Association between the plasma ceramide and coronary microvascular resistance Authors: Lanxin Feng, Xin Zhao, Jianqiao Song, Shuwen Yang, Jianping Xiang, Min Zhang, Chenchen Tu, Xiantao Song Published: 2024 ABSTRACT AbstractBackgroundPlasma ceramide plays a potentially significant role in the pathogenesis of coronary microvascular dysfunction. However, the relationship between plasma ceramide and coronary microvascular resistance in patients remains unclear. This study aimed to evaluate the association between plasma ceramide levels, as well as their distinct ratios, and coronary microvascular resistance.MethodsThis single-center observational study retrospectively enrolled patients who underwent both ceramide measurement and coronary angiography during hospitalization. The microvascular resistance of the coronary arteries was assessed in all patients using the angiography-derived index of microcirculatory resistance (Angio-IMR). The cumulative coronary microvascular resistance was calculated by summing the microvascular resistance of the three main coronary arteries. Multiple linear and logistic regression analyses were employed to evaluate the relationship between plasma ceramide and cumulative coronary microvascular resistance. Restricted cubic spline (RCS) analysis was conducted to investigate the association between plasma ceramide levels and cumulative coronary microvascular resistance. Receiver operating characteristic (ROC) curves were employed to evaluate the predictive value of plasma ceramide for coronary microvascular resistance. Additionally, subgroup analyses and interaction tests were performed.ResultsA total of 225 patients were included in this study, with a median cumulative coronary microvascular resistance of 48.04 (40.32–56.73). After adjusting for potential confounding factors, both plasma 16:0 ceramide and the 16:0/24:0 ceramide ratio were positively associated with cumulative coronary microvascular resistance [standardized β ± standard error: 75.05 ± 8.46 (P < 0.001) and 91.72 ± 20.41 (P < 0.001), respectively]. Similar independent associations were observed in predicting high cumulative microvascular resistance [β = 8.03 ± 1.91 (P < 0.001) and 9.98 ± 3.88 (P = 0.010), respectively]. Additionally, a significant nonlinear relationship was observed between plasma 16:0 ceramide, the 16:0/24:0 ceramide ratio, and cumulative coronary microvascular resistance (P for nonlinear < 0.05). The ROC analysis revealed that the optimal cut-off for plasma 16:0 ceramide is 0.178 µmol/L, with a specificity of 57.1% and a sensitivity of 91.2%. For the 16:0/24:0 ceramide ratio, the optimal cut-off is 0.072, yielding a specificity of 73.2% and a sensitivity of 54.9%. Subgroup analysis indicated that the association between plasma ceramide and coronary microvascular resistance was trending toward non-significance in patients with acute coronary syndrome (ACS).ConclusionsA significant nonlinear relationship exists between plasma ceramide and coronary microvascular resistance, which holds important clinical implications for the risk stratification of coronary microvascular disease. New insights into the potential effects of ceramides enhance our understanding of the complex mechanisms underlying coronary microvascular disease and warrant further investigation in a broader population.Graphical abstract Keywords: Coronary microvascular dysfunction, Ceramides, Angiography-derived index of microcirculatory resistance, Coronary artery disease. Introduction Coronary microvascular dysfunction, characterized by structural and functional abnormalities in the coronary microvasculature, is a critical mechanism underlying impaired myocardial perfusion and ischemia [1]. Coronary microvascular dysfunction is prevalent among patients with cardiovascular risk factors and is closely linked to the progression of cardiovascular disease in these individuals [2, 3]. In recent years, substantial clinical research has demonstrated that coronary microvascular dysfunction is associated with an increased risk of long-term adverse cardiovascular events across various cardiovascular diseases, posing a significant threat to patients’ health [4–6]. However, targeted prevention and treatment strategies for coronary microvascular dysfunction remain lacking in clinical practice. Therefore, early risk stratification of coronary microvascular dysfunction is essential for improving patient prognosis. Ceramide, a bioactive lipid and subclass of sphingolipids, is increasingly recognized as a biomarker for adverse cardiac events [7]. As a non-traditional risk factor for cardiovascular disease, ceramide has demonstrated superior predictive performance compared to conventional cardiovascular