and drafting of the manuscript. Ophir Freund, Shir Frydman, Ariel Banai, Reut Amar Shamir, Jeremy Ben-Shoshan, and Yaron Arbel: analysis and interpretation of data. Shmuel Banai: conceptualization and supervision. Maayan Konigstein: writing, conceptualization, and supervision.

The data that support the findings of this study are not publicly available due to their containing of information that could compromise the privacy of research participants but are available from the corresponding author (L.Z.) upon reasonable request, pending IRB approval. Further inquiries can be directed to the corresponding author.