Uric Acid Level and Prevalence of Atrial Fibrillation in a Japanese General Population of 285,882

Abstract

Background: The association between serum uric acid (UA) levels and atrial fibrillation (AF) in the general population in Japan is not well known.

Methods and Results: In total, 285,882 consecutive subjects (men, 130,897; women, 154,985; age, 58±15 years) not receiving treatment for hyperuricemia who underwent health checkups were enrolled. Subjects were stratified into deciles according to age, body mass index, estimated glomerular filtration rate, systolic blood pressure, and UA level. AF prevalence was calculated for each decile. The odds ratio that defined the decile with the lowest AF prevalence as reference was calculated in each sex. In men, the mean UA was 6.0±1.4 mg/dl; AF prevalence was 1.8% and was lowest in the decile with UA 4.4–4.9 mg/dl. Deciles with both high and low UA (5.4–5.6 mg/dl to >7.8 mg/dl and <4.3 mg/dl) were associated with significantly higher AF prevalence. In women, the mean UA was 4.5±1.1 mg/dl; AF prevalence was 0.7% and was lowest in the decile with UA 3.6–3.8 mg/dl. Deciles with highest UA (5.0–5.2 mg/dl to >5.9 mg/dl) were associated with significantly higher AF prevalence. The analysis adjusted for other clinical covariates demonstrated an independent association between UA and AF in both sexes.

Conclusions: In a representative Japanese general population, UA level was significantly associated with AF, independently of other cardiovascular risk factors.

Atrial fibrillation (AF) is the most common type of arrhythmia and a worldwide health problem. The AF prevalence in the Japanese general population is between 1.35%¹ and 1.6%;² it is increasing and is expected to rise further as life expectancy increases globally. AF is associated with decreased quality of life and increased frequency of prolonged hospitalization because of its clinical consequences such as heart failure and stroke.³ Despite recent advances in therapy, including the use of anticoagulant agents, rate control, and techniques to maintain sinus rhythm, the mortality and morbidity of patients with AF remains unacceptably high. Thus, we need to identify patients at high risk of AF to devise prevention strategies.

Many epidemiological studies have reported risk factors for AF, with aging being one of the key risk factors. Age-related fibrosis or accumulation of accompanying disorders that occur with increasing frequency in older individuals may explain this association.⁴ Another important and well-established risk factor is heart failure,⁵ including with reduced or preserved ejection fraction⁶ and with coronary artery disease.⁷ Other reported risk factors include male sex,^5,7 hypertension,⁸ chronic kidney disease,⁹ sleep apnea syndrome,¹⁰ and metabolic factors such as obesity¹¹ and diabetes mellitus.¹²

Many studies have reported associations between hyperuricemia and cardiovascular diseases, including ischemic heart disease,¹³ heart failure,¹⁴ stroke,¹⁵ and other atherosclerotic disorders.^16,17 Conversely, several studies have reported an association between high serum uric acid (UA) levels and the onset of AF. However, those studies were performed in small populations or in patients at high risk of cardiovascular diseases. Therefore, the correlation between UA and AF needs to be clarified in large, general populations.

Here we aimed to investigate the association between serum UA levels and AF prevalence, including sex specificity, in the general Japanese population.

Methods

Study Population

In total, 325,771 consecutive subjects who underwent routine health checkups at the JA Kagoshima Kouseiren Medical Health Care Center from January, 1979 to December, 2013 were enrolled in this cross-sectional study. Subjects who were receiving medical treatment for hyperuricemia or those with missing serum UA measurement data were excluded, leaving 285,882 included in the final analysis.

This study conformed to the Declaration of Helsinki and was approved by the Institutional Ethics Committee of the Graduate School of Medical and Dental Sciences, Kagoshima University and JA Kagoshima Kouseiren Hospital.

Data Collection

Data regarding medical history, including history of hypertension, diabetes mellitus, dyslipidemia, and cardiovascular diseases, and medication history was obtained through self-administered questionnaires. Body mass index (BMI, kg/m²) was calculated for each subject. Blood pressure was measured using a mercury sphygmomanometer after subjects sat quietly for 5 min. A 12-lead surface ECG was performed. Blood samples were obtained in the overnight non-fasting state.

Serum UA, triglyceride, low-density lipoprotein-cholesterol, high-density lipoprotein-cholesterol, and creatinine levels were measured using standard laboratory procedures. The estimated glomerular filtration rate (eGFR) was determined according to the new Japanese coefficient for the modified isotope dilution mass spectrometry-traceable Modification of Diet in Renal Disease study equation:

eGFR=194×SCr^−1.094×age^−0.287

For women, eGFR was multiplied by a correction factor of 0.739;¹⁸ the cardiovascular risk factors were defined as follows: hypertension: current use of antihypertensive medications, systolic blood pressure (SBP) ≥140 mmHg, and/or diastolic blood pressure (DBP) ≥90 mmHg; diabetes mellitus: active treatment with oral hypoglycemic agents or insulin, or glycosylated hemoglobin ≥6.5%; dyslipidemia: use of lipid-lowering agents, serum triglycerides ≥150 mg/dl, serum low-density lipoprotein-cholesterol ≥140 mg/dl, or serum high-density lipoprotein-cholesterol <40 mg/dl.

