Published online before print August 6, 2007, doi: 10.1161/CIRCULATIONAHA.106.676551
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
(Circulation. 2007;116:714-720.)
© 2007 American Heart Association, Inc.
Epidemiology |
Prevalence and Prognostic Significance of Short QT Interval in a Middle-Aged Finnish Population
From the Department of Medicine (O.A.), Päijät-Häme Central Hospital, Lahti, Finland; National Public Health Institute (H.R., A.R.), Helsinki, Finland; Department of Cardiology (M.V.), University of Helsinki, Helsinki, Finland; and Department of Internal Medicine (M.J.J., H.V.H.), University of Oulu, Oulu, Finland.
Correspondence to Dr Olli Anttonen, Department of Medicine, Päijät-Häme Central Hospital, Keskussairalankatu 7, 15850, Lahti, Finland. E-mail olli.anttonen{at}phsotey.fi
Received November 15, 2006; accepted June 4, 20007.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Background— Short-QT syndrome is an inherited disorder characterized by a short QT interval and an increased risk of sudden cardiac death. The clinical significance of a short QT interval observed in a randomly recorded ECG is not known. Therefore, we assessed the prevalence and prognostic significance of a short QT interval in a general population.
Methods and Results— QT intervals were measured from the 12-lead ECGs of 10 822 randomly selected middle-aged subjects (5658 males, mean age 44±8.4 years) enrolled in a population study and followed up for 29±10 years. The end points were all-cause and cardiovascular mortality. In addition to Bazett’s method (corrected QT interval, or QTc), the Fridericia (QTfc) and nomogram (QTnc) methods were used to correct the QT interval for heart rate. The cutoff values for short QT intervals were defined as 320 ms (very short) and 340 ms (short). The prevalence of QT interval <320 ms based on QTc, QTfc, and QTnc was 0.10%, 0.08%, and 0.06%, and the prevalence of QT interval <340 ms was 0.4%, 0.3%, and 0.3%, respectively. The majority of subjects with short QT intervals were males. All-cause or cardiovascular mortality did not differ between subjects with a very short or short QT interval and those with normal QT intervals (360 to 450 ms). There were no sudden cardiac deaths, aborted sudden cardiac deaths, or documented ventricular tachyarrhythmias among subjects with a QTfc <340 ms.
Conclusion— A short QT interval does not appear to indicate an increased risk for all-cause or cardiovascular mortality in middle-aged nonreferral, community-based individuals.
Key Words: death, sudden • electrocardiography • prognosis • epidemiology
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
A long QT interval measured from a 12-lead ECG has long been regarded as a risk factor for premature death,1–3 but much less is known of the reverse, when the QT interval is abnormally short. The short-QT syndrome has been characterized recently as a novel inherited syndrome that leads to increased susceptibility to arrhythmias and sudden cardiac death.4,5 At present, only a few families and some sporadic individuals have been reported to exhibit the short-QT syndrome.6,7 Major ECG findings include a persistently short QT interval and tall, peaked T waves with virtually no ST segment. There is no definite or even uniformly accepted cutoff value for a short QT interval, but usually, the rate-corrected QT has been <320 ms, which is a clear indicator of an abnormally short ventricular repolarization time.6
Editorial p 686
Clinical Perspective p 720
Population-based studies have mainly concentrated on the long end of the spectrum of QT intervals,1–3 and much less is known of the distribution of a short QT interval and its clinical significance in the general population. Some reports have claimed that a short QT interval in the ECG is associated with an increased risk of dying suddenly, but various cutoff values have been used in those studies to define the short QT interval.8 Therefore, we set out to study the prevalence and prognostic significance of a short QT interval in a large, random, middle-aged Finnish population with a long follow-up.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Study Population
The study population comprised the entire cohort of 10 957 men and women aged 30 to 59 years (mean age 44±8.4 years) drawn from 12 different geographic areas of Finland participating in the Social Insurance Institution’s Coronary Heart Disease Study.9 Health examinations were performed in 1966 to 1972, and the invited individuals were a good geographic representation of the people of that age in Finland. The distribution of the occupations of the invited people was also similar to that of all people in Finland.9 The participation rate was 90%. The invited persons filled in a questionnaire that included questions on health habits, known diseases, and medication before the health examination, and the questionnaire was checked at the examination by a specially trained nurse. Coronary heart disease was diagnosed in 4.9% of the men and 2.1% of the women and heart failure in 3.9% of the men and 3.4% of the women, respectively. The main medications taken for heart diseases at the time of the baseline investigation were nitroglycerin (3.5%) and digitalis (3.6%). The use of β-blockers or other medications causing potential changes in cardiac repolarization was extremely rare at that time. From the original recordings, the QT interval could be measured reliably in 10 822 subjects (98.7%), who formed the present study group.
