Association of Increased QTc Interval With the Cardiometabolic Syndrome

Abstract

J Clin Hypertens (Greenwich).

Recent studies indicate a high prevalence of increased QTc interval length in patients affected by the metabolic syndrome, but there is no data available to demonstrate the correlation of the QTc interval with severity of the cardiometabolic syndrome (CMS). The objective of this study was to estimate the association between increasing number of cardiometabolic abnormalities and increasing QTc interval length. Electrocardiograms were collected from 1420 participants in a cross‐sectional study. The QTc interval lengths were corrected for heart rate using Bazett’s formula. CMS was determined according to the National Cholesterol Education Program Adult Treatment Panel III (NCEP‐ATP III) guidelines. Multiple linear regression models were used examining associations between increasing number of individual components of syndrome with QTc interval length. Participants with CMS had significantly longer QTc interval length, controlling for age, body mass index (BMI), sex, and ethnic group. Increasing number of CMS components was significantly associated with increased QTc interval length, even after adjusting for age, BMI, total cholesterol, fasting C peptide, and history of heart disease. These findings suggest that QTc interval length is increased in the presence of CMS and is linearly related to an increase in number of metabolic abnormalities.

The QTc interval represents the time from onset of depolarization to completion of repolarization of the heart, adjusted for heart rate. Prolonged QTc interval is an indicator of increased risk for cardiac arrhythmia and sudden death. ¹ Although a QTc interval of at least 500 ms generally has been shown to be associated with a higher risk of torsades de pointes, there is no established threshold below which prolongation of the QTc interval is considered free of proarrhythmic risk. ² , ³ , ⁴ Individuals with prolonged QTc interval have higher risk for cardiovascular mortality. ¹ , ⁵ , ⁶ , ⁷ , ⁸ Factors that predispose to QTc prolongation and higher risk of torsades de pointes include older age, female sex, low left ventricular ejection fraction, left ventricular hypertrophy, ischemia, bradycardia, and electrolyte abnormalities such as hypokalemia and hypomagnesemia. ⁴ The QTc interval has also been related to components of the insulin resistance syndrome including increased body mass index (BMI), upper body obesity, increased blood pressure, and increased fasting insulin during an oral glucose tolerance test. ⁹ , ¹⁰ , ¹¹ An association between QTc interval and the cardiometabolic syndrome (CMS) has recently been reported by two separate studies, ¹² , ¹³ but none have yet to examine the association between QTc interval and severity of CMS. We therefore investigated the association between the QTc interval and CMS prevalence and severity.

Methods

A detailed description of the methods of the Kohala Health Research Project has been presented elsewhere. ¹⁴ In brief, the cross‐sectional study surveyed all men and nonpregnant women 18 years and older residing in North Kohala, Hawaii, between 1997 and 2000. The ethnic composition of this community comprises a large Asian/Pacific Islander, including a large contingent of Native Hawaiian and part‐Hawaiians. Multiple methods of recruitment were utilized, in which an estimated 80% to 90% of households were contacted via telephone, using a cross‐reference directory; mail; or a door‐to‐door census. Community support for the research project was fostered through local public television announcements, flyers posted at community centers and stores, and presentations given to community organizations. Participants were eligible for the study if they underwent electrocardiography (ECG) along with corresponding laboratory tests. Participants were excluded if they had a history of arrhythmia.

Personal and family medical histories, an inventory of current medications, and sociodemographic information, including educational attainment, occupational history, household income, and ethnic ancestry, were obtained during an extensive interview. Ethnicity was estimated by self‐report. Participants reporting mixed heritage were each asked to estimate the percentage of each ethnic ancestral group. Native Hawaiian ancestry was defined here as descent from the indigenous Polynesian population residing in the islands of Hawaii prior to initial Western contact in 1778.

