Effect of imaging parameters and clinical characteristics on pericoronary adipose tissue attenuation in patients without coronary plaque

Introduction

Vascular inflammation plays a pivotal role in the formation and rupture of atherosclerotic plaque (1). Consequently, its early detection could facilitate targeted therapies for patients with coronary artery disease (CAD), thus preventing future myocardial infarction (2). The vascular inflammation leads to the release of inflammatory mediators into the pericoronary adipose tissue (PCAT), inhibiting adipogenesis and resulting in smaller adipocytes with reduced intracellular lipid content. This pathophysiological alteration can be quantitatively detected on coronary computed tomography angiography (CCTA) as an increase in PCAT attenuation, measured in Hounsfield units (HUs) (3).

A growing body of evidence has shown that PCAT attenuation enables cardiovascular risk stratification (4-6), the diagnosis and prognosis of ischemic heart disease (5,7,8), and the assessment of cardiac mortality (5,9). However, some studies have failed to confirm these findings (10,11). Ma et al. (10) reported no difference in PCAT attenuation between patients with and without CAD, and Boussoussou et al. (11) found no association between PCAT attenuation and CAD after adjusting for imaging parameters and patient characteristics. These conflicting findings may be attributed to inconsistencies in imaging parameters and clinical characteristics across studies.

Previous studies have demonstrated that imaging parameters and patient characteristics affect the fat attenuation index (FAI) of the PCAT (5,11-15). However, these studies have critical limitations. First, the study conducted by Boussoussou et al. (11) was based on interpatient comparisons among patients with CAD. This approach confounded the effects of imaging and patient-related factors with the variable degrees of coronary inflammation, making it difficult to isolate the specific impact of imaging and patient-related factors on FAI. Second, although the study conducted by Ma et al. (12) enrolled patients without CAD, it only accounted for the influence of tube voltage and did not adequately control for other potential confounders—such as age, sex, and body mass index (BMI). Third, the ex vivo experiments by Etter et al. (14) and Pitteloud et al. (15) were performed on porcine hearts and thus lacked the motion artifacts and physiological variations inherent in clinical CCTA scans.

Therefore, to address these gaps, this study was designed to specifically isolate the effects of imaging parameters and clinical characteristics on FAI in a clinical setting. To achieve this, we included subjects with no evidence of CAD on CCTA, effectively eliminating coronary inflammation as a primary confounding variable. By systematically evaluating these factors across multiple coronary artery segments, this study aims to provide robust, real-world clinical evidence to guide the standardization of FAI measurements, a critical prerequisite for its reliable use as a biomarker of subclinical coronary inflammation. We present this article in accordance with the STROBE reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-2026-0484/rc).

Methods

Study population

The retrospective, single-center study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. This study was approved by the Institutional Review Board of Zhujiang Hospital, Southern Medical University (IRB No. 2025-KY-209-01). Informed consent was waived due to the retrospective nature of the study and the anonymized data collection. Consecutive patients were eligible for inclusion if they were suspected of having CAD and underwent CCTA at Zhujiang Hospital, Southern Medical University between July 2022 and December 2024. The inclusion criterion was the absence of CAD, defined as no coronary plaque (including calcified, non-calcified, and mixed plaques) on CCTA images. Exclusion criteria were as follows (Figure 1): (I) poor image quality of the CCTA scans; (II) anomalous coronary artery origin from the aortic sinus, which could lead to inaccurate measurements; and (III) the presence of other comorbidities, including oncologic, autoimmune, pulmonary, or hematological diseases. These conditions were excluded because they can induce systemic inflammation, which could confound FAI measurements (16). By excluding them, we aimed to isolate the effects of imaging parameters and clinical characteristics on FAI.

Figure 1 Flowchart of patient inclusion. CAD, coronary artery disease; CCTA, coronary computed tomography angiography.

