- Email: [email protected]
Effect of age and overweight on the QT interval and the prevalence of long QT syndrome in children
Effect of Age and Overweight on the QT Interval and the Prevalence of Long QT Syndrome in Children Toshiro Fukushige, MD, Masao Yoshinaga, MD, Atsushi Shimago, MD, Junichiro Nishi, MD, Yukiharu Kono, MD, Yuichi Nomura, MD, Koichiro Miyata, MD, Masato Imamura, MD, Toshimitsu Shibata, MD, Masami Nagashima, MD, and Ichiro Niimura, MD The change in QT interval with age during childhood of normal children and children with long QT syndrome (LQTS) and the effects of body mass index on the QT interval have not been studied in detail. The prevalence of LQTS in children is not well known. We measured 3 consecutive QT and RR intervals in 4,655 children. Their electrocardiograms along with their height and weight were recorded when they were in the first grade in 1994 and again when they were in the seventh grade in 2000. The QT interval was corrected by Bazett’s formula. The longer corrected QT intervals in female subjects than male subjects start at elementary school age, earlier than previously reported. Overweight did not have an impact on the uncorrected or corrected QT
interval. None of the 4 children diagnosed with LQTS in the seventh grade had characteristic electrocardiographic findings of LQTS in the first grade. All 4 are nonfamilial cases. The prevalence of LQTS in children was found to be 1 of 1,164. These data suggest that abnormal electrocardiographic phenotypes in children with nonfamilial LQTS may appear during the elementary school year. The longer QT intervals in female subjects than male subjects start at the same period. No correlation was found between obesity and length of the QT interval. Finally, the prevalence of LQTS in children is greater than previously suspected. 䊚2002 by Excerpta Medica, Inc. (Am J Cardiol 2002;89:395–398)
ong QT syndrome (LQTS) is a rare disease characterized by prolonged ventricular repolarization, L and the clinical presentation of LQTS is the occur-
between QT interval and obesity in school-aged children based on data provided by a large-scale study. In this study we examined changes in the QT interval with age during the childhood of normal and LQTS children, as well as the possible correlation of body mass index (BMI) with the QT interval; our final purpose was to determine the prevalence of LQTS in children.
1
rence of syncope or cardiac arrest in young persons.2,3 The first cardiac event among children with the LQT1 and LQT2 genotypes has been reported to occur between 9 and 12 years of age.4 – 8 ST-T-wave patterns also vary across genotypes and ages8; however, changes in the QT interval with age during childhood in normal children and children with LQTS have not been studied in detail. The prevalence of LQTS in the general population is also not well known, although it is thought to be approximately 1 of 10,000.2,9 A screening program conducted for Japanese children has uncovered large numbers of children with LQTS, but with no family history of the disease, or with any history of LQTS-related cardiac events at the time of diagnosis,10,11 suggesting that the prevalence of LQTS is greater than it may appear. An additional concern is that a prolonged QT interval is sometimes associated with obesity.12–17 However, there has been no report to date concerning the possible association
From the Department of Pediatrics, Faculty of Medicine, Kagoshima University and Kagoshima City Medical Association, Kagoshima, Japan and the Research Committee on Long QT Syndrome in Children in the Japanese Society of Pediatric Cardiology and Cardiac Surgery, Tokyo, Japan. Manuscript received September 7, 2001; revised manuscript received and accepted October 22, 2001. Address for reprints: Masao Yoshinaga, MD, Department of Pediatrics, Faculty of Medicine, Kagoshima University, Sakuragaoka 8-35-1, Kagoshima, 890-8520 Japan. E-mail: [email protected]. kagoshima-u.ac.jp. ©2002 by Excerpta Medica, Inc. All rights reserved. The American Journal of Cardiology Vol. 89 February 15, 2002
METHODS
Subjects: The subjects were 4,655 children whose data (name, sex, weight, and height) and electrocardiograms (ECGs) were obtained both in 1994 when they were first graders, and again in 2000 when they were seventh graders. Data were collected from the Screening Program for Heart Disease in Kagoshima City that has been operated by the Kagoshima City Medical Association and supported by the Board of Education in Kagoshima City. Screening program for heart diseases in Kagoshima City: All resting ECGs were recorded at a speed of 25
mm/s at each school by medical technologists of the Screening Program for Heart Disease in Kagoshima City. Screening for LQTS in this program was based on the electrocardiographic criteria described by Schwartz et al,3 namely, the prolonged corrected QT interval determined by Bazett’s formula (QTc interval) and T-wave morphology. When the heart rate was ⬎75 beats/min, values obtained through an exponential correction of the QT interval (QT/RR0.31) or through Fridericia’s formula (QTFc⫽QT/RR1/3) were also considered.10,18,19 Children diagnosed as having 0002-9149/02/$–see front matter PII S0002-9149(01)02259-7
395
TABLE 1 Mean Values of Heart Rates, Uncorrected QT Intervals, and QT Intervals Corrected by Bazett’s (QTc) Formula in Boys and Girls in the First and Seventh Grades Boys
Heart rate QT interval QTc values
Girls
First
Seventh
First
Seventh
82 ⫾ 12 0.332 ⫾ 0.023 0.386 ⫾ 0.021
77 ⫾ 12 0.353 ⫾ 0.025 0.395 ⫾ 0.022
84 ⫾ 12‡ 0.327 ⫾ 0.022‡ 0.384 ⫾ 0.021*
81 ⫾ 13‡ 0.350 ⫾ 0.025† 0.402 ⫾ 0.021‡
Changes in Uncorrected and Corrected QT Intervals Between the First and Seventh Grades QT interval QTc values
0.021 ⫾ 0.028 0.010 ⫾ 0.026
0.023 ⫾ 0.027* 0.018 ⫾ 0.025‡
*p ⬍0.05; †p ⫽ 0.0004; ‡p ⬍0.0001. The p value between the first and seventh grades of all variables in boys and girls is p ⬍0.0001 (symbols not shown). The difference in the mean values between boys and girls at the first and seventh grades and difference in the changes between the two grades are shown with symbols.
