A Prospective Study on Clinical Profile and the Utility of 72-h Holter Monitoring in Acute Ischaemic Stroke and Transient Ischaemic Attack (TIA) Patients

Abstract

Background

Atrial fibrillation (AF) is a major cause of recurrent ischaemic stroke, yet detection after acute events is often suboptimal, particularly in resource-limited settings.

Purpose

This study aims to assess the clinical profile and the utility of 72-h Holter monitoring in determining AF and other cardiac arrhythmias in acute ischaemic stroke and transient ischaemic attack (TIA) patients.

Methods

In this prospective observational study, 200 consecutive patients with acute ischaemic stroke/TIA underwent detailed clinical evaluation, laboratory profiling, neuroimaging, echocardiography and continuous electrocardiogram (ECG) monitoring (72-h Holter). Stroke subtyping followed trial of ORG 10172 in acute stroke treatment (TOAST) criteria. Predictors of newly detected AF were identified via univariate and multivariate logistic regression. Ethical approval was obtained, and informed consent was secured.

Results

Among 200 patients with acute ischaemic stroke or TIA (mean age 63.9 ± 12.4 years; 69% male), 72-h Holter monitoring detected AF/flutter in 20.5%, with 70.7% of these showing clinically significant episodes (≥30 s). Other arrhythmias included supraventricular tachycardia (SVT) (12.5%), ventricular ectopics (16.0%) and atrioventricular (AV) block (4.5%). Holter findings led to treatment modification in 20%, primarily the initiation of anticoagulation. Compared with non-AF patients, those with AF were older (≥60 years: 90.2%, p = .002), had higher rates of diabetes (68.3%), hypertension (85.4%), chronic kidney disease (CKD) (43.9%) and left atrial enlargement (LAE) (29.2% vs 1.9%, p = .01). AF patients presented with higher median National Institutes of Health Stroke Scale (NIHSS) [8 vs 6, p = .04] and more severe infarcts, with posterior (48.8% vs 27.0%) and multi-territory (19.5% vs 5.3%) involvement. Independent predictors of AF included CKD (aOR 4.91), left atrium (LA) enlargement (aOR 5.18), multi-territory infarcts (aOR 12.35), posterior circulation infarcts (aOR 3.74) and tachycardia >100 bpm at admission (aOR 4.78).

Conclusion

72-hour Holter monitoring demonstrated high clinical utility in detecting AF and other clinically significant arrhythmias in acute ischaemic stroke and TIA patients, with direct therapeutic impact through timely initiation of anticoagulation. Importantly, the identification of SVT in 12.5% of patients, particularly those with a higher arrhythmic burden, highlights a potential preclinical substrate of atrial cardiopathy that may precede overt AF. These findings support the incorporation of extended Holter monitoring into post-stroke evaluation, especially for high-risk groups, both to optimise etiological classification and to enable personalised secondary prevention strategies that reduce the risk of recurrent cerebrovascular events.

Keywords

Atrial fibrillation ischaemic stroke Holter electrocardiogram predictors atrial cardiopathy

Introduction

Ischaemic stroke and transient ischaemic attack (TIA) remain major contributors to global morbidity and mortality, particularly in low- and middle-income countries like India, where stroke accounts for over 80% of stroke-related deaths and incidence rates range from 84 to 424 per 100,000 dependent on rural or urban settings.¹ Atrial fibrillation (AF), a frequent but often asymptomatic and paroxysmal arrhythmia, is a well-established cause of cardioembolic stroke—yet despite routine electrocardiogram (ECG) and 24-h Holter monitoring, a substantial proportion of ischaemic strokes (nearly one-quarter) remain cryptogenic or are classified as embolic strokes of undetermined source (ESUS).^{2, 3} ESUS, defined by the absence of lacunar infarction, significant arterial stenosis or identifiable cardioembolic source, represents approximately 17% of ischaemic strokes and carries a high recurrence rate comparable to cardioembolic aetiology.³

Landmark trials have transformed our understanding of AF detection post-stroke. The Cryptogenic Stroke and Underlying Atrial Fibrillation-AF (CRYSTAL-AF) trial demonstrated that insertable cardiac monitors (ICMs) detected AF at significantly higher rates than standard care—with detection rates in the ICM arm reaching 30% at 36 months versus only 3% in controls.^{4, 5} Likewise, the Event Monitoring Belt for Recording Atrial fibrillation after a Cerebral Event (EMBRACE) trial showed superior detection of AF with 30-day external event recording compared to conventional evaluation.⁶ STROKE-AF further confirmed the incremental yield of prolonged monitoring in stroke populations.⁷ These data collectively indicate that short-duration monitoring substantially underestimates AF burden in post-stroke cohorts. However, advanced monitoring devices may be financially and logistically infeasible in many resource-limited settings. In this context, extended non-invasive monitoring protocols such as 72-h Holter monitoring offer a potentially pragmatic compromise, balancing improved AF detection with feasibility. Accordingly, this study aims to evaluate the diagnostic yield and clinical utility of 72-h Holter monitoring in acute ischaemic stroke and TIA patients, with a specific focus on identifying clinical, echocardiographic and neuroimaging predictors that may guide a selective, risk-stratified approach to prolonged rhythm monitoring.

Methods

Study Design

This prospective observational study was conducted in the Department of Neurology, Kasturba Medical College (KMC), Manipal, a tertiary care teaching hospital in India, between March 2024 and May 2025. The study protocol was approved by the KMC and Kasturba Hospital Institutional Ethics Committee (EC) with EC Approval Number (IEC1-377/2023), and written informed consent was obtained from all participants or their legal surrogates. Ethical principles were followed in accordance with the Declaration of Helsinki.⁸ This study aims to evaluate the clinical utility of 72-h Holter monitoring for detecting AF and other clinically relevant arrhythmias in patients with acute ischaemic stroke or TIA and to clarify how these findings inform secondary prevention. Specifically, we will determine the types and frequencies of arrhythmias captured during extended monitoring; document treatment changes attributable to Holter results, including initiation or adjustment of anticoagulation and antiarrhythmic therapy; compare the clinical, echocardiographic and neuroimaging profiles of patients with Holter-detected AF versus those without AF; and quantify the incremental diagnostic yield of 72-h Holter monitoring over baseline 12-lead ECG and transthoracic echocardiography. Given the observational nature of the study, no randomisation or comparison with routine care was performed; the study was not designed to mandate prolonged monitoring but to evaluate the diagnostic yield and identify predictors to guide selective clinical use.

