---
title: "Atrial fibrillation linked to reduced hyperemic myocardial blood flow in symptomatic patients"
id: "plos-one-21-atrial-fibrillation-is-independently-associated-with-impaired-myocardial"
canonical_url: "https://medichelpline.com/clinical-feed/plos-one-21-atrial-fibrillation-is-independently-associated-with-impaired-myocardial"
content_type: "clinical_feed_article"
specialty: "Cardiology"
source_name: "PLOS ONE (Medicine)"
source_url: "https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0361402"
published_at: "2026-10-27T14:00:00.000Z"
evidence_level: "Journal Feed"
license: "CC-BY-NC-4.0 / Informational Use"
---
# Atrial fibrillation linked to reduced hyperemic myocardial blood flow in symptomatic patients
## Provenance & Clinical Metadata
- **Canonical URL:** https://medichelpline.com/clinical-feed/plos-one-21-atrial-fibrillation-is-independently-associated-with-impaired-myocardial
- **Specialty:** [Cardiology](https://medichelpline.com/clinical-feed/cardiology.md)
- **Primary Source:** PLOS ONE (Medicine)
- **Source URL:** [Original Journal Publication](https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0361402)
- **Published At:** 2026-10-27T14:00:00.000Z
- **Evidence Rating:** Journal Feed
## Executive GIST (TL;DR)
- Investigators performed a retrospective registry-based analysis of patients referred for noninvasive coronary imaging due to suspected obstructive coronary artery disease (CAD). - From an initial 971-patient cohort, 874 met inclusion criteria; 31 had ongoing **atrial fibrillation** (AF) during imaging and 843 had sinus rhythm (SR). - The study matched 31 AF patients to 120 SR patients using propensity scores (1:4 ratio) based on age, sex, body mass index, number of traditional CAD risk factors, calcium score, and total CCTA status. - Imaging combined coronary computed tomography angiography (**CCTA**) and hyperemic [15O]H2O positron emission tomography myocardial perfusion imaging (**PET MPI**) with pharmacologic hyperemia induced by adenosine. - Global hyperemic **myocardial blood flow** (MBF) was significantly lower in AF patients than in matched SR patients (median 1.9 vs. 3.0 mL/g/min; p < 0.001). - Impaired hyperemic MBF (any segment < 2.3 mL/g/min, excluding basal septal segments) occurred in 81% of AF patients vs. 42% of SR patients (p < 0.001). - AF remained independently associated with reduced hyperemic MBF across three multivariate general linear models; p < 0.001 in each model, after adjusting sequentially for matching covariates and additional clinical variables. - Investigators report no between-group differences in total CCTA status or calcium score, indicating observed MBF differences were not explained by epicardial atherosclerotic burden as measured by CCTA. - The study team noted differences in hemoglobin, GFR, NT-proBNP, and medication use (beta-blockers, diuretics) between groups and included relevant covariates in multivariate analyses. - Data cannot be publicly shared due to statutory restrictions; eligible researchers may apply for access through the Wellbeing Services County of North Savo as documented by the authors.
## Clinical Analysis & Structured Key Points
Atrial fibrillation is independently associated with impaired myocardial perfusion | PLOS One Browse Subject Areas ? Click through the PLOS taxonomy to find articles in your field. For more information about PLOS Subject Areas, click here . Article Authors Metrics Comments Media Coverage Reader Comments Figures Figures Abstract A characteristic feature of obstructive coronary artery disease (CAD) is impaired myocardial blood flow (MBF). Previous studies have shown that atrial fibrillation (AF) is also associated with reduced hyperemic MBF. However, it is unknown whether an independent association exists between AF and MBF in symptomatic patients with suspected obstructive CAD. We studied 874 patients at Kuopio University Hospital (Finland) who underwent coronary computed tomography angiography (CCTA) and [ 15 O]H 2 O myocardial perfusion positron emission tomography due to suspected obstructive CAD in 2017–2023. The final cohort comprised 31 AF patients and 120 propensity score-matched sinus rhythm (SR) patients. The matching criteria were age, sex, body mass index, number of traditional risk factors for CAD (0–5), calcium score, and total CCTA status (i.e., the overall severity of CCTA findings). Global hyperemic MBF was lower in AF patients than in SR patients (median 1.9 mL/g/min [IQR 1.4–2.6 mL/g/min] vs. 3.0 mL/g/min [IQR 2.4–3.6 mL/g/min], p < 0.001). Impaired hyperemic MBF (< 2.3 mL/g/min in ≥ 1 segment) was more frequent in AF patients (n = 25/31, 81%) than in SR patients (n = 50/120, 42%, p < 0.001). Furthermore, the presence of AF was independently associated with reduced hyperemic MBF in three multivariate general linear models (p < 0.001 in each model). To summarize, AF was independently associated with