Abnormal signal-averaged electrocardiogram (SAECG) in obesity

Abstract

Objective: The occurrence of small high-frequency electrocardiogram (ECG) potentials (1 to 20 microV) seen at the end of the QRS complex and into the ST segment have been correlated with increased risk for ventricular arrhythmias and sudden cardiac death. Computer-assisted analysis of these "late potentials" by signal-averaged electrocardiography (SAECG) has been studied and utilized to predict the likelihood of ventricular arrhythmias in various clinical states. Obesity is associated with significant cardiovascular morbidity and sudden death. Ventricular arrhythmias are postulated causes. We studied the occurrence of late potentials in a randomly selected group of obese patients and healthy volunteers.

Research methods and procedures: We performed SAECG on 105 subjects. Of these, 62 were obese ambulatory patients with body mass index (BMI) of >30 kg/m2, whereas 43 were healthy asymptomatic volunteers with a BMI of <30 kg/m2. Patients with a history of clinical heart disease and pulmonary disease, electrolyte abnormalities, recent hospitalizations, or abnormal screening ECG or taking medications known to alter the QRS interval were excluded. At least 250 beats were analyzed with a noise level of <0.50 microV. Criteria of a late potential include QRS duration >114 ms, high-frequency low amplitude >38 ms, and root-mean-square voltage <20 microV. Patients were divided into four subgroups based on BMI values. The prevalence of SAECG abnormalities in each BMI subgroup was studied. We utilized multiple logistic regression analysis to study the effect of obesity, hypertension, and diabetes mellitus on abnormal SAECG results.

Results: Compared to age- and sex-matched healthy volunteers with BMI of <30 kg/m2, obese patients with BMI of >30 kg/m2 had significantly more abnormalities on SAECG (4.6% vs. 55%). In the obese group, the prevalence and number of abnormalities increased with increase in BMI (35% in the BMI 31 to 40 kg/m2 subgroup, 86% in the BMI 41 to 50 kg/m2 subgroup, and 100% in patients with BMI of >50 kg/m2). Multiple logistic regression analysis shows that BMI is an independent predictor variable of abnormal SAECG results in obese patients (n = 62) with BMI of >30 kg/m2 as well as in all study subjects (n = 105). BMI also predicts abnormality of each abnormal SAECG criterion in both obese and all subjects. Hypertension was found to influence the QRS duration alone in obese and all subjects.

Discussion: Obesity is associated with increased occurrence of abnormal SAECG results. These abnormalities are found both in obese patients with and without hypertension and/or diabetes. Obesity is an independent predictor variable of abnormal SAECG results. A history of hypertension predicts abnormality of QRS duration only.