TABLE 2 Correlation Coefficients and p Values Between BMI and the RR Interval or QT Intervals QT Interval RR Interval 1st grade boys 1st grade girls 7th grade boys 7th grade girls
0.013 0.036 ⫺0.124 ⫺0.033
(0.5122) (0.0851) (⬍0.0001) (0.1387)
Uncorrected* 0.009 0.024 ⫺0.096 ⫺0.049
(0.6722) (0.2435) (⬍0.0001) (0.0251)
QTc* ⫺0.001 ⫺0.017 0.062 ⫺0.010
*Uncorrected QT intervals and QT intervals corrected by Bazett’s (QTc) formula. The data are expressed as a correlation coefficient and a p value in the bracket.
LQTS on the resting ECG were further examined. Their history was taken and they underwent physical examination and an exercise test (Master’s 2-step or a treadmill exercise test) and/or a face immersion test.11 These tests were performed at the Kagoshima Medical Association Hospital in Kagoshima City or at the outpatient clinic of the Kagoshima University Hospital. Measurements of height and weight: School nurses measured the height and weight of children in their schools in early April of each year. Height was measured without socks or shoes. Weight was measured while the child was wearing only underclothing. The BMI was calculated as (weight in kg)/(height in m)2. Measurement of QT interval: Three consecutive QT and RR intervals in lead V5 were measured retrospectively for each child in this study. The QT interval was measured from the onset of the Q wave to the end of the T wave in lead V5. Bifid T waves, but not U waves, were included in the QT measurement.8 The QT interval measurements were obtained by the first and second investigators (TF and MY). Intra- and interreader variability was tested by triplicated measurements of 100 randomly selected ECGs from the present study. Intrareader coefficients of variability of the first and second investigators were 1.7%, and the 396 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 89
(0.9591) (0.4209) (0.0040) (0.6454)
interreader coefficient of variability was 1.9% of the mean QT interval. Mean values of three corrected QT intervals: We measured 3 consecutive
QT/RR data pairs, and each was corrected using Bazett’s formula (QTc ⫽ QT/RR1/2). The mean values of the 3 consecutive QT/RR data pairs were used in the study. Diagnosis of LQTS: A final diagnosis of LQTS was made when the second investigator (MY) and an additional 3 members of the Japanese Society of Pediatric Cardiology and Cardiac Surgery (TS, MN, and IN [members outside the second investigator’s university]) agreed with the diagnosis, because the genetic abnormality was not determined in the present study. Statistical analysis: Data are expressed as mean ⫾ 1 SD. Statistical testing for correlation coefficients was based on Fisher’s Z transformation. Differences in mean values between the groups were analyzed using a paired or nonpaired Student’s t test. A p value ⬍0.05 was considered statistically significant.
RESULTS
Changes in uncorrected and corrected QT intervals between the first and seventh grades: The mean heart
rate, uncorrected and corrected QT intervals, and the changes in values between the first and seventh grades of the subjects are listed in Table 1. The mean uncorrected and corrected QT intervals of the subjects significantly (p ⬍0.0001) increased from the first to the seventh grades in boys and girls. The mean uncorrected and corrected QT intervals in girls were significantly lower than those in boys at the first screening, i.e., the first grade. However, the changes in the uncorrected and corrected QT intervals between the first and seventh grades were greater in girls than in boys. Relation between body mass index and uncorrected or corrected QT intervals: In the seventh grade boys, the
RR and uncorrected QT intervals were inversely associated with BMI, whereas the QTc value was positively associated with it (Table 2). In the seventh grade girls, the uncorrected and corrected QT intervals were inversely associated with BMI. A stepwise regression analysis was conducted to determine whether body mass affects QT intervals using the QT interval as a dependent variable and the BMI and heart rate as independent variables. This analysis revealed that body mass did not have any impact on the uncorrected or corrected QT intervals in either seventh grade boys or girls (data not shown). Incidence of LQTS in children: In the 1994 screening, no subject in the first grade was diagnosed as having FEBRUARY 15, 2002
FIGURE 1. Representative ECGs of a boy among 4 children diagnosed as having LQTS. Starting from the left, the ECG at the first grade (a), the ECG at the seventh grade (b), the ECG before (c), and after (d) a treadmill exercise test, and the ECG before (e) and after (f) a face immersion test. The paper speed of the ECGs during treadmill exercise testing was at 50 mm/s, and that of remaining ECGs was 25 mm/s. Obvious bifid T waves appeared in leads V3 to V5 after face immersion.