impaired hyperemic MBF in patients with suspected obstructive CAD. This association should be considered in clinical practice when interpreting the results of myocardial perfusion imaging in patients with AF. Citation: Vartiainen N, Hartikainen S, Rissanen A-PE, Kariniemi E, Leikas AJ, Strengell E, et al. (2026) Atrial fibrillation is independently associated with impaired myocardial perfusion. PLoS One 21(10): e0361402. https://doi.org/10.1371/journal.pone.0361402 Editor: Ismaheel Lawal, Emory University, UNITED STATES OF AMERICA Received: August 20, 2026; Accepted: October 11, 2026; Published: October 27, 2026 Copyright: © 2026 Vartiainen et al. This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Data Availability: The data underlying this manuscript cannot be publicly shared since it contains sensitive health information and data access is subject to statutory restrictions and the terms of the relevant data permits. The researchers are not authorized to share individual-level data to third parties. Eligible researchers may apply for data access from the Wellbeing Services County of North Savo (contact email: kirjaamo@pshyvinvointialue.fi ). All applications will be independently evaluated by the data controllers and the data permit authority. Funding: TPL received a grant from the Finnish Foundation for Cardiovascular Research (November 10, 2020). URL: https://www.sydantutkimussaatio.fi/ . The funder played no role role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing interests: The authors have declared that no competing interests exist. Introduction Despite advances in coronary artery disease (CAD) management, it is still a considerable health burden globally [ 1 ]. In recent years, the primary diagnostic methods of CAD have changed. For the initial diagnosis of CAD, the latest American and European guidelines recommend noninvasive imaging in most scenarios [ 2 , 3 ]. Among these methods, coronary computed tomography angiography (CCTA) has emerged as a widely used and practical modality. CCTA enables direct visualization of epicardial coronary arteries and shows the distribution and severity of coronary atherosclerosis. Due to its high sensitivity and negative predictive value, CCTA excels in ruling out CAD even in patients with high pretest probability [ 4 ]. CCTA demonstrates lower specificity for confirming obstructive CAD, however [ 4 ]. To increase diagnostic accuracy, CCTA can be combined with myocardial perfusion imaging (MPI). Among the various MPI modalities, positron emission tomography (PET) with [ 15 O]H 2 O is considered the reference standard for absolute quantification of myocardial blood flow (MBF) [ 5 ]. However, impaired MBF is not specific to obstructed epicardial vessels. Instead, [ 15 O]H 2 O PET MPI also reflects the perfusion of the coronary microvasculature [ 6 ]. Besides CAD, atrial fibrillation (AF), the most common sustained arrhythmia, causes significant morbidity, hospitalization, and mortality [ 7 ]. CAD and AF share risk factors and, additionally, CAD is a risk factor for the development of AF [ 8 , 9 ]. Even patients without CAD may present with symptoms and clinical findings suggesting myocardial ischemia if they have AF [ 10 , 11 ]. This may relate to several mechanisms including coronary microvascular dysfunction [ 12 ]. Previous studies in various patient populations have reported that when compared to patients with sinus rhythm (SR), those with AF have lower myocardial perfusion [ 13 – 18 ]. Still, whether AF is independently associated with myocardial perfusion in patients referred to diagnostic imaging due to suspected obstructive CAD remains unknown. Also, when investigating the association between AF and myocardial perfusion, previous studies have not utilized the severity and extent of coronary atherosclerosis as confounders. We studied whether AF is independently associated with impaired myocardial perfusion in patients referred to noninvasive coronary imaging due to symptoms suggesting the presence of obstructive CAD. To investigate this possible independent association, we compared hyperemic MBF in patients with AF and in comprehensively matched patients with SR. Methods Subjects We conducted a retrospective, registry-based study of patients who underwent CCTA and consecutive [ 15 O]H 2 O PET MPI at Kuopio University Hospital, Finland, from 2017 to 2023. Each patient was clinically referred to the imaging due to symptoms suggesting obstructive CAD. The initial study population comprised 971 patients. Our exclusion criteria encompassed technically unsuccessful CCTA or PET MPI, prior myocardial infarction or coronary revascularization, heart failure, cardiomyopathy, at least moderate valvular heart disease, pacemaker, or any cardiac rhythm other than AF or SR during PET MPI. After applying the exclusion criteria, our final study population included 874 patients. Of these, 31 had ongoing