markers in multiple studies [8–10]. Large-scale clinical studies have shown that plasma ceramide is not only a significant predictor of cardiovascular mortality but is also closely associated with the residual risk in patients with cardiovascular disease [11–13]. Current research evidence indicates that elevated plasma ceramide levels impair coronary endothelial function through multiple pathways, including inflammation, oxidative stress, and apoptosis [14, 15]. Endothelial dysfunction and vascular inflammation are also regarded as significant contributors to coronary microvascular dysfunction [16–18]. Although ceramide plays a crucial role in the pathogenesis of coronary microvascular dysfunction, it remains unclear whether plasma ceramide can help estimate coronary microvascular resistance in patients. Therefore, this study aimed to investigate the association between plasma ceramide and coronary microvascular resistance. Methods Study population and design This study retrospectively enrolled patients who underwent plasma ceramide measurement during their hospitalization at the Coronary Heart Disease Center of Beijing Anzhen Hospital between October 2022 and June 2023. The exclusion criteria were as follows: (1) absence of coronary angiography during hospitalization; (2) presence of chronic total occlusive lesions on coronary angiography; (3) unqualified coronary angiography images; (4) incomplete baseline data. Fasting plasma samples were collected from all patients before coronary angiography and immediately frozen at -80 °C for subsequent ceramide measurement. This study complied with the Declaration of Helsinki and received approval from the Ethics Committee of Beijing Anzhen Hospital, affiliated with Capital Medical University. Plasma ceramide measurement All plasma samples were treated with EDTA anticoagulant before measurement, then frozen at -80 °C for short-term storage (less than two weeks). The LC–MS/MS system (AB Sciex Triple Quad™ 4500MD, Sciex, Framingham, MA, United States) was utilized to quantify the concentrations of four plasma ceramides (Cer16:0, Cer18:0, Cer24:0, and Cer24:1) as described previously [19]. Briefly, during sample preparation, a deuterated internal standard method was employed for protein precipitation. To ensure the accuracy of the quantitative analysis of the target ceramide, liquid chromatography-tandem mass spectrometry with multiple reaction monitoring in positive electrospray ionization mode was employed to screen precursor and product ions with specific mass/charge ratios. Plasma ceramides were measured in a blinded manner, without knowledge of the patients’ clinical characteristics. Angio-IMR computation In this study, the angiography-derived index of microcirculatory resistance (Angio-IMR) was measured for each patient in all three main coronary vessels (left anterior descending, left circumflex, and right coronary artery). Angio-IMR was measured from coronary angiographic images obtained prior to interventional procedures. As previously described, Angio-IMR was calculated by an independent laboratory in a blinded manner using dedicated software (AccuIMR, V1.0; ArteryFlow Technology, Hangzhou, China) [20]. The calculation process of Angio-IMR involves several key steps. First, coronary angiography images from different angles are used to automatically outline and three-dimensionally reconstruct the target vessel lumen contour. Next, the mean blood flow velocity is calculated using the TIMI frame count, and the fractional flow reserve (FFR) is determined. Finally, the Angio-IMR is calculated using the following formula: Angio-IMR = Pa × Angio-FFR × L/V. where Pa is the mean aortic pressure, L is the length of the target vessel, V is the average flow velocity, and Angio-FFR is the calculated FFR value from coronary angiography images. The specific calculation method for Angio-IMR is illustrated in Fig. 1. Data were assessed by three experienced operators, each of whom was blinded to the patients’ clinical characteristics. Schematic diagram of Angio-IMR calculation. A, B Angiograms from 2 projections with automatically delineated lumen contour. C Lumen diameter and computed FFR (Angio-FFR) pullback. D, E Computed Angio-IMR value. Angio-IMR, angiography-derived index of microcirculatory resistance; FFR, fractional flow reserve Statistical analysis All statistical analyses were conducted using R version 4.3.3 (R Foundation for Statistical Computing, Vienna, Austria). Categorical variables were presented as frequencies (percentages) and compared between groups using either the chi-square test or Fisher’s exact test. Continuous variables were presented as mean ± standard deviation (SD) or median with