Investigating the Patent Foramen Ovale–Migraine Link in Southern Saudi Arabia

Summary

Objective:

Migraine is a highly disabling neurological condition, and increasing attention has focused on a possible connection with patent foramen ovale (PFO).

Methods:

This single-center observational study enrolled 45 individuals from January to April 2024: 30 migraine patients with PFO (Group 1) and 15 migraine patients without PFO serving as controls (Group 2). All subjects underwent transthoracic echocardiography and completed standardized instruments including the Migraine Disability Assessment (MIDAS), Headache Impact Test (HIT-6), Migraine Symptom Severity Scale (MSSS), Allodynia Symptom Checklist (ASC-12), Visual Aura Rating Scale (VARS), and Migraine Treatment Optimization Questionnaire (mTOQ-6).

Results:

The participant's mean age was 37.3 ± 9 years. More than half experienced migraine with aura, a proportion notably greater among those with PFO. A positive family history, suboptimal response to therapy, and positional exacerbation were significantly associated with the PFO group. Fibrinogen concentrations and several migraine-related scores including MIDAS, HIT-6, MSSS, ASC-12, and mTOQ-6 differed significantly between groups.

Conclusion:

Migraine linked to PFO are more likely to have an aura, a stronger family patterns, worsening symptoms based on position, and overall, greater disability and increased severity of symptoms. These findings support an association between PFO and migraine characteristics and may have implications for clinical management.

Migraine is a global health issue affecting around 15–20% of the worldwide population. It's characterized by recurring headaches, sensory disturbances, and occasional aura, but the mechanisms driving the condition are not well understood [1].

A small opening called the foramen ovale, which allows blood flow between the right and left atria, is present in the fetal heart. However, the foramen ovale typically seals after birth, remaining open in 25% of the population. This is the leading cause of a right-to-left shunt in adults, and a patent foramen ovale (PFO) acts as a shunt between venous and arterial components of the circulatory system. As a result, it can lead to cross-system paradoxical emboli and strokes, decompression sickness in divers, and migraine with aura. In adults, the main cause of right-to-left shunting is PFO [2].

Patients with migraine with aura (MA) have been reported to have an elevated PFO prevalence in observational studies and clinical trials) [1]. Although the primary endpoints of the PRIMA and PREMIUM trials were not achieved, secondary outcomes were optimistic for MA patients [3,4].

In the European Heart Journal, a meta-analysis reported PFO closure with added medical therapy decreased the frequency of migraine attacks and the total number of migraine days compared to medical therapy alone. The authors acknowledged the absence of complete migraine remission and the presence of selective benefit in subgroups [5]. Currently, the study RELIEF is attempting to identify patients most likely to benefit from closure [6].

This study aimed to evaluate migraine characteristics in patients with PFO. Although there is a growing interest in the association between PFO and migraine, the investigation of migraine characteristics in PFO patients is very limited. The majority of the available research discusses the prevalence of PFO in migraine patients and/or the effects of PFO closure, rather than examining subtypes and other migraine-specific clinical characteristics.

Methods

Participants

This observational study was conducted at a single center at the Armed Forces Hospital in the Southern Region (AFHSR) in Khamis Mushait, Kingdom of Saudi Arabia, recruiting participants from January to April 2024 through its outpatient headache and cardiology clinics.

Participants aged 18-60 with PFO confirmed by transesophageal echocardiography (TEE) were included, while those with hypertension, diabetes, secondary headache disorders, depression, or anxiety were excluded.

This study involved 45 individuals: 30 migraine sufferers with PFO (Group 1) and 15 without PFO as control subjects (Group 2). Every patient completed a comprehensive neurological assessment, which included assessment of headache triggers, headache relief mechanisms, and review of the relevant medical history. Participants were also classified according to the age of migraine onset: before or after 50 years.

No formal sample size determination or statistical power calculation was performed. The investigation was exploratory in nature and relied on the availability of patients during the recruitment period.

Migraine diagnosis, with/without aura, followed ICHD III criteria [7]. Chronic migraine was defined as having headache days ≥15 per month over the preceding three months, with migraine features on ≥8 days.

This study complies with the guidelines laid out by the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) Statement for observational studies. Study protocols were approved by the AFHSR Ethical Committee of the Institutional Review Board of the Armed Forces Hospital in the Southern Region. Informed consent from participants was obtained before the beginning of the study.

Materials

In this research, multiple validated tools were employed to evaluate migraine, we used the Head Impact Test (HIT-6) [8]. HIT-6 is a 6-item questionnaire that measures impact and loss of function due to headaches, weighing pain intensity and impact on cognitive and social roles.

The Migraine Disability Assessment (MIDAS) questionnaire measures the self-reported headache burden over the last 3 months by measuring the degree of functional disability and severity of migraines, providing a rough estimate of the burden migraines impose on daily life and overall functioning [9].

