Helicobacter pylori (HP), a bacterium that colonizes the gastrointestinal mucosa, is prevalent in about 50% of the global population. Generally acquired early in life, it seldom resolves spontaneously, potentially persisting for a lifetime.^1,2 HP plays a significant role in various gastric ailments such as gastritis, peptic ulcers, and gastric cancer. Recent research has broadened the known impact of HP beyond gastrointestinal disorders, associating it with conditions including hematological^3,4 and cardiovascular diseases,^5,6 asthma, allergies,^7,8 and neurological ailments.⁸

Stroke ranks as the second global leading cause of death, characterized by high mortality rates and lasting neurological impairments.⁹ It accounts for approximately 6.5 million deaths annually, significantly affecting the quality of life due to associated disabilities.¹⁰ Inflammation and atherosclerosis largely influence the progression and outcomes of ischemic stroke, a prevalent subtype of stroke.¹¹

Growing evidence suggests that HP might be involved in conditions such as coronary heart disease and ischemic stroke.^12-15 Several reports also indicate that HP eradication could prevent coronary heart disease.^16,17 A common factor in these conditions is atherosclerosis, although the exact mechanism of HP-induced atherosclerosis is yet to be elucidated. The relationship between HP and ischemic stroke is a topic of ongoing debate.^18-24 Variations in existing study results may be attributed to the inherent difficulties in controlling a diverse range of confounding variables that influence ischemic stroke incidences, which are known to fluctuate based on factors such as age and sex. Notably, sex hormones, including estrogen and testosterone, play intricate roles in modulating cardiovascular and neuroendocrine system functions.²⁵ Consequently, the observed discrepancies regarding the influence of HP eradication on ischemic stroke incidence could stem from sex-specific dynamics. Recent data have underscored a decline in coronary heart disease incidence, predominantly among young men and elderly women, post-HP eradication.¹⁶

Given these findings, we hypothesized that sex might be a potential confounding factor in assessing the effects of HP eradication on ischemic stroke prevention. This study aimed to investigate the potential sex-specific influences of HP eradication on ischemic stroke incidence, utilizing data from the Seoul National University Bundang Hospital (SNUBH) ischemic stroke registry.

1. Study population

This prospective observational cohort study included patients who underwent esophagogastroduodenoscopy and were tested for HP infection at SNUBH in Korea, from May 2003 through February 2023. There were two main gastroenterology cohorts. That is, the first one was a prospectively registered group of asymptomatic individuals concerned about their health, who voluntarily participated in a comprehensive HP study, including culture. The second cohort consisted of patients with mild gastritis, simple gastric polyps, a family history of gastric cancer, or mild atrophy/intestinal metaplasia. Patients under the age of 20, those with a history of gastric cancer, those who had previously received HP eradication therapy, or had a history of atrial fibrillation or ischemic stroke were excluded. After applying these exclusion criteria, 6,664 patients remained for analysis. Of these, 4,671 were confirmed to have HP infection and were designated as the HP-positive group. The remaining 1,993 patients, without confirmation of HP infection, were classified as the HP-naïve group.

Subsequent categorization within the HP-positive group was based on the eradication status after treatment, which occurred following enrolment. A total of 3,706 patients who sought and successfully achieved HP eradication posttreatment were classified into the HP-eradicated group. The remaining 965 patients either refused eradication treatment or did not achieve eradication following treatment and were thus included in the non-eradicated group. Approximately 48% were lost to follow-up, about 30% of these patients voluntarily refused treatment, around 17% underwent treatment but did not achieve eradication, and about 5% were not strongly recommended for eradication therapy due to age or urgent health status.

To adjust for potential confounders, propensity score (PS) matching was employed. After PS matching, 2,803 patients were identified in the HP-eradicated group, and 960 in the non-eradicated group (Fig. 1).

Figure 1. Schematic flow of the study. HP, Helicobacter pylori.