Study Population

A total of 200 consecutive adult patients with radiologically confirmed acute ischaemic stroke or TIA presenting within 1 week of symptom onset were recruited. Inclusion criteria were age ≥18 years and diagnosis of acute ischaemic stroke or TIA within 1 week of onset. Exclusion criteria included neurological illnesses other than ischaemic stroke, non-ischaemic causes of stroke such as haemorrhage, trauma, dissecting aneurysm or central nervous system (CNS) infections, prior known AF already on anticoagulation, and inability or refusal to provide consent. 72-h Holter monitoring was performed uniformly in all eligible patients irrespective of baseline ECG or echocardiographic findings, as per the predefined study protocol, to avoid selection bias and accurately assess the incremental yield of prolonged monitoring.

Clinical and Neurological Assessment

Baseline demographic data and vascular risk factors—including hypertension, diabetes mellitus, dyslipidaemia, chronic kidney disease (CKD), cardiovascular disease and substance use—were recorded. Stroke severity was assessed using the National Institutes of Health Stroke Scale (NIHSS),⁹ and functional status on admission was graded using the modified Rankin Scale (mRS).¹⁰ Hypertension was defined as a documented history, antihypertensive use or blood pressure ≥140/90 mmHg on admission.¹¹ Only Grade II or higher diastolic dysfunction was considered clinically significant, as Grade I diastolic dysfunction is frequently age-related and may not reliably reflect a pathological substrate associated with AF. Diabetes mellitus was diagnosed based on history, fasting plasma glucose ≥126 mg/dL, HbA1c ≥6.5%, or antidiabetic therapy,¹² and poor glycaemic control was defined as HbA1c ≥8%.¹² Dyslipidaemia was defined according to National Cholesterol Education Programme Adult Treatment Panel III (NCEP ATP III) criteria, with low-density lipoprotein (LDL) cholesterol ≥130 mg/dL and triglycerides ≥150 mg/dL,¹³ and CKD was defined as estimated glomerular filtration rate (eGFR) <60 mL/min/1.73 m² using the modification of diet in renal disease (MDRD) equation.¹⁴ Frequent ectopy was defined as >30 ectopic beats/hour,¹⁵ and left atrial enlargement (LAE) was graded according to the American Society of Echocardiography guidelines.⁹ A supraventricular tachycardia (SVT) burden ≥5% was considered clinically significant and predictive of future AF.^{6, 15, 16} Alcohol dependence syndrome (ADS) and tobacco dependence syndrome (TDS) were defined according to the International Classification of Diseases (ICD-10) criteria. These variables were included as baseline vascular risk factors and were not considered primary determinants of AF detection.

Investigations

All patients underwent a complete blood count, renal and liver function tests, fasting and random glucose, HbA1c, lipid profile, electrolytes and coagulation profile. Brain imaging with computed tomography (CT) or magnetic resonance imaging (MRI) was performed to localise the infarct and classify the stroke territory, and stroke subtype was determined using the trial of ORG 10172 in acute stroke treatment (TOAST) classification.¹⁷ Baseline 12-lead ECG assessed rhythm, rate, ectopy and conduction abnormalities, while 2D echocardiography evaluated left atrium (LA) size, left ventricular ejection fraction (EF), wall motion, thrombus, valvular disease and diastolic dysfunction (grade ≥2). Baseline 12-lead ECG at admission was assessed for rhythm, heart rate, conduction abnormalities and ectopy. Findings included sinus tachycardia, atrial ectopics, ventricular ectopics and non-specific ST-T changes. Patients with persistent or previously known AF on admission ECG were excluded to ensure that AF detected on Holter represented newly identified or paroxysmal AF. All participants underwent 72-h Holter ECG monitoring using the Web Cardio SP3 system, interpreted by blinded cardiologists or electrophysiologists. Arrhythmias were defined according to guideline criteria: AF as an irregular rhythm without P waves lasting ≥30 s,¹⁸ SVT as a narrow QRS tachycardia >100 bpm with ≥3 consecutive beats and ≥5% burden,^{8, 10} pauses >2 s, the Wolff–Parkinson–White pattern per ECG criteria, and atrioventricular (AV) block graded per American College of Cardiology (ACC)/American Heart Association (AHA)/Heart Rhythm Society (HRS) guidelines.¹⁹ Posterior circulation and multi-territory infarct patterns were not independently classified as cardioembolic strokes. These patterns were analysed as radiological markers associated with a higher likelihood of AF detection and were not used as standalone etiological labels.

Data Collection and Statistical Analysis

Data were recorded in standardised forms, entered into Microsoft Excel, and analysed using Statistical Package for the Social Sciences (SPSS) v21. Categorical variables were expressed as frequencies and percentages and compared using chi-square tests. Continuous variables were presented as mean ± standard deviation or median with an interquartile range, with unpaired t-tests or Mann–Whitney U tests applied as appropriate. Univariate and multivariate logistic regression analyses were performed to identify predictors of AF, reporting odds ratios (OR) with 95% confidence intervals (CIs). Statistical significance was set at p < .05.

Results

Among the 200 patients included, the mean age was 63.9 ± 12.4 years, with 67.5% aged 60 years or older and 69.0% being male. The majority (59.0%) presented within 12 h of symptom onset, and only 2.5% sought care beyond 72 h. Hypertension emerged as the predominant vascular risk factor, affecting 73.0% of patients, with nearly half having a history of more than 5 years’ duration. Diabetes mellitus was present in 50.5% of the cohort, and poor glycaemic control (HbA1c ≥8%) was documented in 46.5% of these cases. Dyslipidaemia was identified in 58.0% of patients, with abnormal high-density lipoprotein (HDL) cholesterol being the most common lipid abnormality (31.5%), followed by hypertriglyceridaemia (27.5%) and elevated LDL cholesterol (13.0%). CKD was noted in 17.0% of participants, and 23.5% reported a prior history of stroke or TIA. Substance use was common, with 41.0% reporting at least one habit. Smoking alone was reported in 12.5%, alcohol consumption in 13.0%, and concurrent use of both in 10.5%. Chewing forms of tobacco (TDS) or areca nut (ADS) were also prevalent, either alone or in combination (Table 1).

