Journal of Human Hypertension (2010), 1–5
& 2010 Macmillan Publishers Limited All rights reserved 0950-9240/10
www.nature.com/jhh
ORIGINAL ARTICLE
How to simplify the diagnostic criteria
of hypertension in adolescents
Q Lu, CM Ma, FZ Yin, BW Liu, DH Lou and XL Liu
Department of Endocrinology, The First Hospital of Qinhuangdao, Qinhuangdao, China
Diagnosis of hypertension in adolescents is complicated because blood pressure values vary with age,
gender and height. How can we simplify the diagnostic
criteria for hypertension in adolescents? In 2006,
anthropometric measurements were assessed in a
cross-sectional population-based study of 3136 Han
adolescents aged 13–17 years. Hypertension was defined according to the 2004 National High Blood
Pressure Education Program Working Group definition.
The following equations for blood pressure-to-height
ratio (BPHR) were used: systolic BPHR (SBPHR) ¼ SBP
(mm Hg)/height (cm) and diastolic BPHR (DBPHR) ¼ DBP
(mm Hg)/height (cm). Receiver-operating characteristic
curve analyses were performed to assess the accuracy
of SBPHR and DBPHR as diagnostic tests for elevated
systolic blood pressure (SBP) and diastolic blood
pressure (DBP), respectively. After the cutoff points
were determined, hypertension was defined by SBPHR/
DBPHR, and the sensitivity and specificity were calculated. The accuracy of SBPHR and DBPHR (assessed by
area under the curve) for identifying elevated SBP and
DBP was 40.85 (0.989–1.000). The optimal thresholds
of SBPHR/DBPHR for defining hypertension (stages 1
and 2) were 0.75/0.48 for boys and 0.78/0.51 for girls, and
for defining hypertension (stage 2) were 0.81/0.57 for
boys and 0.84/0.63 for girls. In identifying hypertension,
the sensitivity and specificity were both 490% (91.0–
99.1%). In identifying stage 2 hypertension, when the
sensitivity was 100%, the specificity was 98.6% for boys
and 99.1% for girls. BPHR is a simple, accurate and nonage-dependent index for screening hypertension in Han
adolescents, especially for stage 2 hypertension.
Journal of Human Hypertension advance online publication,
29 April 2010; doi:10.1038/jhh.2010.46
Keywords: blood pressure-to-height ratio; diagnosis; adolescent
Introduction
The prevalence of hypertension in children and
adolescents seems to be increasing.1 This rise is
partially because of the increasing prevalence of
obesity among children and adolescents, as well as a
growing awareness of this disease. There is evidence
that hypertension in children and adolescents can
lead to adult hypertension.2,3 Hypertension is known
as a risk factor for coronary artery disease in adults,
and the presence of hypertension in children and
adolescents may contribute to the early development
of coronary artery disease. Previous reports have
shown that early development of atherosclerosis
does exist in children and adolescents and may be
associated with childhood hypertension.4,5
Appropriate early-stage diagnosis and intervention of hypertension in children and adolescents are
important for reducing the risk of hypertensionrelated disorders in adults. Identifying hypertension
in children and adolescents is more complicated
than in adults because it is based on a set of age-,
Correspondence: Dr FZ Yin, Department of Endocrinology,
The First Hospital of Qinhuangdao, No 258 Wenhua Road,
Qinhuangdao, 066000, Hebei Province, China.
E-mail: [email protected]
Received 24 January 2010; revised 30 March 2010; accepted 31
March 2010
gender- and height-specific references of systolic
blood pressure (SBP) and diastolic blood pressure
(DBP).6 Although blood pressure has been generally
accepted as a reasonable diagnostic criterion, it still
has potential limitations. The age-, gender- and
height-specific standards of SBP and DBP are less
easily used by parents and non-professionals. More
recently, an increasing number of studies documented that the ratio of waist circumference to height
(waist-to-height ratio) was even superior to waist
circumference and body mass index (BMI) to predict
overweight and obesity in children and adolescents.7,8 It is not known whether the ratio of blood
pressure to height can predict hypertension in
adolescents. This study, for the first time, evaluated
the feasibility and accuracy of the systolic blood
pressure-to-height ratio (SBPHR) and the diastolic
blood pressure-to-height ratio (DBPHR) and proposed the optimal thresholds of SBPHR and DBPHR
for identifying hypertension in Han adolescents
aged 13–17 years.
Subjects and methods
Subjects
After obtaining informed consent from adolescents
and their parents, a cross-sectional, population-based
study was conducted. The study population was
Diagnostic criteria of hypertension in adolescents
Q Lu et al
2
determined according to two-stage cluster sampling.