2. Diagnosis and treatment of HP infection

Patients were diagnosed of HP infection through more than one of histology, the Campylobacter-like organism (CLO) test (Delta West, Bentley, Australia), or culture.²⁶ To ensure consistency, a sole endoscopist (N.K.) collected a total of four biopsy samples: at least two samples in the mid-antrum and another at least two samples in the mid-corpus. Both antral and corpus specimens, at least two from each location, were cultured at 37°C in a microaerophilic environment for a duration of 3 to 5 days. Additionally, two specimens from each section were stained with modified Giemsa and hematoxylin and eosin. These stained samples were preserved in 10% neutral-buffered formalin and embedded in paraffin for HP presence assessment using the updated Sydney system scoring. All other specimens were used for the CLO test.²⁶

If any of the three tests (histology, rapid urease test, and culture) were positive, the patient was said to have an HP infection. HP eradication followed contemporary guidelines.^27,28 Prior to 2012, it involved a 7-day regimen of twice-daily administration comprising 40 mg of esomeprazole, 1,000 mg of amoxicillin, and 500 mg of clarithromycin. Post-2012, patients with HP infection were prescribed a 10-day sequential therapy: the initial 5 days consisted of 40 mg of esomeprazole and 1,000 mg of amoxicillin twice daily, followed by another 5-day course of 40 mg of esomeprazole, 500 mg of clarithromycin, and 500 mg of metronidazole twice daily. Confirmation of HP eradication was determined using the ¹³C-urea breath test (UBT) (UBiT-IR300; Otsuka Electronics, Hirakata, Japan) 4 to 6 weeks after finishing the eradication regimen. Patients who did not respond to the initial therapy were subjected to either a 14-day quadruple therapy or a 14-day moxifloxacin-based triple therapy. The quadruple therapy regimen involved twice-daily doses of 40 mg of esomeprazole, four-times-daily doses of 300 mg of tripotassium dicitrate bismuthate (Denol; Greencross Co., Seoul, Korea), three-times-daily doses of 500 mg of metronidazole, and four-times-daily doses of 500 mg of tetracycline. As an alternative, the 14-day moxifloxacin-based triple therapy included twice-daily administration of 400 mg of moxifloxacin (Avelox; Bayer Health Care, AG, Wuppertal, Germany), 40 mg of esomeprazole, and 1,000 mg of amoxicillin. The evaluation of HP status was performed using histology and/or the CLO test during each follow-up endoscopic surveillance, typically occurring at intervals of 1 or 2 years.

3. Behavioral factors and previous medical history

Information regarding patients' past medical history and behavioral factors was collected through patient-completed questionnaires, electronic medical records from SNUBH, and clinical data repositories. The data sources were used to gather comprehensive details on patients' medical history and behavioral factors. Patients were diagnosed with hypertension if their pretreatment blood pressure was equal to or higher than 140/90 mm Hg or if they received treatment for hypertension according to the guidelines set by the Korean Society of Hypertension. The diagnosis of dyslipidemia was determined based on the serum concentrations meeting the criteria specified in the Korean Guidelines for the Management of Dyslipidemia 4th edition (total cholesterol ≥240 mg/dL, low-density lipoprotein cholesterol ≥160 mg/dL, triglyceride ≥200 mg/dL, or high-density lipoprotein cholesterol [HDLc] <40 mg/dL) or if the patient was currently undergoing treatment for dyslipidemia. The diagnosis of diabetes mellitus was established based on the presence of relevant medical history in the patient's records or the use of hypoglycemic medication to manage blood sugar levels. We also investigated aspirin usage, a factor relevant to ischemic stroke incidence. The majority of participants used aspirin primarily for primary prevention, linked to risk factors like high blood pressure, diabetes, or family cardiovascular history. Around 15% took it for secondary prevention post-cardiovascular procedures, while 4% used it for conditions such as pulmonary or rheumatic diseases. For about 13%, the reasons for aspirin use were unspecified.

