You are here

  • Home
  • Archive
  • Volume 96, Issue 9
  • Prevalence and risk factors for hypertension in Hong Kong Chinese adolescents: waist circumference predicts hypertension, exercise decreases risk

Article Text

Article menu

Download PDFPDF

Original articles
Prevalence and risk factors for hypertension in Hong Kong Chinese adolescents: waist circumference predicts hypertension, exercise decreases risk
  1. Lettie C K Leung1,
  2. Rita Y T Sung2,
  3. Hung-Kwan So2,
  4. Sik Nin Wong3,
  5. Kwok Wai Lee4,
  6. Kwok Piu Lee5,
  7. Man Ching Yam2,
  8. Samantha P S Li3,
  9. So Fun Yuen6,
  10. Stella Chim7,
  11. Keung Kit Chan1,
  12. David Luk8
  1. 1Department of Paediatrics, Kwong Wah Hospital, Kowloon, Hong Kong, China
  2. 2Department of Paediatrics, Prince of Wales Hospital, Shatin, Hong Kong, China
  3. 3Department of Paediatrics, Tuen Mun Hospital, Tuen Mun, Hong Kong, China
  4. 4Department of Paediatrics, Queen Elizabeth Hospital, Kowloon, Hong Kong, China
  5. 5Department of Paediatrics, Alice Ho Miu Ling Nethersole Hospital, Tai Po, Hong Kong, China
  6. 6Department of Paediatrics, Pamela Youde Nethersole Eastern Hospital, Hong Kong, China
  7. 7Department of Paediatrics, Queen Mary Hospital, Hong Kong, China
  8. 8Department of Paediatrics, United Christian Hospital, Kwun Tong, Hong Kong, China
  1. Correspondence to Dr Lettie C K Leung, Department of Paediatrics, Kwong Wah Hospital, 25 Waterloo Road, Kowloon, Hong Kong, China leungckl{at}ha.org.hk

Abstract

Purpose To determine the prevalence, risk factors for and patterns of hypertension in Chinese adolescents based on a territory-wide school based screening programme in Hong Kong.

Methods Cross-sectional anthropometric and oscillometric blood pressure (BP) measurements and lifestyle information were obtained as part of a growth survey of students from randomly selected secondary schools in Hong Kong. Those with blood pressure ≥95th centile were screened a second or third time. Hypertension is defined as elevated blood pressure on three separate occasions. The independent effects of age, sex, body mass index, high waist circumference (≥85th centile), sleep duration, family history of hypertension and frequency of exercise on hypertension were explored by multivariate analysis.

Results Among the 6193 students screened, the prevalence of elevated blood pressure on the first, second and third screens was 9.54%, 2.77% and 1.44% respectively. Hypertension was more likely to be systolic. High waist circumference (≥85th centile) was independently associated with a higher risk of hypertension (adjusted OR 2.4), while exercising twice or more per week was protective (adjusted OR 0.28).

Conclusions The prevalence of hypertension in Hong Kong Chinese adolescents is 1.44%. The current study shows high waist circumference is a predictor of hypertension in adolescents, while increased physical activity is a protective factor. Incorporating waist circumference into screening protocols may increase the sensitivity of cardiovascular risk stratification. Healthcare providers should be strong advocates helping to prevent obesity and promote physical activity in adolescents and children.

https://doi.org/10.1136/adc.2010.202770

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Introduction

    Hypertension in childhood has been shown to be a predictor of hypertension in adulthood1 and is associated with early markers of cardiovascular disease.2,–,4 Recognition of hypertension at an early age may provide an opportunity for risk reduction. With increasing childhood obesity,5 there is a greater need for paediatricians to identify cardiovascular risk in children and target them for intervention.

    Little is known about the epidemiology, risk factors for and patterns of hypertension in Chinese children. A mainland Chinese study used only a single blood pressure measurement to define hypertension.6 Up to now, there have been no published data using the Fourth Report definition of elevated blood pressure on three separate occasions.7 Thus, when a territory-wide school growth study was undertaken by the Chinese University of Hong Kong, we aimed to identify the prevalence of and risk factors for hypertension in Chinese adolescents.