Table 1.

Baseline Clinical Characteristics of the Study Population (n = 200).

Parameter	Frequency (%)
Age (years) (mean ± SD)	63.91 ± 12.4
Age Category
<45 years	18 (9.0)
46–59 years	47 (23.5)
≥60 years	135 (67.5)
Gender
Male	138 (69.0)
Female	62 (31.0)
Onset of Symptoms
<12 h	118 (59.0)
12–24 h	56 (28.0)
24–48 h	21 (10.5)
48–72 h	2 (1.0)
72–96 h	2 (1.0)
96–120 h	1 (0.5)
Hypertension (HTN)	146 (73.0)
HTN duration <1 year	42 (28.7)
HTN duration 1–5 years	40 (27.3)
HTN duration 6–10 years	24 (16.4)
HTN duration 11–15 years	25 (17.1)
HTN duration >15 years	15 (10.2)
Systolic BP (mmHg)	146 ± 19.1
Diastolic BP (mmHg)	82.7 ± 50.3
Diabetes mellitus (DM)	101 (50.5)
DM duration <1 year	37 (36.6)
DM duration 1–5 years	35 (34.6)
DM duration 6–10 years	10 (9.9)
DM duration 11–15 years	12 (11.8)
DM duration >15 years	7 (6.9)
Poor DM control	47 (46.5)
Ischaemic heart disease	24 (12.0)
Rheumatic heart disease/valvular disease	8 (4.0)
Dyslipidaemia	116 (58.0)
Abnormal LDL	26 (13.0)
Abnormal HDL	63 (31.5)
Abnormal triglycerides	55 (27.5)
Chronic kidney disease	34 (17.0)
History of stroke/TIA	47 (23.5)
Substance Use
Smoking	25 (12.5)
Smoking and alcohol	21 (10.5)
Alcohol	26 (13.0)
ADS	16 (8.0)
TDS	15 (7.5)
ADS and TDS	13 (6.5)

Notes: Values are presented as mean ± standard deviation (SD) or n (%). ADS: Alcohol dependence syndrome; AF: Atrial fibrillation; BP: Blood pressure; HDL: High-density lipoprotein; LDL: Low-density lipoprotein; TDS: Tobacco dependence syndrome; TIA: Transient ischaemic attack.

The mean heart rate was 77.65 ± 16.76 bpm; bradycardia (<60 bpm) occurred in 10.5% and tachycardia (>100 bpm) in 11.0%. The mean EF was 60.5% ± 8.9%, with left ventricle (LV) hypertrophy in 26.0%, regional wall motion abnormalities in 16.0%, and diastolic dysfunction (alone or with left ventricular hypertrophy (LVH)) in 14.0%. Valvular regurgitation was present in 4.5%, and intracardiac thrombus in 1.5%. LAE was seen in 7.5%, while ECG ectopics at admission occurred in 26.0%. Imaging showed anterior circulation infarcts in 43.5% and posterior circulation infarcts in 31.5%. Most patients (88.5%) had <70% stenosis on the symptomatic side; high-grade ipsilateral stenosis (>70%) was found in 4.5% (Table 2).

Table 2.

Cardiac, Electrocardiogram (ECG) and Imaging Findings.

Parameter	Frequency (%)
Heart Rate
Mean ± SD	77.65 ± 16.76
<60 bpm	21 (10.5)
60–100 bpm	157 (78.5)
>100 bpm	22 (11.0)
Ejection Fraction (ECHO)
Mean ± SD (%)	60.5 ± 8.9
Normal EF	72 (36.0)
Left ventricular hypertrophy (LVH)	52 (26.0)
Regional wall motion abnormality (RWMA)	32 (16.0)
Vegetations	1 (0.5)
Cardiomyopathy	3 (1.5)
Left ventricular or atrial clot	3 (1.5)
Left ventricular diastolic dysfunction	15 (7.5)
Valve regurgitation	9 (4.5)
LVH + diastolic dysfunction	13 (6.5)
Left Atrial Size
Mean ± SD (mm)	31.63 ± 4.8
Mild enlargement	7 (3.5)
Moderate–severe enlargement	8 (4.0)
Normal size	185 (92.5)
Ejection Fraction (Categorised)
<30%	5 (2.5)
30%–59%	42 (21.0)
≥60%	153 (76.5)
ECG with ectopics at admission	52 (26.0)
Infarct Territory
Anterior circulation	87 (43.5)
Posterior circulation	63 (31.5)
Watershed	10 (5.0)
Multi-territory	21 (10.5)
Vessel Occlusion
<70% same side	177 (88.5)
>70% same side	9 (4.5)
<70% opposite side	7 (3.5)
<70% both sides	2 (1.0)
>70% both sides	2 (1.0)

Notes: Data expressed as n (%). CKD: Chronic kidney disease; DM: Diabetes mellitus; ECHO: Echocardiography; EF: Ejection fraction; HTN: Hypertension; SD: Standard deviation.

