Abstract
Background: The primary prevention of atrial fibrillation (AF), which increases mortality through complications including stroke and heart failure, is important. Excessive salt intake and low potassium intake are risk factors for cardiovascular disease; however, their association with AF remains inconclusive. This study investigated the association between sodium- and potassium-related urinary markers and AF prevalence.
Methods and Results: Data from the Tohoku Medical Megabank Project Community-based Cohort Study were used in this cross-sectional study. The urinary sodium-to-potassium (Na/K) ratio and estimated 24-h sodium and potassium excretion were calculated using spot urine samples and categorized into quartiles (Q1–Q4). The prevalence of AF was the primary outcome. Of the 26,506 participants (mean age 64.8 years; 33.2% males) included in this study, 630 (2.4%) had AF. Using Q1 as the reference group, the odds ratios for AF prevalence in Q4 were 1.35 (95% confidence interval [CI] 1.07–1.73) and 1.59 (95% CI 1.20–2.12) for 24-h estimated urinary Na/K ratio and estimated 24-h sodium excretion, respectively. Estimated 24-h potassium excretion was not associated with AF prevalence.
Conclusions: AF prevalence was positively associated with the urinary Na/K ratio and estimated 24-h urinary sodium excretion, but not with estimated 24-h urinary potassium excretion. Although further prospective studies are warranted, the findings of this study suggest that salt intake may be a modifiable risk factor for AF.
In 2016, it was estimated that atrial fibrillation (AF), the most prevalent type of arrhythmia, affected 46.3 million individuals globally;1 however, with the aging of the population, it is expected that its prevalence will increase.2,3 AF is associated with increased mortality risk due to the incidence of complications such as ischemic stroke and heart failure.3 Thus, the screening and prevention of AF is a major concern in the domain of public health.4,5
Several risk factors have been identified as for AF, including advanced age, obesity, hypertension, diabetes, obstructive sleep apnea, myocardial infarction, heart failure, smoking, alcohol consumption, and family history of AF.3 Lifestyle factors, such as obesity, alcohol consumption, and smoking, can be modified; consequently, these factors have been targeted for AF prevention.
Excessive salt intake and low potassium (vegetables and fruits) intake are modifiable lifestyle factors for cardiovascular diseases (CVDs).6,7 Dietary sodium restriction and the Dietary Approaches to Stop Hypertension diet, which is rich in fruits, vegetables, and low-fat dairy products, have been recommended for the management of hypertension, which is a major risk factor for CVD.8 Hypertension is also a major risk factor for AF; thus, excessive salt intake and low potassium intake may also be modifiable risk factors for AF.
The urinary sodium-to-potassium (Na/K) ratio, which reflects the balance between salt and potassium intake, is a risk factor for hypertension and CVD.9–11 However, to the best of our knowledge, only one study has evaluated the association between the urinary Na/K ratio and AF prevalence.12 An inverse association was observed in that study,12 which is in contrast to the findings of previous studies on the association between the urinary Na/K ratio and CVD.9–11 Another study reported a U-shaped association between urinary sodium excretion and AF prevalence.13 Thus, more studies are needed to evaluate the association between the urinary Na/K ratio and AF prevalence.
The present cross-sectional study used data from the Tohoku Medical Megabank Project Community-Based Cohort Study (TMM CommCohort Study) to investigate the association between the urinary Na/K ratio and AF prevalence and provide new insights into the role of the urinary Na/K ratio as a modifiable risk factor for AF. Estimated 24-h sodium and potassium excretion were also evaluated as components of the urinary Na/K ratio. Spot urine samples were used for analyses because these samples are easy to obtain and applicable to a large population.14
Methods
Study Participants
Data from the secondary survey of the TMM CommCohort Study, a community-based prospective cohort study conducted in Miyagi and Iwate prefectures in Japan,15,16 were used in the present cross-sectional study. The baseline survey of the TMM CommCohort Study was conducted between 2013 and 2016 in Miyagi Prefecture. Three major approaches were used to recruit patients: (1) a survey was conducted at specific health checkup sites (Type 1 survey) to collect basic information, blood and urine samples, questionnaire survey data, and municipal health examination data from sites of annual community health examinations; (2) additional Type 1 surveys were conducted at sites selected by the municipality and the TMM on dates other than those of the specific health examination in the municipality; and (3) a community support center-based survey (Type 2 survey) was conducted at the community support center to collect data regarding the results of physical examinations and blood and urine tests. The secondary survey, comprising examinations similar to the baseline survey and electrocardiography (ECG) performed at the community support centers, was conducted at the community support center between June 2017 and March 2021, approximately 4 years after the baseline survey.
The present study was conducted in accordance with the tenets of the Declaration of Helsinki and was approved by the Institutional Review Board of the Tohoku Medical Megabank Organization (Approval no. 2022-4-160). Written informed consent was obtained from all participants.
The inclusion criteria for the present study were as follows: (1) participation in the secondary survey of the TMM CommCohort Study in Miyagi; (2) age ≥40 years at the time of the secondary survey; (3) data available for the spot urine sample acquired at the time of the secondary survey; (4) no history of CVD based on self-administered questionnaires, including ischemic heart disease, heart failure, pacemaker implantation, and implantation of a implantable cardioverter defibrillator; and (5) no history of dialysis for chronic kidney disease. Patients aged <40 years were excluded from the study because the AF among young individuals is rare and may be attributed to cardiac diseases.