4. Study outcome and follow-up

The primary outcome of this study was the incidence of new-onset ischemic stroke. Patients were prospectively monitored until either the occurrence of an ischemic stroke or until the study's endpoint in February 2023, whichever came first. Those lost to follow-up were excluded from subsequent analyses. The diagnosis of ischemic stroke was confirmed by a board-certified neurologist, based on clinical evaluation of acute cerebral infarction symptoms, and corroborated by neuroimaging techniques such as computed tomography or magnetic resonance imaging. To ensure the exclusion of cardioembolic causes with limited relevance to HP, patients with a history of atrial fibrillation were excluded from the study. Our analysis only included stroke events that occurred at least 30 days post-eradication. To enhance the precision of our findings, we conducted a thorough cross-check of the identified ischemic stroke cases with the SNUBH ischemic stroke registry, which has been built in the Department of Neurology since 2003.

5. Statistical analysis

Two separate analyses were performed: one comparing the HP-eradicated group with the HP-naïve group, and another comparing the HP-eradicated group with the HP non-eradicated group. For each analysis, the two groups were matched using PSs. The HP-eradicated group versus HP-naïve group were matched at a 1:1 ratio, and the HP-eradicated group versus HP non-eradicated group were matched at a 3:1 ratio. The nearest neighbor was selected with a caliber of 0.10. PS was calculated for all participants using logistic regression, adjusting for clinically relevant covariates including age, sex, hypertension, diabetes mellitus, dyslipidemia, aspirin intake, drinking habits, and smoking habits to reduce the effect of potential confounding variables on ischemic stroke. Baseline data were presented as mean ± standard deviation or median (interquartile range) for continuous variables, while categorical variables were presented as n (percentages). Comparative analyses employed the chi-square test or Fisher exact test for categorical variables, and either the Student t-test or the Mann-Whitney U test for continuous variables.

Disease-free survival probabilities were plotted using Kaplan-Meier curves, with statistical comparisons conducted using the log-rank test. The hazard ratio (HR) and its corresponding 95% confidence interval (CI) for the risk of incident ischemic stroke were calculated through both univariable and multivariable Cox proportional hazards models. Subgroup analyses were also conducted based on sex category. Statistical significance was set at a two-tailed p-value of less than 0.05. To ensure robust statistical analysis, we calculated a required sample size of 2,100 participants. This calculation is based on 1.3% of ischemic stroke incidence rate among Koreans aged 19 years or older in 2014, as previously reported,²⁹ aiming for an 80% power at a 0.050 significance level to detect a HR of 0.5270, assuming a control group survival proportion of 0.9505. All statistical analyses were performed using R software version 4.2.3 (R Foundation for Statistical Computing, Vienna, Austria; https://www.r-project.org/).

6. Ethics statement

The study was conducted in accordance with the principles of the Declaration of Helsinki. All participants provided written informed consent in the prospective cohort (B-0602-030-001, B-0903-071-001). In addition, we received approval from the Institutional Review Board of the SNUBH (IRB number: B-2302-810-003).

1. Study cohort and baseline characteristics

The baseline characteristics of HP-naïve, HP eradicated, and non-eradicated group and their PS-matched counterparts are presented in Table 1 and Supplementary Table 1. In this cohort of 6,664 patients evaluated for HP infection, 29.9% (n=1,993) were categorized as HP-naïve, and 70.1% (n=4,671) as HP-positive. Among the HP-positive patients, 79.3% (n=3,706) achieved eradication and were classified as the HP-eradicated group, while 20.7% (n=965) did not achieve eradication, thus constituting the non-eradicated group. At enrolment, the mean age was 56.1±13.0 years, with females representing 56.7% of the cohort (n=3,781).

Table 1 . Baseline Characteristics of the Study Subjects Depending on HP Eradication.