    What is already known on this topic

    • ▶ Obesity is positively related to hypertension in children.

    • ▶ Waist circumference is a known independent risk factor for hypertension in adults but is not well established in children.

    • ▶ Physical activity is a known protective factor for hypertension in adults but is not well established in children.

    What this study adds

    • ▶ The prevalence of hypertension in urban Chinese adolescents is 1.44% as defined by Fourth Report criteria.

    • ▶ High waist circumference is independently associated with a 2.4-fold increased risk for hypertension in Hong Kong Chinese adolescents.

    • ▶ Two or more extra-curricular exercise sessions per week decrease the risk of hypertension by 3.5 times.

    Subjects and methods

    In 2006, a growth survey of over 14 000 school children in Hong Kong aged 6–18 years was undertaken by the Department of Paediatrics of the Chinese University of Hong Kong. It involved 38 secondary and primary schools randomly selected from each of 18 districts in Hong Kong; the sampling details have been described previously.5 8 Our study only recruited secondary school students. Anthropometric and blood pressure measurements and lifestyle information were collected by a team of eight trained research staff.8

    Blood pressure measurements, and the definition and prevalence of hypertension

    Blood pressure was measured using an Accutorr Plus monitor (Datascope, Montvale, New Jersey, USA), an oscillometric device validated with a mercury sphygmomanometer in children.9 At each school screening, blood pressure was measured on the right arm with the students seated and rested for at least 5 min, and the arm supported at heart level. Appropriately sized cuffs were used.7 Two blood pressure measurements were taken 1 min apart and the average was used. Students whose average systolic blood pressure (SBP) or diastolic blood pressure (DBP) reading was ≥95th centile for age and gender8 underwent a second set of blood pressure measurements at least 2 weeks later. Students with SBP or DBP ≥95th centile at the second screen underwent a third set of blood pressure measurements at least another 2 weeks later. Students with elevated SBP or DBP on all three screens were considered hypertensive.7

    Anthropometric, demographic and other information

    Weight and height were measured for each student. Body mass index (BMI) was calculated as weight(kg)/height(m2). BMI z score was calculated by local reference to a 1993 cohort specific for gender and age.10 Overweight and obesity were defined as BMI z scores of >1.0 and <1.65 (equivalent to the 85th–94th centiles) and ≥1.65 (≥95th centile), respectively.11 Waist circumference was measured midway between the lowest rib and the superior border of the iliac crest using an inelastic measuring tape. High waist circumference is defined as a waist circumference ≥85th centile, based on an receiver operating characteristic analysis of metabolic risk factors from a previous study by our group.12

    Information on potential risk factors was collected using a self-administered questionnaire completed by the student and a parent, and checked by research staff. Any queries were resolved directly with the student or their parent by phone. Physical activity was assessed by weekly frequency of regular extracurricular exercise sessions other than school physical education classes. Hong Kong schools generally allocate two 45 min lessons per week to physical education. Exercise during these classes was regarded as a constant and not taken into account in this analysis. For sleep duration, the students were asked what time on average they fell asleep and woke on weekdays. Ethics committee approval was obtained for the study.

    Data analysis

    Prevalence of hypertension

    The prevalence of hypertension was adjusted for loss to follow-up by applying the rate of disease occurrence among those successfully screened to those lost to follow-up, a method used previously.13 The prevalence of elevated blood pressure was calculated by dividing the number of students with elevated blood pressure at screening by the sum of (1) the number of normotensive students at the previous screening(s) and (2) the number of students with elevated blood pressure at the previous screenings(s) who were rescreened according to the protocol.

    The characteristics of students from non-participating schools and schools that participated in the second and third blood pressure screens were compared to determine potential selection bias. Continuous variables were compared using the independent t test. Categorical variables were compared using the χ2 test. We also compared the characteristics of defaulters with those who completed the blood pressure screens to determine if the latter were representative of the whole sample.