ECG Measurements
At entry into the study, standard 12-lead ECGs were recorded at a paper speed of 50 mm/s and were analyzed originally by 9 trained technicians or nurses and stored in an ECG database for possible further analysis. For the present study, QT intervals <360 ms were reanalyzed after a computer search by an experienced cardiologist with special qualification in clinical electrophysiology (O.A.) from the original paper recordings using 2 consecutive cycles in leads II or V2 and a standard tangential method (Figure 1) corrected for heart rate with Bazett’s formula (QTc=QT/
RR),10 a nomogram formula (QTnc=QT+correcting number),11 and the Fridericia method (QTfc=QT/3RR).12 The additional QT measurements other than Bazett’s formula were used because of their better performance at low and high heart rates.11 To assess the concordance of the initial measurements of the QT intervals made by trained technicians or nurses and those measured by an experienced cardiologist, a correlation between these measurements was analyzed in the sample of 366 ECGs, which also included 54 with a normal QTc in the original measurements. A relatively good correlation was observed between these measurements (Pearson correlation coefficient 0.78, P<0.0001). None of the 54 subjects with a QTc interval >360 ms in the initial measurement by a nonexperienced person was reclassified as having a short QTc (<340 ms).
|
In the present study, individuals with QTc, QTnc, or QTfc of less than 340 ms were classified as having a short QT interval, and a value of less than 320 ms was classified as a very short QT interval. These cutoff values were chosen on the basis of data on patients with short-QT syndrome that showed that the majority of patients had QTc <320 ms,6 although a few cases with QTc between 320 and 340 ms have been reported.13 Because of a possible gender difference in cutoff values of short QT intervals, we also used a cutoff value of 360 ms in mortality analyses to include more females in the analyses. All ECG analyses were performed blinded to the follow-up data.
Follow-Up
The end points of the follow-up were death due to all causes (total mortality), death due to all cardiovascular diseases, and death due to coronary artery disease until the end of 2005. The mortality of the study population has been monitored continuously. Death certificates and cause-of-death registers from Statistics Finland, including autopsy records, if available, were used to assess mortality statistics and causes of death. Less than 2% of the population was lost to follow-up after they moved abroad, but the survival status of these individuals was still obtained in the majority of these cases. The validity of these registers has been studied and confirmed previously.2 The mean follow-up was 29±10 years. Among the subjects with a short QT interval, deaths were also classified as sudden or nonsudden by the end-points committee (O.A., M.J.J., and H.V.H.). Sudden cardiac death was defined as a death that occurred within 24 hours after the onset of symptoms. We also screened the National Hospital Discharge Register records for all those study subjects who had died and determined whether they had a history of syncope, aborted sudden cardiac death, atrial fibrillation, or ventricular tachyarrhythmia that necessitated hospitalization. The National Drug Reimbursement Register was investigated to analyze the use of reimbursed medication for cardiac arrhythmias or other heart diseases.
Statistical Analysis
Values are expressed as mean±1 SD or, in case of categorical variables, as prevalences. The 
-coefficient was calculated to assess agreement between the different rate-correction formulas for QT interval. We also calculated the overall -coefficient for very short (<320 ms) and short (<340 ms) QT interval rate-correction formulas. Kaplan-Meier survival curves were plotted for normal QTc (360 to 449 ms) and short QTc (<340 ms) and for normal QTfc (360 to 449 ms) and short QTfc (<340 ms), respectively, and were compared with the log-rank test. The general linear model was used to compare the age- and sex-adjusted means of continuous variables or prevalences of categorical variables of the study subjects between different QT-interval groups. Categorical variables classified for the analyses were smoking (smoker of cigarettes/cigars/pipe, never/ex-smoker), use of chronotropic medication (yes, no), and history of cardiovascular disease at baseline (yes, no).