Patient Health Questionnaire 4 (PHQ-4) was used to assess mood disorders. PHQ-4 is a validated 4-item measure of anxiety and depression that yields a total score of psychological distress [10,11].

The study further evaluated migraine frequency and symptom severity utilizing the Migraine Symptom Severity Score (MSSS) to quantify the overall burden and intensity of migraine-related symptoms [12].

Pain-related cutaneous allodynia was assessed using the Allodynia Symptom Checklist (ASC-12), a 12-item checklist measuring the frequency of hypersensitivity to normally non-painful stimuli and classifying headache episodes in from none to severely painful allodynia [13].

The Visual Aura Rating Scale (VARS) was used to assess visual aura, and VARS is a validated tool to assess visual migraines. A score of 5 or more on VARS highly suggests a migraine aura [14].

Migraine Treatment Optimization Questionnaire (mTOQ-6) is a six-item validated questionnaire designed to assess the effectiveness and optimization of acute migraine treatment. It evaluates key domains such as rapid pain relief, restoration of functional ability, and prevention of headache recurrence. Total scores range from 0 to 6, with higher scores indicating better treatment optimization and greater therapeutic efficacy, whereas lower scores reflect suboptimal response and unmet clinical needs [15].

Brain imaging was performed using a Siemens Somatom Emotion 16-CT scanner (Germany) and/or a Siemens MRI Essenza machine (1.5 Tesla, Germany) to exclude intracranial lesions.

Echocardiography

A comprehensive transthoracic echocardiography was performed using multiple imaging windows and Color Doppler to assess septal anatomy and detect shunts. Agitated saline was injected through a peripheral vein, with coughing or Valsalva maneuvers used to enhance microbubble passage. A PFO was confirmed when bubbles appeared in the left atrium within three cardiac cycles of injection [16].

Controls (Group 2) underwent transthoracic (TTE) rather than transesophageal echocardiography (TEE) to identify patent foramen ovale (PFO), likely underdiagnosing small/occult PFO defects and creating misclassification bias reducing intergroup differences. This decision was made for participant comfort but precludes sensitive detection; future studies should use TEE screening in all participants.

Statistical analysis

Data was analyzed using IBM SPSS Statistics version 27 (IBM Corp., Armonk, NY, USA). Continuous variables were averaged and summarized with standard deviations, and categorical variables were summarized in percentages. Associations between categorical variables were tested with chi square tests. NRS, VARS, MIDAS, and HIT 6 were compared using Student's t test.

All statistical tests were two-tailed, with a significance threshold of p< 0.05. Assumptions for parametric testing (normality and homogeneity of variance) were checked prior to analysis.

As various comparisons were made regarding headache characteristics (Table 2) and PFO size subgroups (Table 4), the Benjamini–Hochberg (BH) False Discovery Rate (FDR) method was used to manage the expected amount of false positive findings. An FDR of 5% was used to adjust the p values. Both raw and FDR adjusted p values are presented. Statistical significance was determined to be the case for the FDR adjusted values.

Table 2.

Headache characteristics in migraine with vs without patent foramen ovale (PFO).

Headache subtypes	Total	Migraine with PFO (N = 30)	Migraine without PFO (control) (N = 15)	P-Value	False Discovery Rate (FDR) adjusted p values
Marital state				0.584	0.584
Single	8	5 (16.7%)	3 (20.0%)
Married	35	23 (76.7%)	12 (80.0%)
Divorced	2	2 (6.7%)	0 (0.0%)
Family history of migraine (First and /or 2nd degree)	20	17 (56.7%)	3 (20.0%)	0.02	0.047
Migraine onset
Early Onset less than 50 years	41	29 (96.7%)	12 (80.0%)	0.064	0.095
Late Onset above 50 years	4	1 (3.3%)	3 (20.0%)
Precipitants
No precipitants	11	5 (16.7%)	6 (40.0%)
Emotional stress	4	3 (10.0%)	1 (6.7%)
Exercise	4	3 (10.0%)	1 (6.7%)
Valsalva Maneuver or related conditions	18	16 (53.3%)	2 (13.3%)	0.068	0.095
Sleep disturbances	7	3 (10.0%)	4 (26.7%)
Odors	1	0 (0.0%)	1 (6.7%)
Positional aggravation (bending forward, straining or lying down)		15 (50.0%)	2 (13.3%)	0.017	0.047
Aura
Migraine without Aura	21	10 (33.3%)	11 (73.3%)
Visual Aura	18	16 (53.3%)	2 (13.3%)	0.018	0.047
Sensory aura	6	4 (13.3%)	2 (13.3%)
Migraine course				0.088	0.103
Episodic Migraine	26	20 (66.7%)	6 (40.0%)
Chronic Migraine	19	10 (33.3%)	9 (60.0%)

Table 4.

