The association between serum uric acid (SUA) levels and atrial fibrillation (AF) is currently poorly known. We examined the association between SUA levels and risk of incident AF in patients with type 2 diabetes mellitus. We followed for 10 years a random sample of 400 type 2 diabetic outpatients, who were free from AF at baseline. A standard 12-lead electrocardiography was undertaken annually and a diagnosis of incident AF was confirmed in affected participants by a single cardiologist. Over 10 years, there were 42 incident AF cases (cumulative incidence of 10.5%). Elevated SUA level was associated with an increased risk of incident AF (odds ratio 2.43, 95% confidence interval 1.8 to 3.4, p <0.0001 for each 1-SD increase in SUA level). Adjustments for age, gender, body mass index, hypertension, chronic kidney disease, electrocardiographic features (left ventricular hypertrophy and PR interval), and use of diuretics and allopurinol did not attenuate the association between SUA and incident AF (adjusted odds ratio 2.44, 95% confidence interval 1.6 to 3.9, p <0.0001). Further adjustment for variables that were included in the 10-year Framingham Heart Study–derived AF risk score did not appreciably weaken this association. Results remained unchanged even when SUA was modeled as a categorical variable (stratifying by either SUA median or hyperuricemia), and when patients with previous coronary heart disease or heart failure were excluded from analysis. In conclusion, our findings suggest that elevated SUA levels are strongly associated with an increased incidence of AF in patients with type 2 diabetes mellitus even after adjustment for multiple clinical risk factors for AF.
Elevated serum uric acid (SUA) level has emerged as an independent predictor of morbidity and mortality in a variety of cardiovascular disease states, including coronary heart disease (CHD) and congestive heart failure (HF). Among patients with HF, it has also been reported that hyperuricemia is correlated to increased atrial filling pressures, ventricular systolic dysfunction, and decreased left atrial work. It would, therefore, be speculated that SUA metabolism is implicated in atrial remodeling, given that increased atrial filling pressures may cause structural and electrophysiologic abnormalities that contribute to the development and perpetuation of atrial fibrillation (AF). In support of this assumption, a number of recent studies have documented a significant association between elevated SUA and increased prevalence of AF. Recently, the Atherosclerosis Risk in Communities study, involving 15,382 AF-free black and white adults, reported that elevated SUA was independently associated with a greater incidence of AF, particularly among blacks and women. In the same line, 2 smaller studies reported that elevated SUA predicted AF recurrence after left atrial catheter ablation in patients with symptomatic, drug-refractory, paroxysmal, or persistent AF. There is currently a lack of available information on the relation between hyperuricemia and risk of incident AF in patients with type 2 diabetes mellitus. This study examines whether elevated SUA predicts subsequent development of incident AF in patients with type 2 diabetes mellitus.
Methods
We followed for 10 years a sample of 400 Caucasian patients with type 2 diabetes, who were clinically free from AF at baseline. As detailed in Figure 1 , these participants were selected by a simple random sampling technique from the whole cohort (n = 1,718) of type 2 diabetic outpatients, who regularly attended the diabetes clinic at the “Sacro Cuore” Hospital of Negrar (Verona) during 2000 to 2001 after excluding patients who did not meet the inclusion criteria for the study.
The sample size of the study was calculated with the specific aim of constructing a confidence interval (CI) around the incidence proportion of AF in patients with analogous characteristics. In a similar patient cohort, the proportion with AF has been estimated to be approximately 7%. Therefore, with a precision of 2.5% and a CI of 95%, we calculated that a sample size of 400 patients would be needed, taking also into account a cumulative proportion of losses to follow-up of 20%. Thus, a sample size of 400 patients from a population of 1,718 patients produces a 95% CI equal to the population proportion ± 2.5%, when the estimated proportion of patients with AF is 7% and the expected cumulative proportion of losses to follow-up is 20%.
All participants were periodically seen at the diabetes clinic (every 6 to 12 months) for medical examinations of glycemic control, chronic diabetic complications, and routine 12-lead electrocardiograms. The ascertainment at the end of the follow-up period (January 2011) for the whole sample was 100%.
The local ethics committee approved the study and all participants gave their informed consent for participation in this medical research.
Body mass index (BMI) was calculated by dividing weight in kilograms by the square of height in meters. Blood pressure was measured with a mercury sphygmomanometer after patient had been seated quietly for >5 minutes. Subjects were considered to have hypertension if their blood pressure was ≥140/90 mm Hg or if they were taking any antihypertensive drugs. Information on medical history, medication use, smoking, and daily alcohol consumption was obtained from all patients by questionnaire.