CCTA acquisition

CCTA was performed using 256-slice CT scanners (Brilliance iCT Elite or Brilliance iCT Elite FHD; Philips Healthcare, Best, the Netherlands). Scanning was performed using retrospective electrocardiogram-gated acquisition in patients with a final heart rate of ≥70 bpm, whereas prospective electrocardiogram-gated sequential acquisition was used in patients with a final heart rate of <70 bpm. To visualize the coronary lumen, a 100-mL bolus of iodinated contrast agent (Ultravist, 370 mg I/mL; Bayer AG, Berlin, Germany) was injected intravenously at a rate of 5.0 mL/s, followed by a 40-mL saline flush. The CCTA protocol parameters were as follows: section collimation of 128×0.625 mm; gantry rotation time of 0.27 s for the Brilliance iCT scanner and 0.33 s for the Brilliance iCT Elite FHD scanner; and tube current and voltage adjusted according to patient size, with tube voltage settings of either 100 or 120 kVp. For the Brilliance iCT scanner, images were reconstructed with a 512×512 matrix, a 0.67-mm slice thickness, and a 0.34-mm increment. Reconstruction was performed using iDose4 (level 5) with Xres Standard, without enhancement. For the Brilliance iCT Elite FHD scanner, images were reconstructed using a 512×512 matrix, a 0.90-mm slice thickness, and a 0.45-mm increment. Reconstruction was performed using Iterative Model Reconstruction (IMR; level 1) with Cardiac Routine, without enhancement.

PCAT analysis

The CCTA dataset with optimal image quality (e.g., minimal motion artifacts and optimal vascular opacification) was selected by a radiologist with 8 years of experience in cardiovascular imaging and transferred to a dedicated workstation (CoronaryDoc; Shukun Technology, Beijing, China) for PCAT evaluation. PCAT was defined as the adipose tissue located within a radial distance from the outer vessel wall equal to the vessel diameter (3). Within this region, the FAI of PCAT was calculated as the mean CT attenuation value of all voxels with attenuation values between −190 and −30 HU (3).

For FAI measurement, we employed two methods (Figure 2): (I) a conventional approach measuring a proximal 40-mm segment of the right coronary artery (RCA), left anterior descending artery (LAD), and left circumflex artery (LCX) (5); if the available length was insufficient due to unilateral dominance, the longest clearly opacified segment of the vessel was measured; and (II) a method measuring a 10-mm segment in each of the three main coronary arteries according to the 18-segment segmentation method proposed by the Society of Cardiovascular Computed Tomography, to minimize interference from side-branch intersections (12). A PCAT volume of interest (VOI) with a length of 10 mm was placed in the proximal, middle, and distal segments of the RCA and LAD, and in the proximal and distal segments of the LCX. Each coronary artery segment exhibiting a clear boundary with the myocardium and avoiding bifurcations was selected. For the PCAT VOI on the RCA and LAD in both measurement methods, the starting point was set 10 mm distal to the vessel origin—to avoid the interference from the aortic wall (for RCA) and to prevent overlap with the LCX measurement (for LAD), respectively. For the PCAT VOI on the LCX, the vessel origin was selected as the start point. Two radiologists, with 8 and 3 years of experience in cardiovascular imaging respectively, who were blinded to clinical data, independently performed FAI measurements. The average of the two readers' measurements was used for data analysis.

Figure 2 Two methods for measuring FAI. (A-C) FAI values for the proximal 40-mm segment of the RCA, LAD, and LCX were −92, −81, and −79 HU, respectively. (D) 10-mm segments in the proximal, middle, and distal RCA and LAD, and in the proximal and distal LCX (red rectangles) are marked on the volume rendering image. (E) FAI value for the 10-mm segment in the proximal LAD was −80 HU. FAI, fat attenuation index; HU, Hounsfield unit; LAD, left anterior descending artery; LCX, left circumflex artery; RCA, right coronary artery.