interquartile range (IQR), depending on the distribution of the data. The t-test was used to compare normally distributed variables, while the Mann-Whitney U test was applied to non-normally distributed variables. The cumulative microvascular resistance of the coronary arteries (the total Angio-IMR across the three main coronary vessels) was assessed for all patients in this study. Based on the median cumulative coronary microvascular resistance in the cohort, patients were categorized into high and low microvascular resistance groups. This method of calculating cumulative microvascular resistance using all main coronary arteries has been extensively utilized in prior studies [21]. Coronary artery disease (CAD) was defined based on a stenosis ≥ 50% of vessel diameter by visual assessment on the coronary angiogram. Multivariate analysis was performed using multiple linear regression for continuous dependent variables and logistic regression for categorical dependent variables. The nonlinear relationship between plasma ceramide levels and coronary microcirculatory resistance was assessed using multivariable restricted cubic splines (RCS). Three models were established to adjust for confounding factors. Receiver operating characteristic (ROC) curves were utilized to determine the optimal cut-off values of plasma ceramide for predicting high coronary microvascular resistance, along with their corresponding sensitivity and specificity values. Additionally, subgroup analyses and interaction tests were conducted based on age, gender, diabetes, hypertension, and the indications for coronary angiography. A two-sided p-value < 0.05 was considered statistically significant. Results A total of 273 patients who underwent plasma ceramide measurement were enrolled in this study, with 225 completing the coronary microvascular resistance assessment and included in the final analysis (Fig. 2). Patients were divided into the high microvascular resistance group (Angio-IMR ≥ 48.04) and the low microvascular resistance group (Angio-IMR < 48.04) based on the median cumulative coronary microvascular resistance in the study cohort. The baseline clinical characteristics of the two groups are shown in Table 1. Patients in the low microvascular resistance group had a higher incidence of previous myocardial infarction (28.6% vs. 15.9%, P = 0.023) and stroke (10.7% vs. 1.8%, P = 0.005). Additionally, a significant difference was observed in the number of CAD vessels between the two groups (P < 0.001). However, the clinical characteristics of the two groups were similar in terms of age, sex ratio, estimated glomerular filtration rate (eGFR), indication for coronary angiography, triglycerides, total cholesterol and highly-sensitive C-reactive protein (P > 0.05). Flowchart of patient enrollment in the study. Angio-IMR, angiography-derived index of microcirculatory resistance Table 1Baseline clinical characteristicsCharacteristicOverall(n = 225)Low microvascular resistance (n = 112)High microvascular resistance (n = 113)p-valueCumulative Angio-IMR48.04 (40.32–56.73)40.19 (34.04–45.42)56.73 (53.00–66.52) < 0.001 Age, years61.00 (55.00–68.00)62.00 (54.75–68.00)60.00 (55.00–67.00)0.348Male, n (%)171(76.0)85 (75.9)86 (76.1)0.970Hypertension, n (%)131(58.2)65 (58.0)66 (58.4)0.955Diabetes mellitus, n (%)70 (31.1)36 (32.1)34 (30.1)0.739Smoking, n (%)74 (32.9)34 (30.4)40 (35.4)0.421Drinking, n (%)56 (24.9)26 (23.2)30 (26.5)0.563Previous MI, n (%)50 (22.2)32 (28.6)18 (15.9) 0.023 Previous stroke, n (%)14 (6.2)12 (10.7)2 (1.8) 0.005 Atrial fibrillation, n (%)4 (1.8)2 (1.8)2 (1.8)1.000Indication for CAG, n (%)0.921ACSSTEMI1 (0.4)1 (0.9)0 (0)NSTEMI12 (5.3)5 (4.5)7 (6.2)UA23 (10.2)12 (10.7)11 (9.7)CCS189 (84.0)94 (83.9)95 (84.1)No. of vessels with CAD, n (%) < 0.001 055 (24.4)13 (11.6)42 (37.2)190 (40.0)47 (42.0)43 (38.1)239 (17.3)25 (22.3)14 (12.4)341 (18.2)27 (24.1)14 (12.4)HbA1c, %6.10 (5.70–6.90)6.10 (5.70–7.00)6.10 (5.70–6.80)0.729Glucose, mmol/l5.39 (4.81–6.18)5.38 (4.76–6.20)5.39 (4.97–6.18)0.789eGFR, mL/min/1.73m288.95 (78.02–96.47)88.42 (76.64–96.26)89.51 (81.30–97.11)0.324Total cholesterol, mmol/l3.62 (3.11–4.28)3.46 (2.96–4.28)3.77 (3.26–4.35)0.067LDL-cholesterol, mmol/l1.87 (1.46–2.45)1.76 (1.37–2.39)1.97 (1.52–2.51)0.098HDL-cholesterol, mmol/l1.03 (0.88–1.22)1.03 (0.86–1.21)1.02 (0.91–1.23)0.279Triglycerides, mmol/l1.37 (1.09–1.74)1.40 (1.04–1.71)1.36 (1.11–1.82)0.725hsCRP, mg/l0.90 (0.51–2.62)1.03 (0.47–2.88)0.82 (0.52–2.11)0.540hs-cTnI, ng/l6.30 (3.10– 8.90)4.65 (2.50–8.83)6.80 (5.30–9.10) 0.002 16:0 ceramide, umol/l0.22 (0.16–0.28)0.17 (0.14–0.24)0.26 (0.21–0.32) < 0.001 18:0 ceramide, umol/l0.06 (0.05–0.08)0.06 (0.04–0.08)0.06 (0.05–0.08)0.12824:0 ceramide, umol/l2.99 (2.39–3.95)2.73 (2.17–3.72)3.35 (2.62–4.24) 0.005 24:1 ceramide, umol/l0.72 (0.58–0.89)0.69 (0.56–0.87)0.75 (0.61–0.96)0.10716:0/24:0 ceramide ratio0.07 (0.05–0.09)0.06 (0.05–0.07)0.07 (0.06–0.11) < 0.001 18:0/24:0 ceramide ratio0.02 (0.01–0.03)0.02 (0.01–0.03)0.02 (0.01–0.03)0.42524:1/24:0 ceramide ratio0.24 (0.18–0.32)0.25 (0.19–0.32)0.24 (0.17–0.30)0.296For categorical variables, absolute numbers (n) and relative proportions (%) are presented. For continuous variables, the median along with the interquartile range is presentedAngio-IMR, Angiography-derived index of microcirculatory resistance; MI, myocardial infarction; ACS, acute coronary syndrome; STEMI, ST-elevation myocardial infarction; NSTEMI, non-ST-elevation myocardial infarction; UA, unstable angina; eGFR, estimated glomerular filtration rate; CAD, coronary artery disease; CAG, coronary angiography; CCS, chronic coronary syndrome; HDL, high-density lipoprotein; LDL, low-density lipoprotein; hsCRP, highly-sensitive C-reactive protein; hs-cTnI, high-sensitivity cardiac troponin I Baseline clinical characteristics For categorical variables, absolute numbers (n) and relative proportions (%) are presented. For continuous variables, the median along with the interquartile range is presented Angio-IMR, Angiography-derived index of microcirculatory resistance; MI, myocardial infarction; ACS, acute coronary syndrome; STEMI, ST-elevation myocardial infarction; NSTEMI, non-ST-elevation myocardial infarction; UA, unstable angina; eGFR, estimated glomerular filtration rate; CAD, coronary artery disease; CAG, coronary angiography; CCS, chronic coronary syndrome; HDL, high-density lipoprotein; LDL, low-density lipoprotein; hsCRP, highly-sensitive C-reactive protein; hs-cTnI, high-sensitivity cardiac troponin I As shown in Table 1, plasma concentrations of ceramides 16:0 and 24:0 were significantly higher in patients with high microvascular resistance compared to those in the low microvascular resistance group. However, no significant differences were observed in the plasma concentrations of 18:0 and 24:1 ceramides between the two groups (P > 0.05). Different multivariate models were employed to investigate the association between plasma ceramide and coronary microvascular resistance. Patient clinical characteristics, including various traditional cardiovascular risk factors and biological markers, were incrementally included in the regression model as potential confounding variables. As shown in Table 2, in the fully adjusted Model 3, the plasma concentration of 16:0 ceramide was positively associated with coronary microvascular resistance (multivariate linear model 3; standardized β = 75.05 ± 8.46, P < 0.001). In contrast, the plasma concentrations of 18:0, 24:0 and 24:1 ceramides were not significantly associated with coronary microvascular resistance (P > 0.05). Table 2Multivariate analysis for angiography-derived IMR (Angio-IMR)Model 1Model 2Model 3β ± SEP-valueβ ± SEP-valueβ ± SEP-valueMultivariate linear regression for Angio-IMR16:0 ceramide79.22 ± 8.63 < 0.001 73.43 ± 8.43 < 0.001 75.05 ± 8.46 < 0.001 18:0 ceramide26.56 ± 31.430.39942.53 ± 30.720.16851.39 ± 37.100.16724:0 ceramide2.34 ± 0.80 0.004 1.89 ± 0.78 0.016 1.70 ± 1.040.10624:1 ceramide5.07 ± 3.240.1185.27 ± 3.090.0905.92 ± 3.600.10116:0/24:0 ceramide ratio71.62 ± 21.07 < 0.001 68.33 ± 20.12 < 0.001 91.72 ± 20.41 < 0.001 18:0/24:0 ceramide ratio-233.79 ± 93.85 0.018 -139.86 ± 93.030.134-83.35 ± 101.220.60524:1/24:0 ceramide ratio-14.78 ± 8.370.079-9.82 ± 8.090.226-4.09 ± 8.390.627Multivariate logistic regression for high microvascular resistance (Angio-IMR ≥ 48.04)16:0 ceramide8.51 ± 1.75 < 0.001 7.90 ± 1.82 < 0.001 8.03 ± 1.91 < 0.001 18:0 ceramide3.94 ± 4.210.3497.74 ± 4.680.0987.51 ± 6.110.21924:0 ceramide0.23 ± 0.11 0.037 0.19 ± 0.120.1080.10 ± 0.160.53924:1 ceramide0.70 ± 0.450.1160.85 ± 0.490.0800.82 ± 0.580.16116:0/24:0 ceramide ratio6.85 ± 3.25 0.035 6.66 ± 3.430.0529.98 ± 3.88 0.010 18:0/24:0 ceramide ratio-17.06 ± 12.790.182-2.95 ± 14.110.8342.09 ± 16.320.89824:1/24:0 ceramide ratio-1.20 ± 1.140.292-0.51 ± 1.210.6720.08 ± 1.300.951Values are presented as standardized β ± standard error for linear regression and raw β ± standard error for logistic regressionModel 1: adjustment for age, male, hypertension and diabetes mellitusModel 2: model 1 + HbA1c, previous MI, previous stroke, atrial fibrillation, and No. of vessels with CADModel 3: model 2 + eGFR, LDL-cholesterol, HDL-cholesterol, triglycerides, hsCRP and hs-cTnIAngio-IMR, Angiography-derived index of microcirculatory resistance; MI, myocardial infarction; eGFR, estimated glomerular filtration rate; CAD, coronary artery disease; HDL, high-density lipoprotein; LDL, low-density