    As normative blood pressure varies with age, height and sex, where there were significant differences in age or sex between groups, we normalised any related blood pressure variation by deriving a BP index,8 where

    Embedded Image

    Risk factors for hypertension

    The subjects were divided into three groups. Group 1 consisted of the normotensive students, whose blood pressure was <95th centile on the first screen. Group 2 were those with transient elevated blood pressure, whose blood pressure was ≥95th centile on the first screen but <95% at either the second or third screen. Group 3 consisted of students with hypertension, whose blood pressures were ≥95th centile on all three blood pressure screens.

    The objective of this analysis was to identify potential risk factors independently associated with hypertension. We viewed our model-building strategy as hypothesis generating because we did not have targeted exposure. The following variables were examined: age, gender, BMI z score, high waist circumference (≥85th centile), sleep duration, history of prematurity, parental history of hypertension, sleep duration and frequency of exercise sessions.

    Univariate analysis

    Continuous data were compared between the three groups using one way analysis of variance. Between-group differences for these variables were analysed post hoc using the Games–Howell method, with adjustments made depending on the agreement of the assumption of variance. Categorical data were compared between the three groups using the χ2 test. Pearson's correlation analysis with adjusted p values (significant at p<0.016) was used to assess the differences in proportion between the three groups.

    Multivariate analysis

    To control for confounding factors, variables with significant differences among the three groups by univariate analysis were included in a multivariable logistic regression model. For this, groups 1 (normotensive) and 2 (transient elevated blood pressure) were combined into a control group and compared to the hypertensive group. Adjusted ORs were computed. All statistical analyses were conducted with SPSS for Windows v 14 (SPSS, Chicago, Illinois, USA).

    Results

    Prevalence of hypertension

    In total, 7918 secondary school students were recruited into the growth survey. Due to workforce limitations, only students from 14 schools were recruited into the study, the particular school chosen being determined by proximity to the eight participating paediatric units. Thus, data on only 6193 students (3074 boys and 3119 girls) were used for prevalence calculation. Table 1 summarises their population characteristics. All were ethnically Chinese and their mean age was 15.2 years. The percentages of overweight and obese children were 11.2% and 2.9%, respectively.

    View this table:
    Table 1

    Population characteristics of students from participating schools compared to students from non-participating schools

    The results of screening for elevated blood pressure are shown schematically in figure 1. Of 6193 students, 9.54% had elevated blood pressure on the first screen. Some students were not rescreened because they were absent or occupied on the day of screening or had left school. Of the 84.6% rescreened students, 33.8% were found to have elevated blood pressure on the second screen. Of these rescreened students, 94.7% had a third screen and 55.0% of them had elevated blood pressure on all three occasions.

    Figure 1
    Figure 1

    Elevated blood pressure at each screening and estimated prevalence of hypertension extrapolated to the total study population. The estimated prevalence of hypertension at the second and third screenings is adjusted for loss to follow-up by applying the rate of disease occurrence among those who were successfully screened to those lost to follow-up.12 This is calculated by dividing the number of students with elevated blood pressure at the screening by the sum of (1) the number of normotensive students at the previous screening(s) and (2) the number of students with elevated blood pressure at the previous screenings(s) who were rescreened according to the protocol. *Estimated prevalence of hypertension after the first screen: 591/6193=9.54%. †Estimated prevalence of hypertension after the second screen: 169/(5602+500)=2.77%. ‡Estimated prevalence of hypertension after the third screen: 88/(5602+331+160)=1.44%. BP, blood pressure; HT, hypertensive; NT, normotensive.

    The prevalence of students with elevated blood pressure after the first screen was 9.54%. The adjusted prevalence was 2.77% after the second screen and 1.44% after the third screen, the latter being the prevalence of hypertension.