Vessel attenuation measurement

Vessel attenuation was defined as the mean attenuation value measured in five circular regions of interest (ROIs) placed within the ascending aorta at the level of the RCA ostium; each ROI had an approximate cross-sectional area of 10 mm². Vessel attenuation was measured independently by two radiologists (with 8 and 3 years of experience in cardiovascular imaging), and the averaged values were used for analysis. A cutoff value of 500 HU was used to classify the groups.

Statistical analysis

Statistical analyses were performed using SPSS software (version 26.0; IBM Corp., Armonk, NY, USA). Continuous variables are expressed as means ± standard deviations (SDs) and compared using one-way repeated measures analysis of variance (ANOVA). Categorical variables are presented as frequencies (with percentages). The intraclass correlation coefficient (ICC) was used to assess interobserver agreement, with 95% confidence intervals (CIs). ICC <0.50 indicates poor agreement; 0.50–0.75, moderate agreement; 0.76–0.90, good agreement; and 0.91–1.00, excellent agreement. Linear regression analyses were performed using the Enter method for each FAI segment, adjusting for imaging parameters (tube voltage, reconstruction method, and vessel attenuation) and clinical characteristics (age, sex, BMI, hypertension, diabetes, family history of CAD, smoking, and dyslipidemia). Overweight was defined as a BMI ≥24 kg/m² according to the Chinese criteria (17). A two-sided P value <0.05 was considered statistically significant.

Results

Study population

Two hundred and forty-three subjects without CAD on CCTA images were initially selected. Forty-seven patients were excluded for the following reasons: poor image quality (n=15), anomalous origin of a coronary artery (n=10), malignancies (n=8), and autoimmune diseases (n=14). Finally, 196 subjects (mean age ± SD, 55.4±9.5 years; 81 males) were included (Figure 1). Subject characteristics are presented in Table 1.

Interobserver agreement in FAI

Interobserver agreement for FAI measurements was excellent, with an ICC of 0.985–0.998 (Table 2).

PCAT analysis

We noted significant regional differences in FAI on a vessel basis. The mean value of all FAI measurements was −82.7±10.6 HU, ranging from −117 to −50 HU. Across the three coronary arteries, FAI values of the 40-mm proximal segments differed significantly (F=40.630, P<0.001), with the RCA showing the lowest value (−84.6±8.7 HU), followed by the LAD (−80.8±8.4 HU) and LCX (−79.9±8.4 HU). Within-vessel segmental analyses revealed significant heterogeneity in FAI values of the RCA segments (F=11.720, P<0.001) and LAD segments (F=39.690, P<0.001), whereas the LCX segments did not differ significantly (F=3.129, P=0.068). All Bonferroni adjusted pairwise results are summarized graphically in Figure 3A-3D.

Figure 3 Bar charts of FAI measurements. Bars represent the mean FAI values. Error bars represent the standard deviation. (A) FAI values for the proximal 40-mm segment of the main coronary arteries. FAI_{_RCA_40mm}, FAI_{_LAD_40mm}, and FAI_{_LCX_40mm} were −84.6±8.7, −80.8±8.4, and −79.9±8.4 HU, respectively (P<0.001). (B) FAI values for the proximal 40-mm segment of the RCA and for its 10-mm segments in the proximal, middle, and distal portions. FAI_{_RCA_40mm}, FAI_{_RCA_prox_10mm}, FAI_{_RCA_mid_10mm}, and FAI_{_RCA_dist_10mm} were −84.6±8.7, −84.3±11.0, −88.2±11.1, and −84.2±12.0 HU, respectively (P<0.001). (C) FAI values for the proximal 40-mm segment of the LAD and for its 10-mm segments in the proximal, middle, and distal portions. FAI_{_LAD_40mm}, FAI_{_LAD_prox_10mm}, FAI_{_LAD_mid_10mm}, and FAI_{_LAD_dist_10mm} were −80.8±8.4, −79.1±9.8, −85.2±9.6, and −85.0±10.9 HU, respectively (P<0.001). (D) FAI values for the proximal 40-mm segment of the LCX and for its 10-mm segments in the proximal and distal portions. FAI_{_LCX_40mm}, FAI_{_LCX_prox_10mm}, and FAI_{_LCX_dist_10mm} were −79.9±8.4, −79.5±10.5 and −78.3±10.8 HU, respectively (P=0.068). *, P<0.05. FAI, fat attenuation index; HU, Hounsfield unit; LAD, left anterior descending artery; LCX, left circumflex artery; RCA, right coronary artery; SD, standard deviation.