Of the 29,383 individuals who participated in the secondary survey of the TMM CommCohort Study, 1,517 aged <40 years, 1,067 with a history of CVDs, 14 who were undergoing dialysis, and 279 with missing urinary Na/K ratio data were excluded from the analysis (Figure).
Sodium- and Potassium-Related Urinary Markers: Urinary Na/K Ratio and Estimated 24-h Sodium and Potassium Excretion
The urinary Na/K ratio, estimated 24-h sodium excretion, and estimated 24-h potassium excretion were set as markers of dietary salt and potassium intake. Spot urine samples were collected at the community support centers. An ion-selective electrode method was used to measure urinary sodium and potassium concentrations. The Tanaka formula was used to calculate estimated 24-h sodium and potassium excretion.17 The urinary Na/K ratio was calculated by dividing the estimated 24-h sodium excretion by potassium excretion. Participants were divided into quartiles of urinary Na/K ratios and estimated 24-h sodium and potassium excretion.
ECG
A 12-lead ECG was recorded for each participant by a trained nurse at the community support centers (ECG-2500; Nihon Kohden Corporation, Tokyo, Japan). ECG findings were diagnosed by cardiologists in accordance with the Minnesota code.
Physical Examinations, Laboratory Data, and Questionnaires
The physical examinations comprised measurement of height (AD-6400; A&D Co., Ltd., Tokyo, Japan) and weight (InBody720; Biospace Co., Ltd., Seoul, Republic of Korea), and the calculation of body mass index. Systolic blood pressure (SBP) and diastolic blood pressure (DBP) were recorded twice by a trained nurse in seated participants after a 1- to 2-min rest. An electronic upper arm cuff device (HEM-9000AI; OMRON Corp., Kyoto, Japan) was used to measure SBP and DBP.18 The mean of the 2 blood pressure measurements was calculated and analyzed.
Data on the estimated glomerular filtration rate (eGFR) and levels of γ-glutamyl transpeptidase, HbA1c, total cholesterol, low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, triglyceride, and creatinine were extracted from blood test results. eGFR (mL/min/1.73 m2) was calculated using following formulas:
eGFR = 194 × Cr−1.094 × age (years)−0.287 in men
eGFR = 194 × Cr−1.094 × age (years)−0.287 × 0.739 in women
where Cr is creatinine (in mg/dL).
Self-administered questionnaires were used to collect data regarding lifestyle habits, such as smoking and drinking habits, as well as medical history and educational background. Participants were categorized as never smokers, former smokers, or current smokers. Drinking habits (including the type, frequency, and amount of alcohol consumed) were used to calculate alcohol intake per day as follows: the frequency of alcohol consumption per week was multiplied by the amount of alcohol consumed on a single occasion and the product was then divided by 7. Data regarding hypertension, diabetes, dyslipidemia, stroke, heart failure, and myocardial infarction were collected from the medical history questionnaire. Individuals who answered that they had hypertension in the questionnaire were defined as having a history of hypertension. Hypertension was defined as a history of hypertension or the blood pressure measured at the research center meeting the criteria for hypertension (SBP ≥140 mmHg or DBP ≥90 mmHg) in the analysis. Participants were divided into 3 groups based on education status: less than high school, high school, or more than high school.
Outcome
The primary outcome measure of this study was the prevalence of AF, defined as the presence of findings suggestive of AF on ECG (Minnesota code: 8-3-1) or a history of AF according to questionnaire responses.
Statistical Analyses
Patients were divided into groups according to urinary Na/K ratio quartiles (Q1–Q4). Normally distributed continuous variables are presented as the mean±SD; skewed continuous variables are presented as the median with interquartile range. Categorical variables are presented as numbers and percentages. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated using multivariable logistic regression models to evaluate the association between AF prevalence and the urinary Na/K ratio, estimated 24-h sodium excretion, and estimated 24-h potassium excretion. A complete case analysis was performed using 3 models. Model 1 included age and sex as covariates. In addition to the covariates included in Model 1, Model 2 included body mass index, history of diabetes, low-density lipoprotein cholesterol and creatinine levels, alcohol consumption, smoking status, and education levels as covariates (estimated 24-h potassium excretion was added to Model 2 for the analysis of estimated 24-h sodium excretion, and estimated 24-h sodium excretion was added to Model 2 for the analysis of estimated 24-h potassium excretion). Hypertension may mediate the association between the sodium- and potassium-related urinary markers and AF prevalence. Therefore, hypertension was not included in Model 2. Model 3 included all the covariates in Model 2 and hypertension. All data used in the models were collected during the secondary survey.
Subgroup analyses were performed according to sex, age (≤65 and > 65 years), hypertension, and eGFR (eGFR <60 and eGFR ≥60 mL/min/1.73 m2), which are important risk factors for AF. Hypertension and eGFR are correlated with the urinary Na/K ratio. The cut-off value for age was the rounded-off median. Model 3 was used to calculate the ORs and 95% CIs for AF prevalence. The subgroup analysis of hypertension included SBP in the model instead of hypertension. The interaction between each category and AF was analyzed by adding variables and multiplying the AF and categories in the model.
Two-tailed P<0.05 was considered statistically significant. All statistical analyses were performed using R version 4.2.1 for Linux.
Results
In all, 26,506 participants (mean age 64.8 years; 33.2% males) who met the eligibility criteria were included in the analysis. Table 1 presents participant characteristics according to urinary Na/K ratio quartiles. Q4 had the oldest participants; moreover, it had the highest proportion of male participants. SBP, DBP, and the prevalence of hypertension were highest in Q4 (Q4 vs. Q1: mean SBP, 134.2 vs. 129.3 mmHg, respectively; mean DBP, 79.4 vs. 77.2 mmHg, respectively; and prevalence of hypertension, 55.8% vs. 46.5%, respectively).