    Validity of prevalence calculation

    Students from participating schools were compared to the whole sample of 18 schools (table 1). Students from non-participating schools were 0.2 years older, less likely to be male and slightly slimmer. However, there was no significant difference in weight class distribution. They had comparable SBP at the first screen but their DBP was 1.2 mm Hg higher. Since we were examining a large population sample, minor differences will reach statistical difference, but significant clinical difference is unlikely since the definition of hypertension takes into account gender and age and there is no significant difference in weight class distribution.

    The characteristics of those who were not rescreened were compared with those who were rescreened (table 2). The defaulters were on average 1.6 years older, possibly because many students defaulted as they had completed high school. There was also a borderline preponderance of females. However, there was no difference in weight class distribution. We normalised the age and sex difference effect on blood pressure by calculating the BP index. After adjustment, the defaulters had a higher first screen SBP. So it is possible that our figures underestimate the prevalence of hypertension, although first blood pressure measurement does not necessarily predict subsequent measurements.

    View this table:
    Table 2

    Comparison of age, gender, BMI z score, weight class and blood pressure in students who did not complete blood pressure rescreens with those who completed all three blood pressure screens

    Characteristics of hypertensive adolescents

    Table 3 summaries the characteristics of students who were hypertensive, normotensive or had transient elevated blood pressure. Hypertension was found in 88 students, with a male to female ratio of 56:32. The mean age was 15.4 years (SD 1.8). Forty (45.5%) were of normal weight, 17 (19.3%) were overweight and 31 (35.2%) were obese. Overall, children with normal weight had a hypertension prevalence of 0.9%, while obese children had a hypertension prevalence of 7.1%.

    View this table:
    Table 3

    Characteristics of normotensive students, hypertensive students and students with transient elevated blood pressure: univariate analysis of anthropometric and demographic variables

    Regarding the pattern of hypertension, 34 (38.6%) had isolated systolic hypertension, 8 (9.1%) had isolated diastolic hypertension and 46 (52.3%) had both systolic and diastolic hypertension. The ratio of systolic to diastolic hypertension was 1:0.68, demonstrating a predominant systolic pattern of hypertension in these students.

    Risk factors for hypertension

    Complete information on the explored variables was available for 5901 students as shown in table 3. Significant variables identified by univariate analysis included age, gender, BMI z score, high waist circumference, sleep duration, parental history of hypertension and frequency of exercise sessions. These variables were entered into a logistic regression model with the results shown in table 4. High waist circumference (adjusted OR 2.378, 95% CI 1.13 to 4.99; p=0.022) was found to be the only independent risk factor for childhood hypertension; two or more exercise sessions a week (adjusted OR 0.282, 95% CI 0.11 to 0.71; p=0.007) were found to be an independent protective factor. Thus, high waist circumference increased the risk of hypertension 2.4-fold, whereas students who exercised for two or more sessions per week had a 3.5-fold lower risk.

    View this table:
    Table 4

    Logistic regression to identify independent risk factors for hypertension, showing high waist circumference is a risk factor for hypertension, and exercise a protective factor

    Discussion

    Hypertension prevalence

    The current study is, to the best of our knowledge, the first large scale study of hypertension prevalence among urban Chinese adolescents, using the criterion of elevated blood pressure on three separate occasions. Studies that define hypertension by blood pressure measured on a single occasion14 15 may over‑estimated the prevalence of hypertension. In the current study, elevated blood pressure was found in 9.54%, 2.77% and 1.44% of students after repeated screens, reflecting the phenomenon of regression to the mean. This shows the importance of confirming elevated blood pressure on multiple occasions before labelling a child as having hypertension.13 16 Also, in our study, BMI, waist circumference percentiles and hypertension definition were based on local data from a large number of subjects, using the same measurement methods, thus maximising its strength.