Factors associated with FAI

The results of univariable linear regression analyses are presented in Tables S1-S4. Multivariable linear regression analyses were performed to identify independent factors associated with FAI values across all measured segments, adjusting for tube voltage, reconstruction method, vessel attenuation, age, sex, BMI, hypertension, diabetes, family history of CAD, smoking, and dyslipidemia (Tables 3-6).

Tube voltage was the most consistent factor, with 120 kVp independently associated with higher FAI values across nearly all coronary segments (β=4.9–8.1, all P<0.001), except for FAI_{_RCA_mid_10mm} (P=0.211) and FAI_{_LCX_dist_10mm} (P=0.310).

Reconstruction method also showed independent effects, with IMR (level 1) with Cardiac Routine yielding significantly lower FAI values than iDose4 (level 5) with Xres Standard in FAI_{_RCA_40mm} (β=−2.9, P=0.014), FAI_{_LAD_40mm} (β=−2.2, P=0.044), FAI_{_LAD_mid_10mm} (β=−3.0, P=0.021), and FAI_{_LCX_prox_10mm} (β=−3.0, P=0.038).

Vessel attenuation was not significantly associated with any FAI values across all measured segments (all P>0.05).

Regarding clinical characteristics, sex was independently associated with FAI only in FAI_{_LCX_dist_10mm}, with males showing lower FAI than females (β=−3.7, P=0.032). BMI (≥24 kg/m²) was associated with lower FAI_{_LAD_40mm} (β=−2.4, P=0.034), FAI_{_LAD_prox_10mm} (β=−3.2, P=0.022), and FAI_{_LAD_dist_10mm} (β=−3.4, P=0.027). Diabetes was negatively associated with FAI_{_LAD_prox_10mm} (β=−3.9, P=0.030). Smoking emerged as a positive predictor for FAI_{_RCA_40mm}, FAI_{_LAD_40mm}, FAI_{_RCA_mid_10mm}, FAI_{_LAD_prox_10mm}, and FAI_{_LCX_dist_10mm} (β=4.1–6.5, P=0.009–0.042). Dyslipidemia showed a negative association with FAI_{_LAD_40mm}, FAI_{_RCA_dist_10mm}, and FAI_{_LCX_prox_10mm} (β=−3.9 to −2.5, P=0.008–0.045). Other factors, including age, hypertension, and family history of CAD, showed no significant independent associations (all P>0.05).

Discussion

PCAT metabolic activity contributes to the progression of coronary atherosclerosis (18,19), and FAI, which reflects PCAT attenuation on CCTA, may serve as a surrogate marker for coronary inflammation (3). In our study, by focusing on subjects without CAD—defined as having no coronary plaque on CCTA—we anticipated uniformly low FAI values with low variability. Contrary to this expectation, we observed considerable variation in FAI values (ranging from −117 to −50 HU), which strongly suggests that factors unrelated to inflammation are at play. Our study found that tube voltage, reconstruction method, sex, BMI, diabetes, smoking, and dyslipidemia were independently associated with FAI in subjects without CAD. This indicates that these factors must be taken into account before drawing conclusions based on PCAT markers.