Migraine severity scores in patients with versus without patent foramen ovale (PFO).

				95% Confidence Interval of the Difference
Headache subtypes	Migraine with PFO (Case) (N = 30)	Migraine without PFO (control) (N = 15)	P-Value	Lower	Upper
Age (mean ± SD)	36.10 ± 8.976	39.80 ± 9.104	0.201	-2.051	9.451
NRS (Mean ± SD)	7.47 ± 1.592	6.276.27 ± 1.438	0.018	-2.184	-0.216
MIDAS (Mean ± SD)	27.13 ± 12.770	12.80 ± 6.293	<0.001	-21.403	-7.264
HIT6 (Mean ± SD)	56.67 ± 12.427	48.20 ± 10.262	0.028	-15.970	-0.963
MSSS (Mean ± SD)	16.43 ± 5.722	6.27 ± 3.173	<0.001	-13.378	-6.955
VARS (Mean ± SD)	2.93 ± 1.946	1.13 ± 1.767	0.004	-3.005	-0.595
PHQ-4 (Mean ± SD)	0.77 ± 1.331	3.73 ± 3.127	<0.001	1.632	4.301
ASC-12 (Mean ± SD)	8.87 ± 4.855	3.60 ± 2.898	<0.001	-8.019	-2.514
mTOQ-6 (Mean ± SD)	3.93 ± 3.016	7.07 ± 3.011	0.002	1.211	5.056
Frequency of headache / month (mean ± SD)	5.00 ± 2.505	3.80 ± 2.242	0.125	-2.745	0.345
Duration of each episode (in days) (mean ± SD)	2.57 ± 1.251	2.40 ± 1.352	0.684	-0.986	0.653
MHDs per Month (Mean ± SD)	11.67 ± 5.833	9.80 ± 5.784	0.316	-5.576	1.843
Fibrinogen (Mean ± SD)	482.4 ± 182.6	331.6 ± 94.32	0.005	-252.4	-49.1

Data are presented as the mean ± standard deviation.

ASC-12: The 12-item Allodynia Symptom Checklist; GAD-7: General Anxiety Disorder-7;HIT-6: Headache Impact Test-6;MHDs: Migraine headache days; MIDAS: Migraine Disability Assessment Scale MSSS: Migraine Symptom Severity Score;mTOQ: Migraine Treatment Optimization Questionnaire; NRS: Numerical Rating Scale;PHQ-4: Patient Health Questionnaire-4; PFO: patent Foramen Ovale; VARS: Visual Aura Rating Scale

Results

The study cohort comprised 45 participants, of whom 30 (66.7%) were classified in Group 1 (migraine with PFO) and 15 (33.3%) in Group 2 (migraine without PFO). The average age of the participants was 37.3 ± 9 years.

The demographic analysis of the participants revealed that the distribution of gender among participants suggested a predominance of females, who constituted 55.6%, whereas males accounted for 44.4% of the sample. Most participants (41, 91.1%) experienced migraine onset at an early age. Migraine with aura were highly prevalent among the participants, with over half reporting aura during attacks; visual aura was the most common type (18 participants, 40%). (Table 1).

Table 1.

Headache characteristics of study participants.

Variables	Frequency
Sex
Male	20 (44.4)
Female	25 (55.6)
Migraine Headache onset
Early onset <50 years	41 (91.1)
Late onset≥50 years	4 (8.9)
Migraine Subtypes
Migraine without Aura	23 (51.1)
Migraine with Aura	22 (48.9)
Precipitants
No Precipitants	11 (24.4)
Emotional stress	4 (8.9)
Exercise	4 (8.9)
Valsalva Maneuver or related conditions	18 (40.0)
Sleep disturbances	7 (15.6)
Odors	1 (2.2)
Positional aggravation (bending forward, straining or lying down)	17(37.8)
Migraine subtypes
Migraine without aura (MO)	21 (46.7)
Visual Aura	18 (40.0)
Sensory aura	6 (13.3)
Patent Foramen Ovale (PFO) categories*
Small to Medium PFO	7 (23.3)
Large PFO	23 (76.7)
Operated PFO	14(46.7)
PFO complicated with ischemic stroke	11 (36.7)

^*Small-to-medium PFO was defined as <4.0 mm, while large PFO was defined as ≥4.0 mm.

Aggravation of symptoms during Valsalva maneuvers or related conditions was reported in 16 patients (53.3%) with migraine and PFO and in 2 patients (13.3%) with migraine without PFO, (P = 0.068 & FDR adjusted p-value = 0.095). (Table 2). In addition, symptom aggravation during Valsalva maneuvers was observed in 2 patients (28.6%) with small to medium PFO and in 14 patients (60.9%) with large PFO, (P = 0.19 & FDR adjusted p value = 0.45) (Table 5).