AF, and 843 had SR during PET MPI. Eventually, we used propensity score matching (see the section Statistical methods for details) to match the 31 AF patients with 120 SR patients and included those 151 patients in this study’s final analyses. Fig 1 presents a flow chart for patient selection. Download: PNG larger image TIFF original image Fig 1. Flow chart for patient selection. AF = atrial fibrillation, CCTA = coronary computed tomography angiography, PET = positron emission tomography, SR = sinus rhythm. https://doi.org/10.1371/journal.pone.0361402.g001 For this research, we secured written authorization from the registrar (the Wellbeing Services County of North Savo) to utilize their data. In Finland, scientific research based only on registry data requires no approval of an ethics committee or informed consent from patients. Therefore, before this study, we received an exemption from ethical approval from the ethics committee of the University of Eastern Finland (description number 16/2023). Data collection took place between February 2024 and September 2024, during which the authors had access to information that could identify individual patients of the study. After the data were collected and analyzed, they were anonymized so that no individual patients could be identified. Clinical data Clinical data were collected from electronic patient records and a questionnaire completed by the patients before CCTA and PET MPI. The data included age, sex, body mass index, medication, angina symptoms, comorbidities, and traditional risk factors for CAD (hypertension, dyslipidemia, diabetes, smoking history, and family history of CAD). Blood test results within three months of PET MPI were collected when available. Ongoing cardiac rhythm was determined based on 12-lead electrocardiography obtained during PET MPI. The type of AF was classified based on electronic patient records. Imaging protocol At Kuopio University Hospital, CCTA serves as the primary noninvasive modality for diagnosing CAD in patients with a low-to-moderate pretest probability. If CCTA is suggestive of obstructive CAD (≥ 50% luminal stenosis in any epicardial coronary artery) or if obstructive CAD cannot be excluded due to suboptimal image quality, patients are referred to PET MPI. Before PET MPI, the patients are instructed to abstain from smoking for at least 4 hours, caffeine for at least 12 hours, and alcohol for at least 48 hours. Also, dipyridamole is discontinued for a minimum of 48 hours before PET MPI. Most patients undergo CCTA and PET MPI consecutively on the same day. Rarely due to scheduling issues or inadequate patient preparation, CCTA and PET MPI are performed on separate days. Coronary computed tomography angiography For this current study, three different CT scanners by Siemens (Siemens Healthineers, Erlangen, Germany) were used for CCTA: Somatom Definition Flash, Somatom Force, and Somatom Edge. If the patient’s heart rate was over 65 beats/min, they were administered up to 20 mg of intravenous metoprolol. Immediately before imaging, the patients also received 1.25 mg of sublingual isosorbide dinitrate. CCTA imaging began with the acquisition of unenhanced CT images, from which the coronary artery calcium score was determined according to Agatston et al . [ 19 ]. Next, contrast-enhanced CCTA scans were performed using prospective electrocardiograph-gated acquisition whenever possible. From CCTA images, we determined findings from the left main (LM), left anterior descending (LAD), left circumflex (LCX), and right (RCA) coronary arteries. A status was categorized for each artery: 0 = vessel not interpretable, 1 = no visible atherosclerosis, 2 = calcified or non-calcified plaque(s) with luminal stenosis < 50%, 3 = calcified plaque(s) with luminal stenosis ≥ 50%, 4 = non-calcified or partially calcified plaque(s) with luminal stenosis ≥ 50%, and 5 = total occlusion. This grading was adapted from the 2022 Coronary Artery Disease-Reporting and Data System [ 20 ]. A total CCTA status of 0–5 was formed based on the highest status from any individual coronary artery. For example, if a patient’s status for RCA, LM, and LAD was 2, and the status for LCX was 4, the total CCTA status was 4. Myocardial perfusion imaging PET MPI was performed with [ 15 O]H 2 O using either Biograph mCT or Biograph Vision PET/CT scanner (Siemens Healthineers, Erlangen, Germany). Imaging was performed during pharmacologically induced hyperemia, achieved with a continuous intravenous infusion of adenosine at a rate of 140 µg/kg/min over 6.5 minutes. During adenosine infusion, an intravenous bolus of [ 15 O]H 2 O with a weight adjusted activity of 500–800 MBq was injected, and dynamic images were acquired over 5 minutes. A low-dose CT was used for attenuation correction. Since our standard clinical protocol included only hyperemic imaging, rest images were unavailable for most patients and