Introduction

Stroke remains the second-leading cause of death and the third-leading cause of death and disability combined in the world1. Research data support a greater focus on stroke prevention and treatment in low-income countries, where there was a much greater burden of stroke in terms of mortality and disability2.

The significant burden of stroke on the individual and society is undisputed3.

Stroke risk is evaluated on an individual and societal basised on the presence of modifiable (e.g., hypertension, atrial fibrillation) and nonmodifiable risk factors (e.g., male sex, age and ethnicity)4.

The significant risk factors for all strokes in a previous study of 3000 patients were: history of hypertension; current smoking; waist-to-hip ratio; diet risk score; regular physical activity; diabetes mellitus; alcohol intake; psychosocial stress and depression; cardiac causes; and ratio of apolipoproteins B to A15. Hypertension is clearly one of the common and important risk factors for stroke.

Arzt et al. study data demonstrate a strong association between moderate to severe sleep-disordered breathing and prevalent stroke, independent of confounding factors6.

Sleep disorders are highly prevalent in patients at risk for stroke, and may be modifiable risk factors for stroke7.

The study by Yaggi et al. indicates that obstructive sleep apnea (OSA) significantly increases the risk of stroke or death from any cause. After adjustment for age, sex, race, smoking status, alcohol-consumption status, body-mass index, and the presence or absence of diabetes mellitus, hyperlipidemia, atrial fibrillation, and hypertension, the obstructive sleep apnea syndrome retained a statistically significant association with stroke or death (hazard ratio, 1.97; 95% confidence interval, 1.12 to 3.48; P = 0.01)8.

The results highlight the independent impact of OSA on stroke risk.

Yaranov et al. also observed a positive dose effect of the apnea-hypopnea index on the rate of stroke, and the results demonstrated that OSA is associated with an increased incidence of stroke, and this association remained significant after accounting for other cardiovascular risk factors9.

This relationship between OSA and stroke was confirmed in a recent meta-analysis of 12 prospective cohort studies by Wang et al. that encompassed 25,760 subjects whereby the relative risk of incident fatal and non-fatal stroke for severe OSA compared to SA was 2.15 (95% confidence interval, 1.42–3.24)10.

Furthermore, their study showed that severe OSA is significantly and independently associated with an increased risk of cardiovascular disease (CVD), stroke, and all-cause mortality10. Several prospective cohort studies have identified OSA as an independent risk factor for stroke after accounting for potential confounding risk factors7,11,12.

It can be seen that OSA should be highlighted as a risk factor for stroke.

Stroke is a multifactorial disease with genetic and environmental components, in which gene–gene and gene–environment interactions are important13. In the other cohort study, the PITX2 and ZFHX3 genes are related to cardioembolic stoke14. This study indicated that there was a significant association between the PITX2 gene and stroke risk15. Other studies have also verified that two single nucleotide polymorphisms (SNPs) rs6843082 (PITX2, 4q25) and rs7193343 (ZFHX3, 16q22) are common loci in many stroke genetic studies16.

However, these studies examined the association of these SNPs with stroke development, but in these genetic locus analyses, no association was indicated between hypertension, OSA, and stroke SNPs. Furthermore, previous research has shown that both OSA and hypertension are independent risk factors for stroke. Therefore, this study aims to investigate the association between OSA, hypertension, and stroke genes.

Materials and methods

Data source

Data for this retrospective cohort study were obtained from the following two sources: (1) the National Health Insurance Research Database (NHIRD, with data to 1998–2015) to identify stroke, OSA, hypertension, diabetes and Atrial Fibrillation Disease) and (2) the Taiwan Biobank (TWB), a national health resource containing basic demographic information, personal health activities, physical examination and genotype data to 2008–2015. These databases were linked and personal identification numbers (PIN) were used to obtain related stroke information. The data used in this study were obtained from the Taiwan Biobank dataset, and informed consent was obtained from all individual participants in the study. This study was approved by the Institutional Review Board of Chung Shan Medical University Hospital (CS1-20009) approved this study in this context.