Table 5.

Clinical, headache phenotype, and laboratory characteristics according to patent foramen ovale (PFO) size.

	Small to Medium PFO N = 7	Large PFO N = 23	P- Value	FDR adjusted p values
Sex			0.6	0.72
Male	3 (42.9%)	12 (52.2%)
Female	4 (57.1%)	11 (47.8%)
Marital state			0.24	0.48
Single	0 (0.0%)	5 (21.7%)
Married	7 (100.0%)	16 (69.6%)
Divorced	0 (0.0%)	2 (8.7%)
Family history of migraine (First and /or 2nd degree)	4 (57.1%)	13 (56.5%)	0.97	0.57
Migraine Onset			0.57	0.72
Early Onset less than 50 years	7 (100.0%)	22 (95.7%)
Late Onset above 50 years	0 (0.0%)	1 (4.3%)
Precipitants			0.374	0.64
No precipitants	2 (28.6%)	3 (13.0%)
Emotional stress	0 (0.0%)	3 (13.0%)
Exercise	2 (28.6%)	1 (4.3%)
Valsalva Maneuvers (VM) related conditions	2 (28.6%)	14 (60.9%)	0.19	0.45
Sleep disturbances	1 (14.3%)	2 (8.7%)
Positional aggravation (bending forward, straining or lying down)	1 (14.3%)	14(60.9%)	0.03	0.18
Migraine Subtype			0.05	0.20
Migraine without Aura	4 (57.1%)	6 (26.1%)
Visual Aura	1 (14.3%)	15 (65.2%)
sensory aura	2 (28.6%)	2 (8.7%)
Migraine course			0.76	0.82
Episodic Migraine	5 (71.4%)	15 (65.2%)
Chronic Migraine	2 (28.6%)	8 (34.8%)
Migraine Disability (by MIDAS)			0.13	0.39
Mild	2 (28.6%)	2 (8.7%)
Moderate	3 (42.9%)	5 (21.7%)
Severe	2 (16)	16 (69.6%)
Response to abortive therapy by (mTOQ)
Poor and very poor response	1 (14.3%)	16 (69.6%)	0.015	0.18
Moderate and max efficacy	6 (85.7%)	7 (30.4%)
PFO complicated with ischemic stroke	2 (28.6%)	9 (39.1%)	0.6	0.72
Fibrinogen level (Mean ± SD)	339.7 ± 339.7	525.8 ± 178.1	0.015	95% CI (-333.8–38.35

Small-to-medium PFO was defined as <4.0 mm, while large PFO was defined as ≥4.0 mm.

MIDAS: Migraine Disability Assessment Scale;mTOQ: Migraine Treatment Optimization Questionnaire; PFO: patent Foramen Ovale.

Positional aggravation (bending forward, straining, or lying down) was reported in 15 patients (50.0%) with migraine and PFO and in 2 patients (13.3%) with migraine without PFO, (P = 0.017 & FDR adjusted p-value = 0.047). (Table 2). Furthermore, positional aggravation was observed in 1 patient (14.3%) with small to medium PFO and in 14 patients (60.9%) with large PFO (P = 0.03 & FDR adjusted p-value = 0.18) (Table 5).

The results indicated a higher prevalence of family migraine history among patients with PFO (56.7%) compared to those without PFO (20%), with a statistically significant difference (p = 0.02 & FDR adjusted p-value = 0.047). A significant difference was also found between migraine patients with and without PFO regarding the presence of visual aura (p = 0.018 &FDR adjusted p-value = 0.047) (Table 2).

Migraine characteristics associated with PFO were analyzed using a Multivariable logistic regression model adjusted for age and gender. The model was significant (χ²(5) = 12.686, p = 0.027), showed good calibration (Hosmer-Lemeshow p = 0.823), 34.1% of variance explained (Nagelkerke R² = 0.341), and 73.3% classification accuracy (83.3% for PFO-positive; Table 3). Family history of migraine trended toward PFO positivity (OR = 4.04, 95% CI: 0.82–19.91, p = 0.087), as did visual aura (OR = 5.22, 95% CI: 0.71–38.54, p = 0.105) and positional aggravation (OR = 2.92, p = 0.259); age (p = 0.515), and gender (p = 0.397) were nonsignificant.

Table 3.

Multivariable logistic regression analysis of factors associated with migraine with PFO versus migraine without PFO.