The health screening examination took approximately 2 to 3 hours. Participants were asked to consume nothing but water for 10 to 14 hours prior to the appointment. Blood was drawn in the fasting state and plasma was separated within 2 hours for assessment of plasma glucose, total cholesterol and lipoproteins, insulin, and C‐peptide. Triglycerides and high‐density lipoprotein cholesterol (HDL‐C) levels were measured in duplicate using a Beckman Synchron CX4 Analyzer (Beckman Coulter, Inc, Brea, CA) using the manufacturer’s enzymatic colorimetric reagents. The plasma concentration of insulin, C‐peptide, and leptin levels were determined by radioimmunoassay in duplicate. Insulin assays were performed using kits from Linco Research Inc (St Charles, MO), whereas C‐peptide was from Diagnostic Products Corporation (Los Angeles, CA). All measurements were performed with quality control procedures in place. Our laboratory also participated in the Centers for Disease Control and Prevention–National Heart Lung Blood Institute Lipid Standardization Program. Intraassay and interassay coefficient of variances were all <10%.

Anthropometric measurements were obtained while standing. Height and weight were measured with participants wearing light‐weight clothing without shoes and used to calculate BMI. Waist circumferences were measured at the level of the navel and used as an estimate of central adiposity. Blood pressure was measured after participants were seated in a quiet area for at least 5 minutes and measured in triplicate from the right arm of each individual per standardized protocol using a standard mercury sphygmomanometer. The mean of the last measurements were used for statistical analyses.

Each participant was evaluated for the presence of CMS, defined by the criteria set forth by the National Cholesterol Education Program Adult Treatment Panel III (NCEP‐ATP III), ¹⁵ with modifications as described by Grundy and colleagues ¹⁶ reflecting the American Diabetes Association recommendations for defining impaired fasting glucose at 100 mg/dL. Participants were considered to exhibit the metabolic syndrome if ≥3 metabolic abnormalities were present. The NCEP‐ATP III definition of CMS and the criteria for these metabolic abnormalities are as follows: (1) abdominal obesity, waist circumference >102 cm (>40 in) for men and >88 cm (35 in) for women; (2) blood pressure ≥130 mm Hg systolic and/or ≥85 mm Hg diastolic; (3) fasting glucose ≥6.1 mmol/L (100 mg/dL); (4) triglycerides ≥1.69 mmol/L (150 mg/dL); and (5) HDL‐C <1.03 mmol/L (40 mg/dL) for men and 1.29 mmol/L (50 mg/dL) for women. Similar to the National Health and Nutrition Examination Survey (NHANES) report, we included use of antihypertensive or antidiabetic medication as indicators of metabolic abnormalities.

Resting 12‐lead ECGs were recorded using a 3‐channel direct writing machine (Marquette MAC PC; GE Healthcare, UK). QTc intervals were electronically measured with patients at rest from multiple leads for at least 3 to 5 cardiac cycles. A sample of 100 tracings was confirmed by a single, blinded rater. The QTc was calculated using Bazett’s formula ¹⁷ and included as a continuous variable in all statistical models.

Linear regression was used to estimate the association between QTc and the number of CMS criteria. General linear model analysis of covariance was used to compare QTc and categoric variables such as sex, ethnicity, impaired glucose metabolism, and ordinal variables such as the number of CMS criteria. Variables thus shown to be associated with both CMS and QTc were included in a multiple linear regression model to adjust for potential confounding effects.

Results

A total of 1447 participants (1420 with acceptable ECGs) completed the entire examination and reported their ethnic ancestry. The 3 largest nonmixed ethnic groups were Caucasians (20%), Japanese Americans (14%), and Filipino Americans (13%). A fifth group, composed of predominantly mixed, non‐Hawaiian ancestry, included 250 individuals.