The hazard ratios and 95% CIs of mortality were computed with Cox proportional hazards models. Models were first adjusted for age and sex and then further adjusted for covariates that were selected on the basis of prior evidence of an association with cardiovascular and overall mortality and statistical evidence from the present study population.9 Age, cholesterol, heart rate, and systolic blood pressure were added as continuous variables in the multivariate model, whereas sex (male, female), smoking (never/ex-smoker, smoker of cigarettes/cigars/pipe), and cardiovascular disease at baseline (yes, no) were added as categorical variables. The statistical analyses were performed with SAS version 9.1.3 (SAS Institute, Cary, NC). P<0.05 was considered statistically significant.
The authors had full access to and take full responsibility for the integrity of the data. All authors have read and agree to the manuscript as written.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Prevalence of Very Short and Short QT Intervals
Eleven of 10 822 subjects (0.10%; 95% CI, 0% to 0.2%) had a QTc interval <320 ms, 9 (0.08%; 95% CI, 0% to 0.15%) had QTfc <320 ms, and 7 (0.06%; 95% CI, 0% to 0.1%) had QTnc <320 ms (Table 1). The range of QTfc or QTnc intervals was 305 to 318 ms for those with a very short QT interval (<320 ms). QTc was <340 ms in 43 of 10 822 subjects (0.4%; 95% CI, 0.3% to 0.5%), whereas QTfc was <340 ms in 33 of 10 822 and QTnc was <340 ms in 34 of 10 822 subjects (0.3%; 95% CI, 0.2% to 0.4% for both).
|
The concordance of short-QT interval measurements calculated with the different methods is shown in Table 1. The overall -score for defining the short QT interval with different correlation formulas was 0.85 for short QT interval (<340 ms) and 0.56 for very short QT interval (<320 ms). As expected, all individuals with discordance of QTc with QTnc or QTfc values had low heart rates, and hence, their QTc value corrected by Bazett’s method was shorter than that measured with the other 2 correcting methods. Bazett’s formula tends to overshorten the QT interval at low heart rates.11 Given the discordant results of the presence of short QT interval (<340 ms) between QTc and the other 2 methods, we decided to use QTfc in addition to QTc in the analysis of the characteristics and mortality of subjects with a short QT interval.
Features of Subjects With a Short QT Interval
There were gender differences in the prevalence of short QTfc and QTc (Table 2). When 2 and 3 SDs of the means of QTc and QTfc intervals were used as cutoff values for short QT interval, respectively, women had significantly higher cutoffs than men, particularly for the short QTc (Table 2). When the proportion of subjects with short QTc was analyzed separately for men and women, 4.4% of males and 1.3% of females had QTc <360 ms.
|
In addition to gender difference, subjects with a short QTfc or QTc were younger than those with a normal QT interval (Table 3). Comparison of baseline variables between age- and gender-matched subjects with normal QT intervals (360 to 450 ms) and those with short QT intervals (<340 ms) showed that individuals with QTfc or QTc <340 ms also had lower heart rates and lower systolic blood pressures (Table 3).
|
Short QT Interval and Mortality
The prognosis of the subjects with a short QT interval was not worse than that of subjects with a normal QT interval (Figures 2 and 3
). There were no differences in all-cause mortality between the 2 groups, even after adjustments for all possible confounding factors (Table 4). In fact, cardiovascular mortality tended to be lower for subjects with a short QT interval, because there was only 1 cardiovascular death among subjects with either QTc <320 ms (relative risk, 0.41; 95% CI, 0.06 to 2.91) or QTfc <320 ms (relative risk, 0.49; 95% CI, 0.07 to 3.51), and there were only 5 cardiovascular deaths among those with QTc <340 ms and 3 among those with QTfc <340 ms (Table 5). Similarly, there were no sudden deaths in the subjects with short QTc or QTfc intervals (Table 5). Neither all-cause mortality nor cardiovascular mortality differed between subjects with QTfc or QTc <360 ms and those with normal QTfc or QTnc (Table 4). Conversely, subjects with a rate-corrected QT interval >450 ms had greater all-cause and cardiovascular mortality than those with normal or short QT intervals (Table 4).
|
|
|
|
The specific causes of deaths of the subjects with short QTfc intervals are shown in Table 5. One subject with a short QT interval was lost to follow-up. Of the 15 deaths among subjects with a short QTfc interval, there were 3 that were attributable to myocardial infarction.