During the first stage, samples of middle schools in
Hebei, China, were randomly obtained; in the
second stage, adolescents aged 13–17 years in these
schools were invited to participate. In the end, 3136
Han adolescents (1601 boys and 1535 girls) participated in 2006.
All participants were required to be healthy. For
this purpose, both a detailed medical history and a
complete physical examination were performed
before the study. Adolescents with a diagnosis of
secondary hypertension, acute or chronic illnesses,
infections, renal or hepatic diseases or neoplasia or
who were under medical treatment were excluded.
Measurements
Anthropometric measurements, including height
and weight, were performed when subjects were
without shoes and in light clothing. Height and
weight were measured to the nearest 0.1 cm and
0.1 kg, respectively. Standing height without shoes
was measured three times with a stadiometer, and
the three measurements were averaged for analysis.
BMI was defined as weight (kg) divided by height
(m) squared. Blood pressure was measured three
times with a mercury sphygmomanometer while the
subjects were seated after 10 min of rest, and the
three measurements were averaged for analysis.
Blood pressure cuff width was 40–50% of the arm
circumference. SBP was determined by the onset of
the ‘tapping’ Korotkoff sounds (K1). The fifth
Korotkoff sound (K5) was the definition of DBP.
Only if a very low K5 persisted was the K4 (muffling
of the sounds) recorded as the DBP. But the guideline of 2004 Working Group did not define the value
of very low K5. In our study, very low K5 was
defined as o40 mm Hg. The following equations for
BPHR were used: SBPHR ¼ SBP (mm Hg)/height
(cm) and DBPHR ¼ DBP (mm Hg)/height (cm).
Definitions
Obesity was defined by BMI of X95th percentiles,
overweight by BMI between the 85th and 95th
percentiles and normal weight by BMI of o85th
percentiles. BMI cutoff values used were age and
gender specific according to the BMI cutoff reference
norm for Chinese children and adolescents.9 The
definition of hypertension was as follows: (1)
hypertension (stages 1 and 2)—mean SBP or DBP
X95th percentile; and (2) hypertension (stage 2)—
mean SBP or DBP 499th percentile þ 5 mm Hg. All
adolescents’ blood pressure measurements were
compared with 2004 Working Group normative tables
to determine the final reported hypertensive status.10
Statistical analysis
All analyses were performed using the SPSS 11.5
statistical software (SPSS 11.5 for Windows; SPSS,
Inc., Chicago, IL, USA). Numerical variables were
Journal of Human Hypertension
reported as the mean±s.d. Comparisons were
performed between groups using Student’s t-test.
The Pearson correlation coefficient was used to
measure the strength of association between two
variables. Multivariate logistic regression analysis
was performed to examine the relationship between
hypertension and obesity. Po0.05 was considered to
be statistically significant.
Using receiver-operating characteristic (ROC) analysis, the ROC curves of SBPHR were drawn to show
how well they could separate subjects into groups
with or without elevated SBP (defined by mean SBP
of X95th percentile and mean SBP of 499th
percentile þ 5 mm Hg, respectively), with age-, gender- and height-specific references of SBP being the
gold standard. The ROC curves of DBPHR were
drawn to show how well they could separate
subjects into groups with and without elevated
DBP (defined by mean DBP of X95th percentile
and mean DBP of 499th percentile þ 5 mm Hg,
respectively), with age-, gender- and height-specific
references of DBP being the gold standard. ROC
curves were plotted using measures of sensitivity
and specificity based on various anthropometric
cutoff values. The ROC curves showed the overall
discriminatory power of a diagnostic test over the
whole range of testing values. A better test shows its
curve skewed closer to the upper left corner. The
area under the curve (AUC) is a measure of the
diagnostic power of a test. A perfect test will have an
AUC of 1.0, and an AUC ¼ 0.5 means the test
performs no better than chance. A test with an
AUC of X0.85 was considered an accurate test.11
Sensitivity and specificity of the anthropometric
measurements have been calculated at all possible
cutoff points to find the optimal cutoff value. The
optimal sensitivity and specificity were the values
yielding maximum sums from the ROC curves. After
the cutoff points were determined, hypertension
was defined by SBPHR and DBPHR, and the
sensitivity and specificity of this method were
calculated.
Results
The prevalence of hypertension was 6.3% (5.5%
stage 1 and 0.8% stage 2). The prevalence of
hypertension was similar in boys and girls (6.7 vs
5.8%, w2 ¼ 5.021, P ¼ 0.081). In this study, 7.6% were
obese, and an additional 11.9% were overweight.
BMI was strongly associated with the risk for
hypertension. In multivariate logistic regression,
the prevalence of hypertension among overweight
(12.0%) and obese (25.9%) adolescents was 4.043
(2.763–5.915) times and 11.423 (7.843–16.637) times
that of normal-weight adolescents (3.6%), respectively, after adjustment for age and gender (Po0.01).