Table 1.
Participant Characteristics According to Urinary Sodium-to-Potassium Ratio Quartiles
| |
Urinary Na/K ratio |
P for
trend |
| Q1 |
Q2 |
Q3 |
Q4 |
| Cut-off |
<3.7 |
≥3.7, <4.3 |
≥4.3, <4.9 |
≥4.9 |
|
| No. patients |
6,239 |
6,966 |
6,510 |
6,791 |
|
| Age (years) |
64.3±10.2 |
64.9±9.7 |
65.2±9.8 |
65.0±9.9 |
<0.001 |
| Male sex |
1,863 (29.9) |
2,212 (31.8) |
2,168 (33.3) |
2,561 (37.7) |
<0.001 |
| BMI (kg/m2) |
23.2±3.5 |
23.2±3.3 |
23.3±3.4 |
23.3±3.6 |
0.059 |
| SBP (mmHg) |
129.3±17.7 |
131.0±17.6 |
132.5±18.0 |
134.2±17.8 |
<0.001 |
| DBP (mmHg) |
77.2±10.8 |
77.8±10.5 |
78.4±10.9 |
79.4±11.0 |
<0.001 |
| Heart rate (beats/min) |
68.3±10.5 |
67.9±10.2 |
67.7±10.1 |
67.6±10.0 |
<0.001 |
| T-Chol (mg/dL) |
211.3±35.0 |
210.2±33.9 |
209.4±34.7 |
206.2±34.7 |
<0.001 |
| LDL-C (mg/dL) |
123.2±30.4 |
122.6±29.5 |
122.2±30.3 |
119.7±30.1 |
<0.001 |
| HbA1c (%) |
5.6±0.6 |
5.7±0.5 |
5.7±0.6 |
5.7±0.5 |
0.137 |
| Cr (mg/dL) |
0.69±0.17 |
0.69±0.17 |
0.70±0.19 |
0.71±0.22 |
<0.001 |
| eGFR |
74.8±13.9 |
74.7±13.8 |
74.5±14.3 |
75.4±14.8 |
0.026 |
| Medical history of hypertension |
1,784 (28.6) |
1,885 (27.1) |
1,857 (28.5) |
2,087 (30.7) |
0.002 |
| HypertensionA |
2,902 (46.5) |
3,346 (48.0) |
3,307 (50.8) |
3,789 (55.8) |
<0.001 |
| Medical history of diabetes |
377 (6.0) |
475 (6.8) |
479 (7.4) |
525 (7.7) |
<0.001 |
| Medical history of dyslipidemia |
1,055 (16.9) |
1,198 (17.2) |
1,140 (17.5) |
1,051 (15.5) |
0.036 |
| Medical history of stroke |
77 (1.2) |
100 (1.4) |
104 (1.6) |
100 (1.5) |
0.195 |
| Urinary Na/K ratio |
3.2±0.4 |
4.0±0.2 |
4.6±0.2 |
5.6±0.6 |
<0.001 |
| Estimated 24-h Na excretion (mg/day) |
3,169±712 |
3,608±690 |
3,913±708 |
4,338±825 |
<0.001 |
| Estimated 24-h K excretion (mg/day) |
1,672±358 |
1,531±293 |
1,451±262 |
1,333±249 |
<0.001 |
| Alcohol consumption (g/day) |
0.0 [0.0–11.0] |
0.0 [0.0–12.3] |
0.0 [0.0–13.5] |
0.0 [0.0–20.0] |
<0.001 |
| Smoking status |
| Never smoked |
4,387 (68.7) |
4,696 (67.4) |
4,222 (64.9) |
4,130 (61.4) |
<0.001 |
| Former smoker |
1,403 (22.5) |
1,638 (23.5) |
1,717 (26.4) |
1,932 (28.4) |
<0.001 |
| Current smoker |
514 (8.2) |
585 (8.4) |
531 (8.2) |
638 (9.4) |
0.033 |
| Education |
| Less than high school |
556 (8.9) |
640 (9.2) |
668 (10.3) |
771 (11.4) |
<0.001 |
| High school |
3,074 (49.3) |
3,592 (51.6) |
3,400 (52.2) |
3,603 (53.1) |
<0.001 |
| More than high school |
2,554 (40.9) |
2,675 (38.4) |
2,395 (36.8) |
2,375 (35.0) |
<0.001 |
Unless indicated otherwise, data are given as the mean±SD, median [interquartile range], or n (%). AHypertension was defined as a history of hypertension or blood pressure (BP) measured at the research center meeting the criteria for hypertension (systolic blood pressure [SBP] ≥140 mmHg or diastolic blood pressure [DBP] ≥90 mmHg). BMI, body mass index; Cr, creatinine; eGFR, estimated glomerular filtration rate; K, potassium; LDL-C, low-density lipoprotein cholesterol; Na, sodium; T-Chol, total cholesterol.