Definition of AF

AF was diagnosed using 12-lead surface ECG; subjects with a history of AF were assigned to the AF group. Although we could recognize the existence of cardiac disease in the medical questionnaire used in the health checkups, we could not study the details of the diseases. The evaluation of cardiac function by echocardiography was not included in the present health check-up, and the subjects of this study did not undergo routine echocardiographic examination. Therefore, we could not determine the number of subjects with existing valvular heart disease.

Statistical Analysis

As serum UA levels differ substantially between the sexes, results were analyzed separately for men and women. Continuous variables, including age, BMI, BP, eGFR, and serum UA concentration, are expressed as mean±standard deviation. Categorical variables, including presence of cardiovascular disease or cardiovascular risk factors, are expressed as number of subjects and proportions (percentages). Differences between men and women for continuous and categorical variables were analyzed using Student’s unpaired t-test and the χ² test, respectively. Subjects were stratified into deciles by each continuous variable, including age, BMI, eGFR, SBP, and UA level, and AF prevalence was calculated for each decile. Furthermore, using logistic regression analysis, the odds ratio (OR) that defined the decile with the lowest AF prevalence as reference was calculated for each sex. Regarding serum UA levels, the results of logistic regression analysis were adjusted for other clinical variables. Univariate and multivariate logistic regression analyses, using clinical variables as continuous parameters, were performed to identify factors associated with AF.

All statistical analyses were performed using JMP Pro version 11 (SAS Institute Inc, Cary, NC, USA) for Windows. P<0.05 was considered significant.

Results

Characteristics of the Study Population

The clinical characteristics of subjects are summarized in Table 1. There were 130,897 men (46%; mean age, 57±16 years) and 154,985 women (54%; mean age, 58±16 years). The average serum UA level was higher in men than in women (men, 6.0±1.4 mg/dl; women, 4.5±1.1 mg/dl; P<0.0001). Furthermore, AF prevalence was higher in men (1.8%) than in women (0.7%; P<0.0001).

Table 1. Characteristics of Health Study Subjects

Variable	Men (n=130,897)	Women (n=154,985)	P value
Age (years)	56.7±15.8	57.7±15.9	<0.0001
BMI (kg/m2)	23.4±3.3	22.8±3.5	<0.0001
Hypertension, n (%)	49,609 (36.4)	52,938 (32.6)	<0.0001
Diabetes mellitus, n (%)	7,178 (10.4)	4,748 (6.6)	<0.0001
Dyslipidemia, n (%)	60,626 (45.9)	67,115 (42.7)	<0.0001
Cardiovascular disease, n (%)
AF	2,395 (1.8)	1,035 (0.7)	<0.0001
Chronic heart failure	355 (0.3)	248 (0.2)	<0.0001
Ischemic heart disease	1,731 (1.3)	1,204 (0.7)	<0.0001
Stroke	1,912 (1.4)	1,258 (0.8)	<0.0001
eGFR (ml/min/1.73 m2)	77.4±17.2	78.3±18.0	<0.0001
UA (mg/dl)	6.0±1.4	4.5±1.1	<0.0001

AF, atrial fibrillation; BMI, body mass index; eGFR, estimated glomerular filtration rate; UA, uric acid.

Association Between AF Prevalence and Other Parameters

Associations between AF prevalence and other parameters categorized into deciles are shown in Table 2.

Table 2. Association Between AF Prevalence and Other Parameters Categorized Into Deciles