Characteristic	Total group				Propensity score-matched group (3:1 ratio)
	HP eradicated (n=3,706)	HP non-eradicated (n=965)	p-value		HP eradicated (n=2,803)	HP non-eradicated (n=960)	p-value
Age, yr	55.0±12.0	55.3±14.3	0.597		54.7±12.6	55.2±14.3	0.365
Male sex	1,740 (47.0)	426 (44.1)	0.128		1,253 (44.7)	426 (44.4)	0.890
Alcohol			0.031				0.298
Non-drinker	2,939 (79.3)	796 (82.5)			2,265 (80.8)	791 (82.4)
Drinker	767 (20.7)	169 (17.5)			538 (19.2)	169 (17.6)
Smoking			0.394				0.867
Non-smoker	3,128 (84.4)	803 (83.2)			2,353 (83.9)	803 (83.6)
Current smoker	578 (15.6)	162 (16.8)			450 (16.1)	157 (16.4)
Comorbidity
Diabetes mellitus	230 (6.2)	70 (7.3)	0.267		184 (6.6)	68 (7.1)	0.631
Hypertension	580 (15.7)	135 (14.0)	0.220		375 (13.4)	135 (14.1)	0.631
Dyslipidemia	860 (23.2)	206 (21.3)	0.237		574 (20.5)	206 (21.5)	0.548
Aspirin intake	339 (9.1)	87 (9.0)	0.949		252 (9.0)	85 (8.9)	0.951

Data are presented as mean±SD or number (%)..

HP, Helicobacter pylori..

In our study cohort, notable differences were observed in the follow-up duration and the number of hospital visits between the HP eradicated and non-eradicated groups. The non-eradicated group exhibited a significantly shorter follow-up duration (32.30±44.29 months) compared to the eradicated group (50.73±35.30 months, p<0.001). Additionally, the number of hospital visits was lower in the non-eradicated group, averaging 14.19±40.42 visits, as opposed to 19.91±23.49 visits in the eradicated group (p<0.001) (Supplementary Table 2).

Prior to PS matching, differences between the HP eradicated and non-eradicated groups were not significant, except for a higher prevalence of alcohol consumption in the eradicated group. The HP-naïve group had a higher mean age and greater frequencies of hypertension, dyslipidemia, and aspirin use, and a lower proportion of males compared to the HP-eradicated group.

After PS matching at ratios of 3:1 for eradicated versus non-eradicated patients and 1:1 for eradicated versus naïve patients, the resulting groups were well-balanced, demonstrating no significant disparities. Baseline characteristics across PS-matched groups were comparable, indicated by a standardized mean difference of less than 0.2 (Supplementary Fig. 1).

2. Association of HP eradication with ischemic stroke

In our study encompassing 6,664 participants, a total of 50 individuals experienced ischemic strokes during the follow-up period. The shortest interval observed in months post-eradication to ischemic stroke was 1.31, and the longest was 192.17 months. The mean interval was 59.74 months, with a standard deviation of 45.50 months, and the 95% CI ranged from 50.93 to 68.54 months. All of them were checked in the SNUBH ischemic stroke registry. Utilizing the Kaplan-Meier method, we assessed the cumulative incidence of ischemic strokes. Our analysis indicates that there was no statistically significant difference in the cumulative incidence of ischemic strokes between the groups who underwent HP eradicated and those who did not, a finding consistent irrespective of the implementation of PS matching, as depicted in Fig. 2. Consistent with this, no statistical significance was observed between the HP-eradicated group and the HP-naïve group, regardless of the application of PS matching (Supplementary Fig. 2).

Figure 2. Comparison of ischemic stroke-free rate depending on Helicobacter pylori (HP) eradication. The data revealed that there were no statistically significant differences in the ischemic stroke-free rates between the patients with and without HP eradication, both before (A) and after (B) propensity score (PS) matching. To evaluate the cumulative ischemic stroke-free rate, Kaplan-Meier estimates were utilized. Subsequently, the log-rank test was employed to compute p-values and determine whether there were any notable differences between the two groups.

3. Univariate and multivariate analyses of potential risk factors for ischemic stroke

We conducted univariate and multivariate analyses to evaluate the potential risk factors for ischemic stroke between the HP eradicated and non-eradicated groups following PS matching, as delineated in Table 2. Preliminary univariate analysis identified age, sex, smoking habits, diabetes mellitus, hypertension, and aspirin intake as variables associated with an augmented risk of ischemic stroke. These variables were subsequently incorporated into the multivariate analysis to further scrutinize their potential roles as independent risk factors.

Table 2 . Univariate and Multivariate Analysis of Potential Risk Factors for Ischemic Stroke in Patients with HP Infection after Propensity Score Matching.