were excluded from this study. Blood pressure (BP) was measured at rest and immediately after the [ 15 O]H 2 O bolus with an automated sphygmomanometer device (SunTech Tango M2, SunTech Medical Inc., Morrisville, North Carolina, USA). A 12-lead electrocardiogram was monitored using the GE Case (General Electric Company, Boston, Massachusetts, USA). Rate pressure product (RPP) was calculated as systolic BP multiplied by heart rate (HR). PET images acquired with Biograph mCT were reconstructed with two iterations and 21 subsets, with a 3.18 mm x 3.18 mm pixel size. For images acquired with Biograph Vision, image reconstruction was performed with eight iterations, five subsets, and a pixel size of 2.83 mm x 2.83 mm. A matrix size of 128 x 128 and Gaussian 6 mm post-filtering were used for all images regardless of the PET/CT scanner. Hyperemic MBF was quantified with Carimas software (Turku PET Centre, Turku, Finland) using a single-tissue compartment model [ 21 , 22 ]. MBF was quantified using the standard 17-segment model [ 23 ], with measurements recorded for each segment in mL/g/min. MBF was also assessed as a global value and for the standardized perfusion territories of LAD, LCX, and RCA. MBF was considered impaired when any individual segment (excluding basal septal segments 2 and 3) had a value of < 2.3 mL/g/min [ 24 ]. Invasive coronary angiography and revascularization Based on the noninvasive imaging protocol and patient-specific evaluation, attending cardiologists considered whether to refer individual patients to ICA. Additionally, if significant stenoses were present in ICA, they decided whether to perform revascularization procedures. Statistical methods We formed two groups for statistical analyses: patients who, during PET MPI, had ongoing AF and those who had ongoing SR. To assess the possible independent association between AF and myocardial perfusion, we formed the groups using propensity score matching. Matching was based on six criteria: age, sex, body mass index, number of traditional risk factors for CAD (0–5), calcium score, and total CCTA status [ 24 – 26 ]. We performed power calculations to assess the required number of patients in each group. The estimated effect size was based on the study by Madsen et al [ 18 ]. In their study, the mean difference in hyperemic MBF between AF and SR patients was 0.70 mL/g/min in patients without CAD and 0.42 mL/g/min in patients with CAD. Because our patient population included those with and without obstructive CAD, we employed an effect size of 0.56 mL/g/min, which represents an average between these two patient groups. According to the power calculations, 28 AF patients and 111 SR patients would be needed to achieve a statistical power greater than 80% with a two-tailed α -level less than 0.05. Therefore, propensity score matching was performed using a 1:4 ratio, i.e., each AF patient was matched with four SR patients. Matching was performed with R 4.3.2 (R Foundation for Statistical Computing, Vienna, Austria) using a logistic regression model. AF and SR patients were matched according to the closest propensity score (within ± 0.2 standard deviations). Apart from propensity score matching, statistical analyses were conducted using IBM SPSS Statistics version 29 (IBM, Armonk, New York, USA). We used the Shapiro-Wilk test to assess the normality of distribution for continuous variables. The results were presented as mean ± standard deviation for normally distributed data, median (interquartile range, IQR) for nonnormally distributed data, or n (%) for categorical data. Finnish Social and Health Data Permit Authority restricts the presentation of data in registry studies. Thus, we were unable to present categorical data with n < 3 in exact numbers. To compare continuous variables between groups, we used the Student’s T-test for normally distributed data and the Mann-Whitney U-test for nonnormally distributed data. Depending on the expected cell count, either the chi-squared test or Fisher’s exact test was used to compare nominal data between groups. Statistical significance was set at p-value < 0.05. To further assess the possible independent association between AF and hyperemic MBF, we performed multivariate general linear model analyses with three models. In each model, the presence of AF was a fixed factor. In the first model, the covariates included each variable used in propensity score matching (i.e., age, sex, body mass index, number of traditional risk factors for CAD [0–5], calcium score, and total CCTA status). In the second model, the covariates included those of the first model as well as hemoglobin and the use of beta-blockers (i.e., patient characteristics that differed statistically significantly between the patients with AF and SR and were statistically significantly associated with global hyperemic MBF). In the third model, the covariates included those of the second