lipoprotein; hsCRP, highly-sensitive C-reactive protein; hs-cTnI, high-sensitivity cardiac troponin I Multivariate analysis for angiography-derived IMR (Angio-IMR) Values are presented as standardized β ± standard error for linear regression and raw β ± standard error for logistic regression Model 1: adjustment for age, male, hypertension and diabetes mellitus Model 2: model 1 + HbA1c, previous MI, previous stroke, atrial fibrillation, and No. of vessels with CAD Model 3: model 2 + eGFR, LDL-cholesterol, HDL-cholesterol, triglycerides, hsCRP and hs-cTnI Angio-IMR, Angiography-derived index of microcirculatory resistance; MI, myocardial infarction; eGFR, estimated glomerular filtration rate; CAD, coronary artery disease; HDL, high-density lipoprotein; LDL, low-density lipoprotein; hsCRP, highly-sensitive C-reactive protein; hs-cTnI, high-sensitivity cardiac troponin I We further examined the relationship between the ceramide ratios (16:0/24:0, 18:0/24:0 and 24:1/24:0) and coronary microvascular resistance. The 16:0/24:0 ceramide ratio was significantly higher in the high microvascular resistance group compared to the low microvascular resistance group (Table 1). As shown in Table 2, across all multivariate linear models, the 16:0/24:0 ceramide ratio was positively correlated with coronary microvascular resistance. In the fully adjusted model 3, no significant associations were found between the 18:0/24:0 and 24:1/24:0 ceramide ratios and coronary microvascular resistance (P > 0.05). As shown in Table 2, multivariate logistic regression analysis revealed that both plasma 16:0 ceramide and the 16:0/24:0 ceramide ratio were predictive of high coronary microvascular resistance. We further assessed the potential nonlinearity in the relationship between plasma 16:0 ceramide, the 16:0/24:0 ceramide ratio, and coronary microvascular resistance using RCS curves, as illustrated in Fig. 3. Our findings indicated a significant nonlinear relationship between plasma 16:0 ceramide, the 16:0/24:0 ceramide ratio, and coronary microvascular resistance (P for nonlinear < 0.05). Nonlinear relationship between coronary microvascular resistance and 16:0 ceramide (A) and 16:0/24:0 ceramide ratio (B). Adjustment for age, male, hypertension, diabetes mellitus, HbA1c, previous MI, previous stroke, atrial fibrillation, No. of vessels with CAD, eGFR, LDL-cholesterol, HDL-cholesterol, triglycerides, hsCRP and hs-cTnI. The solid lines represent β, and blue areas indicate the 95% CIs. MI, myocardial infarction; eGFR, estimated glomerular filtration rate; CAD, coronary artery disease; HDL, high-density lipoprotein; LDL, low-density lipoprotein; hsCRP, highly-sensitive C-reactive protein; hs-cTnI, high-sensitivity cardiac troponin I; CI, confidence interval 792 ROC curves were employed to evaluate the predictive value of plasma ceramides in distinguishing patients with high coronary microvascular resistance (Fig. 4). The analysis indicated that the optimal cut-off for plasma 16:0 ceramide is 0.178 µmol/L, with a specificity of 57.1% and a sensitivity of 91.2%. For the 16:0/24:0 ceramide ratio, the optimal cut-off is 0.072, yielding a specificity of 73.2% and a sensitivity of 54.9% (Table 3). The ROC curves of plasma ceramide for high microvascular resistance (Angio-IMR ≥ 48.04). Angio-IMR, angiography-derived index of microcirculatory resistance; AUC, area under curve; ROC, receiver operating characteristic Table 3The diagnosis values of plasma ceramide for high microvascular resistance (Angio-IMR ≥ 48.04)16:0 ceramide, umol/l16:0/24:0 ceramide ratioCut off value0.1780.072Sensitivity0.9120.549Specificity0.5710.732AUC0.758 (0.693–0.823)0.633 (0.560–0.706)Youden index0.4830.281P value< 0.001< 0.001Angio-IMR, Angiography-derived index of microcirculatory resistance; AUC, area under curve The diagnosis values of plasma ceramide for high microvascular resistance (Angio-IMR ≥ 48.04) Angio-IMR, Angiography-derived index of microcirculatory resistance; AUC, area under curve Figure 5 presents the results of the subgroup analyses and interaction tests between plasma ceramides and coronary microvascular resistance. Notably, significant interaction effects were observed in the indication for coronary angiography regarding both plasma 16:0 ceramide and the 16:0/24:0 ceramide ratio. In comparison to patients with chronic coronary syndrome (CCS), no significant correlation between plasma ceramide and coronary microvascular resistance was observed in patients with acute coronary syndrome (ACS). In the subsequent sensitivity analysis, we further excluded patients with ACS. As shown in Table 4, after fully adjusting the model, both plasma 16:0 ceramide and the 16:0/24:0 ceramide ratio maintained a significant