Variable	Univariate analysis			Multivariate analysis
	HR (95% CI)	p-value		HR (95% CI)	p-value
HP non-eradicated	Reference			Reference
HP eradicated	0.535 (0.225–1.269)	0.156		0.531 (0.221–1.270)	0.157
Age	1.056 (1.018–1.096)	0.004		1.027 (0.986–1.070)	0.199
Sex	3.085 (1.197–7.952)	0.020		1.813 (0.570–5.765)	0.314
Alcohol	1.795 (0.742–4.345)	0.195
Smoking	3.328 (1.412–7.844)	0.006		1.517 (0.525–4.379)	0.441
Diabetes mellitus	6.892 (2.854–16.640)	<0.001		3.076 (1.108–8.538)	0.031
Hypertension	8.399 (3.382–20.860)	<0.001		5.016 (1.802–13.960)	0.002
Dyslipidemia	2.190 (0.922–5.200)	0.076
Aspirin	2.765 (1.072–7.131)	0.035		0.667 (0.217–2.049)	0.479

HP, Helicobacter pylori; HR, hazard ratio; CI, confidence interval..

Upon multivariate analysis, only diabetes mellitus and hypertension emerged as significant determinants, correlating with a 3.076 (95% CI, 1.108 to 8.538; p=0.031) and 5.016 (95% CI, 1.802 to 13.960; p=0.002) fold increase in the risk of incident ischemic stroke, respectively. In contrast, HP eradication did not exhibit a significant association with a reduced risk of ischemic stroke development.

4. Association of HP eradication with ischemic stroke depending on sex

To elucidate the potential sex-specific associations between HP eradication and the incidence of ischemic stroke, we conducted a subgroup analysis of the cumulative incidence rates using the Kaplan-Meier method, as depicted in Fig. 3. In the female group, although not reaching statistical significance, there was a discernible trend indicating a reduced incidence of ischemic stroke in the HP-eradicated group in comparison to the non-eradicated group (p=0.057) (Fig. 3A). Conversely, the male group did not exhibit a significant difference in the cumulative incidence of ischemic stroke between the HP eradicated group and the non-eradicated group (Fig. 3B).

Figure 3. Comparison of ischemic stroke-free rates between the patients depending on Helicobacter pylori (HP) eradication after propensity score matching, with a focus on sex differences. It was found that HP eradication had no observable advantage in terms of the ischemic stroke-free rate, in both female (A) and male (B) subgroups. To assess the cumulative ischemic stroke-free rate, Kaplan-Meier estimates were employed, and the p-value was calculated using the log-rank test.

5. Univariate and multivariate analyses of potential risk factors for ischemic stroke according to sex

Table 3 shows the outcomes of sex-specific subgroup analyses conducted via both univariate and multivariate analyses. In the female group, age, diabetes mellitus, hypertension, and dyslipidemia emerged as statistically significant determinants in the univariate analysis. However, the subsequent post-PS matching multivariate analysis identified only hypertension as an independent risk factor for ischemic stroke (HR, 6.065; 95% CI, 1.354 to 21.157; p=0.018). HP eradication did not manifest a significant effect in reducing the incidence of ischemic stroke.

Table 3 . Univariate and Multivariate Analysis of Potential Risk Factors for Ischemic Stroke in Patients with HP Infection According to Sex after Propensity Score Matching.

Variable	Univariate analysis			Multivariate analysis
	HR (95% CI)	p-value		HR (95% CI)	p-value
Female
HP non-eradicated	Reference			Reference
HP eradicated	1.761 (0.322–9.639)	0.514		0.485 (0.135–1.743)	0.267
Age	1.109 (1.036–1.186)	0.003		1.064 (0.989–1.146)	0.096
Alcohol	0.749 (0.094–5.922)	0.784
Diabetes mellitus	6.088 (1.571–23.620)	0.009		1.194 (0.254–5.608)	0.822
Hypertension	13.130 (3.386–50.960)	<0.001		6.065 (1.354–27.157)	0.018
Dyslipidemia	5.434 (1.532–19.270)	0.009		1.730 (0.394–7.586)	0.467
Aspirin	1.077 (0.136–8.502)	0.944
Male
HP non-eradicated	Reference			Reference
HP eradicated	0.728 (0.240–2.216)	0.576		0.717 (0.232–2.217)	0.563
Age	1.054 (1.012–1.098)	0.012		1.027 (0.979–1.076)	0.267
Alcohol	1.508 (0.597–3.810)	0.385
Smoking	1.946 (0.752–5.035)	0.170
Diabetes mellitus	5.086 (1.970–13.130)	<0.001		1.920 (0.688–5.358)	0.212
Hypertension	15.350 (4.434–53.150)	<0.001		11.396 (3.018–43.026)	<0.001
Dyslipidemia	2.416 (0.958–6.089)	0.062
Aspirin	2.636 (0.939–7.420)	0.066