Predictor	B	SE	Wald	df	P-Value	OR	95% CI for OR
Sex (Male)	0.685	0.810	0.716	1	.397	1.98	0.41 – 9.70
Age	–0.030	0.045	0.424	1	.515	0.97	0.89 – 1.06
Family history of migraine	1.395	0.814	2.937	1	.087	4.04	0.82 – 19.91
Positional aggravation	1.073	0.950	1.275	1	.259	2.92	0.45 – 18.80
Visual aura	1.652	1.020	2.623	1	.105	5.22	0.71 – 38.54
Constant	0.345	1.878	0.034	1

PFO = patent foramen ovale; OR = odds ratio; CI = confidence interval.

Migraine-related scores, including MIDAS, HIT-6, MSSS, ASC-12, and mTOQ-6 differed significantly between patients with and without PFO based on our analysis. Furthermore, fibrinogen levels showed a statistically significant difference between the groups (Table 4).

Poor treatment response (Lower scores mTOQ-6) was observed in 16 of 23 migraine patients with large PFO (69.6%), whereas only 1 of 7 patients with small to medium PFO (14.3%) showed poor response (P = 0.01 & FDR adjusted p value = 0.18) (Table 5).

Moreover, the study suggested that the presence of a large PFO was significantly associated with greater migraine severity and a higher prevalence of visual aura, as reflected by elevated Migraine Symptom Severity Score (MSSS) and Visual Aura Rating Scale (VARS) scores (p = 0.013 and p = 0.032, respectively). Additionally, serum fibrinogen levels were significantly higher among patients with large PFO compared to those with small to medium PFO (p = 0.015) (Table 6).

Table 6.

Migraine severity scores by PFO size.

Migraine Characteristics & fibrinogen level	Small to Medium PFO N = 7	Large PFO N = 23	P value	95% Confidence Interval
				Lower	Upper
Age (Mean ± SD)	38.86 ± 6.793	35.26 ± 9.512	0.362	-4.361	11.553
NRS (Mean ± SD)	7.71 ± 2.215	7.39 ± 1.406	0.646	-1.104	1.750
MIDAS (Mean ± SD)	19.14 ± 14.3	29.57 ± 11.5	0.057	-21.183	0.338
HIT6 (Mean ± SD)	63.71 ± 6.4	54.52 ± 13.1	0.087	-1.409	19.794
MSSS (Mean ± SD)	11.86 ± 8.2	17.83 ± 3.9	0.013	-10.570	-1.368
VARS (Mean ± SD)	1.57 ± 1.8	3.35 ± 1.8	0.032	-3.387	-0.166
PHQ-4 (Mean ± SD)	1.14 ± 1.4	.65 ± 1.301	0.403	-.692	1.673
ASC-12 (Mean ± SD)	6.14 ± 3.8	9.70 ± 4.9	0.090	-7.699	0.594
mTOQ-6 (Mean ± SD)	4.86 ± 2.4	3.65 ± 3.1	0.364	-1.469	3.879
Frequency of headache per month (mean ± SD)	7 ± 3.71	23 ± 5.39	0.123	-3.836	0.482
Duration of each episode (in days) (mean ± SD)	2.29 ± .488	2.65 ± 1.402	0.507	-1.483	0.750
MHDs / Month (Mean ± SD)	8.57 ± 2.992	12.61 ± 6.199	0.110	-9.048	0.974

Data are presented as the mean ± standard deviation.

ASC-12: The 12-item Allodynia Symptom Checklist;GAD-7: General Anxiety Disorder-7;HIT-6: Headache Impact Test-6;MHDs: Migraine headache days; MIDAS: Migraine Disability Assessment Scale;MSSS: Migraine Symptom Severity Score;mTOQ: Migraine Treatment Optimization Questionnaire; NRS: Numerical Rating Scale;PHQ-4: Patient Health Questionnaire-4;VARS: Visual Aura Rating Scale.

Discussion

This study was conducted at a single center and involved a relatively small sample size (30 PFO-positive vs. 15 PFO-negative participants).

The relationship between migraine, more specifically migraine with aura and PFO, has been studied extensively. Most studies show that PFO is more common in migraine populations, particularly those with migraine with aura, however, the exact relationship and the clinical significance remains controversial [2,17].

In migraine and PFO Group (Group 1), early-onset migraine (defined as onset before 50 years of age) was reported in 29 participants (96.7%), compared with 12 participants (80.0%) in Group 2 (migraine without PFO). Conversely, late-onset migraine was observed in 1 participant (3.3%) in Group 1 and in 3 participants (20.0%) in Group 2. The difference between groups did not reach statistical significance (p = 0.06). These findings are consistent with prior literature indicating that migraine typically manifests before the age of 50 in most affected individuals, with late-onset cases being relatively uncommon [18]. The near-significant trend observed here may reflect the established epidemiological pattern, but larger sample sizes would be required to confirm whether group differences are robust.