    The hypertension prevalence in our adolescents is lower than that of their US counterparts. Sorof et al13 found a hypertension prevalence of 4.5% among students after three screens using the Spacelabs oscillometric monitor (Spacelabs, Issaquah, Washington, USA). The higher prevalence may be explained by the skewed distribution to higher BMI percentiles in their study population (20% vs 2.9% obese in our study population). Another reason may be the different blood pressure monitors used. The Spacelabs monitor failed validation in children,17 overestimating SBP by 4–5 mm Hg. The Accutorr oscillometric monitor used in our study has been validated in children.9

    Risk factors for hypertension

    In the current study, 35.2% of the hypertensive adolescents were obese, although only 2.9% of the population were obese.5 Just 0.9% of normal weight adolescents were hypertensive, compared to 7.1% of obese adolescents. Our findings agree with previous studies demonstrating an association between hypertension and obesity in children.13,–,15 18 19

    Among the risk factors examined, only waist circumference emerged as an independent risk factor for hypertension. After adjustment for all other factors, waist circumference ≥85th centile increased the risk of hypertension by 2.4-fold. Our finding that high waist circumference was a better predictor of hypertension than BMI z score is reasonable as muscle and fat are included in the BMI calculation, whereas waist circumference is related to visceral adiposity in adults20 and children21 and thus may be a more sensitive determinant of cardiovascular risk.

    Indeed in adults, abdominal adiposity is a better predictor of health risk than BMI alone.22 23 However, in children, the role of waist circumference as an independent predictor is uncertain. Janssen et al found high waist circumference within weight categories was more closely associated than BMI with lipid profile and the metabolic syndrome, although there was no such effect for high blood pressure.24 Other studies showed waist circumference may be a better anthropometric indicator than BMI of elevated blood pressure and metabolic risk factors.25,–,28 However, these studies on children mainly compared anthropometric measurements and did not adjust for potential confounding factors, in contrast to the current study.

    In showing that waist circumference ≥85th centile is the best positive predictor of hypertension, the current study supplements the knowledge that waist circumference is a predictor of cardiovascular risk. Waist circumference is a simple clinical parameter which can be easily incorporated into screening protocols and may improve risk stratification.

    The current study also found that two or more extracurricular exercise sessions a week reduced the risk of hypertension. We previously reported that higher exercise frequency was associated with lower blood pressure as measured at a single sitting.29 The current study demonstrates that higher exercise frequency lowers the risk of hypertension based on a more robust definition of hypertension. In adults, it is accepted that regular physical activity lowers blood pressure,30 although this is less clear in children. Studies either showed a weak association31 or failed to show an independent association between exercise and blood pressure.32 33 Recently, Maximova et al showed that in adolescents, one less physical activity session per year over 5 years was associated with a 0.18–0.40 mm Hg rise in SBP even after adjustment for BMI, waist circumference and skin-fold thickness.34 Although the current study demonstrates an inverse relationship between exercise and hypertension independent of weight change, more intervention studies need to be conducted to establish causality.

    Regarding sleep duration, adult studies have shown that shorter sleep duration and lower sleep maintenance predict higher blood pressure cross-sectionally and longitudinally.35 We were unable to find this association, probably because sleep duration information was limited to weekdays and adolescents are known to have irregular sleeping patterns, especially at weekends. The information was also limited to students' recall rather than acquired by actigraphy. One recent study using wrist actigraph showed that low weekday sleep efficiency (but not sleep duration) was associated with a higher risk of prehypertension.36 Further studies on sleep and hypertension in children need to address both weekday and weekend sleep duration, preferably using wrist actigraphs.

    There are other limitations to our study. Our default rate was 15.4% for the second screen and 5.3% for the third screen. However, this is still equivalent to or better than default rates in other school-based studies.13 37 Analysis showed defaulters were older and had higher sex adjusted SBP at the first screen. Whether this is reflected by subsequent blood pressure results is unknown, but the true hypertension prevalence rate may have been underestimated.

    We have not followed up those students whose blood pressure at the second or third screens fell below the 95th centile. The phenomenon of repeat blood pressure normalising may not necessarily translate to low risk. It would be worthwhile following such students up to determine increased risk for hypertension.