Venous blood was withdrawn in the morning after an overnight fast. SUA, lipids, creatinine, and other biochemical blood measurements were determined by standard laboratory procedures (DAX 96, Bayer Diagnostics, Milan, Italy). The intra- and interassay coefficients of variation for SUA level ranged from 1.2% to 1.6%. Low-density lipoprotein cholesterol was calculated by Friedewald’s equation, except when plasma triglycerides exceeded 400 mg/dl (n = 11). Hemoglobin A1c (HbA1c) was measured by an automated high-performance liquid chromatography analyzer. Urinary albumin excretion was measured from an early morning urine sample as the albumin-to-creatinine ratio by an immunonephelometric method.
At baseline, hyperuricemia was defined as an SUA level ≥416 μmol/L (≥7 mg/dl) in men and ≥357 μmol/L (≥6 mg/dl) in women and/or allopurinol therapy. Left ventricular hypertrophy (LVH) was diagnosed by a single cardiologist on the basis of a 12-lead electrocardiography taken at rest according to Sokolow-Lyon’s voltage criteria and/or Cornell’s voltage criteria. In all participants, the electrocardiographic PR interval was also recorded. CHD was defined as previous myocardial infarction, angina, or coronary revascularization procedures. Previous HF and mild valvular heart disease were confirmed by reviewing medical records of the hospital, including diagnostic symptoms patterns, echocardiograms and results of other laboratory exams. Chronic kidney disease (CKD) was defined as abnormal albuminuria (urine albumin-to-creatinine ratio ≥30 mg/g) or glomerular filtration rate <60 ml/min/1.73 m 2 as estimated by the 4-variable Modification of Diet in Renal Disease study equation.
The presence of carotid artery stenosis was diagnosed by Doppler echocardiographic scanning, which was performed by a single specialist physician, who was blind to subjects’ characteristics.
At baseline, all participants were free from AF as documented by a standard 12-lead electrocardiography. A 24-hour Holter monitor examination was not routinely performed either at baseline or during the follow-up. During the follow-up, participants were diagnosed with AF if AF or atrial flutter was present on a standard electrocardiogram that was obtained from either a routine clinic examination in our diabetes clinic in about 1/4 of the cases (i.e., a 12-lead electrocardiography at rest was performed annually in all participants) or reviewing hospital and physician charts from the remaining participants. The diagnosis of AF was confirmed in all affected participants by an experienced cardiologist, who was blind to subjects’ characteristics.
Data are presented as mean ± SD, medians (interquartile range), or percentages. Skewed variables (diabetes duration and triglycerides) were transformed using natural logarithmic transformation to improve normality before analysis. The unpaired t test and the chi-square test were used to compare baseline clinical characteristics in patients grouped by either SUA median at baseline or AF status at follow-up ( Tables 1 and 2 ). Binary logistic regression analysis was used to study the association between SUA level and incident AF ( Table 3 ). In this analysis, SUA was modeled as either a continuous variable (per each 1-SD increment, i.e., 83 μmol/L) or a categorical variable (stratifying by either SUA median or hyperuricemia). For prediction of incident AF, men and women were combined and first-order interaction terms for sex-by-SUA interactions on risk for AF were examined. Because the interactions were not statistically significant (p = 0.39), a sex-pooled multivariate logistic regression analysis was used. Four forced-entry logistic regression models were performed: an unadjusted model; a model adjusted for age and gender (model 1); a model further adjusted for BMI, hypertension, CKD, electrocardiographic LVH and PR interval, and current use of diuretics and allopurinol (model 2); and, finally, a regression model (model 3) adjusted for variables included in the 10-year Framingham Heart Study–derived AF risk score (i.e. age, sex, BMI, systolic blood pressure, hypertension treatment, electrocardiographic PR interval, and history of HF) plus allopurinol use. The covariates for multivariate regression analyses were chosen as potential confounding factors on the basis of their significance in univariate analysis or their biological plausibility. A Kaplan-Meier analysis of incidence curves for AF during 10 years of follow-up was undertaken in patients with and without hyperuricemia at baseline. Difference between groups was tested by the log-rank test.