Patient identification

Our initial recruitment included 17,985 participants from the Taiwan Biobank. Demographic information (sex, age, body mass index [BMI]), lifestyle (regular exercise, smoking and alcohol consumption) and genotypic data were retrieved from the biobank database. Patients with incomplete questionnaires (n = 56) and genotyping information (n = 14)were excluded. Finally, this research includes the contrast of 1020 stroke patients and 16,895 control patients. Alcohol drinkers were defined as persons who had consumed more than 150 ml of alcohol per week six months prior to the health examination. Physical activity included any amount of exercise activity at lasted for 3 times a week and lasting at least 30 min each time. Smokers were defined as smoked for more than six months. A second-hand smoker was defined as someone who was currently exposed to the smoke of other people for at least 5 min per day.

Genetic variant selection

A Chromosome 4q25 gene PITX2 SNP(rs6843082) and a Chromosome 16q22 gene ZFHX3 SNP (rs7193343) were selected based on a literature search associated with cardioembolic stroke and genotyping was performed using custom Taiwan biobank chips. SNP genotyping was carried out using custom Taiwan Biobank chips and ran on the Axiom™ Genome-Wide Array Plate System (Affymetrix, Santa Clara, CA, USA). For the TWB SNPs array, we followed a standard quality control procedure to exclude a low recall rate (< 95%), p for the Hardy-Weinberg equilibrium test of < 1.0 × 10 −3,and a minor allele frequency of < 0.05. Multivariate unconditional logistic regression was used, comparing GG or GA with AA for rs6843082 and TT and TC with CC for rs7193343.

Definition of outcomes

We identified patients diagnosed with stroke (ICD-9-CM: 430–432, 433–437), OSA (ICD-9-CM: 327.23, 780.51, 780.53, 780.57), hypertension (ICD-9-CM: 401–405), diabetes mellitus (ICD-9-CM: 250), hyperlipidemia (ICD-9-CM: 272) and atrial fibrillation (ICD-9-CM: 427.3) using two outpatient visits or admission to NHIRD from 1998 to 2015, and the first diagnosis date before acceptance date. Body mass index (BMI) was calculated as weight in kilograms divided by height in meters squared (kg/m2).

Statistical analysis

The chi-square test and the t-test were used to compare the differences between discrete and continuous variables. Logistic regression analysis was used to investigate the effects of OSA and hypertension on stroke in the difference genotype. The odds ratios (OR) were estimated with their 95% confidence intervals. Statistical significance was established with a value of p < 0.05. Statistical analyzes were performed using the statistical analysis system (SAS) (version 9.4) and PLINK.

Result

According to the research conditions and definitions, the study consisted of 1,020 stroke patients and 16,895 control patients.

Among the study participants listed in Table 1, the risk factors for stroke were OSA (OR, 2.182; 95% CI: 1.634 to 2.914), age (OR, 3.032; 95% CI: 2.502 to 3.676), diabetes (OR, 1.521; 95% CI: 1.303 to 1.776), hypertension (OR, 2.443; 95% CI: 2.096 to 2.848), hyperlipidemia (OR, 1.669; 95% CI: 1.429 to 1.948), atrial fibrillation (OR, 3.089; 95% CI: 1.952 to 4.889).

Table 1 Stroke risk factors.
Full size table

Subject characteristics

Table 2 lists the study participants, among stroke patients with the rs6843082 (GG) genotype, there were 587 (5.85%), and 221 (2.2%) were diagnosed with OSA disease. In the genotype rs6843082 (GA + AA), there were 433 (5.5%) stroke patients and 186 (2.36%) were diagnosed with OSA disease. In addition, there are statistically significant differences in education (p = 0.022) and atrial fibrillation (p = 0.018).

Table 2 Descriptive data of the participants.
Full size table

Effect of genotype

Table 3, stratified by genotype, analyses the relationship between people with different genotypes, stroke patients, OSA, and hypertension.