The associations of QTc interval length with selected cardiovascular risk factors and other population characteristics and selected participants are presented in Table I. Presence of CMS was associated with an increased QTc interval length. Sex and ethnic variation in the prevalence of increased QTc in this population were observed, as was previously reported. ¹¹ Women had significantly longer QTc interval lengths than men. As we reported previously, QTc interval lengths also varied significantly between ethnic groups, with longer interval length observed among all non‐Caucasian groups. ¹⁸ The greatest mean QTc interval lengths was observed among Native Hawaiians; however, the mean QTc in this ethnic group was not significantly longer than that observed among Japanese and Filipino ancestry participants. With the exception of current smoking status, all other risk factors were significantly associated with QTc interval length. After adjusting for age, BMI, and sex, however, current smoking was associated with a statistically significant greater QTc interval, while fasting insulin or C‐peptide were not (data not shown).

Table I.

 General Linear Model Estimates of Association Between Selected Population Characteristics and Cardiovascular Risk Factors With QTc Interval Length

Variable	No.a	Mean or Regression Estimate	Standard Error	F Ratio	Probability >F Ratio
Cardiometabolic syndrome
Absent	954	420.2	0.8	77.1	<.01
Present	466	434.2	1.1
Smoking status
Never smoker or quit	1099	424.0	0.8	0.5	.47
Current smoker	241	425.3	1.6
Myocardial infarction
Negative history	1386	423.9	0.7	4.7	.03
Positive history	34	433.2	4.2
Sex
Male	660	419.8	1.0	39.7	<.01
Female	760	428.0	0.9
Ethnic group
Caucasian	295	417.3	1.4	19.1	<.01
Filipino	183	425.5	1.8
Hawaiian/part‐Hawaiian	502	429.4	1.1
Japanese	188	429.0	1.8
Other/mixed ancestry	252	417.2	1.5
Body mass index		0.6	0.1	34.7	<.01
Age		0.3	0.04	67.7	<.01
Fasting insulin		0.3	0.1	18.8	<.01
Fasting C‐peptides		1.6	0.5	11.2	<.01

^aNumbers vary according to available data.

As described above, mean QTc interval length was significantly higher among participants with CMS. Table II demonstrates that each and every one of the individual components of CMS is highly significantly associated with increased QTc interval length. The association between low HDL‐C and high triglycerides was mediated by BMI; however, all the other components of CMS were associated with QTc independent of BMI.

Table II.

 Comparison of Mean (Standard Error) of QTc Interval Length by Individual Components of the Cardiometabolic Syndrome

Variable	Absent	Present	Probability >F Ratio
Central obesity	420.3 (0.82)	430.5 (1.04)	<.01
High blood pressure	417.7 (0.86)	431.7 (0.92)	<.01
High triglycerides	422.4 (0.79)	427.8 (1.15)	<.01
Low high‐density lipoprotein cholesterol	422.3 (0.87)	426.6 (0.98)	<.01
High fasting blood glucose	421.5 (0.72)	433.9 (1.39)	<.01

As illustrated in Table III, QTc interval length was significantly and positively associated with increasing number of metabolic abnormalities. This was true for both men and women, although the mean QTc for women was greater at each level of increasing metabolic abnormalities. A similar monotonic increase in QTc interval length was observed for both men and women.

Table III.

 Least Squared Means and Test for Trend for QTc by Increasing Number of Cardiometabolic Syndrome Features, Estimated by General Linear Model Regression

Number of Abnormalities	Unadjusted Means			Adjusted Meansa
	Men	Women	Combined	Men	Women	Combined
0	409.1	418.7	414.5	417.8	421.7	420.8
1	418.3	425.5	422.0	423.2	425.6	425.3
2	421.9	426.7	424.5	425.9	424.4	426.0
3	425.1	435.1	430.2	427.1	430.9	430.1
4	429.4	435.5	432.6	430.2	430.3	430.8
5	429.8	443.3	438.0	430.3	437.4	436.0
Trend test	<.0001	<.0001	<.0001	.03	<.001	<.001

^aAdjusted for age, body mass index, ethnic group, current smoking, and history of myocardial infarction.