The majority of PCAT studies investigating its relationship with CAD have focused on analyzing the RCA (10–50 mm from the RCA ostium), whereas the LAD and LCX—and their respective segments—could provide additional information and improve the accuracy of outcome prediction (3,5,20). Our study focused on measuring the FAI in the proximal 40 mm of each of the three coronary arteries and in their different segments for a length of 10 mm. The results showed that FAI values obtained using the classical 40-mm measurement length were slightly but significantly different between the RCA and LAD, and between the RCA and LCX. Regarding the 10-mm segmentation analyses, FAI values differed significantly across segments within both the RCA and the LAD. These differences may arise from anatomical variations and differences in surrounding tissues, indicating that FAI values and their corresponding cutoffs—derived from RCA measurements—cannot be directly applied to other coronary arteries.

kVp levels exert a substantial influence on FAI values. In this study, compared with 100, 120 kVp was independently associated with higher FAI values across most coronary segments, except for the middle RCA and distal LCX. Clinically, CCTA acquisitions are performed at various kVp levels, with 100 and 120 kVp being the most commonly used (5,20). Higher kVp increases PCAT attenuation regardless of any pathophysiological process (12,14,21), which is consistent with our findings. Such kVp-related FAI variation may undermine the reliability of the evidence. To address this limitation, Oikonomou et al. (5) corrected the HU value differences between 100 and 120 kVp images using a correction factor of 1.11485—calculated as the ratio of the average attenuation in free-hand ROIs drawn in PCAT on 100 kVp images to that on 120 kVp images.

Two distinct reconstruction methods were used in our study: iDose4 (level 5) with Xres Standard and IMR (level 1) with Cardiac Routine. As the image quality satisfied the clinical and reconstruction requirements, we set the “Enhancement” parameter—which is used for sharpening images—to 0. IMR (level 1) with Cardiac Routine yielded significantly lower FAI values than iDose4 (level 5) with Xres Standard in FAI_{_RCA_40mm}, FAI_{_LAD_40mm}, FAI_{_LAD_mid_10mm}, and FAI_{_LCX_prox_10mm}. The Cardiac Routine filter is specifically designed for heart-focused sharpening, whereas the Xres Standard filter is more general and lacks specificity, leading to less effective sharpening than the Cardiac Routine filter. Our results were consistent with those of previous studies. Lisi et al. (13) discovered that, at the same iterative reconstruction level, the FAI for vascular kernels decreased as kernel sharpness increased. This phenomenon can be attributed to the reduced degree of edge smoothing as kernel sharpness increases. Chen et al. (22) found that reconstruction kernels had a substantial impact on the PCAT, particularly when a sharper kernel was employed. Nevertheless, it should be noted that the reconstruction method might be confounded by scanner type in our study, as the two reconstruction methods were implemented on different scanners. Future studies should compare reconstruction methods using the same scanner.

In the present study, vessel attenuation was not independently associated with FAI in any coronary segment after multivariable adjustment. This contrasts with the weak association reported by Oikonomou et al. (5) (β=0.007, P<0.001) and the poor-to-excellent correlations observed in the ex vivo porcine heart study by Pitteloud et al. (15), in which vessel attenuation influenced FAI measurements via edge smoothing effects of convolution kernels. Several reasons may explain this discrepancy. First, the association observed in the CRISP-CT study was marginal (5); when our more extensive set of confounders—particularly tube voltage, which exerted a dominant effect (β=4.9–8.1)—was adjusted for, the residual influence of vessel attenuation became negligible. Second, Oikonomou’s cohort included patients with CAD (5), in whom coronary inflammation may have confounded the association between vessel attenuation and FAI measurements, whereas we excluded patients with any plaques visible on CCTA to eliminate this bias. Third, our clinical setting inherently contains physiological noise (e.g., motion artifacts, heart rate variability), which attenuates the edge smoothing phenomenon artificially magnified in static ex-vivo models. Therefore, our results suggest that, in a clinical CCTA setting, vessel attenuation may not require correction, provided that tube voltage and reconstruction method are properly controlled.

