The risk and severity of stroke in patients with atrial fibrillation and gout: A National Representative Database study

Abstract Background It has been shown that gout is associated with left atrium remodeling and a pro‐inflammatory state leading to the development of atrial fibrillation (AF). There is limited evidence whether gout increases the risk of stroke in patients with AF. We assessed the incidence of gout and the risk of stroke in patients with AF. Methods This is a retrospective cohort study using the 2016 US National Inpatient Sample (NIS) based on ICD‐10 codes. The outcomes of the study were the risk and severity of new stroke in patients with pre‐existing AF and gout. Results In 2016, we identified 3 844 057 patients admitted to the hospitals in NIS with history of AF, of which 240 875 had history of gout. Patients with AF and gout have higher risk of new stroke (OR 2.07 [1.97‐2.19], P < .001), and this risk remains significantly elevated after adjusting for CHADS2VASC score variables, chronic kidney disease, dyslipidemia, obesity, and race (OR 1.10 [1.01‐1.11], P = .041). However, presence of gout in patients with AF was not associated with all‐cause in‐hospital mortality, need for mechanical ventilation, percutaneous gastrostomy tube insertion, or discharge to skilled nursing facility. Conclusion Subjects with AF and gout compared to AF alone had an increased risk of new stroke, but presence of gout was not associated with stroke severity. There is a potential role of gout as a risk factor or a risk marker for stroke in subjects with AF.

independently associated with premature death and metabolic syndrome. 3,4 Elevated uric acid level in subjects with gout is associated with cardiovascular conditions, such as hypertension (HTN), which are considered risk factors for AF and stroke. 5 Krishnan et al found a 9% increase in the risk for developing arterial HTN in men without metabolic syndrome with each unit increase in serum uric acid. 6 Recent studies have shown that gout is independently associated with a higher risk of AF at diagnosis and the risk remains elevated after the diagnosis. 7 Elevated uric acid in subjects with gout leads to proinflammatory effect and vascular inflammation which may increase the risk of atrial arrhythmias and cardiovascular disease. 2,8 However, since there are many confounding variables, the relationship between gout and AF has remained inconsistent. 9 Whether or not presence of gout increases the risk of cerebrovascular events in those with AF remains less studied. Our study is aimed at re-examining this relationship and the role of gout as a stroke risk factor in AF in a large inpatient North American dataset.

| Data source
This study was based on the National Inpatient Sample (NIS) for the year 2016. The NIS is part of all-payer database devel- Review Board (IRB) review and approval was not required as the NIS is a publicly available dataset that contains de-identified patient information.

| Study population
The NIS provides as many as 30 discharge diagnosis and 15 procedures recorded by using the ICD-10-CM codes. All adults (older than 18 years of age) with history of atrial fibrillation (AF) in 2016 were included in the current study. Risk of new stroke and post-stroke sequalae in patients with AF were stratified based on whether the patient had the diagnosis of gout based on prevalence measurements.
The list of diagnosis ICD-10-CM codes used to identify patients is shown in Table S1.

| Study variables and outcomes
We first compared the baseline variables in all patients with AF, with and without gout. Study variables included age, gender, race,

| Statistical analysis
Categorical variables are presented as frequencies and percentages while continuous variables are presented as mean ± SD.
Differences between groups were assessed with Pearson Chi square test for categorical variables and Student's t-test for continuous variables. We assessed the association between gout and stroke in patients with AF using two models. The first was an unadjusted univariable logistic regression model and the second was a multivariable logistic regression model adjusted for potential confounding factors to calculate the odds ratio and their 95% confidence interval. We adjusted for potential confounders in two steps, the first was for the CHADS2 score variables, which includes CHF, HTN, age 75 years and older, DM and prior history of stroke or TIA. The second step of adjustment included CHADS2VASC score variables, which included CHADS2 score variables in addition to female gender, vascular disease history such as CAD and PVD, and any other significantly different baseline characteristic variable that is not accounted for in the CHADS2VASC score between patients with AF and gout who had new stroke and those with who did not ( Table 2). The second step of adjustment was also used to assess the secondary outcomes of stroke severity and post-stroke sequelae in patient with AF and gout. A P value of less than .05 was considered statistically significant. STATA (IC-15.1 version, STATA Corp) was used for the statistical analyses.