    In conclusion, we believe the current study, with a large sample size of randomly selected schools and repeated blood pressure measurements using a validated oscillometric blood pressure monitor, provides a reliable estimate of the prevalence of hypertension in Hong Kong Chinese adolescents. Systolic hypertension is more common. Of the factors studied, only high waist circumference predicted hypertension on multivariate analysis, with increased exercise frequency being a protective factor. Healthcare providers should be strong advocates helping prevent obesity and promote physical activity in adolescents.

    Acknowledgments

    The authors thank Mr C H Chan for statistical support and Mr C H Loong for clerical support.

    References

    1. ↵
      1. Lauer RM,
      2. Clarke WR
      . Childhood risk factors for high adult blood pressure: the Muscatine Study. Pediatrics 1989;84:633–41.
      OpenUrlAbstract/FREE Full Text
    2. ↵
      1. Sorof JM,
      2. Cardwell G,
      3. Franco K,
      4. et al
      . Ambulatory blood pressure and left ventricular mass index in hypertensive children. Hypertension 2002;39:903–8.
      OpenUrlAbstract/FREE Full Text
    3. ↵
      1. Sanchez A,
      2. Barth JD,
      3. Zhang L
      . The carotid artery wall thickness in teenagers is related to their diet and the typical risk factors of heart disease among adults. Atherosclerosis 2000;152:265–6.
      OpenUrlCrossRefPubMedWeb of Science
    4. ↵
      1. Arnett DK,
      2. Glasser SP,
      3. McVeigh G,
      4. et al
      . Blood pressure and arterial compliance in young adults: the Minnesota Children's Blood Pressure Study. Am J Hypertens 2001;14:200–5.
      OpenUrlAbstract/FREE Full Text
    5. ↵
      1. So HK,
      2. Nelson EA,
      3. Li AM,
      4. et al
      . Secular changes in height, weight and body mass index in Hong Kong Children. BMC Public Health 2008;8:320.
      OpenUrlCrossRefPubMed
    6. ↵
      1. Li Y,
      2. Yang X,
      3. Zhai F,
      4. et al
      . Childhood obesity and its health consequence in China. Obes Rev 2008;9(Suppl 1):82–6.
      OpenUrlCrossRefPubMed
    7. ↵
      National High Blood Pressure Education Program Working Group on High Blood Pressure in Children and Adolescents. The fourth report on the diagnosis, evaluation, and treatment of high blood pressure in children and adolescents. Pediatrics 2004;114(2 Suppl 4th Report):555–76.
      OpenUrlCrossRefPubMedWeb of Science
    8. ↵
      1. Sung RY,
      2. Choi KC,
      3. So HK,
      4. et al
      . Oscillometrically measured blood pressure in Hong Kong Chinese children and associations with anthropometric parameters. J Hypertens 2008;26:678–84.
      OpenUrlCrossRefPubMedWeb of Science
    9. ↵
      1. Wong SN,
      2. Tz Sung RY,
      3. Leung LC
      . Validation of three oscillometric blood pressure devices against auscultatory mercury sphygmomanometer in children. Blood Press Monit 2006;11:281–91.
      OpenUrlCrossRefPubMedWeb of Science
    10. ↵
      1. Leung SS,
      2. Lau JT,
      3. Tse LY,
      4. et al
      . Weight-for-age and weight-for-height references for Hong Kong children from birth to 18 years. J Paediatr Child Health 1996;32:103–9.
      OpenUrlCrossRefPubMedWeb of Science
    11. ↵
      1. Flegal KM,
      2. Ogden CL,
      3. Wei R,
      4. et al
      . Prevalence of overweight in US children: comparison of US growth charts from the Centers for Disease Control and Prevention with other reference values for body mass index. Am J Clin Nutr 2001;73:1086–93.