HP, Helicobacter pylori; HR, hazard ratio; CI, confidence interval..

Similarly, the univariate analysis in the male group indicated age, diabetes mellitus, and hypertension as statistically significant contributors to ischemic stroke. In alignment with the findings in women, the post-PS matching multivariate analysis underscored hypertension as the sole independent risk factor for ischemic stroke (HR, 11.396; 95% CI, 3.018 to 43.026; p<0.001). Here too, HP eradication did not significantly diminish the incidence of ischemic stroke.

6. Impact of new comorbidities on HP eradication and ischemic stroke incidence

During the extended follow-up period of our study, we rigorously assessed the emergence and impact of newly diagnosed comorbidities, namely hypertension, diabetes mellitus, and dyslipidemia, as well as the potential sex-based disparities. Our analysis indicated that the incidence rates of hypertension, diabetes mellitus, and dyslipidemia were uniformly low and exhibited consistency across all study groups and within each sex subgroup. Through chi-square testing, we ascertained that the p-values consistently exceeded the 0.05 threshold. This consistency, thoroughly documented in Supplementary Tables 3 and 4, reinforces the premise that newly diagnosed comorbidities during the follow-up period and any potential sex-based differences had a minimal influence on the correlation between HP eradication and the risk of ischemic stroke.

7. HP eradication and its impact on lipid and inflammatory markers

In the retrospective analysis undertaken to augment our understanding of the impact of HP eradication on lipid and inflammatory markers, we observed significant changes in HDLc and C-reactive protein (CRP) levels post-eradication. Specifically, our study revealed a statistically significant increase in HDLc levels within the first year following HP eradication, with a p-value of 0.016. This trend, however, did not persist beyond the 1-year mark. Further stratification of the data in a subgroup analysis based on sex indicated notable increases in HDLc levels for both male (p=0.044) and female (p=0.038) participants within the same timeframe. Furthermore, in females, there was a significant rise in CRP levels 1 year after eradication (p=0.010). The results of this additional analysis which aims to describe trends in these parameters are detailed in Supplementary Table 5 and the subgroup analysis by sex is detailed in Supplementary Table 6.

As a prospective observational study encompassing 6,664 participants over a longitudinal span of up to 20 years, we found that the eradication of HP does not significantly alter the incidence of ischemic stroke during a long-term follow-up. However, there was a suggestive trend towards reduced ischemic stroke incidence among females post-HP eradication, though this did not reach statistical significance. This tendency highlights a possible sex-specific effect, which may have profound implications for understanding the role of HP management in stroke prevention strategies for females.

Although the precise mechanism remains speculative, the role of HP infection as a potential contributing factor in the onset of ischemic stroke cannot be disregarded. This potential causality might be attributed to its capacity to incite inflammation, engender chronic infectious states, induce endothelial injury, and alter lipid metabolism, supplemented by an elevation in homocysteine levels.^30,31 The prevailing hypothesis suggests that chronic HP infection may activate immune responses, triggering the release of inflammatory agents such as interleukin-6, tumor necrosis factor-α, and fibrinogen. This persistent inflammation might foster a series of changes conducive to atherosclerotic progression.^6,32-36 Furthermore, prolonged infection seems to facilitate platelet aggregation and exacerbate atherosclerotic instability, potentially escalating the levels of von Willebrand factor, thereby modulating lipid profiles and possibly heightening the risk of ischemic stroke.^37-39