LQTS. In 2000, on the other hand, 9 children in the seventh grade were suspected of having LQTS from the resting ECG, 8 of whom were further examined. One child was unable to visit the hospital. Finally, 4 children (2 boys and 2 girls) were diagnosed as having LQTS. None of the diagnosed children had a family history of the disease or any past history of cardiac events. The first-grade ECGs of these 4 children did not show the typical characteristics of LQTS. Representative ECGs of 1 of these 4 children are shown in Figure 1. The ECG shows obvious bifid T waves in leads V3 to V5 after face immersion, mimicking the findings of genotype LQT2 described by Zhang et al.8 The remaining 3 patients had obvious bifid T waves in leads V4 to V6 (1 boy)—subtle bifid T waves with a second component on top of the T wave. These findings mimic those of genotype LQT2 (2 girls). The electrocardiographic findings of one girl mimicked the infantile pattern of genotype LQT1, although the child in this study was 12 years old. The incidence of long QT syndrome of children in the seventh grade was 4 of 4,655 (1 of 1,164).
DISCUSSION The present study indicates that longer QT intervals in females begin to be apparent in elementary school, earlier than previously reported.20 Obesity did not have any impact on the uncorrected or corrected QT intervals. Four children diagnosed with LQTS in the seventh grade had not had electrocardiographic patterns characteristic of LQTS in the first grade. The prevalence of LQTS in children was found to be 1 of 1,164.
Changes in the QT interval in childhood: The uncorrected and corrected QT intervals calculated by Bazett’s formula are the longest in adult women.21 However, these values and changes in these values in patients are not well known. Eberle et al20 and Pearl22 report that there is no significant influence of gender on QT interval up to the age of 14 years. The present study showed that uncorrected and corrected QT intervals were shorter in girls in the first grade, and that the difference in these values between the first and seventh grades was greater in girls than in boys. The reason for this longer QT interval may be the earlier appearance of puberty in girls.20 These data indicate that the longer QT intervals in females start at elementary school age and that its onset is earlier than previously reported. Relation between QT interval and overweight: Several reports claim that short-term weight loss is associated with shortening of the QT interval in children, adolescents, and adults.12–16 Abdominal obesity has also been associated with a prolonged QTc interval.17 In the present study, a stepwise regression analysis did not indicate any correlation between BMI and the uncorrected and corrected QT intervals, strongly suggesting that overweight is not associated with the length of the QT interval in children. The association of obesity and the QT interval in adults may be due to autonomic factors between obese adults and the QT interval,13 or may be the result of having the disease over a long period of time. Effect of age on diagnosing LQTS: None of the 4 children in the seventh grade diagnosed with nonfamilial LQTS had electrocardiographic patterns char-
ARRHYTHMIAS AND CONDUCTION DISTURBANCES/CHANGES IN CHILDHOOD QT INTERVALS
397
acteristic of LQTS in the first grade in the present study. The median age at the first cardiac event of LQTS children of the LQT1 and LQT2 genotypes has been reported to be from 9 to 12 years.4 –7 That of children of the LQT3 genotype has been reported at as high as 16 to 18 years,5,7 as has the median age for auditory-related events associated with the LQT2 genotype. These data indicate that the risk of cardiac events in children with LQTS starts at approximately 9 to 12 years. The present study revealed that LQTS children had characteristic electrocardiographic findings in the seventh grade (12 or 13 years old), but not in the first grade (6 or 7 years old). Thus, abnormal electrocardiographic phenotypes appear in many children with LQTS when they are between these ages. The present data support the data about the onset time of cardiac events of children with LQTS. Prevalence of LQTS in children: The specific prevalence of LQTS in the general population is not known, although it is thought to be approximately 1 of 10,000.2,9 These data may be based on the familial and/or symptomatic cases of LQTS. The present data were based on the screening program, and showed a prevalence of nonfamilial LQTS to be 4 of 4,655 (1 of 1,164) in the seventh grade, greater than previously suspected. Study limitations: The present study did not include genetic diagnosis for LQTS children. We recommend that our data be compared with a follow-up study of individuals diagnosed as children with genotype-identified LQTS. 1. Roden DM, Lazzara R, Rosen M, Schwartz PJ, Towbin J, Vincent GM, for the
SADS Foundation Task Force on LQTS. Multiple mechanisms in the long-QT syndrome. Current knowledge, gaps, and future directions. Circulation 1996;94: 1996 –2012. 2. Schwartz PJ. The long QT syndrome. Curr Prob Cardiol 1997;226:302–351. 3. Schwartz PJ, Moss AJ, Vincent GM, Crampton RS. Diagnostic criteria for the long QT syndrome. An update. Circulation 1993;88:782–784. 4. Rosero SZ, Zareba W, Moss AJ, Robinson JL, Hajj-Ali RH, Locati EH, Benhorin J, Andrews ML. Asthma and the risk of cardiac events in the Long QT syndrome. Long QT Syndrome Investigative Group. Am J Cardiol 1999;84: 1406 –1411.