correlation with coronary microvascular resistance (p < 0.05). Subgroup analysis of the association between coronary microvascular resistance and 16:0 ceramide (A) and 16:0/24:0 ceramide ratio (B). Adjustment for age, male, hypertension, diabetes mellitus, HbA1c, previous MI, previous stroke, atrial fibrillation, No. of vessels with CAD, eGFR, LDL-cholesterol, HDL-cholesterol, triglycerides, hsCRP and hs-cTnI. Variables used as stratification criteria were not adjusted. MI, myocardial infarction; eGFR, estimated glomerular filtration rate; CAD, coronary artery disease; HDL, high-density lipoprotein; LDL, low-density lipoprotein; hsCRP, highly-sensitive C-reactive protein; hs-cTnI, high-sensitivity cardiac troponin I; CI, confidence interval Table 4Multivariate analysis for angiography-derived IMR (Angio-IMR) in patients with chronic coronary syndromeModel 1Model 2Model 3β ± SEP-valueβ ± SEP-valueβ ± SEP-valueMultivariate linear regression for Angio-IMR16:0 ceramide92.38 ± 8.67 < 0.001 86.41 ± 8.62 < 0.001 86.75 ± 8.81 < 0.001 18:0 ceramide11.11 ± 35.050.75233.61 ± 34.570.33247.05 ± 39.770.23824:0 ceramide2.21 ± 0.85 0.010 1.93 ± 0.82 0.021 1.80 ± 1.130.11224:1 ceramide2.98 ± 3.470.3923.65 ± 3.330.2753.40 ± 3.850.37716:0/24:0 ceramide ratio108.47 ± 22.98 < 0.001 98.88 ± 22.22 < 0.001 119.00 ± 22.25 < 0.001 18:0/24:0 ceramide ratio-289.44 ± 101.96 0.005 -197.84 ± 102.600.055-101.49 ± 110.150.35824:1/24:0 ceramide ratio-20.45 ± 8.90 0.023 -15.45 ± 8.680.077-9.33 ± 8.910.296Multivariate logistic regression for high microvascular resistance (Angio-IMR ≥ 48.04)16:0 ceramide9.84 ± 2.00 < 0.001 8.86 ± 2.09 < 0.001 9.56 ± 2.35 < 0.001 18:0 ceramide0.96 ± 4.620.8345.39 ± 5.160.2968.26 ± 6.700.21724:0 ceramide0.20 ± 0.120.0940.17 ± 0.130.1850.15 ± 0.180.40724:1 ceramide0.37 ± 0.460.4230.54 ± 0.500.2800.68 ± 0.630.28016:0/24:0 ceramide ratio11.00 ± 4.15 0.008 9.56 ± 4.41 0.030 11.99 ± 5.04 0.017 18:0/24:0 ceramide ratio-25.22 ± 14.260.077-9.12 ± 15.810.5642.16 ± 18.700.90824:1/24:0 ceramide ratio-1.94 ± 1.270.127-1.23 ± 1.340.358-0.54 ± 1.530.726Values are presented as standardized β ± standard error for linear regression and raw β ± standard error for logistic regressionModel 1: adjustment for age, male, hypertension and diabetes mellitusModel 2: model 1 + HbA1c, previous MI, previous stroke, atrial fibrillation, and No. of vessels with CADModel 3: model 2 + eGFR, LDL-cholesterol, HDL-cholesterol, triglycerides, hsCRP and hs-cTnIAngio-IMR, Angiography-derived index of microcirculatory resistance; MI, myocardial infarction; eGFR, estimated glomerular filtration rate; CAD, coronary artery disease; HDL, high-density lipoprotein; LDL, low-density lipoprotein; hsCRP, highly-sensitive C-reactive protein; hs-cTnI, high-sensitivity cardiac troponin I Multivariate analysis for angiography-derived IMR (Angio-IMR) in patients with chronic coronary syndrome Values are presented as standardized β ± standard error for linear regression and raw β ± standard error for logistic regression Model 1: adjustment for age, male, hypertension and diabetes mellitus Model 2: model 1 + HbA1c, previous MI, previous stroke, atrial fibrillation, and No. of vessels with CAD Model 3: model 2 + eGFR, LDL-cholesterol, HDL-cholesterol, triglycerides, hsCRP and hs-cTnI Angio-IMR, Angiography-derived index of microcirculatory resistance; MI, myocardial infarction; eGFR, estimated glomerular filtration rate; CAD, coronary artery disease; HDL, high-density lipoprotein; LDL, low-density lipoprotein; hsCRP, highly-sensitive C-reactive protein; hs-cTnI, high-sensitivity cardiac troponin I Discussion The findings of this study demonstrate that both plasma 16:0 ceramide and the 16:0/24:0 ceramide ratio are independent biomarkers significantly associated with coronary microvascular resistance. These markers effectively predict the risk of elevated coronary microvascular resistance. This is the first study to investigate the relationship between plasma ceramide and coronary microvascular resistance, which is significant for the in-depth exploration of the complex mechanisms underlying their relationship. Low-density lipoprotein cholesterol and triglycerides, as the most prominent circulating lipids, are widely recognized biomarkers for adverse cardiac events [22]. Recent advances in lipidomic technologies have led to several large-scale clinical studies identifying ceramide as an emerging non-traditional risk factor for cardiovascular disease [9]. Extensive clinical evidence indicates that ceramide is closely linked to the onset and progression of cardiovascular diseases, making it a crucial marker for identifying high-risk patients [23, 24]. Ceramide is a highly heterogeneous bioactive sphingolipid, with