A review of the medical records of those subjects whose QTfc was <340 ms revealed no hospital visits due to syncope, serious ventricular tachyarrhythmia, or even atrial fibrillation. None of the already deceased individuals or those still living who had a short QT interval had received reimbursement for antiarrhythmic medication.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The present data show that a short QT interval is not associated with an increased risk of cardiovascular mortality or sudden death during a long follow-up in a population of nonreferral middle-aged subjects. Furthermore, no hospitalizations due to syncope or ventricular tachyarrhythmia had occurred among the subjects with a short QT interval, and none had been prescribed antiarrhythmic drugs. These observations challenge the recently introduced claim that a short QT interval as a sole determinant of abnormal ventricular repolarization is a serious condition associated with a high risk of life-threatening arrhythmias.14 However, the present findings concur with those observed recently by Gallagher et al,15 who studied healthy young subjects with a relatively short follow-up, and by Viskin et al,16 both of whom found that a short QT interval may not always indicate an increased risk for cardiac mortality.
Measurement and Definition of Short QT Interval
When ECG recording is performed in a random population with a wide age range and with different clinical conditions, a wide range of heart rates is usually encountered. Because cardiac cycle length is not the only predictor of the QT interval, most equation formulas have failed to work well with heart rates in the high and low ranges. Although Bazett’s formula has its recognized drawbacks, it is rooted so strongly in medical practice that it was chosen for rate correction in the present study. Because of the methodological problems related to its accuracy at low and high heart rates,11 we also made calculations with the other equations mentioned above to overcome these limitations and to diminish the influence of bradycardia and tachycardia. In the present study, we noticed that many of the subjects with a short QT interval measured by Bazett’s method were in fact different individuals from those with a short QT interval measured by the QTfc or QTnc methods. Therefore, it was crucial to also correct the QT interval with the actual heart rate in the present study, although we realize that there is no consensus within the scientific community on the ideal correction method.
Rautaharju et al17 studied >14 000 healthy individuals to assess the deviation of QT intervals in the general population. They also set limits for the low and high boundaries of QT intervals. In general, QT values below and beyond 2 SD from the mean were considered as either short or long. In their study population, 360 subjects (0.4%) had a QT interval above or below 2 SD from the mean QT interval. The limit of short QT intervals in the present study was selected as 320 ms for a very short QT interval and 340 ms for a short QT interval, because these values clearly represent abnormally short QT intervals in the present study population (3.3 SD and 2.4 SD of the mean, respectively), and because there are no reported cases of short-QT syndrome with a QTc >340 ms. In fact, the majority of patients with short-QT syndrome have had a QTc interval <320 ms.6
The boundaries of corrected short QT intervals (320 and 340 ms) were very short in the present study. There were many subjects who had a QT interval >340 ms but <360 ms. According to many criteria,16 these subjects could also be considered to have a shorter than normal QT interval. However, even with the inclusion of these subjects in the analysis, the results with respect to mortality did not change. The cutoff value of 360 ms also enabled us to include more women in the analyses.
Determinants of a Short QT Interval
The length of the QT interval is influenced at least in part by the sex hormones, mainly testosterone. During puberty, the QT interval tends to shorten in males but remains almost constant in females.17 Individuals with a short QT interval in the present study were almost exclusively younger males. It is possible that a short QT interval in young males is a reflection of the greater influence of testosterone on repolarization, representing only the tail of the normal gaussian distribution of QT intervals in the general population, and that occurrence of the actual short-QT syndrome is very rare. Another possibility is that the subjects with a very short QT interval (<320 ms) identified in the present study do have a true short-QT syndrome, with an inherited or sporadic mutation in the gene(s) regulating the ion channels, but the prognosis of this syndrome is benign, at least in the present middle-aged Finnish population.