Of the 26,506 participants, 630 (2.4%), comprising 437 (5.0%) men and 193 (1.1%) women, had AF. A total of 336 participants had findings suggestive of AF on the ECG, and 462 participants indicated that they had AF on the questionnaire. Table 2 presents the prevalence of AF and ORs according to urinary Na/K ratio, estimated 24-h sodium excretion, and estimated 24-h potassium excretion quartiles. AF prevalence increased along the quartiles in all categories. Multivariable logistic regression analyses performed using Model 2 revealed a positive association between the urinary Na/K ratio and AF prevalence (P for trend=0.013). Using Q1 as the reference, the OR of the urinary Na/K ratio for the prevalence of AF in Q4 was 1.37 (95% CI 1.08–1.75). Estimated 24-h sodium excretion was positive associated with AF prevalence (P for trend=0.005). Using Q1 as the reference, the OR for the prevalence of AF in Q4 was 1.60 (95% CI 1.20–2.13) in Model 2. No statistically significant association was observed between estimated 24-h potassium excretion and AF prevalence. No significant changes in the findings were observed when hypertension was included as a covariate in Model 3.
Table 2.
ORs and 95% CIs for the Prevalence of AF According to Urinary Sodium-to-Potassium Ratio, 24-h Sodium Excretion, and 24-h Potassium Excretion Quartiles (Q1–Q4)
| |
Q1 |
Q2 |
Q3 |
Q4 |
P for
trend |
| Urinary Na/K ratio |
| Cut-off |
<3.7 |
≥3.7, <4.3 |
≥4.3, <4.9 |
≥4.9 |
|
| AF prevalence |
120/6,239 (1.9) |
156/6,966 (2.2) |
153/6,510 (2.4) |
201/6,791 (3.0) |
|
| |
OR (95% CIs) |
|
| Crude |
Ref. |
1.18 (0.94–1.49) |
1.31 (1.05–1.65)* |
1.56 (1.26–1.94)*** |
<0.001 |
| Model 1 |
Ref. |
1.12 (0.89–1.41) |
1.18 (0.94–1.48) |
1.36 (1.09–1.70)** |
0.006 |
| Model 2 |
Ref. |
1.14 (0.89–1.47) |
1.13 (0.88–1.45) |
1.37 (1.08–1.75)** |
0.013 |
| Model 3 |
Ref. |
1.14 (0.90–1.47) |
1.12 (0.87–1.44) |
1.35 (1.07–1.73)* |
0.017 |
| Estimated 24-h Na excretion |
| Cut-off (mg/day) |
<3,171 |
≥3,171, <3,699 |
≥3,699, <4,284 |
≥4,284 |
|
| AF prevalence |
85/6,627 (1.3) |
147/6,620 (2.2) |
170/6,627 (2.6) |
228/6,632 (3.4) |
|
| |
OR (95% CIs) |
|
| Crude |
Ref. |
1.75 (1.34–2.30)*** |
2.03 (1.56–2.64)*** |
2.74 (2.14–3.54)*** |
<0.001 |
| Model 1 |
Ref. |
1.41 (1.08–1.86)* |
1.39 (1.07–1.82)* |
1.61 (1.25–2.09)*** |
<0.001 |
| Model 2 |
Ref. |
1.44 (1.09–1.91)* |
1.39 (1.05–1.84)* |
1.60 (1.20–2.13)** |
0.005 |
| Model 3 |
Ref. |
1.45 (1.10–1.92)* |
1.39 (1.05–1.85)* |
1.59 (1.20–2.12)** |
0.007 |
| Estimated 24-h K excretion |
| Cut-off (mg/day) |
<1,271 |
≥1,271, <1,449 |
≥1,449 , <1,673 |
≥1,673 |
|
| AF prevalence |
110/6,605 (1.7) |
139/6,661 (2.1) |
183/6,612 (2.8) |
198/6,628 (3.0) |
|
| |
OR (95% CIs) |
|
| Crude |
Ref. |
1.26 (0.98–1.62) |
1.68 (1.33–2.14)*** |
1.82 (1.44–2.31)*** |
<0.001 |
| Model 1 |
Ref. |
1.06 (0.82–1.37) |
1.27 (0.99–1.62) |
1.17 (0.92–1.49) |
0.11 |
| Model 2 |
Ref. |
1.00 (0.77–1.31) |
1.15 (0.88–1.49) |
0.97 (0.74–1.28) |
0.98 |
| Model 3 |
Ref. |
1.00 (0.77–1.31) |
1.16 (0.89–1.50) |
0.98 (0.75–1.30) |
0.92 |
Atrial fibrillation (AF) prevalence data are given as n/N (%). *P<0.05, **P<0.01, ***P<0.001 compared with Q1. Hypertension was defined as a medical history of hypertension or blood pressure measured at the research center meeting the criteria for hypertension (systolic blood pressure ≥140 mmHg or diastolic blood pressure ≥90 mmHg). Model 1 included age and sex. Model 2 included Model 1, body mass index, a history of diabetes, low-density lipoprotein cholesterol and creatinine levels, alcohol consumption, smoking status, and education level (estimated 24-h potassium [K] excretion was added to Model 2 for estimated 24-h sodium [Na] excretion, and estimated 24-h Na excretion was added to Model 2 for estimated 24-h K excretion). Model 3 included Model 2 and hypertension. CI, confidence interval; OR, odds ratio.
Tables 3 and 4 present the results of subgroup analyses of the association between AF prevalence and the urinary Na/K ratio and estimated 24-h sodium excretion, respectively. Although the OR was higher in the female, without hypertension, and normal eGFR subgroups for both urinary estimated 24-h Na/K ratio and estimated 24-h sodium excretion, an interaction effect with AF prevalence was not observed in any subgroup.
Table 3.