its biological effects varying based on the length of the acyl chains [25]. Different long-chain and very long-chain sphingolipids contribute to the pathogenesis of cardiovascular diseases through distinct pathophysiological roles, involving mechanisms such as oxidation and inflammation [26, 27]. Specific acyl species may be more linked to pathogenic pathways related to cardiovascular diseases. Plasma 16:0 ceramide has been identified in multiple studies as the most potent ceramide marker for predicting acute coronary syndrome and adverse cardiovascular prognosis [13, 28, 29]. In patients with type 2 diabetes, plasma 16:0 ceramide and the 16:0/24:0 ceramide ratio have been also identified as the biomarkers associated with cardiovascular disease risk, independent of traditional cardiovascular risk factors, and more predictive than other ceramide indicators [30]. Although the mechanisms underlying the association between ceramide and cardiovascular disease are becoming increasingly clear, it remains uncertain whether ceramide is independently associated with coronary microvascular disease. The ATHEROREMO-IVUS study, which included 581 patients with coronary artery disease, demonstrated that the 16:0/24:0 ceramide ratio was associated with coronary plaque vulnerability characteristics as determined by intravascular ultrasound [31]. Considering that plaque vulnerability is closely related to endothelial dysfunction, this study suggests that plasma ceramide may be an important contributor to coronary endothelial dysfunction. Akhiyat et al. investigated the relationship between plasma ceramides and coronary endothelial dysfunction in patients with nonobstructive epicardial coronary artery disease [15]. Their study revealed that plasma ceramides 16:0 and 24:0 were independently associated with coronary endothelial dysfunction. Although increasing evidence indicates that ceramide may be a significant independent driver of coronary endothelial dysfunction, further direct clinical studies are required to elucidate the relationship between plasma ceramide and coronary microvascular resistance. Direct assessment of coronary microvascular resistance is clinically challenging due to the small size of microvessels, which impedes the advancement of related clinical research. IMR is an invasive technique that independently measures coronary microvascular resistance, regardless of epicardial vascular stenosis, and provides an accurate and practical method for clinical assessment [32]. However, its clinical application is relatively limited due to high technical demands and complex operation. The new generation of Angio-IMR is a method for assessing coronary microvascular resistance using coronary angiography images without the pressure wire and demonstrates high consistency with IMR [20]. In this study, we introduced Angio-IMR as the method for measuring coronary microvascular resistance. Since Angio-IMR doesn’t require additional technical procedures during coronary angiography, it allows for the offline measurement of microvascular resistance across all coronary arteries without introducing extra risks to the patient. Inspired by previous clinical research designs, this study calculated the cumulative microvascular resistance of the main coronary arteries to achieve a more comprehensive physiological assessment of patients’ coronary microvascular resistance [33, 34]. Ceramide exerts multiple effects on coronary microvascular function. Prolonged exposure of arterioles to ceramide has been shown to promote the conversion of endothelial-derived nitrous oxide to hydrogen peroxide, thereby preserving flow-induced dilation (FID) in coronary microvessels [35–37]. Although this shift temporarily benefits the regulation of local myocardial blood perfusion, it compromises the long-term integrity of coronary microvascular function. Moreover, ceramide contributes to the activation of inflammasomes in endothelial cells [38]. Targeted reduction of ceramide can decrease vascular inflammation and potentially improve coronary microvascular endothelial function [39]. However, evidence suggests that ceramide may also play a beneficial role in coronary microvascular function and be necessary for maintaining normal microvascular endothelial function in humans [40]. Specific ceramide metabolites have been shown to stimulate nitrous oxide production in endothelial cells via the sphingosine-1-phosphate receptor 1 (S1PR1) signaling pathway [41]. Our study also found a significant nonlinear relationship between ceramide levels and coronary microvascular resistance, suggesting that lower ceramide levels may have a protective effect on coronary microvascular function. Although current