model as well as diastolic blood pressure after [ 15 O]H 2 O bolus, glomerular filtration rate (GFR), and the use of diuretics (i.e., variables that differed statistically significantly between AF and SR patients, but which were not statistically significantly associated with global hyperemic MBF). In the multivariate general linear model analyses, we observed no unequal variances (i.e., p-value in Levene’s test was consistently ≥ 0.05) or nonnormally distributed residuals. Also, we found no significant multicollinearity between the covariates, i.e., the variance inflation factor was consistently < 2.5 [ 27 , 28 ]. Results Clinical characteristics Thirty-one AF patients were matched with 120 SR patients. Two of the AF patients could be matched with only two SR patients within the specified propensity score range (± 0.2 standard deviations). The patients’ clinical characteristics are shown in Table 1 . Most AF patients had persistent or permanent AF. Among SR patients, 16 (13%) had previously documented AF paroxysms but had ongoing SR during PET MPI. No statistically significant differences existed between the groups in terms of age, sex, or body mass index. Also, no differences between the groups were observed in the five traditional risk factors for CAD. The use of an angiotensin-converting enzyme (ACE) inhibitor or an angiotensin II receptor (AR) blocker, a calcium channel blocker, and cholesterol medication were similar between the AF and SR patients. Beta-blocker and diuretic medications were more common in AF patients than in SR patients. Each of the three blood test results differed between the groups: GFR was lower, whereas hemoglobin and N-terminal pro-B-type natriuretic peptide (NT-proBNP) were higher in AF patients than in SR patients. Download: PNG larger image TIFF original image Table 1. Patient characteristics. https://doi.org/10.1371/journal.pone.0361402.t001 Coronary computed tomography angiography Table 2 presents the CCTA results. Premedication with beta-blockers was equally common between AF and SR patients. Calcium score and total CCTA status showed no difference between the groups. Also, no between-group differences were observed in the status of any individual epicardial coronary artery. Download: PNG larger image TIFF original image Table 2. CCTA results. https://doi.org/10.1371/journal.pone.0361402.t002 Myocardial perfusion imaging The hemodynamic parameters during PET MPI are shown in Table 3 . At rest, systolic BP was similar between the groups, while diastolic BP, HR, and RPP were higher in AF patients than in SR patients. After the [ 15 O]H 2 O bolus, systolic BP, HR, and RPP were similar between the groups, whereas compared to SR patients, those with AF had higher diastolic BP. Download: PNG larger image TIFF original image Table 3. Hemodynamics during PET MPI. https://doi.org/10.1371/journal.pone.0361402.t003 Fig 2 presents the results of PET MPI. Global hyperemic MBF was lower in AF patients than in SR patients (median: 1.9 mL/g/min, IQR: 1.4–2.6 mL/g/min vs. 3.0 mL/g/min, IQR: 2.4–3.6 mL/g/min, p < 0.001). Also, hyperemic MBF in each standardized perfusion territory was lower in AF patients (
## Related Clinical Research

- [Comparing BMI, Waist Circumference, and BRI for Predicting ASCVD, AF, and Mortality in 1.2M Koreans](https://medichelpline.com/clinical-feed/plos-one-14-prognostic-associations-of-anthropometric-indices-with-major-health-outcomes-a.md)
- [Impact of Nighttime Light Exposure on Heart Health and Disease Risk](https://medichelpline.com/clinical-feed/medical-news-today-0-nighttime-light-exposure-may-alter-the-heart-and-increase-disease-risk.md)
- [Frailty prediction in heart failure with acute infection: role of thiazide diuretics and an interp](https://medichelpline.com/clinical-feed/plos-one-9-frailty-prediction-in-heart-failure-patients-with-acute-infections-the.md)
- [Cardiovascular risk and mortality after statin initiation: achieved LDL and non‑HDL levels in a 3‑](https://medichelpline.com/clinical-feed/plos-medicine-0-risk-of-cardiovascular-events-and-mortality-associated-with-ldl-and-non-hdl.md)
- [GLP-1 drugs tied to lower diabetes progression and cardiovascular events than metformin](https://medichelpline.com/clinical-feed/aha-news-4-glp-1-meds-linked-to-lower-diabetes-heart-risks-than-standard-diabetes.md)

## Navigation
- [← Back to Cardiology Feed](https://medichelpline.com/clinical-feed/cardiology.md)
- [← All Clinical Specialties](https://medichelpline.com/clinical-feed.md)
## Medical & Regulatory Disclaimer

> [!CAUTION]
> MedicHelpline content is structured for research, educational, and professional discovery purposes. It does not constitute individual medical advice, clinical diagnosis, or treatment recommendations.
> Always verify dosing, contraindications, and regulatory alerts against official product labeling and primary regulatory sources before clinical decision-making.