Subgroup Analyses of ORs and 95% CIs for the Prevalence of AF According to Urinary Sodium-to-Potassium Ratio Quartiles (Q1–Q4)
| |
Q1 |
Q2 |
Q3 |
Q4 |
P for
trend |
P for
interaction |
| Sex |
| Male sex |
93/2,167 (4.3) |
113/2,281 (5.0) |
100/2,060 (4.9) |
131/2,296 (5.7) |
|
0.19 |
| Ref. |
1.09 (0.82–1.46) |
1.04 (0.77–1.40) |
1.30 (0.98–2.72) |
0.089 |
|
| Female sex |
37/4,376 (0.8) |
55/4,754 (1.4) |
41/4,342 (0.9) |
60/4,230 (1.4) |
|
|
| Ref. |
1.47 (0.95–2.31) |
1.21 (0.77–1.96) |
1.72 (1.12–2.70)* |
0.015 |
|
| Age |
| ≤65 years |
13/2,774 (0.5) |
17/2,854 (0.6) |
30/2,841 (1.1) |
27/2,736 (1.0) |
|
0.88 |
| Ref. |
1.16 (0.56–2.46) |
1.67 (0.87–3.39) |
1.57 (0.80–3.22) |
0.130 |
|
| >65 years |
111/4,037 (2.7) |
121/3,454 (3.5) |
135/3,856 (3.5) |
176/3,954 (4.5) |
|
|
| Ref. |
1.27 (0.97–1.67) |
1.21 (0.93–1.59) |
1.50 (1.17–1.93)** |
0.003 |
|
| Hypertension |
| Without |
37/3,337 (1.1) |
43/3,014 (1.4) |
49/3,335 (1.5) |
67/3,476 (1.9) |
|
0.22 |
| Ref. |
1.46 (0.92–2.35) |
1.40 (0.89–2.23) |
1.76 (1.15–2.75)* |
0.017 |
|
| With |
96/3,374 (2.8) |
113/3,479 (3.2) |
90/3,130 (2.9) |
135/3,361 (4.0) |
|
|
| Ref. |
1.04 (0.78–1.39) |
0.91 (0.67–1.23) |
1.28 (0.97–1.69) |
0.130 |
|
| eGFR |
| <60 mL/min/1.73 m2 |
34/834 (3.1) |
42/917 (4.6) |
45/879 (5.1) |
49/853 (5.7) |
|
0.77 |
| Ref. |
1.06 (0.65–1.75) |
1.15 (0.71–1.87) |
1.22 (0.76–1.98) |
0.380 |
|
| ≥60 mL/min/1.73 m2 |
86/5,411 (1.6) |
114/6,048 (1.9) |
107/5,627 (1.9) |
152/5,936 (2.6) |
|
|
| Ref. |
1.17 (0.87–1.57) |
1.10 (0.82–1.49) |
1.44 (1.09–1.90)* |
0.015 |
|
Unless indicated otherwise, data are given as n/N (%) or as the OR (95% CI). *P<0.05, **P<0.01, ***P<0.001 compared with Q1. Model 3, which included age, sex, body mass index, diabetes, low-density lipoprotein cholesterol and creatinine levels, alcohol consumption, smoking status, education level, and hypertension, was used for analysis. In the subgroup analysis of hypertension, systolic blood pressure was included in the model instead of hypertension. Hypertension was defined as a medical history of hypertension or blood pressure measured at the research center meeting the criteria for hypertension (systolic blood pressure ≥140 mmHg or diastolic blood pressure ≥90 mmHg). eGFR, estimated glomerular filtration rate. Other abbreviations as in Table 2.
Table 4.
Subgroup Analyses of ORs and 95% CIs for the Prevalence of AF According to Estimated 24-h Sodium Excretion Quartiles (Q1–Q4)
| |
Q1 |
Q2 |
Q3 |
Q4 |
P for
trend |
P for
interaction |
| Sex |
| Male |
80/2,200 (3.6) |
105/2,202 (4.8) |
110/2,198 (5.0) |
142/2,204 (6.4) |
|
0.18 |
| Ref. |
1.16 (0.86–1.59) |
1.11 (0.81–1.53) |
1.44 (1.05–1.98)* |
0.038 |
|
| Female |
30/4,425 (0.7) |
41/4,426 (0.9) |
61/4,427 (1.4) |
61/4,424 (1.4) |
|
|
| Ref. |
1.37 (0.83–2.28) |
1.82 (1.14–2.98)* |
1.60 (0.97–2.72) |
0.049 |
|
| Age |
| ≤65 years |
14/2,797 (0.5) |
16/2,807 (0.6) |
29/2,798 (1.0) |
28/2,803 (1.0) |
|
0.70 |
| Ref. |
0.95 (0.46–1.99) |
1.25 (0.64–2.52) |
1.06 (0.52–2.24) |
0.712 |
|
| >65 years |
96/3,824 (2.5) |
130/3,821 (3.4) |
128/3,838 (3.3) |
189/3,828 (4.9) |
|
|
| Ref. |
1.34 (1.02–1.78)* |
1.22 (0.92–1.64) |
1.69 (1.27–2.26)*** |
0.001 |
|
| Hypertension |
| Without |
20/3,291 (0.6) |
41/3,288 (1.2) |
53/3,294 (1.6) |
82/3,289 (2.5) |
|
0.10 |
| Ref. |
1.82 (1.04–3.29)* |
2.09 (1.22–3.75)** |
2.60 (1.50–4.71)*** |
0.001 |
|
| With |
77/3,334 (2.3) |
103/3,333 (3.1) |
110/3,345 (3.3) |
144/3,332 (4.3) |
|
|
| Ref. |
1.26 (0.92–1.73) |
1.23 (0.89–1.70) |
1.54 (1.11–2.14)** |
0.015 |
|
| eGFR |
| <60 mL/min/1.73 m2 |
27/868 (3.1) |
50/869 (5.8) |
38/867 (4.4) |
55/868 (6.3) |
|
0.60 |
| Ref. |
1.74 (1.06–2.93)* |
1.02 (0.59–1.78) |
1.32 (0.77–2.30) |
0.887 |
|
| ≥60 mL/min/1.73 m2 |
60/5,757 (1.0) |
98/5,758 (1.7) |
129/5,748 (2.2) |
172/5,759 (3.0) |
|
|
| Ref. |
1.33 (0.96–1.87) |
1.52 (1.10–2.12)* |
1.75 (1.26–2.47)** |
<0.001 |
|
Unless indicated otherwise, data are given as n/N (%) or as the OR (95% CI). *P<0.05, **P<0.01, ***P<0.001 compared with Q1. Model 3, which included age, sex, body mass index, diabetes, low-density lipoprotein cholesterol and creatinine levels, alcohol consumption, smoking status, education level, and hypertension, was used for analysis. In the subgroup analysis of hypertension, systolic blood pressure was included in the model instead of hypertension. Hypertension was defined as a medical history of hypertension or blood pressure measured at the research center meeting the criteria for hypertension (systolic blood pressure ≥140 mmHg or diastolic blood pressure ≥90 mmHg). eGFR, estimated glomerular filtration rate. Other abbreviations as in Table 2.