398 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 89
5. Moss AJ, Robinson JL, Gessman L, Gillespie R, Zareba W, Schwartz PJ, Vincent GM, Benhorin J, Heilbron EL, Towbin JA, et al. Comparison of clinical and genetic variables of cardiac events associated with loud noise versus swimming among subjects with the long QT syndrome. Am J Cardiol 1999;848:876 – 879. 6. Moss AJ, Zareba W, Hall J, Schwartz PJ, Crampton RS, Benhorin J, Vincent M, Locati EH, Priori SG, Napolitano C, et al. Effectiveness and limitation of -blocker therapy in congenital long-QT syndrome. Circulation 2000;101:616 – 623. 7. Wilde AAM, Roden DM. Predicting the long-QT genotype from clinical data. From sense to science. Circulation 2000;102:2796 –2798. 8. Zhang L, Timothy KW, Vincent GM, Lehmann MH, Fox J, Giuli LC, Shen J, Splawski I, Priori SG, Compton SJ, et al. Spectrum of ST-T-wave patterns and repolarization parameters in congenital long-QT syndrome. ECG findings identify genotypes. Circulation 2000;102:2849 –2855. 9. Zupancic JA, Triedman JK, Alexander M, Walsh EP, Richardson DK, Berul CI. Cost-effectiveness and implications of newborn screening for prolongation of QT interval for the prevention of sudden infant death syndrome. J Pediatr 2000;1364:481–489. 10. Aihoshi S, Yoshinaga M, Nakamura M, Oku S, Haraguchi T, Nishibatake M. Screening for QT prolongation using a new exponential formula. Jpn Circ J 1995;59:185–189. 11. Yoshinaga M, Kamimura K, Fukushige T, Kusubae R, Shimago A, Nishi J, Kono Y, Nomura Y, Miyata K. Face immersion in cold water induces prolongation of the QT interval and T-wave changes in children with non-familial long QT syndrome. Am J Cardiol 1999;83:1494 –1497. 12. Frank S, Colliver JA, Frank A. The electrocardiogram in obesity: statistical analysis of 1,029 patients. J Am Coll Cardiol 1986;7:295–299. 13. el-Gamal A, Gallagher D, Nawras A, Gandhi P, Gomez J, Allison DB, Steinberg JS, Shumacher D, Blank R, Heymsfield SB. Effects of obesity on QT, RR, and QTc intervals. Am J Cardiol 1995;75:956 –959. 14. Pietrobelli A, Rothacker D, Gallagher D, Heymsfield SB. Electrocardiographic QTc interval: short term weight loss effects. Int J Obes Relat Metab Disord 1997;212:110 –114. 15. Pidlich J, Pfeffel F, Zwiauer K, Schneider B, Schmidinger H. The effect of weight reduction on the surface electrocardiogram: a prospective trial in obese children and adolescents. Int J Obes Relat Metab Disord 1997;21:1018 –1023. 16. Carella MJ, Mantz SL, Rovner DR, Willis PW, Gossain VV, Bouknight RR, Ferenchick GS. Obesity, adiposity, and lengthening of the QT interval: improvement after weight loss. Int J Obes Relat Metab Disord 1996;20:938 –942. 17. Park JJ, Swan PD. Effect of obesity and regional adiposity on the QTc interval in women. Int J Obes Relat Metab Disord 1997;21:1104 –1110. 18. Yoshinaga M, Tomari T, Aihoshi S, Kawashita T, Nishi J, Tanaka Y, Takezaki T, Kono Y, Yuasa Y, Nakamura M, et al. Exponential correction of QT interval to minimize the effect of heart rate in children. Jpn Circ J 1993;57:102– 108. 19. Aihoshi S, Yoshinaga M, Tomari T, Nakamura M, Nomura Y, Oku S, Haraguchi T, Osawa N, Miyata K. Correction of the QT interval in children. Jpn Circ J 1995;59:190 –197. 20. Eberle T, Hessling G, Ulmer HE, Brockmeier K. Prediction of normal QT intervals in children. J Electrocardiol 1998;31:121–125. 21. Sagie A, Larson MG, Goldberg RJ, Bengstson JR, Levy D. An improved method for adjusting the QT interval for heart rate (the Framingham Heart Study). Am J Cardiol 1992;70:797–801. 22. Pearl W. Effect of gender, age, and heart rate on QT intervals in children. Pediatr Cardiol 1996;17:135–136.