Men
Age (years)	<37	38–44	45–50	51–55	56–60	61–64	65–67	68–71	72–76	≥77
AF (%)	0.1	0.2	0.5	0.8	1.3	2.0	2.2	3.2	3.6	5.0
OR	Ref.	2.74	7.10	12.73	21.28	31.01	34.97	51.43	58.86	81.86
95% CI		1.31– 6.22	3.71– 15.33	6.81– 27.12	11.58– 44.81	16.94– 65.11	19.06– 73.59	28.27– 107.66	32.4– 123.10	45.11– 171.09
P value		0.0063	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001
BMI (kg/m2)	<19.4	19.5–20.6	20.7–21.6	21.7–22.4	22.5–23.1	23.2–23.9	24.0–24.8	24.9–25.8	25.9–27.4	≥27.5
AF (%)	1.6	1.7	1.6	1.8	1.8	1.7	2.0	1.9	1.9	2.3
OR	Ref.	1.06	1.02	1.10	1.12	1.09	1.23	1.21	1.19	1.45
95% CI		0.87–1.29	0.84–1.25	0.94–1.30	0.89–1.33	0.89–1.32	1.02–1.48	1.03–1.52	0.97–1.42	1.19–1.72
P value		0.5595	0.8064	0.3367	0.2299	0.3597	0.0247	0.0431	0.0658	<0.0001
eGFR (ml/min/1.73 m2)	<57.6	57.7–64.0	64.1–68.7	68.8–72.7	72.8–76.6	76.7–80.5	80.6–84.9	85.0–90.3	90.4–98.4	≥98.5
AF (%)	5.1	3.2	2.6	1.9	1.5	1.2	1.2	0.8	0.5	0.5
OR	11.55	7.19	5.75	4.24	3.22	2.61	2.50	1.66	Ref.	1.05
95% CI	9.10–15.62	5.70–9.91	4.51–7.90	3.31–5.89	2.43–4.40	1.91–3.54	1.90–3.51	1.22–2.35		0.75–1.54
P value	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	0.0017		0.7771
SBP (mmHg)	<104	105–110	111–115	116–120	121–123	124–128	129–132	133–138	139–148	≥149
AF (%)	1.0	1.0	1.0	1.1	0.8	1.0	1.2	1.2	1.4	1.3
OR	1.20	1.20	1.13	1.28	Ref.	1.15	1.47	1.37	1.71	1.58
95% CI	0.73–2.01	0.73–2.02	0.67–1.92	0.80–2.11		0.70–1.92	0.90–2.46	0.84–2.30	1.06–2.81	0.98–2.63
P value	0.4708	0.4632	0.6452	0.3039		0.5771	0.1192	0.2032	0.0250	0.0601
UA (mg/dl)	<4.3	4.4–4.9	5.0–5.3	5.4–5.6	5.7–6.0	6.1–6.3	6.4–6.6	6.7–7.1	7.2–7.7	≥7.8
AF (%)	1.5	1.2	1.3	1.5	1.7	1.9	1.7	2.1	2.5	3.0
OR	1.25	Ref.	1.09	1.25	1.39	1.57	1.42	1.72	2.10	2.49
95% CI	1.01–1.53		0.88–1.35	1.01–1.54	1.14–1.68	1.28–1.91	1.15–1.74	1.42–2.08	1.73–2.54	2.07–2.99
P value	0.0332		0.3999	0.0362	0.0007	<0.0001	0.0011	<0.0001	<0.0001	<0.0001
Women
Age (years)	<37	38–45	46–52	53–57	58–61	62–65	66–69	70–72	73–76	≥77
AF (%)	0.04	0.05	0.1	0.2	0.4	0.4	0.8	1.1	1.6	2.2
OR	Ref.	1.13	2.61	4.42	9.12	9.07	19.02	25.04	35.64	51.12
95% CI		0.40– 3.24	1.14– 6.68	2.04– 10.97	4.46– 21.94	4.47– 21.73	9.62– 44.87	12.64– 59.13	18.15– 83.71	26.16– 119.71
P value		0.8045	0.0214	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001
BMI (kg/m2)	<18.6	18.7–19.8	19.9–20.7	20.8–21.6	21.7–22.4	22.5–23.3	23.4–24.2	24.3–25.4	25.5–27.3	≥27.4
AF (%)	0.5	0.6	0.6	0.6	0.6	0.7	0.7	0.7	0.8	0.9
OR	Ref.	1.28	1.22	1.10	1.28	1.39	1.40	1.31	1.61	1.82
95% CI		0.93–1.72	0.89–1.67	0.77–1.47	0.90–1.69	1.03–1.89	1.02–1.89	0.97–1.78	1.16–2.08	1.30–2.30
P value		0.0957	0.1923	0.5272	0.0958	0.0261	0.0220	0.0706	0.0008	<0.0001
eGFR (ml/min/1.73 m2)	<57.5	57.6–64.0	64.1–68.7	68.8–72.8	72.9–76.9	77.0–81.1	81.2–85.9	86.0–91.9	92.0–101.0	≥101.1
AF (%)	2.6	1.2	0.7	0.6	0.4	0.3	0.3	0.3	0.2	0.2
OR	16.4	7.19	4.27	3.83	2.29	1.87	2.00	1.60	Ref.	1.20
95% CI	11.12– 25.09	4.76– 11.05	2.67– 6.47	2.34– 5.73	1.43– 3.69	1.21– 3.21	1.24– 3.27	0.96– 2.64		0.70– 2.04
P value	<0.0001	<0.0001	<0.0001	<0.0001	0.0003	0.0093	0.0034	0.0593		0.4940
SBP (mmHg)	<98	99–105	106–110	111–115	116–120	121–124	125–130	131–136	137–144	≥145
AF (%)	0.2	0.1	0.2	0.2	0.2	0.3	0.4	0.3	0.4	0.3
OR	2.46	Ref.	2.62	2.55	2.08	3.89	4.37	4.06	5.06	3.70
95% CI	0.80–9.10		0.87–9.55	0.80–9.58	0.67–7.71	1.33–14.07	1.65–15.03	1.43–14.44	1.85–17.69	1.31–13.16
P value	0.1169		0.0862	0.1124	0.2032	0.0122	0.0020	0.0071	0.0010	0.0123
UA (mg/dl)	<3.1	3.2–3.5	3.6–3.8	3.9–4.1	4.2–4.3	4.4–4.6	4.7–4.9	5.0–5.2	5.3–5.8	≥5.9
AF (%)	0.4	0.4	0.3	0.4	0.4	0.4	0.4	0.8	1.1	2.0
OR	1.19	1.14	Ref.	1.11	1.18	1.37	1.40	2.44	3.37	6.59
95% CI	0.80–1.77	0.76–1.69		0.76–1.63	0.78–1.77	0.95–1.97	0.97–2.03	1.74–3.48	2.48–4.69	4.92–9.04
P value	0.3811	0.5088		0.5697	0.4170	0.0834	0.0691	<0.0001	<0.0001	<0.0001

CI, confidence interval; OR, odds ratio; SBP, systolic blood pressure. Other abbreviations as in Table 1.

In men, AF prevalence was lowest in the age decile under 37 years (0.1%), and it increased significantly with increasing age. In the decile above 77 years, AF prevalence was 5.0%, and OR was 81.86 (P<0.0001) compared with the decile under 37 years. Regarding BMI, AF prevalence was lowest in the decile with BMI <19.4 kg/m² (1.6%); the prevalence increased with increasing BMI; particularly in the decile >27.5 kg/m², it significantly increased to 2.3% compared with that of the decile under 19.4 kg/m² (OR=1.45, P<0.0001). AF prevalence was lowest in the decile with eGFR 90.4–98.4 ml/min/1.73 m² (0.5%) and significantly high in those with lower eGFR. No clear association was found between SBP and AF prevalence.