Discussion
This cross-sectional study, which included 26,506 participants from a community-based cohort, revealed that a higher urinary Na/K ratio and higher estimated 24-h sodium excretion using spot urine samples were positively associated with a higher AF prevalence. The finding suggests that salt intake may affect AF prevalence. In contrast, there was no significant association between estimated 24-h potassium excretion and AF prevalence.
Previous studies have evaluated the association between sodium intake or sodium-related urinary markers and AF. A study conducted in Finland revealed that the incidence of new-onset AF increased when dietary sodium intake increased.19 However, this measure may be biased, because dietary sodium intake was estimated based on self-reported 7-day food consumption records, highlighting the need for objective biomarkers. Two studies by the UK Biobank evaluated sodium-related urinary markers derived from spot urine samples.12,13 One study found an inverse association between the urinary Na/K ratio and the incidence of new-onset AF,12 whereas the other observed a U-shaped association between estimated 24-h sodium excretion (based on spot urine samples) and the incidence of new-onset AF, specifically in men.13 The inconsistent findings between these studies may stem from differences in urinary markers and covariates. The first study assessed the urinary Na/K ratio (sodium concentration divided by potassium concentration), whereas the second estimated 24-h sodium excretion using the Kawasaki formula. In addition, hypertension was included as a covariate in the latter study, but not in the former.
Building on previous studies, we assessed the urinary Na/K ratio along with estimated 24-h sodium and potassium excretion. In addition, to examine the impact of hypertension, we used several models both with and without hypertension as a covariate. Our results demonstrated a positive association between sodium-related urinary markers and AF prevalence, suggesting that calculating the urinary Na/K ratio by dividing sodium excretion by potassium excretion may weaken the observed association with AF prevalence. However, unlike previous studies, the present study is cross-sectional; therefore, we evaluated the association with the prevalence of AF rather than with new-onset AF.
The development of AF can be attributed to left atrial structural remodeling and electrophysiological disturbances.20 The association between sodium-related urinary markers and AF prevalence can be explained by the mechanisms described below.
First, excessive salt intake and low potassium intake increase BP; thus, hypertension may partially mediate the association between the sodium-related urinary markers and AF prevalence. Hypertension can induce stiffening of the blood vessels, which leads to increased left ventricular diastolic pressures and overload and retrograde dilation of the left atrium.21 Angiotensin II and inflammation, which are activated in patients with hypertension, promote fibrosis in the myocardium. This results in the dilation of the left atrium, thereby inducing electrical disturbances.22,23 To assess the effect of hypertension, we used Model 2, which did not include hypertension as a covariate. However, an association between the sodium-related urinary markers and AF prevalence was observed even after adjusting for hypertension in Model 3, suggesting the presence of a mechanism without hypertension. Although no interaction effect was observed, the association between the sodium-related urinary markers and AF development was prominent in the normotensive subgroup.
Second, aldosterone levels may explain the association between the sodium-related urinary markers and AF prevalence. Previous studies reported that high sodium intake increased the cardiac synthesis of aldosterone independent of the circulating renin-angiotensin-aldosterone system in an animal model.24,25 Aldosterone can induce electrophysiological disturbances by modulating myocardial calcium channels and facilitating arrhythmia.26 Aldosterone was not measured in the present study, and further studies are needed to clarify the mechanism.
Third, salt intake may induce electrophysiological disturbances. A study on humans reported that a loading dose of 18 g salt administered for 7 days induced QT interval prolongation on ECG.27 A prolonged QT interval, which reflects the ventricular effective refractory period, is associated with an increased risk of developing AF.28,29
Potassium plays an important role in cardiac electrophysiology, and hypokalemia induces AF.30 However, in the present study, there was no significant association between estimated 24-h potassium excretion and AF prevalence. Homeostasis may have contributed to these results. However, data regarding the serum potassium levels were unavailable; therefore, the true mechanism remains unclear.