Based on these theories, we hypothesized that HP eradication might temper these cytokine dynamics, consequently reducing ischemic stroke incidence. However, our findings did not corroborate this hypothesis, indicating no significant decrement in ischemic stroke incidence post-eradication. This outcome invites a more detailed discussion through the contrasting results from the meta-analyses by Yu et al.¹⁹ and Wang et al.⁴⁰ While the meta-analysis of Yu et al., which included only 10 prospective observational studies, found no significant association between HP infection and ischemic stroke, Wang et al.'s analysis of 13 case-control studies indicated a significant relationship. This divergence likely stems from the methodological differences in the included studies. Case-control studies, with their retrospective nature, are susceptible to selection bias and may face challenges in establishing a precise causal link between HP exposure and stroke outcomes. In contrast, prospective observational studies, despite certain limitations when compared to randomized control trials, offer a more reliable framework for examining how risk factors such as HP infection may influence specific outcomes due to their ability to track participants over time and observe the emergence of outcomes.

Meanwhile, despite the prevalent hypothesis that HP infection contributes to stroke risk through mechanisms like chronic inflammation and atherosclerosis, both the study by Yu et al.¹⁹ and our research have presented unexpected results. These findings suggest that the pathogenesis and prevention of ischemic stroke are influenced by a multi factors,⁴¹ implying that HP eradication may be inadequate to reverse chronic metabolic disturbances often associated with conditions such as diabetes mellitus or hypertension. Consequently, the clinical impact of HP eradication might be insufficient to significantly alter the trajectory or prevention of ischemic stroke. Additionally, it is speculated that the influence of HP infection is particularly pronounced during the initial atherosclerotic phase, predating the clinical onset of ischemic stroke.^42-44 Regrettably, the exact onset of atherosclerosis is difficult to detect, hindering its inclusion as a definitive outcome in our study.

Nonetheless, our study unveiled a notable trend: HP eradication appeared to exhibit a sex-specific impact on the risk of ischemic stroke, particularly in females. This finding aligns with growing attention to sex disparities in cerebrovascular diseases. Specifically, the incidence of stroke varies across different life stages for males and females. At young age, females experience a higher incidence of stroke compared to their male counterparts. This trend reverses during middle age, when males exhibit a higher incidence than females. Intriguingly, the risk for females escalates rapidly upon entering menopause, ultimately becoming comparable to or exceeding that of males in later life. This disparity may be influenced by hormonal differences.^45,46 In addition, recent research has revealed that HP infection may serve as an important independent risk factor for intracranial atherosclerosis in females aged 60 or younger.⁴⁷ Furthermore, given that females generally have a longer life expectancy compared to males, they are predisposed to an increased cumulative risk of ischemic stroke over their lifespan.⁴⁸

Interestingly, existing literature suggests that HP eradication efforts tend to foster a rise in HDLc levels in female, a trend not mirrored in their male counterparts.^49,50 The clinical implications of elevated HDLc levels are profound, with high HDLc concentrations known to facilitate the repair of endothelial cell damage by rejuvenating endothelial progenitor cells and enhancing their functionality.^51,52 Thus, in light of the hormonal influence of estrogen, it seems plausible that females might derive greater benefits from HP eradication, potentially owing to favorable modifications in lipid metabolism that could potentially mitigate the risk of ischemic stroke.

This study boasts several significant strengths that underscore the reliability of its findings. Firstly, our study, to the best of our knowledge, holds significant value as the first prospective observational study meticulously utilizing data from the ischemic stroke registry at a tertiary medical center to explore whether HP eradication can influence the reduction of ischemic stroke risk. Furthermore, it delves into analyzing potential sex differences in this effect, offering a comprehensive view of the impact of HP eradication on ischemic stroke across different demographics. In an attempt to specifically pinpoint ischemic strokes that are associated with atherosclerosis, we took the prudent step of excluding patients exhibiting atrial fibrillation–a known predisposing factor for cardioembolic events⁵³–right from the enrolment phase. Notably, preceding research efforts exploring the nexus between HP and ischemic stroke have predominantly leaned on health insurance claim data, deciphered through International Classification of Diseases codes.^18,19,23,5 This approach, however, encounters limitations in terms of reliability, primarily because it sometimes encompasses individuals who have been assigned diagnostic codes based on suspicion rather than concrete clinical confirmation. Additionally, a substantial number of these studies do not distinctly segregate the cohort of ischemic stroke patients whose condition originates from cardioembolic sources, a category that maintains a tenuous link with HP infection.^, Secondly, the expansive scope of this study, involving a cohort of 6,664 patients monitored over a substantial duration–extending up to 20 years, with an average span of 64.4 months–lends considerable weight to the robustness of our findings. Lastly, the authenticity of our results is further cemented by the stringent diagnostic criteria employed to ascertain HP eradication, which involved repetitive assessments via urea breath test and histology/CLO test, thereby ensuring the veracity of the outcome data.

