Arterial hypertension (HBP) and atrial fibrillation (AF) are closely related and often coexist clinically (1). About 60–80% of AF patients have HBP (2), and persistently elevated blood pressure (BP) is one of the important risk factors for AF (3–5). Conversely, patients with HBP are about several times more likely to develop AF than those with normal BP (6, 7). Studies have also shown that high normal BP increases the risk of AF in middle-aged individuals (6, 8).

However, it is uncertain above which BP level the risk of AF is significantly increased. Also, the extent to which BP control reduces AF in hypertensive patients is unknown, despite clear evidence that BP control reduces target organ damage and improves overall survival (9–11). Though some researches demonstrated an ideal BP in patients with AF (12), most current HBP guidelines do not specify BP-lowering goals for patients with AF (13–16). Considering that the target organ damage in HBP patients with AF is significantly more severe than that in patients with HBP alone (17), it is of great significance to establish BP control targets to reduce the risk of AF in HBP patients.

In many diseases, clinical manifestations vary by sex. Specifically, females have a lower overall rate of HBP than males (13), but post-menopausal females experience a faster rise in BP (18). On the other hand, AF is more common in males (19). However, few studies have explored sex differences in the relationship between BP and AF.

In this study, we aimed to examine in detail the association of BP and its control status with AF using a large case-control patient population. We also investigated sex differences in the relationship of BP and AF to provide more personalized suggestions for clinical treatment.

There is a strong relationship between HBP and AF, and the two share many common risk factors. HBP is also a key risk factor for AF. However, neither previous studies nor current guidelines clarify what high levels of BP should be controlled to reduce the occurrence of AF (13–16). In this study of a large case-control population, we provided a BP cut-off that was significantly associated with increased prevalence of AF and a cut-off that was associated with reduced prevalence of AF in hypertensive individuals. We demonstrated lower BP levels associated with an increased prevalence of AF in women than in men, indicating that women with HBP may need to control BP to a stringent target to reduce the prevalence of AF.

Studies have shown that multiple risk factors are closely related to the occurrence of AF, including uncontrollable factors such as age, sex, family history, race, height, and genetics (19, 22), controllable factors such as HBP, diabetes, myocardial infarction, valvular heart disease, COPD, chronic kidney disease, obesity, endurance exercise, sleep apnea, thyroid dysfunction, smoking, and intemperance (19, 22, 23), as well as some laboratory parameters such as left ventricular hypertrophy, left atrial enlargement, reduced left ventricular short axis shortening rate, C-reactive protein, and plasma brain natriuretic peptide (19). These findings are largely consistent with our results. After adjusting by multivariate regression, we observed that age, HBP, valvular heart disease, COPD, and renal insufficiency were risk factors for AF (Supplementary Figure 1). However, BMI, smoking, alcohol consumption, diabetes, and coronary artery disease were not significantly associated with AF risk (Supplementary Figure 1). This observation may be due to the nature of the patients we included. As all participants were from our cardiology unit, the risk factors we identified were highly related to most cardiovascular diseases. These risk factors may not differ between groups with and without AF in patients admitted for other cardiovascular diseases. Specifically for coronary disease, we excluded patients with acute myocardial infarction who are more likely to develop AF than those with common coronary heart disease (24, 25). It is worth mentioning that female gender appears to be a risk factor for AF in this study, unlike what we know (19). The reason may be that our study subjects were all patients, unlike a comprehensive community survey, which recruits healthy persons.

In the general population, the occurrence and recurrence of AF were associated with BP (26, 27). Even high normal BP is an important predictor of new-onset AF (6, 8). Our results confirm this. Compared with the non-HBP patients, the AF ORs of patients with grades 1, 2, and 3 HBP increased by 1.75, 2.66, and 4.30 times, respectively. The BP cut-off for significantly increasing the OR of AF was approximately 130/75 mmHg. Not only that, but the influence of BP on the risk of AF in females was more pronounced, especially when BP was significantly elevated. We demonstrated that the OR for AF in females with grade 3 HBP was 7.15 times higher than in non-hypertensive females. Meanwhile, the BP cut-off for females (127.1/75.1 mmHg) to develop AF was lower than that for males (133.1/79.9 mmHg), suggesting that more attention should be paid to women, even in the early stages of high normal BP.

In hypertensive patients, BP control can significantly reduce the incidence of AF. Compared with the non-hypertensive patients, the OR of AF in patients with uncontrolled BP was 3.00 times, which decreased to 1.62 times after BP control, fully demonstrating the importance of BP control. In hypertensive patients, lowing BP to 145.1/85.8 mmHg had a significant effect on the OR of AF, whereas in female patients, it should be lowered to 140.5/82.5 mmHg.