Table 3 Association of stroke with OSA and hypertension by rs6843082.
Full size table

OSA disease (OR, 2.31; 95% CI, 1.60 to 3.36) with hypertension (OR, 2.46; 95% CI, 2.01 to 3.02), were both statistically significant as a risk factor compared to the control group in patients with the genotype rs6843082(GG). OSA disease (OR, 1.98; 95% CI, 1.24 to 3.16) with hypertension (OR, 2.42; 95% CI, 1.92 to 3.07), both showed statistical significance when compared to the control group in patients with Genotype rs6843082(GA + AA).

Effect of hypertension

Additionally, genotype grouping and hypertension stratification were used to explore the relationship between OSA, hypertension, and stroke in people with different genotypes.

Table 4 shows the results of people with genotype rs6843082 (GG) grouped by hypertension, compared to patients without OSA, those with OSA without hypertension (OR, 2.75; 95% CI: 1.24 to 3.16), with OSA with hypertension (OR, 2.17; 95% CI: 1.41 to 3.34), all were statistically significant, suggesting that OSA was associated with stroke. The statistical results showed that there was no interaction between OSA and hypertension (P = 0.627).

Table 4 Association of stroke with OSA based on the hypertension in rs6843082(GG) genotype.
Full size table

Table 5 shows the results of people with the genotype rs6843082 (GA + AA) grouped by hypertension. Compared to patients without OSA, there were no statistically significant differences between patients with OSA and without hypertension (OR, 0.53; 95% CI: 0.13 to 2.25)., but in those with OSA and hypertension (OR, 2.57; 95% CI: 1.53 to 4.33), there was a statistically significant difference, showing that only those with hypertension had OSA as a risk factor for stroke. The statistical results showed that there was an interaction between OSA and hypertension (P = 0.0429).

Table 5 Association of stroke with OSA based on the hypertension in rs6843082(GA + AA) genotype.
Full size table

Subjective outcomes

Table 6 further analyzed the relationship between different genotypes of rs6843082 and stroke based on the combination of OSA and hypertension. The results showed that people with genotype rs6843082(GG), regardless of whether they had no OSA and hypertension (OR = 2.49, 95% CI: 2.02 to 3.06), had OSA and no hypertension (OR,2.68, 95% CI: 1.35 to 5.33), OSA and hypertension (OR,5.46, 95% CI: 3.46 to 8.60), showed statistically significant differences from the control group and the risk probability is significantly increased. This indicates that no matter whether there is hypertension or not, OSA was a risk factor for stroke.

Table 6 Odds ratios (95% CI) for stroke after combining OSA and hypertension in the different genotype (rs6843082).
Full size table

The genotype rs6843082 (GA + AA) showed that people without OSA and hypertension (OR, 2.30, 95% CI: 1.81 to 2.92), or with OSA and hypertension (OR,6.25, 95% CI: 3.63 to 8.60) had a statistically significant difference, but there was no statistically significant difference between those with OSA and those without hypertension (OR,0.57, 95% CI: 0.14 to 2.36).

Discussion

The pathogenesis of stroke is complex. Epidemiological studies have identified various risk factors for stroke, including obstructive sleep apnea (OSA), hypertension, dietary risks, diabetes, obesity, smoking, air pollution, alcohol consumption, hypercholesterolemia, physical inactivity, aging, and gender7,17.

Stroke may also be caused by genetic factors; the interaction between genetic and disease factors may lead to stroke in certain individuals. To our knowledge, this is the first cohort study investigating the association between OSA, hypertension, the PITX2 SNP rs6843082, and stroke using data from the Taiwan Biobank and clinical research databases. The combined results from these data sources indicate that different genotypes of SNP rs6843082, in combination with OSA and hypertension, have different impacts on the risk of stroke. Understanding these interactions is crucial for developing targeted prevention strategies and personalized treatments for high-risk stroke populations, as well as for public health implications.