TA B L E 2
Basal characteristics compared between patients with AF and gout who had a stroke and who did not have a stroke

| RE SULTS
During the year 2016, we identified 3 844 057 admissions among patients with AF, of whom 240 875 patients had history of gout. Table 1 describes a comparison of baseline characteristics between patients with AF, with and without a history of gout. Patients with AF who had gout compared to those without gout had lower proportion of females (31.9% vs 48.4%, P < .001), Whites (73.6% vs 78.7%, P < .001) and Hispanics (3.6% vs 5.3%, P < .001). There was a higher proportion of Blacks (13.8% vs 8.3%, P < .001) and Asian/Pacific Islander (3.3% vs 1.8%, P < .001) among patients AF and gout compared to patients with AF without gout. There was no significant difference in the proportion of Native Americans.
In terms of past medical illnesses, patients with AF and gout compared to without gout had higher prevalence of obesity (13.1% vs 9%, P < .001), DM (20.9% vs 19.1%, P < .001), CHF (11.7% vs 9.1%, P < .001), dyslipidemia (55.5% vs 46.4%, P < .001), CAD (47.8% vs 40.2%, P < .001), PVD (9.8% vs 7.6%, P < .001), and CKD/ESRD (57.2% vs 39.1%, P < .001). The prevalence of hypertension and tobacco smoking was lower in patients with AF and gout compared to AF with gout (31.9% vs 48.4% and 5.6% vs 9.1%, respectively, P < .001). There was no significant difference in age (75.5 ± 0.1 vs 75.6 ± 0.05 years, P = .162) between the two groups. The prevalence of a prior history of stroke or TIA was slightly higher in patients with AF and gout compared to AF patients without gout (12.5% vs 12.6%, P = .642), however, the difference did not reach statistical significance.  We also adjusted for CHADS2VASC variables along with any other significantly different baseline characteristic variable that is not accounted for in the CHADS2VASC score between patients with AF and gout who had new stroke and those who did not have stroke (Table 2). These variables were CKD/ESRD, dyslipidemia, obesity, and race. The risk remained significantly elevated (OR 1.10 [1.01-1.11], P = .041) as shown in Table 3.
The association between gout and post-stroke sequelae (secondary outcomes) among patients with AF is shown in

| D ISCUSS I ON
The current study focused on the association between gout and cardiovascular disease, specifically AF and stroke. Among patients with AF, presence of gout was associated with a higher risk of stroke not worse post-stroke outcomes.
The relationship between gout and cardiovascular comorbid conditions is complex. Many cardiovascular illnesses predispose to gout and vice versa. 10 Recent evidence has shown that different comorbid conditions such as cluster together in people with gout, such as obesity, DM, heart failure, dyslipidemia, CAD, PVD, COPD, OSA, and CKD/ESRD. [11][12][13][14][15][16][17][18] The current study is in line with these associations as shown in Table 1. Despite the fact that HTN is the most common risk factor for the development of AF, our results showed a paradoxically lower prevalence of HTN in the gout group compared to subjects without gout. 19 This lower prevalence of HTN in subjects with AF and gout in our study may be due to the higher prevalence of other risk factors of AF, such as DM, CKD, and obesity, in subjects with gout resulting in a lower relative contribution of HTN to the development of AF as compared to subjects without gout. Our results show a higher risk of new stroke on admission in patients with AF and gout before and after adjustment to CHADS2 score, CHADS2VASC score, CKD/ESRD, dyslipidemia, obesity, and race (Table 3). It has been shown that hyperuricemia increases the risk of left atrial thrombus, which raises the possibility of cardioembolic events. [21][22][23][24] The association between stroke and gout in subjects with AF was dampened significantly by multivariable adjustment although gout remained significantly associated with stroke. The increased risk of stroke in patients with gout and AF appears to be predominantly mediated by a higher prevalence of traditional risk factors in those with gout although traditional risk factors may not fully account for this increased risk. from the database especially in terms of anticoagulation and gout medications. The uric acid level is not provided in this large administrative database which precluded studying the association between severity of gout and stroke. Finally, the severity of stroke in the current study was based on in-hospital outcomes and discharge to a skilled nursing facility but was not based on actual imaging data.

| Clinical implication
• In the current study, gout was an independent risk factor in predicting new stroke in patients with AF but did not predict worse stroke outcomes or severity.
• The current study adds to the limited but growing evidence on the role of gout and hyperuricemia in refining the risk stratification of stroke in patients with AF.

| CON CLUS ION
Presence of gout in patients with AF is associated with stroke but not poor in-hospital outcomes. It remains to be seen if future prospective studies would further investigate the role of gout in predicting the risk of stroke in patients with AF and whether gout and hyperuricemia can be integrated into future risk assessment scores to improve the prediction of stroke in AF patients.