FEBRUARY 15, 2002
interval. None of the 4 children diagnosed with LQTS in the seventh grade had characteristic electrocardiographic findings of LQTS in the first grade. All 4 are nonfamilial cases. The prevalence of LQTS in children was found to be 1 of 1,164. These data suggest that abnormal electrocardiographic phenotypes in children with nonfamilial LQTS may appear during the elementary school year. The longer QT intervals in female subjects than male subjects start at the same period. No correlation was found between obesity and length of the QT interval. Finally, the prevalence of LQTS in children is greater than previously suspected. 䊚2002 by Excerpta Medica, Inc. (Am J Cardiol 2002;89:395–398)
ong QT syndrome (LQTS) is a rare disease characterized by prolonged ventricular repolarization, L and the clinical presentation of LQTS is the occur-
between QT interval and obesity in school-aged children based on data provided by a large-scale study. In this study we examined changes in the QT interval with age during the childhood of normal and LQTS children, as well as the possible correlation of body mass index (BMI) with the QT interval; our final purpose was to determine the prevalence of LQTS in children.
1
rence of syncope or cardiac arrest in young persons.2,3 The first cardiac event among children with the LQT1 and LQT2 genotypes has been reported to occur between 9 and 12 years of age.4 – 8 ST-T-wave patterns also vary across genotypes and ages8; however, changes in the QT interval with age during childhood in normal children and children with LQTS have not been studied in detail. The prevalence of LQTS in the general population is also not well known, although it is thought to be approximately 1 of 10,000.2,9 A screening program conducted for Japanese children has uncovered large numbers of children with LQTS, but with no family history of the disease, or with any history of LQTS-related cardiac events at the time of diagnosis,10,11 suggesting that the prevalence of LQTS is greater than it may appear. An additional concern is that a prolonged QT interval is sometimes associated with obesity.12–17 However, there has been no report to date concerning the possible association
From the Department of Pediatrics, Faculty of Medicine, Kagoshima University and Kagoshima City Medical Association, Kagoshima, Japan and the Research Committee on Long QT Syndrome in Children in the Japanese Society of Pediatric Cardiology and Cardiac Surgery, Tokyo, Japan. Manuscript received September 7, 2001; revised manuscript received and accepted October 22, 2001. Address for reprints: Masao Yoshinaga, MD, Department of Pediatrics, Faculty of Medicine, Kagoshima University, Sakuragaoka 8-35-1, Kagoshima, 890-8520 Japan. E-mail: [email protected]. kagoshima-u.ac.jp. ©2002 by Excerpta Medica, Inc. All rights reserved. The American Journal of Cardiology Vol. 89 February 15, 2002
METHODS
Subjects: The subjects were 4,655 children whose data (name, sex, weight, and height) and electrocardiograms (ECGs) were obtained both in 1994 when they were first graders, and again in 2000 when they were seventh graders. Data were collected from the Screening Program for Heart Disease in Kagoshima City that has been operated by the Kagoshima City Medical Association and supported by the Board of Education in Kagoshima City. Screening program for heart diseases in Kagoshima City: All resting ECGs were recorded at a speed of 25
mm/s at each school by medical technologists of the Screening Program for Heart Disease in Kagoshima City. Screening for LQTS in this program was based on the electrocardiographic criteria described by Schwartz et al,3 namely, the prolonged corrected QT interval determined by Bazett’s formula (QTc interval) and T-wave morphology. When the heart rate was ⬎75 beats/min, values obtained through an exponential correction of the QT interval (QT/RR0.31) or through Fridericia’s formula (QTFc⫽QT/RR1/3) were also considered.10,18,19 Children diagnosed as having 0002-9149/02/$–see front matter PII S0002-9149(01)02259-7
395
TABLE 1 Mean Values of Heart Rates, Uncorrected QT Intervals, and QT Intervals Corrected by Bazett’s (QTc) Formula in Boys and Girls in the First and Seventh Grades Boys
Heart rate QT interval QTc values
Girls
First
Seventh
First
Seventh
82 ⫾ 12 0.332 ⫾ 0.023 0.386 ⫾ 0.021
77 ⫾ 12 0.353 ⫾ 0.025 0.395 ⫾ 0.022
84 ⫾ 12‡ 0.327 ⫾ 0.022‡ 0.384 ⫾ 0.021*
81 ⫾ 13‡ 0.350 ⫾ 0.025† 0.402 ⫾ 0.021‡
Changes in Uncorrected and Corrected QT Intervals Between the First and Seventh Grades QT interval QTc values
0.021 ⫾ 0.028 0.010 ⫾ 0.026
0.023 ⫾ 0.027* 0.018 ⫾ 0.025‡
*p ⬍0.05; †p ⫽ 0.0004; ‡p ⬍0.0001. The p value between the first and seventh grades of all variables in boys and girls is p ⬍0.0001 (symbols not shown). The difference in the mean values between boys and girls at the first and seventh grades and difference in the changes between the two grades are shown with symbols.