Our findings have several implications for clinical practice. First, we observed that individuals with the rs6843082 genotype (GG) are at an increased risk of stroke when affected by OSA, regardless of their hypertension status, indicating that OSA is a risk factor for stroke. In contrast, individuals with the rs6843082 genotype (GA + AA) and hypertension show an increased risk of stroke when affected by OSA, whereas OSA does not significantly impact the stroke risk in those without hypertension.

Interestingly, our study found that individuals with the rs6843082 genotype (GG) have a higher risk of stroke compared to those with OSA or hypertension alone. This suggests that the presence of the rs6843082 genotype (GG) may synergistically increase the risk of stroke when combined with OSA and hypertension.

Furthermore, our study found that individuals with the rs6843082 genotype (GA + AA) have a significantly higher stroke risk compared to those with the GG genotype. When stratifying by the presence of hypertension and OSA, we found that the combination of hypertension, OSA, and the rs6843082 genotype (GA + AA) significantly increases stroke risk. This indicates a potential gene-environment interaction, where the genetic susceptibility conferred by the rs6843082 SNP amplifies the risk of stroke in the presence of hypertension and OSA.

Stroke is a leading cause of global mortality and morbidity, with hypertension being one of its most significant risk factors18,19. OSA is also considered an independent risk factor for stroke, with this relationship being independent of other risk factors8,20. The direct and profound cardio-respiratory effects of OSA, including hypoxemia, changes in heart rate and blood pressure, altered cerebral or other regional blood flow, intrathoracic pressure fluctuations, and increased sympathetic activity, provide multiple potential pathophysiological mechanisms leading to cardiovascular dysfunction21,22.

OSA is common among hypertensive patients, and these conditions often coexist. Hypertension and OSA share common pathophysiological mechanisms, such as endothelial dysfunction, inflammation, and sympathetic overactivity, which synergistically increase the risk of stroke23,24. Additionally, hypertension and OSA are related to similar risk factors for stroke, including obesity, metabolic syndrome, and insulin resistance25.

Several studies have shown a potential association between genetic variations in the PITX2 gene and the risk of stroke. PITX2 is a transcription factor involved in the development of the heart and brain, with previous research identifying SNPs related to cardioembolic stroke, including PITX2 rs6843082 on 4q25 and ZFHX3 rs879324 on 16q2226,27.

In individuals with the rs6843082 genotype (GG), OSA appears to be a risk factor for stroke, independent of hypertension status. Previous studies have found that the presence of the G allele may increase the risk of stroke. The potential mechanisms may involve the exacerbation of arrhythmias, endothelial dysfunction, and cardiovascular complications due to intermittent hypoxia and sympathetic activation associated with OSA16,28. Conversely, in individuals with the rs6843082 genotype (GA + AA) and hypertension, the combination of OSA and hypertension synergistically increases the risk of stroke. However, individuals with OSA alone and without hypertension exhibit a lower stroke risk, suggesting that the presence of the A allele may alter stroke risk. The rs6843082 (GA + AA) genotype may help maintain intracellular redox balance, thereby reducing the risk of stroke and other cardiovascular diseases29. However, more research is needed to further elucidate the specific roles and mechanisms of these genetic variations.

The results also indicate that individuals with the rs6843082 (GA + AA) genotype and OSA should pay more attention to the prevention of hypertension, as the combination of hypertension significantly increases the risk of stroke by 6.25 times compared to those without OSA and without hypertension.

Additionally, our study found that individuals with the rs6843082 genotype (GA + AA) have a significantly higher stroke risk compared to those with the GG genotype. This may be attributed to a more pronounced adverse effect on vascular function and a higher burden of comorbidities. This suggests a potential gene-environment interaction, where the genetic susceptibility conferred by the rs6843082 SNP amplifies the risk of stroke in the presence of hypertension and OSA30.