      OpenUrlAbstract/FREE Full Text
    12. ↵
      1. Sung RY,
      2. Yu CC,
      3. Choi KC,
      4. et al
      . Waist circumference and body mass index in Chinese children: cutoff values for predicting cardiovascular risk factors. Int J Obes (Lond) 2007;31:550–8.
      OpenUrlCrossRefPubMed
    13. ↵
      1. Sorof JM,
      2. Lai D,
      3. Turner J,
      4. et al
      . Overweight, ethnicity, and the prevalence of hypertension in school-aged children. Pediatrics 2004;113:475–82.
      OpenUrlAbstract/FREE Full Text
    14. ↵
      1. Urrutia-Rojas X,
      2. Egbuchunam CU,
      3. Bae S,
      4. et al
      . High blood pressure in school children: prevalence and risk factors. BMC Pediatr 2006;6:32.
      OpenUrlCrossRefPubMed
    15. ↵
      1. Wan Y-P,
      2. Zhang X-M,
      3. Lu L-P,
      4. et al
      . Distribution tendency of 7326 primary and middle school students in Shanghai and analysis of the associated factors. J Clin Pediatr 2007;25:1005–8.
      OpenUrl
    16. ↵
      1. Adrogué HE,
      2. Sinaiko AR
      . Prevalence of hypertension in junior high school-aged children: effect of new recommendations in the 1996 Updated Task Force Report. Am J Hypertens 2001;14:412–14.
      OpenUrlCrossRefPubMedWeb of Science
    17. ↵
      1. Belsha CW,
      2. Wells TG,
      3. Bowe Rice H,
      4. et al
      . Accuracy of the SpaceLabs 90207 ambulatory blood pressure monitor in children and adolescents. Blood Press Monit 1996;1:127–33.
      OpenUrlPubMed
    18. ↵
      1. Harding S,
      2. Maynard MJ,
      3. Cruickshank K,
      4. et al
      . Overweight, obesity and high blood pressure in an ethnically diverse sample of adolescents in Britain: the Medical Research Council DASH study. Int J Obes (Lond) 2008;32:82–90.
      OpenUrlCrossRefPubMed
    19. ↵
      1. Li YP,
      2. Yang XG,
      3. Zhai FY,
      4. et al
      . Disease risks of childhood obesity in China. Biomed Environ Sci 2005;18:401–10.
      OpenUrlPubMedWeb of Science
    20. ↵
      1. Clasey JL,
      2. Bouchard C,
      3. Teates CD,
      4. et al
      . The use of anthropometric and dual-energy X-ray absorptiometry (DXA) measures to estimate total abdominal and abdominal visceral fat in men and women. Obes Res 1999;7:256–64.
      OpenUrlPubMedWeb of Science
    21. ↵
      1. Taylor RW,
      2. Jones IE,
      3. Williams SM,
      4. et al
      . Evaluation of waist circumference, waist-to-hip ratio, and the conicity index as screening tools for high trunk fat mass, as measured by dual-energy X-ray absorptiometry, in children aged 3–19 y. Am J Clin Nutr 2000;72:490–5.
      OpenUrlAbstract/FREE Full Text
    22. ↵
      1. Ardern CI,
      2. Katzmarzyk PT,
      3. Janssen I,
      4. et al
      . Discrimination of health risk by combined body mass index and waist circumference. Obes Res 2003;11:135–42.
      OpenUrlPubMedWeb of Science
    23. ↵
      1. Janssen I,
      2. Katzmarzyk PT,
      3. Ross R
      . Waist circumference and not body mass index explains obesity-related health risk. Am J Clin Nutr 2004;79:379–84.
      OpenUrlAbstract/FREE Full Text
    24. ↵
      1. Janssen I,
      2. Katzmarzyk PT,
      3. Srinivasan SR,
      4. et al
      . Combined influence of body mass index and waist circumference on coronary artery disease risk factors among children and adolescents. Pediatrics 2005;115:1623–30.
      OpenUrlAbstract/FREE Full Text
    25. ↵
      1. Watts K,
      2. Bell LM,
      3. Byrne SM,
      4. et al
      . Waist circumference predicts cardiovascular risk in young Australian children. J Paediatr Child Health 2008;44:709–15.