TABLE 2 Correlation Coefficients and p Values Between BMI and the RR Interval or QT Intervals QT Interval RR Interval 1st grade boys 1st grade girls 7th grade boys 7th grade girls
0.013 0.036 ⫺0.124 ⫺0.033
(0.5122) (0.0851) (⬍0.0001) (0.1387)
Uncorrected* 0.009 0.024 ⫺0.096 ⫺0.049
(0.6722) (0.2435) (⬍0.0001) (0.0251)
QTc* ⫺0.001 ⫺0.017 0.062 ⫺0.010
*Uncorrected QT intervals and QT intervals corrected by Bazett’s (QTc) formula. The data are expressed as a correlation coefficient and a p value in the bracket.
LQTS on the resting ECG were further examined. Their history was taken and they underwent physical examination and an exercise test (Master’s 2-step or a treadmill exercise test) and/or a face immersion test.11 These tests were performed at the Kagoshima Medical Association Hospital in Kagoshima City or at the outpatient clinic of the Kagoshima University Hospital. Measurements of height and weight: School nurses measured the height and weight of children in their schools in early April of each year. Height was measured without socks or shoes. Weight was measured while the child was wearing only underclothing. The BMI was calculated as (weight in kg)/(height in m)2. Measurement of QT interval: Three consecutive QT and RR intervals in lead V5 were measured retrospectively for each child in this study. The QT interval was measured from the onset of the Q wave to the end of the T wave in lead V5. Bifid T waves, but not U waves, were included in the QT measurement.8 The QT interval measurements were obtained by the first and second investigators (TF and MY). Intra- and interreader variability was tested by triplicated measurements of 100 randomly selected ECGs from the present study. Intrareader coefficients of variability of the first and second investigators were 1.7%, and the 396 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 89
(0.9591) (0.4209) (0.0040) (0.6454)
interreader coefficient of variability was 1.9% of the mean QT interval. Mean values of three corrected QT intervals: We measured 3 consecutive
QT/RR data pairs, and each was corrected using Bazett’s formula (QTc ⫽ QT/RR1/2). The mean values of the 3 consecutive QT/RR data pairs were used in the study. Diagnosis of LQTS: A final diagnosis of LQTS was made when the second investigator (MY) and an additional 3 members of the Japanese Society of Pediatric Cardiology and Cardiac Surgery (TS, MN, and IN [members outside the second investigator’s university]) agreed with the diagnosis, because the genetic abnormality was not determined in the present study. Statistical analysis: Data are expressed as mean ⫾ 1 SD. Statistical testing for correlation coefficients was based on Fisher’s Z transformation. Differences in mean values between the groups were analyzed using a paired or nonpaired Student’s t test. A p value ⬍0.05 was considered statistically significant.
RESULTS
Changes in uncorrected and corrected QT intervals between the first and seventh grades: The mean heart
rate, uncorrected and corrected QT intervals, and the changes in values between the first and seventh grades of the subjects are listed in Table 1. The mean uncorrected and corrected QT intervals of the subjects significantly (p ⬍0.0001) increased from the first to the seventh grades in boys and girls. The mean uncorrected and corrected QT intervals in girls were significantly lower than those in boys at the first screening, i.e., the first grade. However, the changes in the uncorrected and corrected QT intervals between the first and seventh grades were greater in girls than in boys. Relation between body mass index and uncorrected or corrected QT intervals: In the seventh grade boys, the
RR and uncorrected QT intervals were inversely associated with BMI, whereas the QTc value was positively associated with it (Table 2). In the seventh grade girls, the uncorrected and corrected QT intervals were inversely associated with BMI. A stepwise regression analysis was conducted to determine whether body mass affects QT intervals using the QT interval as a dependent variable and the BMI and heart rate as independent variables. This analysis revealed that body mass did not have any impact on the uncorrected or corrected QT intervals in either seventh grade boys or girls (data not shown). Incidence of LQTS in children: In the 1994 screening, no subject in the first grade was diagnosed as having FEBRUARY 15, 2002
FIGURE 1. Representative ECGs of a boy among 4 children diagnosed as having LQTS. Starting from the left, the ECG at the first grade (a), the ECG at the seventh grade (b), the ECG before (c), and after (d) a treadmill exercise test, and the ECG before (e) and after (f) a face immersion test. The paper speed of the ECGs during treadmill exercise testing was at 50 mm/s, and that of remaining ECGs was 25 mm/s. Obvious bifid T waves appeared in leads V3 to V5 after face immersion.