The present study demonstrated that sodium-related urinary markers may aid in identifying populations at high risk of developing AF, thereby enhancing the effectiveness of AF prevention and screening efforts. In addition, they may aid in promoting reduced dietary salt intake. The estimated 24-h sodium excretion and urinary Na/K ratio can be derived from blood and urine tests, facilitating their adoption in clinical settings. Further studies must be conducted to confirm the effectiveness of reducing dietary salt intake in preventing AF development.
The present study has some limitations. First, AF was diagnosed using ECGs, not Holter ECGs, and self-report questionnaires. A previous study reported that the sensitivity and specificity of self-reported AF for AF registered in medical records were 49.6% and 99.2%, respectively.31 The prevalence of AF was higher in the present study than in a previous study conducted in Japan.32 However, undiagnosed AF may have been included in the group without AF, which may have led to the underestimation of the results. Second, information regarding the use of diuretics and AF treatments was not available. Some diuretics inhibit sodium and potassium excretion;33 thus, the effect of diuretics on the urinary Na/K ratio is unknown. Third, although the gold standard is 24-h urinary collection, in the present study we estimated 24-h urinary sodium and potassium excretion using spot urine samples acquired at only one point.34 Fourth, average salt intake in Japan is higher than in other countries,35 and it is unclear whether these findings can be generalized to other populations. Finally, the causal association was unclear because this was a cross-sectional study. The participants will be followed up in the future to address this problem.
In conclusion, the present study revealed that a high urinary Na/K ratio and high sodium excretion in spot urine samples were positively associated with a high prevalence of AF, regardless of the presence of other risk factors, including hypertension. Although further prospective studies are warranted, the urinary Na/K ratio and estimated 24-h sodium excretion, as markers of salt intake, may be useful for preventing AF.
Acknowledgments
The authors thank the members of the Tohoku Medical Megabank Organization, including the Genome Medical Research Coordinators and the office and administrative personnel, for their assistance. A complete list of members is available at https://www.megabank.tohoku.ac.jp/english/a230901/.
Sources of Funding
This research was supported, in part, by the Japan Agency for Medical Research and Development (AMED; Grant no. JP21tm0124005). This research used a supercomputer system provided by the Tohoku Medical Megabank Project (funded by AMED under Grant no. JP21tm0424601).
Disclosures
S.Y. is a member of Circulation Journal’s Editorial Team. The authors do not have any conflicts of interest to disclose with respect to this manuscript.
IRB Information
This study was approved by the Institutional Review Board of the Tohoku Medical Megabank Organization (Approval no. 2022-4-160).
References
- 1.
Benjamin EJ, Muntner P, Alonso A, Bittencourt MS, Callaway CW, Carson AP, et al. Heart disease and stroke statistics – 2019 update: A report from the American Heart Association. Circulation 2019; 139: e56–e528.
- 2.
Naccarelli GV, Varker H, Lin J, Schulman KL. Increasing prevalence of atrial fibrillation and flutter in the United States. Am J Cardiol 2009; 104: 1534–1539.
- 3.
Kornej J, Börschel CS, Benjamin EJ, Schnabel RB. Epidemiology of atrial fibrillation in the 21st century: Novel methods and new insights. Circ Res 2020; 127: 4–20.
- 4.
Watanabe E, Takahashi N, Aronson R, Ohsawa A, Ishibashi Y, Murakawa Y; the SCAN-AF Investigators. Systematic screening for atrial fibrillation in patients at moderate-to-high risk of stroke: Potential to increase the atrial fibrillation detection rate (SCAN-AF). Circ J 2022; 86: 1245–1251.
- 5.
Arafa A, Kokubo Y, Kashima R, Teramoto M, Sakai Y, Nosaka S, et al. Association between white blood cell count and atrial fibrillation risk: A population-based prospective cohort study. Circ J 2023; 87: 41–49.
- 6.
He FJ, Tan M, Ma Y, MacGregor GA. Salt reduction to prevent hypertension and cardiovascular disease: JACC state-of-the-art review. J Am Coll Cardiol 2020; 75: 632–647.
- 7.
Aune D, Giovannucci E, Boffetta P, Fadnes LT, Keum N, Norat T, et al. Fruit and vegetable intake and the risk of cardiovascular disease, total cancer and all-cause mortality: A systematic review and dose-response meta-analysis of prospective studies. Int J Epidemiol 2017; 46: 1029–1056.
- 8.
Juraschek SP, Miller ER, Weaver CM, Appel LJ. Effects of sodium reduction and the DASH diet in relation to baseline blood pressure. J Am Coll Cardiol 2017; 70: 2841–2848.
- 9.
Ma Y, He FJ, Sun Q, Yuan C, Kieneker LM, Curhan GC, et al. 24-Hour urinary sodium and potassium excretion and cardiovascular risk. N Engl J Med 2022; 386: 252–263.
- 10.
Hirata T, Kogure M, Tsuchiya N, Miyagawa K, Narita A, Nochioka K, et al. Impacts of the urinary sodium-to-potassium ratio, sleep efficiency, and conventional risk factors on home hypertension in a general Japanese population. Hypertens Res 2021; 44: 858–865.
- 11.
Muroya T, Satoh M, Murakami T, Nakayama S, Asayama K, Hirose T, et al. Association between urinary sodium-to-potassium ratio and home blood pressure and ambulatory blood pressure: The Ohasama study. J Hypertens 2022; 40: 862–869.
- 12.
Zanetti D, Bergman H, Burgess S, Assimes TL, Bhalla V, Ingelsson E. Urinary albumin, sodium, and potassium and cardiovascular outcomes in the UK Biobank: Observational and Mendelian randomization analyses. Hypertension 2020; 75: 714–722.