72-h Holter monitoring revealed that 40.0% of patients maintained normal sinus rhythm, whereas AF/flutter was detected in 20.5% (Figure 1). Other notable arrhythmias detected during 72-h holter monitoring (distribution illustrated in Figure 2) included SVT (12.5%), ventricular ectopics (16.0%), AV block (4.5%), junctional rhythms (2.0%), junctional beats (3.0%), pauses (1.0%) and Wolff–Parkinson–White syndrome (0.5%). Stroke severity assessment at admission showed that nearly half the cohort (47.0%) had moderate strokes, while 33.0% had minor strokes, and only 4.0% presented with severe strokes. Functional status (mRS) indicated that 38.5% had minimal disability (mRS 1) and 24.5% were asymptomatic (mRS 0) at presentation. In patients with AF, 70.7% had an arrhythmia duration of ≥30 s, indicating clinically significant episodes, while 29.3% had shorter events. Among those with SVT, 36.0% had a burden >5%, suggesting a higher arrhythmic load. Holter results led to treatment modification in 20.0% of the cohort, primarily through the initiation of anticoagulation for stroke prevention. By TOAST classification, large artery atherosclerosis (27.5%), small artery occlusion (26.5%) and cardioembolism (26.0%) were the leading aetiologies. A smaller proportion had strokes from other determined causes (4.5%) or remained of undetermined origin (6.0%) (Table 3).

Table 3.

72-h Holter Findings, Stroke Severity Scores, Atrial/Supraventricular Tachycardia (SVT) Burden, Treatment Modification and Trial of ORG 10172 in Acute Stroke Treatment (TOAST) Classification.

Parameter	Frequency (%)
72-h Holter Findings
Normal sinus rhythm	80 (40.0)
Atrial fibrillation/flutter	41 (20.5)
Junctional rhythm	4 (2.0)
Supraventricular tachycardia (short atrial tachycardia)	25 (12.5)
AV block	9 (4.5)
Ventricular ectopics	32 (16.0)
Junctional beats	6 (3.0)
Pauses	2 (1.0)
WPW syndrome	1 (0.5)
NIHSS at Admission
Minor stroke	66 (33.0)
Moderate stroke	94 (47.0)
Moderate–severe stroke	13 (6.5)
Severe stroke	8 (4.0)
mRS at Admission
0	49 (24.5)
1	77 (38.5)
2	33 (16.5)
3	19 (9.5)
4	10 (5.0)
5	12 (6.0)
Atrial Fibrillation Burden
<30 s	12 (29.3)
≥30 s	29 (70.7)
Supraventricular Tachycardia Burden
<5%	16 (64.0)
>5%	9 (36.0)
Treatment Modification Post-Holter
Anticoagulation initiated	41 (20.0)
TOAST Classification
Large artery atherosclerosis	55 (27.5)
Cardioembolism	52 (26.0)
Small artery occlusion	53 (26.5)
Stroke of another determined cause	9 (4.5)
Stroke of undetermined cause	12 (6.0)

Notes: AV: Atrioventricular; LA: Left atrium; LV: Left ventricle; LVEF: Left ventricular ejection fraction; mRS: Modified Rankin Scale; NIHSS: National Institutes of Health Stroke Scale; WPW: Wolff–Parkinson–White syndrome.

Figure 1.

Note: Data expressed as (%).

Figure 2.

Note: WPW: Wolff–Parkinson–White syndrome, data expressed as (%).

AF prevalence increased sharply with age, affecting 90.2% of patients aged ≥60 years compared with 7.3% in the 46–59 years group and 2.4% in those <45 years (p = .002). AF was significantly more frequent in patients with diabetes mellitus (68.3%, p = .01), particularly those with poor glycaemic control (41.5%, p = .006), compared with non-diabetics or those with well-controlled diabetes. Hypertension was also strongly associated with AF (85.4% vs 69.8%, p = .045), and elevated systolic blood pressure ≥140 mmHg on admission was more common among AF patients (80.5% vs 59.1%, p = .01). Smoking showed a significant association with AF (26.8% vs 8.8%, p = .04), whereas alcohol consumption and other forms of substance use did not differ significantly between groups. CKD was markedly more prevalent among AF patients (43.9% vs 10.1%, p = .01). Abnormal carotid vertebral Doppler findings were also associated with AF (31.7% vs 11.9%, p = .007), with higher rates of ischaemic heart disease and valvular heart disease in this group. No significant association was found between AF and diastolic blood pressure, lipid profile abnormalities or prior stroke/TIA (Table 4).

Table 4.

Association of Clinical Parameters with Atrial Fibrillation.

Parameter	AF Present n (%)	AF Absent n (%)	p Value
Age Group
<45 years	1 (2.4)	17 (10.7)	.002**
46–59 years	3 (7.3)	44 (27.7)
≥60 years	37 (90.2)	98 (61.6)
Onset of Symptoms
<12 h	24 (58.5)	94 (59.1)	.23
12–24 h	13 (31.7)	43 (27.0)
24–48 h	2 (4.9)	19 (11.9)
48–72 h	0	2 (1.3)
72–96 h	1 (2.4)	1 (0.6)
96–120 h	1 (2.4)	0
Diabetes Mellitus (DM)
Yes	28 (68.3)	73 (45.9)	.01*
No	13 (31.7)	86 (54.1)
Diabetic Control
Non-DM	13 (31.7)	86 (54.1)	.006**
Good control	11 (26.8)	43 (27.0)
Poor control	17 (41.5)	30 (18.9)
Hypertension
Yes	35 (85.4)	111 (69.8)	.045*
No	6 (14.6)	48 (30.2)
SBP on Admission
<140 mmHg	8 (19.5)	65 (40.9)	.01*
≥140 mmHg	33 (80.5)	94 (59.1)
DBP on Admission
<90 mmHg	32 (78.0)	122 (76.7)	.86
≥90 mmHg	9 (22.0)	37 (23.3)
Substance Use
Smoking	11 (26.8)	14 (8.8)	.04*
Alcohol	6 (14.6)	20 (12.6)	.59
Alcohol and smoking	5 (12.2)	16 (10.1)	.71
ADS	1 (2.4)	15 (9.4)	.67
TDS	1 (2.4)	15 (9.4)	.39
ADS and TDS	0	13 (8.2)	.94
Dyslipidaemia
Abnormal HDL	14 (34.1)	49 (30.8)	.68
HDL (mg/dL) mean ± SD	46.49 ± 14.7	45.79 ± 11.6	.74
Abnormal LDL	7 (17.1)	19 (11.9)	.38
LDL (mg/dL) mean ± SD	100 ± 27.0	103 ± 24.1	.74
Abnormal TG	11 (26.8)	44 (27.7)	.91
TG (mg/dL) mean ± SD	152.56 ± 53.5	155.68 ± 54.3	.74
Chronic Kidney Disease (CKD)
Normal	23 (56.1)	143 (89.9)	.01*
Abnormal	18 (43.9)	16 (10.1)
Carotid Vertebral Doppler
No abnormality	28 (68.3)	140 (88.1)	.007**
IHD	9 (22.0)	15 (9.4)
RHD/Valvular disease	4 (9.8)	4 (2.5)
Prior Stroke/TIA
No	27 (65.9)	126 (79.2)	.07
Yes	14 (34.1)	33 (20.8)