Deciles with higher UA levels (5.4–5.6 mg/dl to >7.8 mg/dl) showed significantly higher AF prevalence than those with UA levels 4.4–4.9 mg/dl, with the latter having the lowest AF prevalence. In addition, AF prevalence in the decile with lowest UA levels (UA <4.3 mg/dl) was significantly higher than that in the decile with levels 4.4–4.9 mg/dl (OR=1.25, P=0.0332).

In women, the relationships of AF prevalence with age, BMI, and eGFR were similar to those observed in men. However, in deciles with higher SBP of 121–124 mmHg, AF prevalence was significantly higher than in the reference SBP decile (SBP, 99–105 mmHg). Furthermore, deciles with higher UA levels (5.0–5.2 mg/dl to >5.9 mg/dl) had a significantly higher AF prevalence than those with UA levels 3.6–3.8 mg/dl (OR=2.44 to 6.59, P<0.0001 for all), with the latter having the lowest prevalence. As observed in men, there was a trend towards higher AF prevalence in deciles with lower UA levels (<3.1 mg/dl and 3.2–3.5 mg/dl) than in the decile with UA levels 3.6–3.8 mg/dl; however, this trend did not reach significance.

Independent Association Between UA Level and AF Prevalence

The Figure shows the OR obtained from the adjusted logistic regression analysis for each UA decile compared with the lowest AF prevalence decile as the reference. ORs were adjusted for clinical covariates, including age, eGFR, BMI, and coexisting diseases, such as hypertension, diabetes mellitus, dyslipidemia, heart failure, ischemic heart disease, and stroke.

Figure.

Adjusted odds ratios (ORs) for the prevalence of atrial fibrillation (AF) according to serum uric acid (UA) decile in men (A) and women (B). ORs were adjusted for clinical covariates, including age; eGFR; BMI; and coexisting disease such as hypertension, diabetes mellitus, dyslipidemia, heart failure, ischemic heart disease, and stroke. *P<0.05; **P<0.001.

In men, deciles with higher UA levels (5.4–5.6 mg/dl to >7.8 mg/dl) had a significantly higher AF prevalence than those with UA levels 4.4–4.9 mg/dl. Furthermore, the decile with the lowest UA levels (<4.3 mg/dl) had a higher AF prevalence than that with UA levels 4.4–4.9 mg/dl (OR=1.42, P=0.0171).

In women, deciles with higher UA levels (5.0–5.2 mg/dl to >5.9 mg/dl) had a significantly higher AF prevalence than those with UA levels 3.6–3.8 mg/dl. Furthermore, deciles with lower UA levels (<3.1 mg/dl and 3.2–3.5 mg/dl) had a higher AF prevalence than that with UA levels 3.6–3.8 mg/dl; however, these differences did not reach significance (decile with UA levels <3.1 mg/dl, OR=1.84, P=0.0702; decile with UA levels 3.2–3.5 mg/dl, OR=1.82, P=0.0737).

Multivariate Analyses of Factors Associated With AF

Table 3 shows the results of logistic regression analysis for factors associated with AF. In men, the univariate analysis revealed that all variables were significantly associated with AF; the multivariate analysis revealed that age, BMI, eGFR, UA, and the presence of diabetes mellitus, dyslipidemia, chronic heart failure, ischemic heart disease, and stroke were significantly associated with AF, but not with hypertension. In women, univariate analysis revealed that all variables, except dyslipidemia, were significantly associated with AF; multivariate analysis revealed that all variables, except BMI and the presence of hypertension or ischemic heart disease, were significantly associated with AF.

Table 3. Factors Associated With AF

	Univariate			Multivariate
	OR	95% CI	P value	OR	95% CI	P value
Men
Age	1.07	1.06–1.07	<0.0001	1.06	1.05–1.06	<0.0001
BMI	1.02	1.01–1.04	<0.0001	1.08	1.06–1.09	<0.0001
Hypertension	1.67	1.54–1.81	<0.0001	0.99	0.89–1.10	0.8882
Diabetes mellitus	1.34	1.15–1.56	0.0002	1.40	1.19–1.63	<0.0001
Dyslipidemia	0.66	0.61–0.72	<0.0001	0.69	0.61–0.77	<0.0001
Chronic heart failure	9.11	6.66–12.20	<0.0001	6.27	4.42–8.69	<0.0001
Ischemic heart disease	3.27	2.63–4.01	<0.0001	1.70	1.34–2.12	<0.0001
Stroke	3.98	3.29–4.77	<0.0001	2.10	1.71–2.55	<0.0001
eGFR	0.95	0.95–0.96	<0.0001	0.98	0.98–0.98	<0.0001
UA	1.20	1.17–1.23	<0.0001	1.19	1.14–1.24	<0.0001
Women
Age	1.09	1.08–1.10	<0.0001	1.07	1.06–1.08	<0.0001
BMI	1.03	1.01–1.05	<0.0001	1.00	0.97–1.03	0.7225
Hypertension	2.27	2.01–2.56	<0.0001	1.17	0.97–1.41	0.0858
Diabetes mellitus	2.14	1.63–2.77	<0.0001	1.80	1.36–2.35	<0.0001
Dyslipidemia	0.92	0.81–1.04	0.2043	0.68	0.56–0.82	<0.0001
Chronic heart failure	17.37	11.16–25.86	<0.0001	9.91	6.10–15.48	<0.0001
Ischemic heart disease	2.58	1.58–3.88	0.0003	1.31	0.80–2.03	0.2577
Stroke	3.89	2.65–5.48	<0.0001	2.10	1.40–3.02	0.0005
eGFR	0.94	0.94–0.95	<0.0001	0.98	0.97–0.98	<0.0001
UA	1.72	1.65–1.80	<0.0001	1.44	1.34–1.55	<0.0001

Abbreviations as in Tables 1,2.

