Impact of diabetes development on atrial fibrillation in hypertensive patients

November 17, 2008
Tonje A. Aksnes, MD1, Trygve B. Tjugen, MD1, Roland E. Schmieder, MD, PhD2, Sverre E. Kjeldsen, MD, PhD1,3, Stevo Julius, MD, ScD3

From the 1Department of Cardiology, Ullevaal University Hospital, O

Cardiology Review® Online, November 2008, Volume 25, Issue 11

Strategic Alliance Partnership | <b>Lowe Syndrome Association</b>

Patients with new-onset diabetes mellitus in the VALUE (Valsartan Antihypertensive Long-term Use Evaluation) trial had an increased incidence of atrial fibrillation compared with patients without diabetes. Clustering of risk factors or the presence of dysglycemia may make the heart more susceptible to arrhythmias.

Patients with hypertension and diabetes mellitus have a higher risk of developing atrial fibrillation (AF), which increases the risk of cardiovascular events, especially stroke. Identifying patients at risk of developing AF may be important for preventing AF and stroke caused by a cerebral embolus.

Methods and results

We recently reported that patients from the VALUE (Valsartan Antihypertensive Long-term Use Evaluation) trial who developed diabetes mellitus during the average 4.2 years of follow-up had an increased incidence of AF compared with patients without diabetes mellitus.1 The VALUE trial was a prospective randomized clinical trial that investigated the effects of antihypertensive treatment with valsartan (Diovan) versus amlodipine (Norvasc) on cardiac morbidity and mortality.2 A total of 15,245 patients from 31 countries were randomized into the trial; these patients were older than 50 years and were treated (92%) or untreated for essential hypertension.2 Additional inclusion criteria were the presence of predefined combinations of cardiovascular risk factors or disease factors according to an algorithm based on age and sex.3 New-onset diabetes mellitus was prespecified as a secondary end point and defined as patients without diabetes mellitus at the outset of the trial who later developed diabetes as evidenced by study investigators, or who were started on antidiabetic drugs or insulin during the trial or had a fasting glucose concentration of 7.0 mmol/L (126 mg/dL) or higher in a blood sample at the end of the trial.4

The incidence of new-onset AF was also prespecified as an investigator-initiated end point.5 Electrocardiograms (ECGs) were obtained at baseline every year and centrally analyzed. New-onset AF was defined as any ECG recording of this arrhythmia in patients without AF at baseline. The patients with new-onset and baseline diabetes mellitus were compared with patients without diabetes using a Cox regression model with adjustment for prespecified covariates.

At baseline, 13,760 patients in the VALUE trial (4634 with baseline diabetes, 1252 with new-onset diabetes, and 7874 without diabetes throughout the trial) were without AF and had follow-up ECGs.1 Of these patients, 398 (159 with diabetes at baseline, 35 with new-onset diabetes, and 204 without diabetes) had evidence of AF on the baseline ECG, and those with diabetes at baseline had significantly more AF at baseline compared with patients without diabetes (P = .0137). No significant difference was found between the patients who developed diabetes later on and those without diabetes throughout the trial.

During the average 4.2 years of the VALUE trial, 551 patients with sinus rhythm at baseline had at least 1 repeat ECG recording demonstrating AF. The patients with new-onset diabetes developed significantly more AF compared with patients without diabetes (hazard ratio, 1.49; Figure). There was also a trend toward more frequent AF in patients with diabetes at baseline, but this did not prove statistically significant. This may be because patients with diabetes at baseline already had an increased incidence of AF, which suggests the risk of AF development starts in the early phases of hyperglycemia.

The results were adjusted for prespecified covariates in the VALUE trial, including age, diabetes status, history of coronary heart disease, left ventricular hypertrophy, and randomized study treatment. Because the risk of developing AF may also be related to differences in potassium levels, body mass index, blood pressure, and heart rate, additional adjustment for these variables was undertaken. The result did not change significantly after these adjustments, indicating that the increased risk of new-onset AF in patients with hypertension relates to dysglycemia rather than to other factors.


AF is known to increase cardiac morbidity, which was confirmed by our study. Cardiac morbidity and stroke occurred more frequently in patients who had new-onset AF regardless of their diabetes status. The development of new-onset AF in the patients with new-onset diabetes mellitus significantly increased the risk of heart failure and stroke, with a hazard ratio of 3.56 and 1.96, respectively; however, the development of AF had no significant effect on the event rate of myocardial infarction (MI). When AF and heart failure are both present, patient prognosis worsens. AF contributes to the progression of heart failure, and heart failure increases the chance of developing AF.

Hypertension and diabetes mellitus were the only significant independent cardiovascular risk factors for AF when controlling for age and other predisposing conditions in the 38-year follow-up data from the Framingham Heart Study.6 Other observational studies have also recognized glucose level or diabetes mellitus as important risk factors for AF.7-11 A recently published prospective community-based study from Japan showed that individuals who met standard criteria for the metabolic syndrome had at least a 60% increased age- and sex-adjusted risk of new-onset AF over 4.5 years.12 Hypertension and diabetes mellitus or dysglycemia are important components of metabolic syndrome, and patients with both diseases have the added risk of AF.

The pathophysiologic relationship between diabetes mellitus and AF is not known, but it is likely complex and results through various pathways, including coronary artery disease, hypertension, abnormal sympathetic tone, or the direct effect of diabetes on atrial tissue.13 “Diabetic cardiomyopathy”14 may be an entity that increases risk for AF, but diabetes mellitus is also associated with many systemic illnesses, such as infection, electrolyte abnormalities, or renal failure, which may lower the resistance for arrhythmias.11 Arrhythmias and AF may also be caused by hypoglycemia in diabetic patients, and we could speculate if it is the fluctuations in metabolic control rather than the chronic state of hyperglycemia that may cause the arrhythmia.13,15


Hypertensive patients who developed diabetes mellitus during the VALUE trial had more AF than patients without diabetes. Clustering of risk factors or the presence of dysglycemia may make the heart more susceptible to developing arrhythmias.


The VALUE trial was supported by Novartis Pharma AG. Drs Schmieder, Kjeldsen, and Julius were lead investigators in the study and have served as consultants or received grants from Novartis and other major pharmaceutical companies. Drs Aksnes and Tjugen have no affilliations to disclose that might represent a conflict of interest with the content of this article.