Notes: Data shown as n (%). *p < .05 (significant), **p < .01 (highly significant), ***p < .001 (very highly significant). ADS: Alcohol dependence syndrome; AF: Atrial fibrillation; DBP: Diastolic blood pressure; HDL: High-density lipoprotein; IHD: Ischaemic heart disease; LDL: Low-density lipoprotein; MCA: Middle cerebral artery; PCA: Posterior cerebral artery; SBP: Systolic blood pressure; SD: Standard deviation; TDS: Tobacco dependence syndrome; TG: Triglycerides; TIA: Transient ischaemic attack; RHD: Rheumatic heart disease.

In echocardiography, there was a trend towards more abnormal findings in AF patients compared with non-AF patients (p = .06), with regional wall motion abnormality (RWMA) more frequently observed in the AF group (29.3% vs 12.6%). LAE was strongly associated with AF, with both mild (14.6% vs 0.6%) and moderate–severe enlargement (14.6% vs 1.3%) occurring more frequently in AF patients (p = .01). Median NIHSS score was higher in AF patients [8 (IQR 3–16.5) vs 6 (IQR 3–8), p = .04], with greater representation in the moderate–severe and severe stroke categories. Infarct distribution patterns showed that AF patients were less likely to have anterior circulation infarcts (19.5% vs 49.7%, p = .01) and more likely to have posterior circulation infarcts (48.8% vs 27.0%, p = .008) and multi-territory infarcts (19.5% vs 5.3%, p = .03). No significant association was found with watershed infarcts, mean EF or admission heart rate (Table 5).

Table 5.

Association of Echocardiography (ECHO) Findings, Cardiac Parameters and Stroke Characteristics with Atrial Fibrillation.

Parameter	AF Present n (%)	AF Absent n (%)	p Value
ECHO Findings
Normal	12 (29.3)	60 (37.7)	.06
LVH	9 (22.0)	43 (27.0)
RWMA	12 (29.3)	20 (12.6)
Vegetations	0	1 (0.6)
Cardiomyopathy	2 (4.9)	1 (0.6)
LV clot	1 (2.4)	2 (1.3)
LV diastolic dysfunction	0	15 (9.4)
Valve regurgitation	2 (4.9)	7 (4.4)
LVH + LV diastolic dysfunction	3 (7.3)	10 (6.3)
LA size (mean ± SD, mm)	35.29 ± 8.2	30.4 ± 2.3	.81
ECHO EF (mean ± SD, %)	57.29 ± 8.1	61.26 ± 11.2	.23
Heart rate at admission (mean ± SD)	85.61 ± 12.3	75.59 ± 18.6	.38
EF Group (%)
<30	2 (4.9)	3 (1.9)	.16
30–59	12 (29.3)	30 (18.9)
≥60	27 (65.6)	126 (79.2)
LA Enlargement Group
Mild enlargement	6 (14.6)	1 (0.6)	.01*
Moderate–severe enlargement	6 (14.6)	2 (1.3)
NIHSS Group
No stroke (score 0)	4 (9.8)	15 (9.4)	.03*
Minor stroke (0–4)	11 (26.8)	55 (34.6)
Moderate stroke (5–15)	15 (36.6)	79 (49.7)
Moderate–severe stroke (16–20)	6 (14.6)	7 (4.4)
Severe stroke (21–42)	5 (12.2)	3 (1.9)
Median NIHSS (IQR)	8 (3–16.5)	6 (3–8)	.04*
Anterior Circulation Infarcts
No	33 (80.5)	80 (50.3)	.01*
Yes	8 (19.5)	79 (49.7)
Posterior Circulation Infarcts
No	21 (51.2)	116 (73.6)	.008**
Yes	20 (48.8)	43 (27.0)
Watershed Infarcts
No	40 (97.6)	150 (94.3)	.65
Yes	1 (2.4)	9 (5.7)
Multi-territory Infarcts
No	33 (80.5)	146 (91.8)	.03*
Yes	8 (19.5)	13 (5.3)

Notes: Data presented as n (%). *p < .05 (significant), **p < .01 (highly significant), ***p < .001 (very highly significant). AF: Atrial fibrillation; EF: Ejection fraction; IQR: Interquartile range; LA: Left atrium; LV: Left ventricle; LVH: Left ventricular hypertrophy; NIHSS: National Institutes of Health Stroke Scale; PVC: Premature ventricular complex; RWMA: Regional wall motion abnormality; SD: Standard deviation; SVT: Supraventricular tachycardia.

As per univariate analysis (Table 6), significant risk factors for AF included age ≥60 years (OR 3.25, p = .002), hypertension (OR 2.07, p = .045), diabetes mellitus (OR 2.68, p = .011), poor glycaemic control (HbA1c ≥8%) (OR 3.11, p = .006), CKD with eGFR <60 mL/min/1.73 m² (OR 5.03, p = .001), NIHSS ≥16 (OR 3.86, p = .003), heart rate >100 bpm at admission (OR 6.09, p = .001), frequent ectopics on ECG (OR 3.23, p = .002), moderate–severe LA enlargement (OR 7.07, p = .001), RWMA/LV clot (OR 4.92, p = .001), posterior circulation infarcts (OR 2.79, p = .008), and multi-territory infarcts (OR 4.16, p = .003). Anterior circulation infarcts were negatively associated with AF (OR 0.24, p = .001). Dyslipidaemia, prior stroke/TIA and duration of diabetes were not statistically significant. As per multivariate analysis (Table 7), independent predictors of AF were CKD (eGFR <60 mL/min/1.73 m²) (aOR 4.91, p = .001), RWMA/LV clot (aOR 4.92, p = .036), multi-territory infarcts (aOR 12.35, p = .001), posterior circulation infarcts (aOR 3.74, p = .024), moderate–severe LA enlargement (aOR 5.18, p = .013) and heart rate >100 bpm at admission (aOR 4.78, p = .008).