| Variable | SUA (μmol/L) ∗ | p | |
|---|---|---|---|
| ≤300 (n = 212) | >300 (n = 188) | ||
| Men/women | 104/108 | 131/57 | <0.0001 |
| Age (yrs) | 64 ± 9 | 63 ± 10 | 0.44 |
| BMI (kg/m 2 ) | 29 ± 4 | 31 ± 5 | <0.0001 |
| Diabetes duration (yrs) | 6 (1–13) | 5 (1–12) | 0.84 |
| Systolic blood pressure (mm Hg) | 139 ± 14 | 141 ± 15 | 0.35 |
| Diastolic blood pressure (mm Hg) | 80 ± 7 | 81 ± 8 | 0.37 |
| Hemoglobin A1c (%) | 7.9 ± 1.6 | 7.5 ± 1.6 | 0.09 |
| LDL cholesterol (mmol/L) | 3.25 ± 0.8 | 3.12 ± 0.8 | 0.12 |
| LDL cholesterol (mg/dl) | 125 ± 31 | 120 ± 31 | 0.12 |
| HDL cholesterol (mmol/L) | 1.30 ± 0.3 | 1.17 ± 0.3 | <0.001 |
| HDL cholesterol (mg/dl) | 50 ± 11 | 45 ± 11 | <0.001 |
| Triglycerides (mmol/L) | 1.37 (1.03–1.9) | 1.58 (1.19–2.3) | <0.001 |
| Triglycerides (mg/dl) | 121 (91–167) | 139 (105–202) | <0.001 |
| Estimated GFR (ml/min/1.73 m 2 ) | 87 ± 20 | 80 ± 20 | <0.001 |
| PR interval (ms) | 168 ± 25 | 174 ± 30 | <0.05 |
| Obesity | 35 | 49 | <0.005 |
| Hypertension | 66 | 76 | <0.05 |
| Electrocardiographic LVH | 20 | 29 | <0.05 |
| Current smoker | 16 | 26 | 0.15 |
| Carotid artery stenoses ≥30% | 52 | 55 | 0.54 |
| CKD | 18 | 32 | <0.001 |
| Previous mild valvular disease | 1 | 1 | 0.92 |
| Previous CHD | 6 | 13 | <0.05 |
| Previous HF | 0.5 | 3.5 | <0.05 |
| Diuretics | 23 | 33 | <0.05 |
| ACE inhibitors or sartans | 60 | 64 | 0.47 |
| Calcium channel blockers | 26 | 32 | 0.15 |
| β Blockers | 10 | 15 | 0.18 |
| α Blockers | 3 | 8 | 0.10 |
| Lipid-lowering drugs | 25 | 29 | 0.43 |
| Antiplatelet drugs | 57 | 67 | 0.25 |
| Oral hypoglycemic drugs | 84 | 83 | 0.89 |
| Insulin therapy | 23 | 21 | 0.73 |
| Allopurinol therapy | 1 | 9 | <0.001 |
| SUA (μmol/L) | 246 ± 43 | 378 ± 56 | ND |
| Variable | AF at follow-up | p | |
|---|---|---|---|
| No (n = 358) | Yes (n = 42) | ||
| Men/Women | 211/147 | 24/18 | 0.85 |
| Age (yrs) | 63 ± 9 | 69 ± 9 | <0.0001 |
| BMI (kg/m 2 ) | 29 ± 4.7 | 30 ± 5 | 0.54 |
| Diabetes duration (yrs) | 5 (1–17) | 9 (1–24) | <0.01 |
| Systolic blood pressure (mm Hg) | 139 ± 15 | 147 ± 15 | <0.0001 |
| Diastolic blood pressure (mm Hg) | 81 ± 7 | 80 ± 8 | 0.81 |
| Hemoglobin A1c (%) | 7.7 ± 1.6 | 7.7 ± 1.7 | 0.92 |
| LDL cholesterol (mmol/L) | 2.84 ± 0.8 | 2.81 ± 1.3 | 0.82 |
| LDL cholesterol (mg/dl) | 110 ± 31 | 108 ± 50 | 0.82 |
| HDL cholesterol (mmol/L) | 1.24 ± 0.3 | 1.32 ± 0.3 | 0.16 |
| HDL cholesterol (mg/dl) | 48 ± 11 | 51 ± 11 | 0.16 |
| Triglycerides (mmol/L) | 1.45 (0.41–2.49) | 1.41 (0.52–2.42) | 0.20 |
| Triglycerides (mg/dl) | 128 (36–219) | 124 (46–213) | 0.20 |
| PR interval (ms) | 166 ± 23 | 210 ± 36 | <0.0001 |
| Current smoker | 21 | 17 | 0.45 |
| Obesity | 43 | 33 | 0.32 |
| Hypertension | 68 | 90 | <0.01 |
| Electrocardiographic LVH | 21 | 52 | <0.0001 |
| Carotid artery stenoses ≥30% | 50 | 81 | <0.005 |
| CKD | 24 | 35 | 0.08 |
| Previous mild valvular disease | 1 | 2.5 | 0.37 |
| Previous CHD | 9 | 10 | 0.96 |
| Previous HF | 1 | 10 | <0.0001 |
| Diuretics | 26 | 41 | <0.05 |
| ACE inhibitors or sartans | 61 | 71 | 0.19 |
| Calcium channel blockers | 27 | 41 | 0.10 |
| β Blockers | 12 | 14 | 0.71 |
| α Blockers | 6 | 14 | 0.08 |
| Lipid-lowering drugs | 28 | 19 | 0.23 |
| Antiplatelet drugs | 61 | 75 | 0.41 |
| Oral hypoglycemic drugs | 84 | 81 | 0.65 |
| Insulin therapy | 21 | 26 | 0.33 |
| Allopurinol therapy | 3 | 19 | <0.0001 |
| Hyperuricemia | 15 | 45 | <0.0001 |
| SUA (μmol/L) | 300 ± 77 | 375 ± 94 | <0.0001 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