In our participants, a family history of migraine involving first- and/or second-degree relatives was reported in 20 participants. This was significantly more frequent among individuals with migraine and PFO, where 17 participants (56.7%) had a positive family history, compared with only 3 participants (20.0%) in the migraine without PFO group (p = 0.02 & FDR adjusted p value = 0.047). This finding is supported by studies suggesting that familial aggregation of migraine may be more pronounced in PFO-associated cases, potentially reflecting an interaction between genetic predisposition and right-to-left shunt mechanisms [1,19].

Visual aura was present in 16 patients (53.3%) in Group 1, compared to only 2 patients (13.3%) in Group 2. This difference was statistically significant (p = 0.018 & FDR adjusted p-value = 0.047). In addition, mean Visual Aura Rating Scale (VARS) score was significantly lower in patients without PFO (1.13 ± 1.767) compared to those with PFO (2.93 ± 1.946), with statistical significance (p = 0.004). This finding is consistent with prior reports demonstrating a strong association between migraine with aura and the presence of PFO, particularly in cases involving large shunts [20].

Out of 30 migraine patients with PFO, 23 (76.7%) had large PFO (≥4.0 mm) and 7 (23.3%) had small-to-medium PFO (<4 mm). Within the subgroup of patients with large PFO, visual aura was observed in 65.2%, compared with 14.3% among those with smaller PFO (p = 0.05 & FDR adjusted p value = 0.2). While this comparison approached conventional statistical thresholds, the limited sample size likely constrained the study's statistical power, increasing the risk of Type II error. Consequently, even moderate differences between groups may not have achieved statistical significance, highlighting the importance of larger, adequately powered investigations to validate these preliminary findings.

Several mechanisms have been proposed to explain the potential association between PFO and migraine, but these remain speculative and were not directly assessed in this study. Observational evidence suggests that PFO is more prevalent among patients with migraine with aura, and larger right-to-left shunts appear to correlate more strongly with migraine occurrence. For instance, a community-based study in China reported a significant association between large shunts and migraine without aura [21]. However, randomized controlled trials evaluating PFO closure as a therapeutic strategy have consistently failed to demonstrate meaningful reductions in migraine frequency or complete remission, leaving the clinical relevance of the PFO-migraine link unresolved [3,4,22].

There are several biological rationales that have been posited to explain the PFO and migraine link. Right to left shunting allows venous microthrombi and platelet clumps to escape the filtering of the lungs and to enter the systemic arterial circulation. If these micro-emboli reach the cerebral circulation, they may result in micro-infarcts or areas of under perfusion which trigger cortical spreading depression (CSD). This is a phenomenon that is well accepted and is involved in the pathogenesis of migraine auras and may even induce the headache itself [23]. The lungs also filter and metabolize vasoactive substances, (serotonin, bradykinin, and prostaglandins). A PFO allows these substances which have not been metabolized to reach the brain in higher concentrations which may increase neuronal hyper-excitability, alter the tone of cerebral vessels and activate inflammatory pathways that are associated with the pathophysiology of migraine. Unfiltered serotonin is known to directly impact cerebral vessels and pathways associated with migraine [24]. Research suggests that there may be a genetic basis for the association of PFO and migraine in certain families, although the association in the study is not indicative of causation [17].

Concerning migraine course, Group 1 had episodic migraine in 66.7% (20 cases) while Group 2 had it in 40% (6 cases). Chronic migraine was noted in 10 cases (33.3%) in Group 1 and 9 cases (60%) in Group 2 (p = 0.08). Previous research reported that PFO occurred more in individuals with episodic migraine with aura (MA) than in chronic migraine sufferers or those without migraine [25]. The existing literature provides minimal evidence regarding the association between PFO size and the incidence of migraine. One such study suggested that there was a lack of evidence to support proof of the claim that patients with larger PFOs have higher incidences of migraine, and were associated in no way with migraine with aura (MA) [26]. On the other hand, one study described patients with chronic migraine and found there to be a higher incidence of PFOs, regardless of the presence of an aura. However, the lack of uniformity in the methodology of studies, and the inconsistency in the classification and description of the characteristics of the shunt, limits the ability to interpret and compare these findings [27].

Symptom aggravation associated with Valsalva maneuver or related conditions was more common in patients with PFO than those without, even if the observed difference was not statistically significant (53.3% vs. 13.3%; p = 0.068 & FDR adjusted p-value = 0.095). Furthermore, this tendency was more evident when the analysis was focused on PFO size, with aggravation reported in 60.9% of patients with large PFOs vs. 28.6% of patients with small to medium-sized PFOs (p-value = 0.19 & FDR adjusted p-value = 0.45). Even though statistically significant differences were not found, the evident tendency suggests that larger PFOs associate with more pronounced aggravation to symptoms during Valsalva maneuver. Along similar lines, earlier investigations demonstrated that 45.8% of migraine with aura patients cited Valsalva maneuver as one of the triggering factors [28]. Valsalva maneuver are known to increase intrathoracic as well as right atrial pressure [29], and thus, may lead to a temporary right to left shunt. Taking these observations, it is fair to assume that hemodynamic alterations during Valsalva maneuver may interact with PFOs to influence migraine symptoms.