This study has also limitations. Firstly, the study’s single-center cohort design introduces an inherent limitation, as it could not account for ischemic stroke events occurring outside the purview of the designated institution, possibly resulting in an underestimation of event rates. Secondly, the absence of detailed, longitudinal data on lifestyle factor changes, such as medication, smoking habits, and alcohol consumption. Furthermore, we did not track changes in quantitative parameters such as body mass index, hemoglobin A1c, serum glucose, and lipid profile, which has inevitably constrained the depth of our findings in these areas. This issue parallels challenges seen in research exploring the link between blood pressure variability and stroke recurrence.⁵⁶ We acknowledge this data collection gap and its potential implications for the analysis conducted using the Cox model. Thirdly, to ascertain the impact of eradication on ischemic stroke, we analyzed alterations in lipid profile and CRP post-eradication. This observation partially concurs with previous research that documented an increase in HDLc following eradication.⁴⁹ However, our study did not identify a sex-specific difference in HDLc increase post-eradication. The retrospective nature of lipid profile and CRP data collection introduces additional layers of complexity and potential bias. The variability in follow-up periods among participants, even within the same subgroup, and the reliance on non-standardized criteria for conducting tests further complicates the interpretation of our findings. These factors contribute to constrain the broader applicability of our study outcomes. Lastly, the lower follow-up duration and fewer hospital visits in the non-eradicated group suggest a potential for selection bias. However, even if this bias led to underreported ischemic stroke events in the non-eradicated group, our main conclusion—that HP eradication does not significantly reduce stroke risk—remains supported by the overall data trend. This factor should be considered in the context of our main conclusions. Additionally, it is anticipated that drug adherence might be lower in the non-eradicated group, which could lead to an underestimation of the impact of dyslipidemia, hypertension, and diabetes on ischemic stroke.

In conclusion, our comprehensive prospective observational cohort study indicates that the eradication of HP may not significantly alter the risk of ischemic stroke development. Nonetheless, a suggestive trend emerged in the data, pointing towards a potential diminution in ischemic stroke risk among females post-HP eradication. This finding warrants further in-depth investigation to elucidate a more definitive correlation and potentially unlock new avenues in stroke prevention.

This work was supported by grants (06-2020-0184 and 02-2020-0041) from the Seoul National University Bundang Hospital Research Fund. In addition, this research was supported by the Korea Center for Gendered Innovations for Science and Technology Research (GISTeR), through the Center for Women In Science, Engineering and Technology (WISET) funded by the Ministry of Science and ICT (No. WISET202203GI01).

The authors would like to thank the Division of Statistics in the Medical Research Collaborating Center at Seoul National University Bundang Hospital for help with the statistical analyses.

No potential conflict of interest relevant to this article was reported.

Study concept and design: N.K. Data acquisition: N.K. Data analysis and interpretation: E.B.J. Drafting of the manuscript: E.B.J. Critical revision of the manuscript for important intellectual content: B.J.K., I.C.H., S.B.K. Statistical analysis: H.Y., C.M.S., Y.C., J.H.K., S.A., D.H.L. Obtained funding: N.K. Administrative, technical, or material support; study supervision: Y.K.J., Y.S.P., N.K. Approval of final manuscript: all authors.

Supplementary materials can be accessed at https://doi.org/10.5009/gnl230458.