After adjusting for all confounding variables described in Table I, trend tests revealed that this increase continued to be statistically significant for both men and women (Table III); however, the differences across levels of CMS abnormalities appeared less distinct. Interestingly, the adjusted overall QTc means were still significantly higher among women (422.8 ms) than among men (417.0 ms).

Discussion

In this cross‐sectional study, all components of the CMS criteria were correlated with increasing QTc interval length. This is the first study showing a correlation between the number of CMS abnormalities and QTc interval length. The results showed a marked increased in QTc interval with higher numbers of metabolic abnormalities in women. In comparison, men had a less pronounced but still significant correlation between the QTc interval and number of CMS abnormalities. Adjustment for age, BMI, ethnicity, history of heart disease, and cardiac medications including β‐blockers, calcium channel blockers, and angiotensin‐converting enzyme inhibitors did not result in any change regarding the relationship and only slightly reduced the statistical significance of these findings.

In this study, women had markedly pronounced increase in QTc length at each level of increasing number of metabolic abnormalities. Although the parallel increases were less dramatic after adjustments for other cardiovascular disease risk factors, the overall mean QTc continued to be significantly greater among women than men. This is consistent with previous studies indicating that women have longer QTc intervals. ¹⁹ , ²⁰ Sex hormones, including male androgens such as testosterone or 5‐dehydrotestosterone, play important roles in cardiac repolarization and may provide a reason for the difference seen between men and women. ²¹ , ²² Correspondingly, the JT interval is shorter in virilized women. ²³ Such studies suggest that hormones such as testosterone may have a protective role in QTc prolongation.

This finding of increasing QTc interval in association with increasing number of metabolic abnormalities is consistent with many previous studies. ²⁴ , ²⁵ Studies have shown an increased prevalence of prolonged QTc interval in type I and II diabetes and patients with insulin resistance syndrome. ⁹ These findings suggest that insulin resistance or a component of insulin resistance syndrome may influence myocardial membrane activity and thus QTc interval length. Hyperglycemia can potentially alter cardiac repolarization through the formation of advanced glycation end products, endothelial dysfunction, and oxidative stress. ²⁶ , ²⁷ , ²⁸ In addition, a role for autonomic neuropathy, possibly as a result of impaired glucose disposal, can lead to impaired sympathovagal balance and increased cardiac sympathetic activation. ⁸ An earlier report from our study revealed a similar association with blood glucose levels and blood pressure ¹⁸ but did not examine the effect on QTc interval length after accounting for the clustering of risk factors associated with CMS.

There is controversy regarding the method of measurement for QTc interval. Although QTc interval experts argue that intraobserver and interobserver variability and measurement error are higher when the QTc interval is taken from computerized ECG algorithms rather than from careful high‐resolution manual measurements, automated reading is the most widely used and practical method for clinicians. ⁴ , ²⁹ The use of automated ECG measurement of QTc was chosen in this study because of its practical implications to practicing community clinicians. Of note, there were also a limited number of individuals taking medication in this study. Exclusion of patients taking medications for hypertension or heart disease did not alter the statistical significance of these findings.

Regardless of what mechanisms contribute to an increased QTc interval, this finding has two major clinical impacts. First, patients with CMS and a high number of metabolic abnormalities need an ECG for evaluation of the QTc interval. This would be necessary before prescribing medications that could potentially further prolong the QTc interval in order to avoid life‐threatening cardiac arrhythmias. Second, the QTc interval may be used as an adjunctive clinical tool for providing risk stratification for CMS and cardiovascular mortality in the general population. The results of our study suggest that individuals with CMS are at risk for prolongation of the QTc interval. This may be particularly true for women, who were observed to have longer intervals than men with each level of metabolic abnormality. Future prospective studies will be necessary to address this issue.

Disclosures:  This research was supported by grants from the National Institutes of Health, National Center for Research Resources, Research Centers in Minority Institutions program (grant 12RR03061) and the National Institute of Neurological Disorders and Stroke (grants 1 R25 RR019321 and S11 1 NS43364).