Interestingly, positional aggravation of migraine symptoms (such as bending forward, straining, or lying down) appeared to be more prevalent among patients with PFO compared to those without. While only 13.3% of patients without PFO reported such aggravation, the proportion rose to 50% among patients with PFO (P = 0.017& FDR adjusted p value = 0.047). Moreover, the likelihood of positional aggravation seemed to increase with PFO size, being reported in 14.3% of patients with small to medium PFOs and in 60.9% of those with large PFOs (p = 0.03& FDR adjusted p value = 0.18). These findings suggest that PFO presence and particularly larger defects may contribute to positional sensitivity in migraine, potentially reflecting altered hemodynamic or neurovascular mechanisms. A resting-state fMRI study demonstrated altered neural network connectivity in patients with migraine and PFO compared to those without, suggesting that PFO-related physiological changes may influence migraine symptom expression. This supports the idea that positional or hemodynamic stressors could exacerbate symptoms in PFO-positive patients [30].

In the primary comparison of migraineurs with vs. without PFO (n = 45; 30 vs. 15, respectively), family history of migraine, positional aggravation and visual aura remained statistically significant after Benjamini–Hochberg FDR correction (p = 0.047 for all), suggesting that the observed associations were methodologically and statistically consistent. By contrast, in the PFO-positive subgroup (n = 30; 7 small-to-medium vs. 23 large), these associations did not remain statistically significant after FDR adjustment, perhaps reflecting lack of statistical power (small sample sizes). These findings support our overall findings linking PFO with migraine features and highlight the need for larger adequately powered studies to clarify these associations across shunt size.

Across all evaluated scales including the Numerical Rating Scale (NRS), Migraine Disability Assessment (MIDAS), Headache Impact Test (HIT-6), Migraine Severity Scale Score (MSSS), and Allodynia Symptom Checklist (ASC-12) patients with PFO exhibited notable distinctions compared with those without PFO. The broader literature examining the relationship between PFO and migraine intensity remains limited and somewhat inconsistent: while certain studies have reported reductions in pain intensity following PFO closure, others have found no association between PFO status and self-reported headache severity [31].

Standardized tools for measuring migraine impact include MIDAS and HIT-6, and in patients with moderate to large PFOs these tend to be higher than in patients with small PFOs or no PFO [24]. PFO closure is increasingly recognized as an effective treatment for certain patients with MA as it has been reported to significantly decrease the frequency of headaches and the number of migraine days [32]. It has been shown that patients with moderate or large PFOs have an increase in migraine attack frequency and elevated MIDAS and HIT-6 scores; set closure is associated with improvement in VARS, HIT-6, MIDAS, and reduced attack duration in some studies [1].

In interpreting the subgroup analyses according to PFO size (Table 5 and Table 6), it should be emphasized that the small-to-medium PFO subgroup comprised only seven participants. The observed differences in MSSS and VARS scores between small-to-medium and large PFOs, although statistically significant, are therefore highly vulnerable to both type I (false-positive) and type II (false-negative) errors and should be regarded strictly as preliminary and hypothesis-generating. These findings require cautious interpretation and confirmation in larger, adequately powered cohorts before any firm conclusions can be drawn about the impact of PFO size on migraine characteristics.

The ASC-12 showed that allodynia impacts about 63% of migraine sufferers, and this tracks with severity, disability, and symptom burden, which correlates migraine pathophysiology [33].

Group 1 participants had significantly lower mTOQ-6 scores than Group 2, with means of 3.9 ± 3 vs. 7 ± 3, respectively (p = 0.002). This result is in line with previous study, in a cohort of 239 migraineurs, the subgroup with PFO showed worse headache prognosis at 1 month (17.5% vs. 47.5% good response) and 3 (20.0% vs. 45.0%; P < 0.05) due to right-to-left shunt microembolism resistant to β-blockers or CGRP antibodies [34].

Multivariable logistic regression analysis showed a trend toward higher odds of a patent foramen ovale (PFO) in migraine patients with a positive family history (OR = 4.04, 95% CI: 0.82–19.91, p = 0.087), visual aura (OR = 5.22, 95% CI: 0.71–38.54, p = 0.105), and positional aggravation (OR = 2.92, p = 0.259), after adjusting for age and sex. Although these associations were not statistically significant, the effect sizes suggest the possibility of a relationship between migraine phenotype and presence of PFO. Such a trend should be interpreted with caution, however, especially in light of the small sample size in this pilot study (n = 45), there were only 30 with PFO versus 15 without. This may have contributed to the limited power of this study to detect a statistically significant difference.