Heart rate was also lower among subjects with a short QT interval. This may be due in part to inaccuracy of all correction formulas for heart rate or to the fact that a short QT interval and attenuated sinus node activity share common alterations in ion channel function. Another possibility is that there are other common factors, such as hormonal effects or effects of the autonomic nervous system, that influence both cardiac repolarization and sinus node activity. In this respect, the lower systolic blood pressure found in subjects with a short QT interval points to a role of the autonomic nervous system.
Study Limitations
Only middle-aged subjects were included in the present study. Therefore, it may not be appropriate to generalize the results to younger subjects. The number of subjects with a short QT interval was also relatively small. Therefore, the study may be underpowered to give a definitive answer regarding the prognostic significance of a short QT interval. On the other hand, to the best of our knowledge, this is one of the largest studies reported so far in which the prognostic significance of the QT interval, especially a short QT interval, has been assessed in a random population with a lengthy follow-up. We also used only leads II and V2 in the measurement of QT intervals. The prevalence of a short QT interval might have been different if all 12 leads had been used in the measurements.
One potential limitation is the gender difference in the QT intervals. Unlike for the long QT interval, there have been no generally established separate cutoffs for abnormally short QT intervals for men and women. The cutoff value for an abnormally short QT interval appeared to be longer for women than for men. This became especially evident when Bazett’s formula was used for rate correction. Therefore, different cutoff values for short QT intervals should perhaps also be used for both sexes. We attempted to avoid this bias in the present study by using a higher cutoff value (360 ms) that also included women.
Implications and Conclusions
There is a general belief that a short QT interval observed in the ECG is a serious condition that confers a major risk for sudden cardiac death, and hence, that aggressive preventive therapy (ie, implantable cardioverter-defibrillators, antiarrhythmic medication, or at least screening and counseling of families) is justified. The present study shows that an abnormally short QT interval in a middle-aged population may not be a life-threatening condition that requires extensive risk stratification or antiarrhythmic treatment, at least in nonreferral, community-based individuals. The results might be different in a highly selected referral sample of individuals. Furthermore, it is fundamentally important to make careful note of family history, because these recommendations do not necessarily apply to familial cases, and further studies are needed to undertake risk stratification of subjects with short QT intervals to be able to define diagnostic criteria for the short-QT syndrome and individual proposals with regard to therapy. Simply put, however, the presence of a short QT interval in the ECG does not appear to justify the diagnosis of short-QT syndrome.
|
Acknowledgments |
|---|
Sources of Funding
Dr Anttonen received funding from the Finnish Cultural Foundation, Helsinki, Finland. The research group of Dr Huikuri is funded by the Finnish Academy of Science, Sigrid Juselius Foundation, and the Finnish Foundation for Cardiovascular Research, Helsinki, Finland.
Disclosures
None.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
1. Goldberg RJ, Bengtson J, Chen ZY, Anderson KM, Locati E, Levy D. Duration of the QT interval and total and cardiovascular mortality in healthy persons (the Framingham Heart Study experience). Am J Cardiol. 1991; 67: 55–58.[CrossRef][Medline] [Order article via Infotrieve]
2. Karjalainen J, Reunanen A, Ristola P, Viitasalo M. QT interval as a cardiac risk factor in a middle aged population. Heart. 1997; 77: 543–548.
3. Rautaharju PM, Kooperberg C, Larson JC, LaCroix A. Electrocardiographic abnormalities that predict coronary heart disease events and mortality in postmenopausal women: the Women’s Health Initiative. Circulation. 2006; 113: 473–480.
4. Gussak I, Brugada P, Brugada J, Wright RS, Kopecky SL, Chaitman BR, Pjerregaard P. Idiopathic short QT interval: a new clinical syndrome? Cardiology. 2000; 94: 99–102.[CrossRef][Medline] [Order article via Infotrieve]
5. Gaita F, Giustetto C, Bianchi F, Wolpert C, Schrimpf R, Riccardi R, Grossi S, Richiardi E, Borgreffe M. Short QT syndrome: a familial cause of sudden death. Circulation. 2003; 108: 965–970.