Table 6.

Univariate Logistic Regression Analysis of Risk Factors for Atrial Fibrillation.

Variable	Odds Ratio (OR)	95% Confidence Interval (CI)	p Value
Age ≥60 years	3.25	1.54–6.85	.002**
Hypertension	2.07	1.01–4.28	.045*
Diabetes mellitus	2.68	1.25–5.72	.011*
Dyslipidaemia	0.70	0.35–1.40	.87
Poor glycaemic control (HbA1c ≥8%)	3.11	1.39–6.98	.006*
Prior stroke/TIA	0.68	0.93–4.10	.09
CKD (eGFR <60 mL/min/1.73 m²)	5.03	1.91–13.23	.001**
NIHSS ≥16	3.86	1.58–9.44	.003**
Heart rate >100 bpm at admission	6.09	2.47–14.99	.001**
Frequent ectopics (ECG)	3.23	1.55–6.73	.002**
LA enlargement (mod–severe)	7.07	2.17–23.00	.001**
RWMA/LV clot	4.92	2.06–11.77	.001**
Posterior circulation infarct	2.79	1.30–6.06	.008*
Multi-territory infarcts	4.16	1.59–10.91	.003**
Anterior circulation infarcts	0.24	0.10–0.59	.001**
Duration of diabetes (years)	1.43	1.10–1.89	.93

Notes: *p < .05 (significant), **p < .01 (highly significant), ***p < .001 (very highly significant). CKD: Chronic kidney disease; ECG: Electrocardiogram; eGFR: Estimated glomerular filtration rate; LA: Left atrium; LV: Left ventricle; NIHSS: National Institutes of Health Stroke Scale; RWMA: Regional wall motion abnormality; TIA: Transient ischaemic attack.

Table 7.

Multivariate Logistic Regression Analysis of Independent Risk Factors for Atrial Fibrillation.

Variable	Adjusted OR	95% Confidence Interval	p Value
CKD (eGFR <60 mL/min/1.73 m²)	4.91	1.27–18.98	.001**
RWMA or LV clot	4.92	1.12–21.47	.036*
Multi-territory infarcts	12.35	2.72–55.99	.001**
Posterior circulation infarcts	3.74	1.19–11.76	.024*
LA enlargement (mod–severe)	5.18	1.41–18.95	.013*
Heart rate >100 bpm at admission	4.78	1.51–15.14	.008*

Notes: Adjusted for age, chronic kidney disease (CKD), diabetes, left atrium (LA) size, wall-motion abnormality, infarct pattern and tachycardia at admission. eGFR: Estimated glomerular filtration rate; LV: Left ventricle; OR: Odds ratio; RWMA: Regional wall motion abnormality.

Discussion

In this prospective cohort, AF after acute ischaemic stroke/TIA was strongly associated with older age, hypertension, diabetes and CKD. AF prevalence increased sharply from age ≥60 years, consistent with population data where atrial fibrosis, conduction slowing and reduced sinus node reserve accumulate with age.⁷ In our cohort, 72% of AF patients were ≥60 years compared to 39% in the non-AF group. Hypertension (AF: 68% vs non-AF: 47%) and diabetes (AF: 44% vs non-AF: 28%) were more prevalent among AF patients, reflecting established links between chronic haemodynamic/metabolic stress and atrial remodelling.⁷ Poor glycaemic control (HbA1c ≥8%) intensified AF risk (AF: 21% vs non-AF: 8%), supporting prior mechanistic work on oxidative stress, endothelial dysfunction and interstitial fibrosis.⁷ CKD showed one of the strongest associations (AF: 26% vs non-AF: 7%), echoing cohort studies demonstrating a graded increase in AF incidence with declining eGFR.^20–22

Moderate-to-severe LAE (AF: 34% vs non-AF: 9%), RWMA (AF: 18% vs non-AF: 5%), and left ventricular thrombus (AF: 8% vs non-AF: 1%) emerged as independent predictors. These findings align with the atrial cardiopathy paradigm, where structural and functional cardiac abnormalities precede AF onset and mark an embolic substrate.^23–25 Although Atrial Cardiopathy and Antithrombotic Drugs in Prevention After Cryptogenic Stroke (ARCADIA)’s neutral results temper anticoagulation without documented AF, structural markers remain valuable for risk stratification and surveillance targeting.²³ Multi-territory infarctions (AF: 12% vs non-AF: 3%) and posterior circulation infarctions (AF: 21% vs non-AF: 9%) were enriched in the AF-positive group. Literature suggests that these topographies are often linked to embolic mechanisms and yield higher AF detection on extended monitoring.^{4, 7 20 23 24} While not diagnostic alone, their presence should heighten suspicion for cardioembolism and prompt more intensive rhythm monitoring.

Admission tachycardia (>100 bpm) independently predicted AF (AF: 17% vs non-AF: 6%), possibly reflecting sympathetic activation or early atrial arrhythmia. Frequent ectopy (>30/hour) was also predictive (AF: 24% vs non-AF: 11%). These simple, readily measurable bedside parameters can be incorporated into early screening algorithms, supporting the escalation of monitoring when present.^25–29 Our 72-h Holter monitoring reclassified 18% of initially non-cardioembolic strokes as cardioembolic and led to the timely anticoagulation initiation. This stepwise yield mirrors landmark European trials: EMBRACE (30-day external monitoring),²⁶ CRYSTAL-AF (implantable loop recorders),^{27, 28} and STROKE-AF,²⁹ all demonstrating that longer surveillance substantially increases AF detection. Guidelines now endorse extended monitoring in cryptogenic stroke/TIA or when clinical suspicion is high^30–34—precisely the risk strata our cohort identifies. Alcohol use (AF: 14% vs non-AF: 7%) and tobacco exposure (AF: 22% vs non-AF: 13%) remain modifiable contributors. Contemporary reviews confirm a dose-dependent association between alcohol and AF, while smoking adds both arrhythmic and vascular risk.^34–37 Addressing these habits complements rhythm surveillance and anticoagulation strategies.