- 13.
Wuopio J, Orho-Melander M, Ärnlöv J, Nowak C. Estimated salt intake and risk of atrial fibrillation in a prospective community-based cohort. J Intern Med 2021; 289: 700–708.
- 14.
Kogure M, Nakaya N, Hirata T, Tsuchiya N, Nakamura T, Narita A, et al. Sodium/potassium ratio change was associated with blood pressure change: Possibility of population approach for sodium/potassium ratio reduction in health checkup. Hypertens Res 2021; 44: 225–231.
- 15.
Hozawa A, Tanno K, Nakaya N, Nakamura T, Tsuchiya N, Hirata T, et al. Study profile of the Tohoku Medical Megabank Community-Based Cohort Study. J Epidemiol 2021; 31: 65–76.
- 16.
Hozawa A, Nakaya K, Nakaya N, Nakamura T, Kogure M, Hatanaka R, et al. Progress report of the Tohoku Medical Megabank Community-Based Cohort Study: Study profile of the repeated center-based survey during second period in Miyagi Prefecture. J Epidemiol 2024; 34: 434–443.
- 17.
Tanaka T, Okamura T, Miura K, Kadowaki T, Ueshima H, Nakagawa H, et al. A simple method to estimate populational 24-h urinary sodium and potassium excretion using a casual urine specimen. J Hum Hypertens 2002; 16: 97–103.
- 18.
Usuzaki T, Ishikuro M, Metoki H, Murakami K, Noda A, Ueno F, et al. Comparison among research, home, and office blood pressure measurements for pregnant women: The TMM BirThree Cohort Study. J Clin Hypertens (Greenwich) 2020; 22: 2004–2013.
- 19.
Pääkkö TJW, Perkiömäki JS, Silaste ML, Bloigu R, Huikuri HV, Antero Kesäniemi Y, et al. Dietary sodium intake is associated with long-term risk of new-onset atrial fibrillation. Ann Med 2018; 50: 694–703.
- 20.
Wijesurendra RS, Casadei B. Mechanisms of atrial fibrillation. Heart 2019; 105: 1860–1867.
- 21.
Dzeshka MS, Shahid F, Shantsila A, Lip GYH. Hypertension and atrial fibrillation: An intimate association of epidemiology, pathophysiology, and outcomes. Am J Hypertens 2017; 30: 733–755.
- 22.
González A, Ravassa S, López B, Moreno MU, Beaumont J, San José G, et al. Myocardial remodeling in hypertension. Hypertension 2018; 72: 549–558.
- 23.
Chae CU, Lee RT, Rifai N, Ridker PM. Blood pressure and inflammation in apparently healthy men. Hypertension 2001; 38: 399–403.
- 24.
Silvestre JS, Robert V, Heymes C, Aupetit-Faisant B, Mouas C, Moalic JM, et al. Myocardial production of aldosterone and corticosterone in the rat: Physiological regulation. J Biol Chem 1998; 273: 4883–4891.
- 25.
Copyright E, Society TE. Sodium-induced cardiac aldosterone synthesis causes cardiac hypertrophy. Endocrinology 2000; 141: 1901–1904.
- 26.
Buffolo F, Tetti M, Mulatero P, Monticone S. Aldosterone as a mediator of cardiovascular damage. Hypertension 2022; 79: 1899–1911.
- 27.
He M, Mu J, Liu F, Ren K, Wang Y, Guo T, et al. Effects of a high salt intake and potassium supplementation on QT interval dispersion in normotensive healthy subjects. Intern Med 2015; 54: 295–301.
- 28.
Mandyam MC, Soliman EZ, Alonso A, Dewland TA, Heckbert SR, Vittinghoff E, et al. The QT interval and risk of incident atrial fibrillation. Heart Rhythm 2013; 10: 1562–1568.
- 29.
Zhang N, Gong M, Tse G, Zhang Z, Meng L, Yan BP, et al. Prolonged corrected QT interval in predicting atrial fibrillation: A systematic review and meta-analysis. Pacing Clin Electrophysiol 2018; 41: 321–327.
- 30.
Tazmini K, Frisk M, Lewalle A, Laasmaa M, Morotti S, Lipsett DB, et al. Hypokalemia promotes arrhythmia by distinct mechanisms in atrial and ventricular myocytes. Circ Res 2020; 126: 889–906.
- 31.
Malmo V, Langhammer A, Bønaa KH, Loennechen JP, Ellekjaer H. Validation of self-reported and hospital-diagnosed atrial fibrillation: The HUNT-study. Clin Epidemiol 2016; 8: 185–193.
- 32.
Inoue H, Fujiki A, Origasa H, Ogawa S, Okumura K, Kubota I, et al. Prevalence of atrial fibrillation in the general population of Japan: An analysis based on periodic health examination. Int J Cardiol 2009; 137: 102–107.
- 33.
Bernstein PL, Ellison DH. Diuretics and salt transport along the nephron. Semin Nephrol 2011; 31: 475–482.
- 34.
Iwahori T, Ueshima H, Miyagawa N, Ohgami N, Yamashita H, Ohkubo T, et al. Six random specimens of daytime casual urine on different days are sufficient to estimate daily sodium/potassium ratio in comparison to 7-day 24-h urine collections. Hypertens Res 2014; 37: 765–771.
- 35.
Brown IJ, Tzoulaki I, Candeias V, Elliott P. Salt intakes around the world: Implications for public health. Int J Epidemiol 2009; 38: 791–813.