      OpenUrlCrossRefPubMedWeb of Science
    26. ↵
      1. Chuang SY,
      2. Pan WH
      . Predictability and implications of anthropometric indices for metabolic abnormalities in children: nutrition and health survey in Taiwan elementary children, 2001–2002. Asia Pac J Clin Nutr 2009;18:272–9.
      OpenUrlPubMed
    27. ↵
      1. Meininger JC,
      2. Brosnan CA,
      3. Eissa MA,
      4. et al
      . Overweight and central adiposity in school-age children and links with hypertension. J Pediatr Nurs 2010;25:119–25.
      OpenUrlCrossRefPubMed
    28. ↵
      1. Viggiano D,
      2. De Filippo G,
      3. Rendina D,
      4. et al.
      ; O.Si.M.E. Study Group. Screening of metabolic syndrome in obese children: a primary care concern. J Pediatr Gastroenterol Nutr 2009;49:329–34.
      OpenUrlCrossRefPubMedWeb of Science
    29. ↵
      1. So HK,
      2. Sung RY,
      3. Li AM,
      4. et al
      . Higher exercise frequency associated with lower blood pressure in Hong Kong adolescents: a population-based study. J Hum Hypertens 2010;24:646–51.
      OpenUrlCrossRefPubMed
    30. ↵
      1. Whelton SP,
      2. Chin A,
      3. Xin X,
      4. et al
      . Effect of aerobic exercise on blood pressure: a meta-analysis of randomized, controlled trials. Ann Intern Med 2002;136:493–503.
      OpenUrlCrossRefPubMedWeb of Science
    31. ↵
      1. Leary SD,
      2. Ness AR,
      3. Smith GD,
      4. et al
      . Physical activity and blood pressure in childhood: findings from a population-based study. Hypertension 2008;51:92–8.
      OpenUrlAbstract/FREE Full Text
    32. ↵
      1. Gaya AR,
      2. Alves A,
      3. Aires L,
      4. et al
      . Association between time spent in sedentary, moderate to vigorous physical activity, body mass index, cardiorespiratory fitness and blood pressure. Ann Hum Biol 2009;36:379–87.
      OpenUrlCrossRefPubMed
    33. ↵
      1. Kelley GA,
      2. Kelley KS,
      3. Tran ZV
      . The effects of exercise on resting blood pressure in children and adolescents: a meta-analysis of randomized controlled trials. Prev Cardiol 2003;6:8–16.
      OpenUrlPubMed
    34. ↵
      1. Maximova K,
      2. O'Loughlin J,
      3. Paradis G,
      4. et al
      . Declines in physical activity and higher systolic blood pressure in adolescence. Am J Epidemiol 2009;170:1084–94.
      OpenUrlAbstract/FREE Full Text
    35. ↵
      1. Knutson KL,
      2. Van Cauter E,
      3. Rathouz PJ,
      4. et al
      . Association between sleep and blood pressure in midlife: the CARDIA sleep study. Arch Intern Med 2009;169:1055–61.
      OpenUrlCrossRefPubMedWeb of Science
    36. ↵
      1. Javaheri S,
      2. Storfer-Isser A,
      3. Rosen CL,
      4. et al
      . Sleep quality and elevated blood pressure in adolescents. Circulation 2008;118:1034–40.
      OpenUrlAbstract/FREE Full Text
    37. ↵
      1. McNiece KL,
      2. Poffenbarger TS,
      3. Turner JL,
      4. et al
      . Prevalence of hypertension and pre-hypertension among adolescents. J Pediatr 2007;150:640–4, 644.e1.
      OpenUrlCrossRefPubMedWeb of Science

    Footnotes

    • Funding This study was supported by the Hong Kong Paediatric Nephrology Society with a research grant from the Children's Kidney Trust Fund.

    • Competing interests None.

    • Ethics approval This study was conducted with the approval of the Hospital Authority Kowloon West Cluster Ethics Committee, Hong Kong.

    • Provenance and peer review Not commissioned; externally peer reviewed.