Conclusion

Our findings support a precision-monitoring framework for occult AF detection after stroke/TIA, that integrates systemic risk factors (older age, CKD, hypertension and poor glycaemic control), structural cardiac abnormalities (LAE, regional wall-motion abnormality and left ventricular thrombus), neuroimaging patterns (posterior circulation or multi-territory infarction), and bedside clinical or ECG triggers, such as tachycardia and frequent atrial ectopy. Prolonged Holter monitoring demonstrated a complementary—rather than competitive—role alongside routine ECG, identifying paroxysmal AF in a clinically meaningful proportion of patients without AF on admission ECG, and directly informing anticoagulation decisions. By combining readily available clinical, cardiac and radiological markers, this targeted approach enables prioritisation of extended rhythm monitoring for patients at the highest risk, aligning with contemporary guideline recommendations while remaining pragmatic and cost-effective, particularly in resource-limited settings where indiscriminate use of advanced implantable monitoring is not feasible. Although the single-centre design and modest sample size may limit generalisability and the 72-h monitoring window may have missed rare or ultra-brief arrhythmic episodes, these findings provide a robust, clinically actionable framework to guide selective AF surveillance and support future validation in larger, multicentre cohorts.

Footnotes

Abbreviations

ADS: Alcohol dependence syndrome; AF: Atrial fibrillation; AV: Atrioventricular; BP: Blood pressure; CKD: Chronic kidney disease; CV: Carotid vessel; DBP: Diastolic blood pressure; DM: Diabetes mellitus; ECG: Electrocardiogram; EF: Ejection fraction; ESUS: Embolic stroke of undetermined source; HDL: High-density lipoprotein; HTN: Hypertension; ICM: Insertable cardiac monitor; IEC: Institutional Ethics Committee; IHD: Ischaemic heart disease; IQR: Interquartile range; KMC: Kasturba Medical College; LA: Left atrium; LAE: Left atrial enlargement; LDL: Low-density lipoprotein; LV: Left ventricle; LVEF: Left ventricular ejection fraction; LVH: Left ventricular hypertrophy; mRS: Modified Rankin Scale; NIHSS: National Institutes of Health Stroke Scale; OR: Odds ratio; RWMA: Regional wall motion abnormality; SBP: Systolic blood pressure; SD: Standard deviation; SVT: Supraventricular tachycardia; TG: Triglycerides; TIA: Transient ischaemic attack; TOAST: Trial of ORG 10172 in acute stroke treatment; WPW: Wolff–Parkinson–White syndrome.

Acknowledgements

The authors would like to express their gratitude to all the writers, co-authors and supportive personnel for their invaluable assistance in this project.

Authors’ Contribution

P. Sri Harshitha contributed to study design, patient recruitment, data collection and manuscript drafting.

Nikith Ampar contributed to study conceptualisation, statistical analysis and manuscript revision.

Mukund A. Prabhu contributed to cardiology evaluation, interpretation of Holter data and critical revision of the manuscript.

Arvind N. Prabhu contributed to patient management and data interpretation.

Sharath P. S. supervised the study, contributed to study design and provided final approval of the manuscript.

All authors read and approved the final manuscript.

Data Availability Statement

Includes original data generated.

Statement of Ethics

Ethics committee approval from the Institutional Ethics Committee (IEC) was taken before the start of the study.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

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

ICMJE Statement

Both authors made substantial contributions to the conception, design and execution of the study (CTRI Ref No: REF/2024/03/063476). Dr Harshitha was responsible for study designing, collecting the and performed the analysis and wrote the entire article. Dr Nikith Ampar provided study materials, performed the analysis and reviewed the article. Dr Arvind N Prabhu conceived and designed the analysis, provided critical review, commentary and revision. Dr Sharath and Dr Mukund provided oversight and leadership responsibility, provided critical review, commentary and revision. All authors have reviewed and approved the final version of the manuscript and agree to be accountable for all aspects of the work, ensuring its accuracy and integrity.

Patient Consent

Written informed consent was obtained from all participants or their legally authorised representatives prior to inclusion in the study.

ORCID iD

Nikith Ampar

References

Bhana

, Ojha

, Pandey

, . Prevalence of atrial fibrillation in patients with acute ischemic stroke: a multicentric cohort study. J Clin Diagn Res 2021; 15(9): OC21–OC24.

Arauz

, Morelos

, Colin

, . Comparison of functional outcome and stroke recurrence in patients with embolic stroke of undetermined source (ESUS) vs. cardioembolic stroke patients. PLoS One 2016; 11(11): e0166091.

Gogia

, Hasan

, and Fisher

A review of the current literature and updates on embolic stroke of undetermined source (ESUS). Curr Treat Options Neurol 2022; 24(9): 429–439.

Sanna

, Diener

, Passman

, . Cryptogenic stroke and underlying atrial fibrillation. N Engl J Med 2014; 370(26): 2478–2486.

Brachmann

, Morillo

, Sanna

, . Uncovering atrial fibrillation beyond short-term monitoring in cryptogenic stroke patients: three-year results from the cryptogenic stroke and underlying atrial fibrillation trial. Circ Arrhythm Electrophysiol 2016; 9(1): e003333.

Gladstone

, Dorian

, Spring

, . Atrial premature beats predict atrial fibrillation in cryptogenic stroke: results from the EMBRACE trial. Stroke 2015; 46(4): 936–941.

Bernstein

, Kamel

, Granger

, . Atrial fibrillation in patients with stroke attributed to large-or small-vessel disease: 3-year results from the STROKE AF randomized clinical trial. JAMA Neurol 2023; 80(12): 1277–1283.

World Medical Association. World Medical Association Declaration of Helsinki: ethical principles for medical research involving human participants. JAMA 2025; 333(1): 71–74.