In summary, obstructive sleep apnea and hypertension are significantly associated with an increased risk of stroke, independent of other risk factors. The severity of OSA is linked to an increased risk of this composite outcome.

Our findings show an association between OSA and the PITX2 gene with the risk of stroke. Recent studies have shown that severe obstructive sleep apnea increases the risk of fatal and non-fatal cardiovascular events31. Studies indicate that OSA independently increases stroke risk, with an unadjusted relative risk of 5.16 (95% CI: 3.72–6.60)11. In the study population of Yaranov et al., OSA was significantly associated with stroke in multivariate logistic regression analysis, with an adjusted odds ratio of 3.65 (95% CI: 1.252 to 10.623), showing that OSA patients have a risk of stroke9. Another GWAS analysis reported that the rs6843082 variant near PITX2 showed a strong association with severe cardiac embolic stroke risk (p = 2.8 × 10–16) 16. Moreover, our study also shows a statistically significant association between the PITX2 SNP rs6843082, OSA, hypertension, and stroke, consistent with previous studies showing this association between PITX2 and ischemic stroke32,33. Therefore, the PITX2 gene may be involved in the circulatory system to some extent and may act on stroke-related risk factors. We speculate that the rs6843082 SNP and PITX2 are directly related to the expression of OSA and hypertension26.

Overall, our findings about the increased risk of stroke in patients with obstructive sleep apnea are consistent with previous reports.

However, in an observational cohort study by Yaggi HK et al., obstructive sleep apnea syndrome significantly increased the risk of stroke or death from any cause, and this increase was independent of other risk factors, including hypertension. Conversely, our study found that different genotypes and the presence of hypertension have different impacts on stroke risk8.

Finally, the data analysis shows that SNP 7,193,343 does not show a statistically significant association with OSA, hypertension, and stroke, and no statistical difference was observed after adjusting for relevant factors.

Despite the important findings, some limitations should be acknowledged. First, the study relies on retrospective data from the Taiwan Biobank and the National Health Insurance database, and due to the sample size of stroke patients, this study did not specifically investigate stroke subtypes. Secondly, considering the sample size distribution of the rs6843082 allele, according to SNPedia, the allele distribution of rs6843082 among the Han Chinese is (A; A) 10.2%, (A; G) 42.3%, (G; G) 47.4%34,35. Due to the smaller number of individuals with the rs6843082 A allele in our data, we combined the SNP rs6843082 (G; A) and (A; A) for observation.

Additionally, our analysis considered all known major risk factors for stroke. Although we attempted to control for major cardiovascular risk factors, residual confounding factors may still influence our adjusted risk ratios.

In addition, The AUC values, likelihood ratio chi-square values, and p-values for each tables are all shown in Supplementary Table 1. In this study, although there is a significant disparity in the number of cases and controls, we employed cross-validation to assess the robustness and reliability of our analysis. We randomly split both cases and controls into ten equal parts and conducted the analysis by removing one part at a time, utilizing the remaining nine parts for the analysis. The results are presented in Supplementary Table 2.

In the cross-validation, the OR values ranged from a minimum of 2.031 to a maximum of 2.396, each achieving statistical significance (p value < 0.0001). These results are similar to those obtained using the complete dataset, with an OR value of 2.180.

Finally, the analysis is limited to Taiwanese individuals, and the results may not be generalizable to other populations with different genetic backgrounds and environmental exposures.

The genetics of multifactorial diseases like stroke remain a complex field, and research into new or emerging risk factors remains an active area of investigation. Therefore, variations in the PITX2 gene can be used to identify the risk of stroke in patients. Timely screening for OSA, hypertension, and genetic testing can help individuals better prevent the occurrence of stroke, thereby reducing its impact on individuals and society36.

Conclusion

Our findings showed that people with genotype rs6843082 (GG), with or without hypertension had OSA as a risk factor for stroke. For individuals with the genotype rs6843082 (GA + AA), those with hypertension, OSA is a risk factor for stroke, and for those without hypertension, OSA is not associated with stroke.