Table 1 presents the clinical characteristics of
the study subjects. Boys had significantly higher
average height, weight, BMI and SBP than did
Diagnostic criteria of hypertension in adolescents
Q Lu et al
3
Table 1 Clinical characteristics of the study subjects
Variable
Height (cm)
Weight (kg)
BMI (kg m–2)
SBP (mm Hg)
DBP (mm Hg)
SBPHR (mm Hg cm–1)
DBPHR (mm Hg cm–1)
Boys (n ¼ 1601)
Girls (n ¼ 1535)
t-test
P-value
167.2±9.0
59.1±14.5
20.9±4.0
109.8±12.8
69.0±8.7
0.66±0.07
0.41±0.05
159.3±5.7
52.2±9.8
20.5±3.4
105.4±12.0
68.7±8.0
0.66±0.08
0.43±0.05
29.400
15.653
3.257
9.873
1.257
1.903
10.090
0.000
0.000
0.001
0.000
0.209
0.057
0.000
Abbreviations: BMI, body mass index; DBP, diastolic blood pressure; DBPHR, diastolic blood pressure-to-height ratio; SBP, systolic blood
pressure; SBPHR, systolic blood pressure-to-height ratio.
Values are expressed as mean±s.d. Comparisons were performed between the two groups using Student’s t-test.
Table 2 Areas under ROC curve of SBPHR and DBPHR for diagnosing elevated SBP and DBP defined by age-, gender- and height-specific
references
Gender
Area under the curve
P-value
95% confidence interval
SBP X95th percentile
Boys
Girls
0.992
0.996
0.000
0.000
0.988–0.996
0.993–0.998
DBP X95th percentile
Boys
Girls
0.993
0.989
0.000
0.000
0.989–0.997
0.984–0.995
SBP 499th percentile+5 mm Hg
Boys
Girls
0.998
0.998
0.000
0.000
0.996–1.000
0.996–1.000
DBP 499th percentile+5 mm Hg
Boys
Girls
1.000
1.000
0.014
0.014
1.000–1.000
1.000–1.000
Abbreviations: DBP, diastolic blood pressure; DBPHR, diastolic blood pressure-to-height ratio; ROC, receiver-operating characteristic; SBP,
systolic blood pressure; SBPHR, systolic blood pressure-to-height ratio.
girls (Po0.01). The levels of DBP and SBPHR were
similar between boys and girls (P40.05). The level
of DBPHR was lower in boys than in girls (Po0.01).
The correlation between SBPHR and age (r ¼ 0.046, P ¼ 0.066 for boys and r ¼ 0.036, P ¼ 0.155
for girls) was much weaker than the correlation
between SBP and age for boys, but not for girls
(r ¼ 0.213, P ¼ 0.000 for boys and r ¼ 0.055, P ¼ 0.031
for girls). The correlation between DBPHR and age
(r ¼ 0.018, P ¼ 0.481 for boys and r ¼ 0.022,
P ¼ 0.387 for girls) was much weaker than the
correlation between DBP and age (r ¼ 0.224, P ¼ 0.000
for boys and r ¼ 0.067, P ¼ 0.009 for girls). SBPHR was
positively correlated with SBP in boys (r ¼ 0.890,
P ¼ 0.000) and girls (r ¼ 0.947, P ¼ 0.000). DBPHR
was positively correlated with DBP in boys
(r ¼ 0.904, P ¼ 0.000) and girls (r ¼ 0.950, P ¼ 0.000).
The abilities of SBPHR and DBPHR to accurately
define elevated SBP and DBP were assessed by
AUC. Table 2 shows that for both genders, the
accuracy levels of SBPHR and DBPHR for identifying elevated SBP and DBP (as assessed by AUC)
were both 40.85 (0.989–1.000).
The optimal thresholds of SBPHR and DBPHR for
identifying elevated SBP and DBP for boys and girls
are shown in Table 3. The optimal thresholds of
SBPHR for defining SBP X95th percentile were 0.75
in boys and 0.78 in girls. The optimal thresholds of
SBPHR for defining SBP 499th percentile þ 5 mm Hg were 0.81 in boys and 0.84 in girls.
The optimal thresholds of DBPHR for defining DBP
X95th percentile were 0.48 in boys and 0.51 in girls.
The optimal thresholds of DBPHR for defining DBP
499th percentile þ 5 mm Hg were 0.57 in boys and
0.63 in girls. The optimal thresholds were selected
by ensuring both the best sensitivity and specificity
of the diagnostic test.