In terms of clinical characteristics, we found that sex, BMI, diabetes, smoking, and dyslipidemia were independently associated with FAI in specific coronary segments. These factors are known to exert metabolic or inflammatory effects on PCAT, which may translate into measurable changes in FAI (23-27). Age and hypertension were respectively associated with FAI_{_LCX_dist_10mm} (β=−3.8, P=0.014) and FAI_{_LAD_mid_10mm} (β=3.3, P=0.023) in univariable analyses, but neither showed an independent association with FAI after adjustment for other confounders. Family history of CAD was not associated with any FAI in univariable or multivariable analyses.

Our study has several limitations. First, the diagnosis of no CAD was based on the absence of coronary plaques on CCTA, and the coronary artery calcium score was not assessed. Therefore, some small calcifications might have been missed. To mitigate this limitation, we analyzed only CCTA images of optimal quality. However, even when CCTA showed no evidence of CAD, we cannot exclude the possibility of PCAT inflammation in some patients without CAD. Second, although this study controlled for the major confounders, other potential factors—such as heart rate and tube current—were not accounted for. Third, although we achieved excellent interobserver agreement for FAI measurements, the placement of the 10-mm segment may still be subject to some variability. Finally, this study was a single-center study; further multicenter, large-sample studies are needed to validate the clinical applicability of our findings.

Conclusions

Tube voltage, reconstruction method, sex, BMI, diabetes, smoking, and dyslipidemia significantly affect PCAT attenuation, even in the absence of CAD. This systematic quantification of multiple confounders within a single cohort highlights the need to consider both imaging and clinical variables when interpreting FAI measurements.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://qims.amegroups.com/article/view/10.21037/qims-2026-0484/rc

Data Sharing Statement: Available at https://qims.amegroups.com/article/view/10.21037/qims-2026-0484/dss