After BP control, the risk of AF decreases. On the one hand, the remodeling of the atrial myocardium by HBP is fundamentally released after BP reduction, (28). On the other hand, antihypertensive drugs such as angiotensin-converting enzyme inhibitors, angiotonin receptor blockers, and angiotensin receptor neprilysin inhibitor can significantly improve cardiac remodeling and reverse the structural basis of AF (29), which also help relieve AF.

Left atrial remodeling is the main cause of AF in hypertensive patients. Verdecchia et al. conducted a 16-year follow-up of 2,482 patients with primary HBP and found that the risk of AF increases with patient age and left ventricular size, and that increased left atrial volume predisposes to chronic AF (30). This is because elevated BP leads to an increase in left ventricular end-diastolic pressure, resulting in left ventricular hypertrophy and diastolic dysfunction, which in turn increase left atrial pressure. Left atrial remodeling induced by long-standing elevated left atrial pressure is the mechanistic basis of AF (31, 32). In animal models, Kistler et al. performed electrophysiological and pathological examinations on chronic hypertensive sheep and found that elevated BP was significantly positively correlated with atrial electrical and anatomical remodeling (33). Lau et al. also used a sheep model of HBP to study the effect of HBP on atrial remodeling and demonstrated that early antihypertensive therapy was effective in delaying this process (34). In a follow-up of 112 patients with AF, Tanabe et al. found that the left atrium was reversed in hypertensive patients with controlled BP compared with those with uncontrolled BP (35), supporting that BP control in hypertensive patients is of great significance.

Sex differences in the relationship of HBP and AF is the highlight of this study. On the one hand, women tend to have lower BP for reasons such as estrogen (36), weight loss, dieting, and bland diets. Many women have low BP but no clinic signs of shock. This means that normal or high normal BP may be a hypertensive state in women. Other possible reasons include perimenopause and post-menopause. Studies have found that perimenopausal and post-menopausal women are at increased risk of various cardiovascular diseases, especially early menopause, which is associated with estrogen (37–39). In this study, we added to analyze age as a sex-disaggregated risk factor for AF and found that the OR of AF increased significantly in women after 44.6 or 48.6 years (Figure 6; Supplementary Figure 6), which roughly coincides with the timing of menopause in women, especially those with earlier menopause. Relatively, the cut-off age was older for men. However, further studies are needed to consider information such as the timing of menopause.

In this study, to minimize the information loss caused by data grouping, in addition to logistic regression, we also conducted RCS to directly process the data. The RCS results showed that unexpectedly, the BP cut-off values for a significantly increased OR of AF due to elevated BP and a reduced OR of AF after BP control in hypertensive patients were not 140/90 mmHg and 130/80 mmHg, respectively (40), and there were significant differences in the cut-off value of males and females. Meanwhile, the results before and after PSM were basically the same, indicating the stability of the data.

This study has some limitations. First, the collected data came from a single-center retrospective study, and further verification is needed in prospective studies with larger samples and more centers. Second, this study did not collect data on echocardiogram, HBP medication, and menstruation and menopause status of female patients. The interpretation of relevant conclusions is limited to our current findings and previous literature reports, and further confirmation is needed in follow-up work. Overall, however, these do not affect the clinical significance of our findings, which convey to clinicians that BP targets for AF risk reduction should include the sex factor.

Blood pressure (BP) over 133.1/79.9 mmHg in male patients or 127.1/75.1 mmHg in female patients was significantly associated with the prevalence of AF. In HBP patients, a BP lower than 145.1/85.8 mmHg in men or 140.5/82.5 mmHg in women was associated with reduced prevalence of AF. These data suggest that BP targets to reduce the risk of AF may consider sex differences, and that a lower and individualized BP target may be established to protect against AF in women.

The datasets used and analyzed during the current study are available from the corresponding author on reasonable request. Requests to access these datasets should be directed to XT, tangxh007007@163.com.

The studies involving human participants were reviewed and approved by IRB of The Third Xiangya Hospital of Central South University. Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.

XZ and QF established the idea and helped in wrote main ideas for this research, main results, and discussion of the findings. AW was a major contributor in wrote the manuscript. XZ and WJ interpreted statistical analysis to prove the main findings of this project. QF, XW, and JW collected data. XT contributed on editing this manuscript and giving advice for the main authors to organized the manuscript and ideas of the project. All authors contributed to the article and approved the submitted version.