Discussion

In this cross-sectional study of a large sample, we found an interesting association between serum UA levels and AF prevalence. In men, AF prevalence increased in both deciles with high and low UA levels. Although a similar trend was observed in women, high serum UA levels had a stronger association with AF prevalence in women than in men.

Hyperuricemia and AF

The association between UA level and cardiovascular diseases, such as cardiovascular events, heart failure, and carotid atherosclerosis, has been reported,¹⁹ and a few studies have reported the association between hyperuricemia and AF. Tamariz et al reported a significant association between hyperuricemia and AF in their prospective study of 15,382 ethnically-diverse subjects.²⁰ Suzuki et al demonstrated a similar significant association in a cross-sectional study of approximately 7,155 cardiovascular outpatients with high cardiovascular risk.²¹ On the other hand, the evidence of an independent association between serum UA levels and cardiovascular diseases is weak and less consistent in healthy populations. We demonstrated an independent association between serum UA levels and AF prevalence in 285,882 subjects from a Japanese general population with a mean age of 57 years (range, 19–99 years).

Serum UA is the final product of purine metabolism catalyzed by xanthine oxidase. Xanthine oxidase has been reported to be correlated with oxidative stress²² and elevation of systemic inflammatory markers. Thus, serum UA levels are considered a marker of tissue oxidative stress and inflammation.²³ Electrical and structural remodeling of the left atrium is an important process in the pathogenesis of AF.^24–26 Onset and persistence of AF is posited to be associated with oxidative stress²⁷ and inflammation.^28,29 A previous study reported that serum UA levels significantly correlated with the left atrial diameter and hyperuricemia, and that UA was an important risk factor for new-onset AF.³⁰ Recently, a further explanation of the association between high UA levels and AF has emerged. Intracellular accumulation of UA via activation of urate transporters (UATs) is posited to cause cell injury through several signaling pathways.³¹ UATs are expressed not only in renal tubular cells but also vascular smooth muscle cells and endothelial cells.^32,33 Intracellular UA uptake by UATs reportedly enhances Kv1.5 protein expression, which may be attributable to shortening of the action potential duration (APD), resulting in the initiation or sustainment of AF.³⁴

Hyperuricemia is caused by accelerated generation of UA and/or impaired excretion in the kidney and is classified into 3 types: overproduction, underexcretion, and mixed types. We lacked data on urinary excretion of UA, and therefore we could not determine the type of hyperuricemia in the present subjects. There are no reports that show AF prevalence related to the type of hyperuricemia, and further study is needed to clarify this point, including the underlying mechanisms.

Hypouricemia and AF

The association between serum UA levels and the risk of cardiovascular events or death reportedly follows a J-shaped curve. A previous study reported that serum UA levels <4.5 mg/dl in men and <3.2 mg/dl in women were associated with increased risk cardiovascular diseases.³⁵ Another study showed that subjects with serum UA levels <4.9 mg/dl were at increased risk of death from cardiovascular diseases.³⁶

In this study, we first demonstrated that extremely low UA levels were associated with an increased risk of AF in the general population. The AF prevalence increased at serum UA levels <4.3 mg/dl in men and <3.5 mg/dl in women. UA is a powerful antioxidant and scavenger of singlet oxygen and radicals;³⁷ it has been shown to contribute as much as 60% of free radical scavenging in human serum.³⁸ Previous study reported that elevated reactive oxygen species resulted in shortening of the atrial APD by increasing the transient outward current (Ito)³⁹ and producing delayed afterdepolarizations by increasing sarcoplasmic reticulum Ca²⁺ release via enhanced ryanodine receptor activity.⁴⁰ The relationship between renal hypouricemia related to SLC22A12 (URAT1) or SLC2A9 (GLUT9) gene mutations and exercise-induced acute renal failure has been reported in several studies.^41,42 Murakami et al proposed that increased oxidative stress because of renal hypouricemia induced renal arterial spasms after exercise, resulting in loss of kidney function.⁴³ Although extremely low UA levels may increase oxidative stress in the left atrium, increasing the risk of AF, further studies are warranted to elucidate the currently poorly-understood association between hypouricemia and AF.

In the present study, there were 1,214 subjects (men, 296 and women, 918) whose UA values were <2.0 mg/dl, and 2 male subjects (0.7%) and 5 female subjects (0.5%) had AF. In women, the AF prevalence in those with UA <2.0 mg/dl was slightly higher than that in women with a UA value in the 1st decile (UA <3.1 mg/dl, 0.4%) or in women with a value in the reference range (UA=3.6–3.8 mg/dl, 0.3%), but the result did not reach statistical significance. In men, the AF prevalence in subjects with UA <2.0 mg/dl was not higher than that in men with a UA value in the 1st decile (UA <4.3 mg/dl, 1.5%) or in those with a value in the reference range (UA=4.4–4.9 mg/dl, 1.2%). Consequently, we could not demonstrate higher AF prevalence in subjects with UA <2.0 mg/dl. The reason for this may be that the subjects with UA <2.0 mg/dl were younger in the female group and had lower BMI and higher eGFR in both sexes compared with the other groups. Further studies are needed to clarify the association between extremely low levels of UA and AF.