clinical evidence highlights ceramide as a key driver of cardiovascular disease, treatment strategies aimed at its targeted reduction require further investigation. This new perspective on the potential dual role of ceramide not only enhances our understanding of coronary microvascular dysfunction but also opens new avenues for therapeutic interventions aimed at modulating ceramide levels. Further research is needed to explore the impact of lowering plasma ceramide levels on coronary microvascular function in patients. Furthermore, while the association between plasma ceramide and coronary microvascular resistance was consistent across most subgroups, a trend appeared in the analyses with weak association. In patients with ACS, the relationship between coronary microvascular resistance and plasma ceramides was not significant. Multiple clinical studies have demonstrated that plasma ceramide levels in patients with ACS are significantly higher than those in patients with CCS [42, 43]. Abnormal regulation of ceramide may be associated with plaque instability, potentially triggering ACS. This could be a significant factor contributing to the heterogeneity observed in the relationship between ceramides and coronary microvascular resistance in ACS patients. Nevertheless, given the relatively small number of ACS patients in this study, the associated research conclusions should be interpreted with caution, and further investigation is warranted. Several limitations of this study should be discussed. First, as a retrospective study, this research could not establish a causal relationship between ceramide and coronary microvascular resistance, nor did it explore the underlying pathophysiological mechanisms behind this association. Prospective studies are necessary to confirm these findings and further elucidate the underlying mechanisms. Second, the sample size of this study was relatively small, which limited its statistical power. Consequently, we could evaluate only a limited number of subgroups in the sensitivity analyses, as some subgroups were too underrepresented for meaningful analysis. Additionally, as this study was conducted at a single center, the credibility of its conclusions may be limited. Large-scale multicenter studies are needed in the future to further validate the findings of this study. Conclusions In conclusion, this study demonstrates a significant association between plasma ceramide and coronary microvascular resistance. Plasma 16:0 ceramide and the 16:0/24:0 ceramide ratio are emerging biomarkers of coronary microvascular resistance and may hold significant clinical value for the early detection and prevention of coronary microvascular disease. Further studies are essential to determine whether potential heterogeneity exists in this association among different populations. Acknowledgements We are indebted to all members who contributed to this work. Glossary Abbreviations Acute coronary syndrome Angiography-derived index of microcirculatory resistance Chronic coronary syndrome Fractional flow reserve Standard deviation Interquartile range Restricted cubic splines Estimated glomerular filtration rate Flow-induced dilation Sphingosine-1-phosphate receptor 1 Author contributions FLX designed the research, analyzed the data and wrote the manuscript. ZX, SJQ, YSW, XJP, ZM collected the data and performed the statistical analysis. TCC and SXT critically revised the manuscript for important intellectual content. All authors read and approved the final manuscript. Funding This study was supported by the Capital’s Funds for Health Improvement and Research (2024-2-2066), the Coordinated innovation of scientific and technological in Beijing-Tianjin-Hebei Region (Z231100003923008), Beijing Nova Program (20220484222), and the Beijing Hospitals Authority’s Ascent Plan (DFL20220603), High-level public health technical talent construction project of Beijing Municipal Health Commission (Leading Talent-02-01). Data availability No datasets were generated or analysed during the current study. Declarations Competing interests The authors declare no competing interests. Ethics approval and consent to participate This study complied with the Declaration of Helsinki and received approval from the Ethics Committee of Beijing Anzhen Hospital, affiliated with Capital Medical University. Consent for publication All authors have reviewed and approved the final version of the manuscript. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Contributor Information Chenchen Tu, Email: tcc2033@163.com. Xiantao Song, Email: song0929@mail.ccmu.edu.cn. References Associated Data Data Availability Statement No datasets were generated or analysed during the current study.