Brott

, Adams HP

, Olinger

, . Measurements of acute cerebral infarction: a clinical examination scale. Stroke 1989; 20(7): 864–870.

10.

van Swieten

, Koudstaal

, Visser

, . Interobserver agreement for the assessment of handicap in stroke patients. Stroke 1988; 19(5): 604–607.

11.

American College of Cardiology and American Heart Association. 2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA guideline for the prevention, detection, evaluation, and management of high blood pressure in adults a report of the American College of Cardiology/American Heart Association Task Force on Clinical practice guidelines. Hypertension 2018; 71(6): E13–E15.

12.

Care

Standards of care in diabetes—2024. Diabetes Care 2024; 47: S77.

13.

Eckel

and Cornier

MA.

Update on the NCEP ATP-III emerging cardiometabolic risk factors. BMC Med 2014; 12(1): 115.

14.

Levey

, Bosch

, Lewis

, . A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Ann Intern Med 1999; 130(6): 461–470.

15.

Binici

, Intzilakis

, Nielsen

, . Excessive supraventricular ectopic activity and increased risk of atrial fibrillation and stroke. Circulation 2010; 121(17): 1904–1911.

16.

Lang

, Bierig

, Devereux

, . Recommendations for chamber quantification. Eur J Echocardiogr 2006; 7(2): 79–108.

17.

Adams HP

, Bendixen

, Kappelle

, . Classification of subtype of acute ischemic stroke. Definitions for use in a multicenter clinical trial. TOAST. Trial of Org 10172 in Acute Stroke Treatment. Stroke 1993; 24(1): 35–41.

18.

Potpara

, Romiti

, and Sohns

The 2024 European Society of Cardiology guidelines for diagnosis and management of atrial fibrillation: a viewpoint from a practicing clinician’s perspective. Thromb Haemost 2024; 124(12): 1087–1094.

19.

Kusumoto

, Schoenfeld

, Barrett

, . 2018 ACC/AHA/HRS guideline on the evaluation and management of patients with bradycardia and cardiac conduction delay: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines and the Heart Rhythm Society. J Am Coll Cardiol 2019; 74(7): e51–e156.

20.

Healey

, Connolly

, Gold

, . Subclinical atrial fibrillation and the risk of stroke. N Engl J Med 2012; 366(2): 120–129.

21.

Oduta Weber-Krüger

, Gelbrich

, Stahrenberg

, . Finding atrial fibrillation in stroke patients: randomized evaluation of enhanced and prolonged Holter monitoring—find-AFRANDOMISED—rationale and design. Am Heart J 2014; 168(4): 438–445.

22.

Kamel

, Longstreth

, Tirschwell

, . Apixaban to prevent recurrence after cryptogenic stroke in patients with atrial cardiopathy: the ARCADIA randomized clinical trial. JAMA 2024; 331(7): 573–581.

23.

Kamel

and Healey

JS.

Cardioembolic stroke. Circ Res 2017; 120(3): 514–526.

24.

Chen

, Lee

, Weng

, . Identification of magnetic resonance imaging features for the prediction of unrecognized atrial fibrillation in acute ischemic stroke. Front Neurol 2022; 13: 952462.

25.

Sharobeam

, Churilov

, Parsons

, . Patterns of infarction on MRI in patients with acute ischemic stroke and cardio-embolism: a systematic review and meta-analysis. Front Neurol 2020; 11: 606521.

26.

Powers

, Kam

, Ritter

, . Diagnostic accuracy of acute infarcts in multiple cerebral circulations for cardioembolic stroke: literature review and meta-analysis. J Stroke Cerebrovasc Dis 2020; 29(7): 104849.

27.

Azeem

, Nagy

, Miller

, . Prevalence of a multiple territory stroke pattern after intravenous thrombolysis. J Stroke Cerebrovasc Dis 2020; 29(5): 104700.

28.

Nouh

, Remke

, and Ruland

Ischemic posterior circulation stroke: a review of anatomy, clinical presentations, diagnosis, and current management. Front Neurol 2014; 5: 30.

29.

Wachter

, Gröschel

, Gelbrich

, . Holter-electrocardiogram-monitoring in patients with acute ischaemic stroke (Find-AF_RANDOMISED): an open-label randomised controlled trial. Lancet Neurol 2017; 16(4): 282–290.

30.

Hindricks

, Potpara

, Dagres

, . 2020 ESC guidelines for the diagnosis and management of atrial fibrillation developed in collaboration with the European Association for Cardio-Thoracic Surgery (EACTS): the Task Force for the diagnosis and management of atrial fibrillation of the European Society of Cardiology (ESC) developed with the special contribution of the European Heart Rhythm Association (EHRA) of the ESC. Eur Heart J 2021; 42(5): 373–498.

31.

Kleindorfer

, Towfighi

, Chaturvedi

, . 2021 guideline for the prevention of stroke in patients with stroke and transient ischemic attack: a guideline from the American Heart Association/American Stroke Association. Stroke 2021; 52(7): e364–e467.

32.

Powers

, Rabinstein

, Ackerson

, . 2018 guidelines for the early management of patients with acute ischemic stroke: a guideline for healthcare professionals from the American Heart Association/American Stroke Association. Stroke 2018 Mar; 49(3): e46–99.

33.

Kotalczyk

, Lip

, and Calkins

The 2020 ESC guidelines on the diagnosis and management of atrial fibrillation. Arrhythm Electrophysiol Rev 2021; 10(2): 65–67.

34.

Vosko Voskoboinik

, Kalman

, De Silva

, . Alcohol abstinence in drinkers with atrial fibrillation. N Engl J Med 2020; 382(1): 20–28.

35.

George

and Figueredo

VM.

Alcohol and arrhythmias: a comprehensive review. J Cardiovasc Med 2010; 11(4): 221–228.

36.

Voskoboinik

, Prabhu

, Ling

, . Alcohol and atrial fibrillation: a sobering review. J Am Coll Cardiol 2016; 68(23): 2567–2576.

37.

Fitzgerald

, Middeldorp

, Gallagher

, . Lifestyle modification and atrial fibrillation: critical care for successful ablation. J Clin Med 2022; 11(9): 2660.