LQTS. In 2000, on the other hand, 9 children in the seventh grade were suspected of having LQTS from the resting ECG, 8 of whom were further examined. One child was unable to visit the hospital. Finally, 4 children (2 boys and 2 girls) were diagnosed as having LQTS. None of the diagnosed children had a family history of the disease or any past history of cardiac events. The first-grade ECGs of these 4 children did not show the typical characteristics of LQTS. Representative ECGs of 1 of these 4 children are shown in Figure 1. The ECG shows obvious bifid T waves in leads V3 to V5 after face immersion, mimicking the findings of genotype LQT2 described by Zhang et al.8 The remaining 3 patients had obvious bifid T waves in leads V4 to V6 (1 boy)—subtle bifid T waves with a second component on top of the T wave. These findings mimic those of genotype LQT2 (2 girls). The electrocardiographic findings of one girl mimicked the infantile pattern of genotype LQT1, although the child in this study was 12 years old. The incidence of long QT syndrome of children in the seventh grade was 4 of 4,655 (1 of 1,164).
DISCUSSION The present study indicates that longer QT intervals in females begin to be apparent in elementary school, earlier than previously reported.20 Obesity did not have any impact on the uncorrected or corrected QT intervals. Four children diagnosed with LQTS in the seventh grade had not had electrocardiographic patterns characteristic of LQTS in the first grade. The prevalence of LQTS in children was found to be 1 of 1,164.
Changes in the QT interval in childhood: The uncorrected and corrected QT intervals calculated by Bazett’s formula are the longest in adult women.21 However, these values and changes in these values in patients are not well known. Eberle et al20 and Pearl22 report that there is no significant influence of gender on QT interval up to the age of 14 years. The present study showed that uncorrected and corrected QT intervals were shorter in girls in the first grade, and that the difference in these values between the first and seventh grades was greater in girls than in boys. The reason for this longer QT interval may be the earlier appearance of puberty in girls.20 These data indicate that the longer QT intervals in females start at elementary school age and that its onset is earlier than previously reported. Relation between QT interval and overweight: Several reports claim that short-term weight loss is associated with shortening of the QT interval in children, adolescents, and adults.12–16 Abdominal obesity has also been associated with a prolonged QTc interval.17 In the present study, a stepwise regression analysis did not indicate any correlation between BMI and the uncorrected and corrected QT intervals, strongly suggesting that overweight is not associated with the length of the QT interval in children. The association of obesity and the QT interval in adults may be due to autonomic factors between obese adults and the QT interval,13 or may be the result of having the disease over a long period of time. Effect of age on diagnosing LQTS: None of the 4 children in the seventh grade diagnosed with nonfamilial LQTS had electrocardiographic patterns char-
ARRHYTHMIAS AND CONDUCTION DISTURBANCES/CHANGES IN CHILDHOOD QT INTERVALS
397
acteristic of LQTS in the first grade in the present study. The median age at the first cardiac event of LQTS children of the LQT1 and LQT2 genotypes has been reported to be from 9 to 12 years.4 –7 That of children of the LQT3 genotype has been reported at as high as 16 to 18 years,5,7 as has the median age for auditory-related events associated with the LQT2 genotype. These data indicate that the risk of cardiac events in children with LQTS starts at approximately 9 to 12 years. The present study revealed that LQTS children had characteristic electrocardiographic findings in the seventh grade (12 or 13 years old), but not in the first grade (6 or 7 years old). Thus, abnormal electrocardiographic phenotypes appear in many children with LQTS when they are between these ages. The present data support the data about the onset time of cardiac events of children with LQTS. Prevalence of LQTS in children: The specific prevalence of LQTS in the general population is not known, although it is thought to be approximately 1 of 10,000.2,9 These data may be based on the familial and/or symptomatic cases of LQTS. The present data were based on the screening program, and showed a prevalence of nonfamilial LQTS to be 4 of 4,655 (1 of 1,164) in the seventh grade, greater than previously suspected. Study limitations: The present study did not include genetic diagnosis for LQTS children. We recommend that our data be compared with a follow-up study of individuals diagnosed as children with genotype-identified LQTS. 1. Roden DM, Lazzara R, Rosen M, Schwartz PJ, Towbin J, Vincent GM, for the
SADS Foundation Task Force on LQTS. Multiple mechanisms in the long-QT syndrome. Current knowledge, gaps, and future directions. Circulation 1996;94: 1996 –2012. 2. Schwartz PJ. The long QT syndrome. Curr Prob Cardiol 1997;226:302–351. 3. Schwartz PJ, Moss AJ, Vincent GM, Crampton RS. Diagnostic criteria for the long QT syndrome. An update. Circulation 1993;88:782–784. 4. Rosero SZ, Zareba W, Moss AJ, Robinson JL, Hajj-Ali RH, Locati EH, Benhorin J, Andrews ML. Asthma and the risk of cardiac events in the Long QT syndrome. Long QT Syndrome Investigative Group. Am J Cardiol 1999;84: 1406 –1411.