Funding: This work was supported by Clinical Research Special Fund Project of Guangdong Medical Association (No. A202301044, Grantee: D.S.), and National Natural Science Foundation of China (NSFC) (No. U21A6005, Grantee: Z.W.).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-2026-0484/coif). D.S. received support from the Clinical Research Special Fund Project of Guangdong Medical Association. Z.W. received support from the National Natural Science Foundation of China (NSFC). The other authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. This study was approved by the Institutional Review Board of Zhujiang Hospital, Southern Medical University (IRB No. 2025-KY-209-01). Informed consent was waived due to the retrospective nature of the study and the anonymized data collection.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Libby P, Tabas I, Fredman G, Fisher EA. Inflammation and its resolution as determinants of acute coronary syndromes. Circ Res 2014;114:1867-79. [Crossref] [PubMed]
2. Ridker PM, Everett BM, Thuren T, MacFadyen JG, Chang WH, Ballantyne C, et al. Antiinflammatory Therapy with Canakinumab for Atherosclerotic Disease. N Engl J Med 2017;377:1119-31. [Crossref] [PubMed]
3. Antonopoulos AS, Sanna F, Sabharwal N, Thomas S, Oikonomou EK, Herdman L, et al. Detecting human coronary inflammation by imaging perivascular fat. Sci Transl Med 2017;9:eaal2658. [Crossref] [PubMed]
4. Klüner LV, Oikonomou EK, Antoniades C. Assessing Cardiovascular Risk by Using the Fat Attenuation Index in Coronary CT Angiography. Radiol Cardiothorac Imaging 2021;3:e200563. [Crossref] [PubMed]
5. Oikonomou EK, Marwan M, Desai MY, Mancio J, Alashi A, Hutt Centeno E, et al. Non-invasive detection of coronary inflammation using computed tomography and prediction of residual cardiovascular risk (the CRISP CT study): a post-hoc analysis of prospective outcome data. Lancet 2018;392:929-39. [Crossref] [PubMed]
6. Baichoo R, Fu J, Xiong X, Zhou Q, Guan X, Xu X. The role of cardiac adipose tissue in coronary artery disease: a narrative review of cardiac imaging studies. Quant Imaging Med Surg 2024;14:9685-97. [Crossref] [PubMed]
7. Lu ZF, Yin WH, Schoepf UJ, Abrol S, Ma JW, Zhao L, Su XM, An YQ, Xiao ZC, Lu B. Prediction value of pericoronary fat attenuation index for coronary in-stent restenosis. Eur Radiol 2024;34:4950-9. [Crossref] [PubMed]
8. Xi Y, Huang L, Hao J, Luo C, Li J, Dong Y, Lin G, Zhuang J, Jia Q, Huang M. Predictive performance of the perivascular fat attenuation index for interventional antegrade percutaneous coronary intervention for chronic total occlusion. Eur Radiol 2023;33:3041-51. [Crossref] [PubMed]
9. Moser PT, Schernthaner R, Loewe C, Strassl A, Denzinger F, Faby S, Wels M, Nizhnikava V, Uyanik-Uenal K, Zuckermann A, Stelzmueller ME, Beitzke D. Evaluation of perivascular fat attenuation with coronary CT angiography in cardiac transplantation patients: an imaging biomarker candidate for prediction of cardiac mortality and re-transplantation. Eur Radiol 2023;33:6299-307. [Crossref] [PubMed]
10. Ma R, van Assen M, Ties D, Pelgrim GJ, van Dijk R, Sidorenkov G, van Ooijen PMA, van der Harst P, Vliegenthart R. Focal pericoronary adipose tissue attenuation is related to plaque presence, plaque type, and stenosis severity in coronary CTA. Eur Radiol 2021;31:7251-61. [Crossref] [PubMed]
11. Boussoussou M, Vattay B, Szilveszter B, Simon J, Lin A, Vecsey-Nagy M, Konkoly G, Merkely B, Maurovich-Horvat P, Dey D, Kolossváry M. The effect of patient and imaging characteristics on coronary CT angiography assessed pericoronary adipose tissue attenuation and gradient. J Cardiovasc Comput Tomogr 2023;17:34-42. [Crossref] [PubMed]
12. Ma R, Ties D, van Assen M, Pelgrim GJ, Sidorenkov G, van Ooijen PMA, van der Harst P, van Dijk R, Vliegenthart R. Towards reference values of pericoronary adipose tissue attenuation: impact of coronary artery and tube voltage in coronary computed tomography angiography. Eur Radiol 2020;30:6838-46. [Crossref] [PubMed]