Sex Differences in the Association Between UA and AF

Several studies have demonstrated that hyperuricemia in women is more strongly associated with cardiovascular events⁴⁴ and death⁴⁵ than in men. Increased UA levels are reportedly associated with a greater risk of AF among women.²⁰ Corroborating those studies, we demonstrated that high UA levels were more strongly associated with AF prevalence in women compared with men. Although the mechanism underlying this sex difference remains unknown, it may be attributable to differences in the sex hormones. Menopause is associated with increased serum UA levels because of the uricosuric effect of estrogen that induces UA excretion from kidney’s proximal tubules. Further studies are warranted to elucidate the underlying mechanism of sex-related differences in the association between serum UA levels and AF prevalence.

Beneficial Effects of UA-Lowering Therapy

We demonstrated an independent relationship between increased serum UA levels and AF prevalence. UA-lowering therapy has been shown to be beneficial in patients with coronary artery disease⁴⁶ and heart failure,⁴⁷ and early intervention is expected to reduce the possibility of AF. However, the efficacy of such therapy in lowering the AF risk and for cardiovascular mortality prevention remains unknown.

The AF prevalence increased at serum UA levels of 5.3 mg/dl in men and 5.0 mg/dl in women. Despite the differences in the distribution of serum UA levels between men and women, limited dissociation between serum UA levels increasing AF prevalence was observed in each sex. Hyperuricemia was generally defined as serum UA level ≥7.0 mg/dl based on the limit of serum-solubility in both sexes. Even in patients with gout, the recommended serum UA level for recurrence prevention and tophi resolution is <6.0 mg/dl.⁴⁸ Even after assuming the efficacy of aggressive UA-lowering therapy, treatment goals need to be determined. Further studies are warranted to elucidate the association between hyperuricemia and AF prevalence.

Study Limitations

First, we were unable to clarify the causal relationship between serum UA levels and AF prevalence and determine the predictive value of UA on incident AF because of the cross-sectional design. Therefore, a longitudinal study should be performed. Second, the subjects were limited to participants in health checkups at a single facility in Japan. Third, we might have underestimated AF prevalence by failing to notice paroxysmal or persistent AF without symptoms, as the diagnosis of AF was made on the basis of 12-lead ECG or self-reported medical history. Finally, we did not include data regarding medications, such as diuretics or renin-angiotensin system inhibitors, that may affect serum UA levels and AF prevalence, or regarding menopause, which may influence serum UA levels in women. We also lacked data about lifestyle factors, including drinking, smoking and exercise, which are reported to affect serum UA levels and AF prevalence. Therefore, further research that includes these factors should be performed.

Conclusions

In a large Japanese population of health check-up subjects, serum UA levels were significantly associated with AF prevalence in both men and women, an association that was independent of other cardiovascular risk factors.