In migraine patients with PFO, PHQ-4 scores were lower (less psychological distress) (0.77 ± 1.33 vs. 3.73 ± 3.13, p<0.001) despite higher severity/disability of headache. This is in contrast to the known comorbidity of migraine and anxiety/depression [35,36]. As few studies have stratified psychiatric comorbidity by PFO status our data may suggest that the PFO-migraine might be a phenotype and biologically distinct (aura-rich, episodic, Valsalva/positional, hypercoagulable) with less affective component (psychiatric comorbidities) than non-PFO migraine where disability is more associated with psychiatric characteristics, in part, suggested by higher prevalence of episodic migraine in PFO (66.7% vs. 40.0%) and literature of chronic migraine and psychiatric disorders [37]. Small sample (n = 45) and limited PHQ-4 data warrant confirmation by larger studies, preferably with longitudinal design and more sophisticated instruments (e.g. PHQ-9, GAD-7).

The mean fibrinogen level for PFO patients (482.4 ± 182.6) was significantly higher than for non-PFO patients (331.6 ± 94.3) (p = 0.005). Coagulation and fibrinogen, an acute-phase inflammatory marker, have been linked. A number of studies have investigated PFO patients and the triad of inflammation, hypercoagulability, and migraine [38]. There are a few studies that show migraine and circulating levels of fibrinogen have a positive correlation. More specifically, migraine with aura patients show increased levels of fibrinogen, along with other biomarkers that indicate the activation of the coagulation pathway. The information given complies with further support for the proposed theory of migraine pathophysiology. Migraine are accompanied by injury and inflammation to the tissues, and that is reflected by the levels of inflammatory indices such as fibrinogen and other inflammatory markers that are also elevated [39,40].

Study limitations include: (1) single-center design, small sample (n = 30 PFO+ vs. 15 PFO-), no power calculation, risk of Type II errors; (2) transthoracic echocardiography only, potential occult PFO in controls and misclassification/selection bias; (3) Subgroup analyses comparing small-to-medium (n = 7) versus large PFOs (n = 23) warrant cautious interpretation due to limited statistical power, elevating risks of Type I/II errors. Nominal associations (e.g., MSSS, VARS) likely reflect chance, as none survived Benjamini-Hochberg FDR adjustment, rendering them exploratory and hypothesis-generating for larger studies.

Future research should therefore focus on larger, multi-center cohorts with systematic PFO screening, formal power calculations, and correction for multiple testing. Longitudinal designs incorporating advanced imaging modalities, particularly 4D flow MRI, may provide deeper insights into shunt hemodynamics and their relationship to migraine burden, building directly on the associations observed in this study.

Regardless of these limitations, the study provides novel regional data from Southern Saudi Arabia, thus characterizing migraine features within a Middle Eastern region and adding to the international data. In addition to prevalence estimates, the study translates migraine burden functionally with validated instruments (MIDAS, HIT-6, MSSS, ASC-12, mTOQ-6) that assess disability, burden, and response to treatment in migraine patients with and without PFO. Interestingly, prothrombotic or inflammatory mechanisms coupled with migraine burden were identified with elevated fibrinogen levels in PFO positive migraine patients.

In conclusion, the present study supports that patients with migraine who have a PFO especially those who experience visual auras and have a positive family history, are more likely to suffer greater symptom aggravation with physical activities that are associated with increased intrathoracic pressure and changes in venous return. Patients with both PFO and migraine experience more headache-related pain and disability as seen by increased scores in (MIDAS, HIT-6, MSSS). Even with the standard abortive treatments, many patients are still unable to attain satisfactory symptom relief, and a considerable number describe symptoms of cutaneous allodynia. In addition, migraine patients who also have a PFO have elevated levels of fibrinogen, indicating that there might be a prothrombotic state or inflammation that contributes to the clinical symptoms.

Acknowledgments

The authors would like to thank Falcon Scientific Editing (https://falconediting.com) for proofreading the English language in this paper.

Disclosure

In this manuscript, artificial intelligence (AI) tools were used the least amount possible. More specifically, any AI-assisted tools were used for technical formatting, editing, and minor language revisions. AI tools were not used to: (1) obtain any significant scientific content or clinical insight; (2) obtain, alter, or produce any of the images or figures included in this work; (3) gather, organize, or administrate any of the data pertaining to the study; (4) conduct any statistical testing or data review. All of the authors’ work that went into the research design, data management, statistical testing, clinical insights, and writing the manuscript was done without the use of AI tools. The authors take complete accountability for the scientific credibility and integrity of this manuscript.