13. Lisi C, Klambauer K, Moser LJ, Mergen V, Manka R, Flohr T, Eberhard M, Alkadhi H. The pericoronary adipose tissue attenuation in CT strongly depends on kernels and iterative reconstructions. Eur Radiol 2025;35:2866-76. [Crossref] [PubMed]
14. Etter D, Warnock G, Koszarski F, Niemann T, Mikail N, Bengs S, Buechel RR, Kaufmann P, Gebhard C, Rossi A. Towards universal comparability of pericoronary adipose tissue attenuation: a coronary computed tomography angiography phantom study. Eur Radiol 2023;33:2324-30. [Crossref] [PubMed]
15. Pitteloud J, Moser LJ, Klambauer K, Mergen V, Flohr T, Eberhard M, Alkadhi H. Effect of Vessel Attenuation, VMI Level, and Reconstruction Kernel on Pericoronary Adipose Tissue Attenuation for EID CT and PCD CT: An Ex Vivo Porcine Heart Study. AJR Am J Roentgenol 2024;223:e2431607. [Crossref] [PubMed]
16. Patel NH, Dey AK, Sorokin AV, Teklu M, Petrole R, Zhou W, Mehta NN. Chronic inflammatory diseases and coronary heart disease: Insights from cardiovascular CT. J Cardiovasc Comput Tomogr 2022;16:7-18. [Crossref] [PubMed]
17. Zhou BF. Predictive values of body mass index and waist circumference for risk factors of certain related diseases in Chinese adults--study on optimal cut-off points of body mass index and waist circumference in Chinese adults. Biomed Environ Sci 2002;15:83-96.
18. Margaritis M, Antonopoulos AS, Digby J, Lee R, Reilly S, Coutinho P, Shirodaria C, Sayeed R, Petrou M, De Silva R, Jalilzadeh S, Demosthenous M, Bakogiannis C, Tousoulis D, Stefanadis C, Choudhury RP, Casadei B, Channon KM, Antoniades C. Interactions between vascular wall and perivascular adipose tissue reveal novel roles for adiponectin in the regulation of endothelial nitric oxide synthase function in human vessels. Circulation 2013;127:2209-21. [Crossref] [PubMed]
19. Mazurek T, Zhang L, Zalewski A, Mannion JD, Diehl JT, Arafat H, Sarov-Blat L, O'Brien S, Keiper EA, Johnson AG, Martin J, Goldstein BJ, Shi Y. Human epicardial adipose tissue is a source of inflammatory mediators. Circulation 2003;108:2460-6. [Crossref] [PubMed]
20. Goeller M, Tamarappoo BK, Kwan AC, Cadet S, Commandeur F, Razipour A, Slomka PJ, Gransar H, Chen X, Otaki Y, Friedman JD, Cao JJ, Albrecht MH, Bittner DO, Marwan M, Achenbach S, Berman DS, Dey D. Relationship between changes in pericoronary adipose tissue attenuation and coronary plaque burden quantified from coronary computed tomography angiography. Eur Heart J Cardiovasc Imaging 2019;20:636-43. [Crossref] [PubMed]
21. Pan Y, Gao Y, Wang Z, Dou Y, Sun X, Yang Z, Pan S, Jia C. Effects of low-tube voltage coronary CT angiography on plaque and pericoronary fat assessment: intraindividual comparison. Eur Radiol 2024;34:5713-23. [Crossref] [PubMed]
22. Chen L, Cao L, Liu B, Li J, Qu T, Li Y, Li Y, Pan N, Cheng Y, Fan G, Jian Z, Guo J. Relationship between pericoronary adipose tissue attenuation value and image reconstruction parameters. Heliyon 2024;10:e34763. [Crossref] [PubMed]
23. Arnold AP, Cassis LA, Eghbali M, Reue K, Sandberg K. Sex Hormones and Sex Chromosomes Cause Sex Differences in the Development of Cardiovascular Diseases. Arterioscler Thromb Vasc Biol 2017;37:746-56. [Crossref] [PubMed]
24. Perdomo CM, Avilés-Olmos I, Dicker D, Frühbeck G. Towards an adiposity-related disease framework for the diagnosis and management of obesities. Rev Endocr Metab Disord 2023;24:795-807. [Crossref] [PubMed]
25. Zheng P, Yang MC, Zhou GH, Cao HL, Wang SF, Shi SH, Ma GS, Yao YY. Diminished lymphangiogenesis in pericoronary adipose tissue mediates diabetes-aggravated coronary artery atherosclerosis. European Heart Journal 2025;46:ehaf784.4877.
26. Rossi C, Santini E, Chiarugi M, Salvati A, Comassi M, Vitolo E, Madec S, Solini A. The complex P2X7 receptor/inflammasome in perivascular fat tissue of heavy smokers. Eur J Clin Invest 2014;44:295-302. [Crossref] [PubMed]
27. Niculescu R, Stoian A, Arbănași EM, Russu E, Babă DF, Manea A, Stoian M, Gliga FI, Cocuz IG, Sabău AH, Szabo DA, Cotoi OS. The Dual Role of Perivascular Adipose Tissue in Vascular Homeostasis and Atherogenesis: From Physiology to Pathological Implications. Int J Mol Sci 2025;26:8320. [Crossref] [PubMed]