References

• 1.    Nakayama  T,  Date  C,  Yokoyama  T,  Yoshiike  N,  Yamaguchi  M,  Tanaka  H. A 15.5-year follow-up study of stroke in a Japanese provincial city: The Shibata Study. Stroke 1997; 28: 45–52.
• 2.    Iguchi  Y,  Kimura  K,  Aoki  J,  Kobayashi  K,  Terasawa  Y,  Sakai  K, et al. Prevalence of atrial fibrillation in community-dwelling Japanese aged 40 years or older in Japan: Analysis of 41,436 non-employee residents in Kurashiki-city. Circ J 2008; 72: 909–913.
• 3.    Alonso  A,  Norby  FL. Predicting atrial fibrillation and its complications. Circ J 2016; 80: 1061–1066.
• 4.    Spach  MS,  Heidlage  JF,  Dolber  PC,  Barr  RC. Mechanism of origin of conduction disturbances in aging human atrial bundles: Experimental and model study. Heart rhythm 2007; 4: 175–185.
• 5.    Krummen  DE. Atrial fibrillation and heart failure: Identification of patients at risk. Circ J 2016; 80: 587–589.
• 6.    Tsang  TS,  Gersh  BJ,  Appleton  CP,  Tajik  AJ,  Barnes  ME,  Bailey  KR, et al. Left ventricular diastolic dysfunction as a predictor of the first diagnosed nonvalvular atrial fibrillation in 840 elderly men and women. J Am Coll Cardiol 2002; 40: 1636–1644.
• 7.    Schnabel  RB,  Sullivan  LM,  Levy  D,  Pencina  MJ,  Massaro  JM,  D’Agostino  RB Sr, et al. Development of a risk score for atrial fibrillation (Framingham Heart Study): A community-based cohort study. Lancet 2009; 373: 739–745.
• 8.    Conen  D,  Tedrow  UB,  Koplan  BA,  Glynn  RJ,  Buring  JE,  Albert  CM. Influence of systolic and diastolic blood pressure on the risk of incident atrial fibrillation in women. Circulation 2009; 119: 2146–2152.
• 9.    Baber  U,  Howard  VJ,  Halperin  JL,  Soliman  EZ,  Zhang  X,  McClellan  W, et al. Association of chronic kidney disease with atrial fibrillation among adults in the United States: REasons for Geographic and Racial Differences in Stroke (REGARDS) Study. Circ Arrhythm Electrophysiol 2011; 4: 26–32.
• 10.    Iwasaki  YK,  Kato  T,  Xiong  F,  Shi  YF,  Naud  P,  Maguy  A, et al. Atrial fibrillation promotion with long-term repetitive obstructive sleep apnea in a rat model. J Am Coll Cardiol 2014; 64: 2013–2023.
• 11.    Pathak  RK,  Elliott  A,  Middeldorp  ME,  Meredith  M,  Mehta  AB,  Mahajan  R, et al. Impact of CARDIOrespiratory FITness on arrhythmia recurrence in obese individuals with atrial fibrillation: The CARDIO-FIT Study. J Am Coll Cardiol 2015; 66: 985–996.
• 12.    Marcus  GM,  Alonso  A,  Peralta  CA,  Lettre  G,  Vittinghoff  E,  Lubitz  SA, et al. European ancestry as a risk factor for atrial fibrillation in African Americans. Circulation 2010; 122: 2009–2015.
• 13.    Kim  SY,  Guevara  JP,  Kim  KM,  Choi  HK,  Heitjan  DF,  Albert  DA. Hyperuricemia and coronary heart disease: A systematic review and meta-analysis. Arthritis Care Res 2010; 62: 170–180.
• 14.    Tamariz  L,  Harzand  A,  Palacio  A,  Verma  S,  Jones  J,  Hare  J. Uric acid as a predictor of all-cause mortality in heart failure: A meta-analysis. Congest Heart Fail 2011; 17: 25–30.
• 15.    Mazza  A,  Pessina  AC,  Pavei  A,  Scarpa  R,  Tikhonoff  V,  Casiglia  E. Predictors of stroke mortality in elderly people from the general population: The CArdiovascular STudy in the ELderly. Eur J Epidemiol 2001; 17: 1097–1104.
• 16.    Takayama  S,  Kawamoto  R,  Kusunoki  T,  Abe  M,  Onji  M. Uric acid is an independent risk factor for carotid atherosclerosis in a Japanese elderly population without metabolic syndrome. Cardiovasc Diabetol 2012; 11: 2.
• 17.    Neogi  T,  Ellison  RC,  Hunt  S,  Terkeltaub  R,  Felson  DT,  Zhang  Y. Serum uric acid is associated with carotid plaques: The National Heart, Lung, and Blood Institute Family Heart Study. J Rheumatol 2009; 36: 378–384.
• 18.    Matsuo  S,  Imai  E,  Horio  M,  Yasuda  Y,  Tomita  K,  Nitta  K, et al. Revised equations for estimated GFR from serum creatinine in Japan. Am J Kidney Dis 2009; 53: 982–992.
• 19.    Huang  WM,  Hsu  PF,  Cheng  HM,  Lu  DY,  Cheng  YL,  Guo  CY, et al. Determinants and prognostic impact of hyperuricemia in hospitalization for acute heart failure. Circ J 2016; 80: 404–410.
• 20.    Tamariz  L,  Agarwal  S,  Soliman  EZ,  Chamberlain  AM,  Prineas  R,  Folsom  AR, et al. Association of serum uric acid with incident atrial fibrillation (from the Atherosclerosis Risk in Communities [ARIC] study). Am J Cardiol 2011; 108: 1272–1276.
• 21.    Suzuki  S,  Sagara  K,  Otsuka  T,  Matsuno  S,  Funada  R,  Uejima  T, et al. Gender-specific relationship between serum uric acid level and atrial fibrillation prevalence. Circ J 2012; 76: 607–611.
• 22.    Dawson  J,  Walters  M. Uric acid and xanthine oxidase: Future therapeutic targets in the prevention of cardiovascular disease? Br J Clin Pharmacol 2006; 62: 633–644.
• 23.    Kanellis  J,  Kang  DH. Uric acid as a mediator of endothelial dysfunction, inflammation, and vascular disease. Semin Nephrol 2005; 25: 39–42.
• 24.    Nattel  S,  Harada  M. Atrial remodeling and atrial fibrillation: Recent advances and translational perspectives. J Am Coll Cardiol 2014; 63: 2335–2345.