6. Giustetto C, Di Monte F, Wolpert C, Borgreffe M, Schimpf R, Spragia P, Leone G, Maury P, Anttonen O, Haissaguerre M, Gaita F. Short QT syndrome: clinical findings and diagnostic-therapeutic implications. Eur Heart J. 2006; 27: 2440–2447.
7. Bellocq C, van Ginneken AC, Bezzina CR, Alders M, Escande D, Mannens MM, Baró I, Wilde AA. Mutation in the KCNQ1 gene leading to the short QT interval syndrome. Circulation. 2004; 109: 2394–2397.
8. Algra A, Tijssen JG, Roelandt JR, Pool J, Lubsen J. QT interval variables from 24 hour electrocardiography and the two year risk of sudden death. Br Heart J. 1993; 70: 43–48.
9. Reunanen A, Aromaa A, Pyörälä K, Punsar S, Maatela J, Knekt P. The Social Insurance Institution’s Coronary Heart Study: baseline data and 5-year mortality experience. Acta Med Scand Suppl. 1983; 673: 1–120.[Medline] [Order article via Infotrieve]
10. Bazett HC. An analysis of time relations of electrocardiogram. Heart. 1920; 7: 353–370.
11. Karjalainen J, Viitasalo M, Mänttäri M, Manninen V. Relation between QT intervals and heart rates from 40 to 120 beats/min in rest electrocardiograms of men and a simple method to adjust QT interval values. J Am Coll Cardiol. 1994; 23: 1547–1553.[Abstract]
12. Fridericia LS. Die Systolendauer im Elektrokardiogramm bei normalen Menschen und bei Herzkranken. Acta Med Scand. 1920; 53: 469–486.
13. Maury P, Hollington L, Duparc A, Brugada R. Short QT syndrome: should we push the frontier forward? Heart Rhythm. 2005; 2: 1135–113.[CrossRef][Medline] [Order article via Infotrieve]
14. Bjerregaard P, Gussak I. Short QT syndrome. Ann Noninvasive Electrocardiol. 2005; 10: 436–440.[CrossRef][Medline] [Order article via Infotrieve]
15. Gallagher M, Magliano G, Yap YG, Padula M, Morgia V, Postorino C, Di Liberato F, Leo R, Borzi M, Romeo F. Distribution and prognostic significance of QT intervals in lowest half centile in 12,012 apparently healthy persons. Am J Cardiol. 2006; 98: 933–935.[CrossRef][Medline] [Order article via Infotrieve]
16. Viskin S, Zeltser D, Ish-Shalom M, Katz A, Glickson M, Justo D, Tekes-Manova D, Belhassen B. Is idiopathic ventricular fibrillation a short QT syndrome? Comparison of QT intervals of patients with idiopathic ventricular fibrillation and healthy controls. Heart Rhythm. 2004; 1: 587–591.[CrossRef][Medline] [Order article via Infotrieve]
17. Rautaharju PM, Zhou SH, Wong S, Calhoun HP, Berenson GS, Prineas R, Davignon A. Sex differences in the evolution of the electrocardiographic QT interval with age. Can J Cardiol. 1992; 8: 690–695.[Medline] [Order article via Infotrieve]
 |
|
 |
|
  J. R. Silva, H. Pan, D. Wu, A. Nekouzadeh, K. F. Decker, J. Cui, N. A. Baker, D. Sept, and Y. Rudy A multiscale model linking ion-channel molecular dynamics and electrostatics to the cardiac action potential PNAS, July 7, 2009; 106(27): 11102 - 11106. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Ding, W. Hua, H. Niu, K. Chen, and S. Zhang Primary prevention of sudden cardiac death using implantable cardioverter defibrillators Europace, September 1, 2008; 10(9): 1034 - 1041. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. N. Viswanathan and R. L. Page Short QT: When Does It Matter? Circulation, August 14, 2007; 116(7): 686 - 688. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||