398 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 89
5. Moss AJ, Robinson JL, Gessman L, Gillespie R, Zareba W, Schwartz PJ, Vincent GM, Benhorin J, Heilbron EL, Towbin JA, et al. Comparison of clinical and genetic variables of cardiac events associated with loud noise versus swimming among subjects with the long QT syndrome. Am J Cardiol 1999;848:876 – 879. 6. Moss AJ, Zareba W, Hall J, Schwartz PJ, Crampton RS, Benhorin J, Vincent M, Locati EH, Priori SG, Napolitano C, et al. Effectiveness and limitation of -blocker therapy in congenital long-QT syndrome. Circulation 2000;101:616 – 623. 7. Wilde AAM, Roden DM. Predicting the long-QT genotype from clinical data. From sense to science. Circulation 2000;102:2796 –2798. 8. Zhang L, Timothy KW, Vincent GM, Lehmann MH, Fox J, Giuli LC, Shen J, Splawski I, Priori SG, Compton SJ, et al. Spectrum of ST-T-wave patterns and repolarization parameters in congenital long-QT syndrome. ECG findings identify genotypes. Circulation 2000;102:2849 –2855. 9. Zupancic JA, Triedman JK, Alexander M, Walsh EP, Richardson DK, Berul CI. Cost-effectiveness and implications of newborn screening for prolongation of QT interval for the prevention of sudden infant death syndrome. J Pediatr 2000;1364:481–489. 10. Aihoshi S, Yoshinaga M, Nakamura M, Oku S, Haraguchi T, Nishibatake M. Screening for QT prolongation using a new exponential formula. Jpn Circ J 1995;59:185–189. 11. Yoshinaga M, Kamimura K, Fukushige T, Kusubae R, Shimago A, Nishi J, Kono Y, Nomura Y, Miyata K. Face immersion in cold water induces prolongation of the QT interval and T-wave changes in children with non-familial long QT syndrome. Am J Cardiol 1999;83:1494 –1497. 12. Frank S, Colliver JA, Frank A. The electrocardiogram in obesity: statistical analysis of 1,029 patients. J Am Coll Cardiol 1986;7:295–299. 13. el-Gamal A, Gallagher D, Nawras A, Gandhi P, Gomez J, Allison DB, Steinberg JS, Shumacher D, Blank R, Heymsfield SB. Effects of obesity on QT, RR, and QTc intervals. Am J Cardiol 1995;75:956 –959. 14. Pietrobelli A, Rothacker D, Gallagher D, Heymsfield SB. Electrocardiographic QTc interval: short term weight loss effects. Int J Obes Relat Metab Disord 1997;212:110 –114. 15. Pidlich J, Pfeffel F, Zwiauer K, Schneider B, Schmidinger H. The effect of weight reduction on the surface electrocardiogram: a prospective trial in obese children and adolescents. Int J Obes Relat Metab Disord 1997;21:1018 –1023. 16. Carella MJ, Mantz SL, Rovner DR, Willis PW, Gossain VV, Bouknight RR, Ferenchick GS. Obesity, adiposity, and lengthening of the QT interval: improvement after weight loss. Int J Obes Relat Metab Disord 1996;20:938 –942. 17. Park JJ, Swan PD. Effect of obesity and regional adiposity on the QTc interval in women. Int J Obes Relat Metab Disord 1997;21:1104 –1110. 18. Yoshinaga M, Tomari T, Aihoshi S, Kawashita T, Nishi J, Tanaka Y, Takezaki T, Kono Y, Yuasa Y, Nakamura M, et al. Exponential correction of QT interval to minimize the effect of heart rate in children. Jpn Circ J 1993;57:102– 108. 19. Aihoshi S, Yoshinaga M, Tomari T, Nakamura M, Nomura Y, Oku S, Haraguchi T, Osawa N, Miyata K. Correction of the QT interval in children. Jpn Circ J 1995;59:190 –197. 20. Eberle T, Hessling G, Ulmer HE, Brockmeier K. Prediction of normal QT intervals in children. J Electrocardiol 1998;31:121–125. 21. Sagie A, Larson MG, Goldberg RJ, Bengstson JR, Levy D. An improved method for adjusting the QT interval for heart rate (the Framingham Heart Study). Am J Cardiol 1992;70:797–801. 22. Pearl W. Effect of gender, age, and heart rate on QT intervals in children. Pediatr Cardiol 1996;17:135–136.
FEBRUARY 15, 2002