• 25.    Mahajan  R,  Lau  DH,  Brooks  AG,  Shipp  NJ,  Manavis  J,  Wood  JP, et al. Electrophysiological, electroanatomical, and structural remodeling of the atria as consequences of sustained obesity. J Am Coll Cardiol 2015; 66: 1–11.
• 26.    Dzeshka  MS,  Lip  GY,  Snezhitskiy  V,  Shantsila  E. Cardiac fibrosis in patients with atrial fibrillation: Mechanisms and clinical implications. J Am Coll Cardiol 2015; 66: 943–959.
• 27.    Lenaerts  I,  Driesen  RB,  Hermida  N,  Holemans  P,  Heidbüchel  H,  Janssens  S, et al. Role of nitric oxide and oxidative stress in a sheep model of persistent atrial fibrillation. Europace 2013; 15: 754–760.
• 28.    Yamashita  T,  Sekiguchi  A,  Iwasaki  YK,  Date  T,  Sagara  K,  Tanabe  H, et al. Recruitment of immune cells across atrial endocardium in human atrial fibrillation. Circ J 2010; 74: 262–270.
• 29.    Guo  Y,  Lip  GY,  Apostolakis  S. Inflammation in atrial fibrillation. J Am Coll Cardiol 2012; 60: 2263–2270.
• 30.    Chao  TF,  Hung  CL,  Chen  SJ,  Wang  KL,  Chen  TJ,  Lin  YJ, et al. The association between hyperuricemia, left atrial size and new-onset atrial fibrillation. Int J Cardiol 2013; 168: 4027–4032.
• 31.    Wright  AF,  Rudan  I,  Hastie  ND,  Campbell  H. A ‘complexity’ of urate transporters. Kidney Int 2010; 78: 446–452.
• 32.    Price  KL,  Sautin  YY,  Long  DA,  Zhang  L,  Miyazaki  H,  Mu  W, et al. Human vascular smooth muscle cells express a urate transporter. J Am Soc Nephrol 2006; 17: 1791–1795.
• 33.    Sugihara  S,  Hisatome  I,  Kuwabara  M,  Niwa  K,  Maharani  N,  Kato  M, et al. Depletion of uric acid due to SLC22A12 (URAT1) loss-of-function mutation causes endothelial dysfunction in hypouricemia. Circ J 2015; 79: 1125–1132.
• 34.    Maharani  N,  Ting  YK,  Cheng  J,  Hasegawa  A,  Kurata  Y,  Li  P, et al. Molecular mechanisms underlying urate-induced enhancement of Kv1.5 channel expression in HL-1 atrial myocytes. Circ J 2015; 79: 2659–2668.
• 35.    Verdecchia  P,  Schillaci  G,  Reboldi  G,  Santeusanio  F,  Porcellati  C,  Brunetti  P. Relation between serum uric acid and risk of cardiovascular disease in essential hypertension: The PIUMA study. Hypertension 2000; 36: 1072–1078.
• 36.    Mazza  A,  Zamboni  S,  Rizzato  E,  Pessina  AC,  Tikhonoff  V,  Schiavon  L, et al. Serum uric acid shows a J-shaped trend with coronary mortality in non-insulin-dependent diabetic elderly people: The CArdiovascular STudy in the ELderly (CASTEL). Acta Diabetol 2007; 44: 99–105.
• 37.    Ames  BN,  Cathcart  R,  Schwiers  E,  Hochstein  P. Uric acid provides an antioxidant defense in humans against oxidant- and radical-caused aging and cancer: A hypothesis. Proc Natl Acad Sci USA 1981; 78: 6858–6862.
• 38.    Maxwell  SR,  Thomason  H,  Sandler  D,  Leguen  C,  Baxter  MA,  Thorpe  GH, et al. Antioxidant status in patients with uncomplicated insulin-dependent and non-insulin-dependent diabetes mellitus. Eur J Clin Invest 1997; 27: 484–490.
• 39.    Sridhar  A,  Nishijima  Y,  Terentyev  D,  Khan  M,  Terentyeva  R,  Hamlin  RL, et al. Chronic heart failure and the substrate for atrial fibrillation. Cardiovasc Res 2009; 84: 227–236.
• 40.    Vest  JA,  Wehrens  XH,  Reiken  SR,  Lehnart  SE,  Dobrev  D,  Chandra  P, et al. Defective cardiac ryanodine receptor regulation during atrial fibrillation. Circulation 2005; 111: 2025–2032.
• 41.    Enomoto  A,  Kimura  H,  Chairoungdua  A,  Shigeta  Y,  Jutabha  P,  Cha  SH, et al. Molecular identification of a renal urate anion exchanger that regulates blood urate levels. Nature 2002; 417: 447–452.
• 42.    Dinour  D,  Gray  NK,  Campbell  S,  Shu  X,  Sawyer  L,  Richardson  W, et al. Homozygous SLC2A9 mutations cause severe renal hypouricemia. J Am Soc Nephrol 2010; 21: 64–72.
• 43.    Murakami  T,  Kawakami  H,  Fukuda  M,  Furukawa  S. Patients with renal hypouricemia are prone to develop acute renal failure: Why? Clin Nephrol 1995; 43: 207–208.
• 44.    Høieggen  A,  Alderman  MH,  Kjeldsen  SE,  Julius  S,  Devereux  RB,  De Faire  U, et al. The impact of serum uric acid on cardiovascular outcomes in the LIFE study. Kidney Int 2004; 65: 1041–1049.
• 45.    Ioannou  GN,  Boyko  EJ. Effects of menopause and hormone replacement therapy on the associations of hyperuricemia with mortality. Atherosclerosis 2013; 226: 220–227.
• 46.    Rajendra  NS,  Ireland  S,  George  J,  Belch  JJ,  Lang  CC,  Struthers  AD. Mechanistic insights into the therapeutic use of high-dose allopurinol in angina pectoris. J Am Coll Cardiol 2011; 58: 820–828.
• 47.    Doehner  W,  Schoene  N,  Rauchhaus  M,  Leyva-Leon  F,  Pavitt  DV,  Reaveley  DA, et al. Effects of xanthine oxidase inhibition with allopurinol on endothelial function and peripheral blood flow in hyperuricemic patients with chronic heart failure: Results from 2 placebo-controlled studies. Circulation 2002; 105: 2619–2624.
• 48.    Suresh  E. Diagnosis and management of gout: A rational approach. Postgrad Med J 2005; 81: 572–579.