Hypertension (stages 1 and 2) was defined by
SBPHR of X0.75 and/or DBPHR of X0.48 in boys,
and SBPHR of X0.78 and/or DBPHR of X0.51 in
girls. Hypertension (stage 2) was defined by SBPHR
of X0.81 and/or DBPHR of X0.57 in boys and by
SBPHR of X0.84 and/or DBPHR of X0.63 in girls.
The sensitivity and specificity of this method was
490% (Table 4).
Discussion
Prevention of hypertension in children and adolescents during the period of active growth and
development is feasible, effective and safe, and can
Journal of Human Hypertension
Diagnostic criteria of hypertension in adolescents
Q Lu et al
4
Table 3 Optimal thresholds of SBPHR and DBPHR for defining elevated SBP and DBP, and corresponding sensitivity and specificity in
Han adolescents aged 13–17 years
SBP X95th percentile
DBP X95th percentile
Boys
Girls
Boys
Girls
Thresholds
Sensitivity
Specificity
0.75
0.986
0.937
0.78
0.984
0.979
0.48
1.000
0.943
0.51
0.976
0.954
Thresholds
Sensitivity
Specificity
SBP 499th percentile+5 mm Hg
0.81
0.84
1.000
1.000
0.986
0.991
DBP 499th percentile+5 mm Hg
0.57
0.63
1.000
1.000
1.000
1.000
Abbreviations: DBP, diastolic blood pressure; DBPHR, diastolic blood pressure-to-height ratio; SBP, systolic blood pressure; SBPHR, systolic
blood pressure-to-height ratio.
Table 4 Optimal thresholds of SBPHR/DBPHR for defining prehypertension and hypertension and the corresponding sensitivity
and specificity in Han adolescents aged 13–17 years
Hypertension
(stages 1 and 2)
Thresholds
Sensitivity
Specificity
Hypertension
(stage 2)
Boys
Girls
Boys
Girls
0.75/0.48
0.991
0.910
0.78/0.51
0.989
0.949
0.81/0.57
1.000
0.986
0.84/0.63
1.000
0.991
Abbreviations: DBPHR, diastolic blood pressure-to-height ratio;
SBPHR, systolic blood pressure-to-height ratio.
decrease the levels of these factors in adults in the
future.12 To obtain effective prevention efficacy,
appropriate early-stage diagnosis of hypertension in
adolescents is necessary. In a cohort study of 14 187
children and adolescents aged from 3 to 18 years, 507
(3.6%) had hypertension, and 131 of these 507 (26%)
had a diagnosis of hypertension or elevated blood
pressure documented in the electronic medical
record.6 One important reason for the high rate of
undiagnosed hypertension in this paediatric population was the complicated diagnostic criteria for
hypertension. Age and gender are closely associated
with blood pressure in adolescents. In addition, it is
necessary to include the adolescent height percentile
to determine whether blood pressure is normal,
because body size is an essential determinant of
blood pressure in adolescents.13 Blood pressure for
age, gender and height has been widely accepted as
the most appropriate and useful measure for defining
hypertension in adolescents. However, blood pressure cutoff points for age, gender and height may be
appropriate for professionals, but it is too complicated for self-assessment and monitoring by adolescents and their parents.
An ideal measurement method for adolescent
hypertension should meet the following criteria:
the method should be simple, inexpensive, easy to
use and acceptable to the subjects. SBPHR and
DBPHR have several advantages over SBP and DBP
Journal of Human Hypertension
in practice for identifying hypertension in Han
adolescents. First, SBPHR and DBPHR are significantly related to SBP and DBP, respectively. Second,
height is simultaneously taken into account. Referring to age-, gender- and height-specific references
should prevent one from mislabelling tall adolescents who are not overweight as hypertensive or
from missing a diagnosis of high-normal blood
pressure or hypertension in short, heavy adolescents, as was the case when only a single value for
blood pressure was used for each age.14 SBPHR and
DBPHR can also prevent misdiagnosis, as the
SBPHR and DBPHR of adolescents with the same
gender, age and blood pressure were lower in tall
adolescents than in those of short adolescents.
Third, SBPHR and DBPHR were not correlated with
age, which makes it possible to propose non-agedependent cutoff points (as we did in our study) that
are easy and feasible to manipulate for both
professionals and lay people.
In this study, we found that BPHR had a reliable
accuracy for the diagnosis of hypertension in
adolescents. The areas under the ROC curve of
0.989–1.000 were consistent with robust diagnostic
performance, and indicated that SBPHR and DBPHR
have powerful discriminative abilities to identify
those with and without elevated SBP and DBP in
boys and girls. Further research confirmed that
SBPHR and DBPHR were accurate surrogate indexes
of SBP and DBP. When hypertension was defined by
SBPHR and DBPHR, we found that both the
sensitivities and specificities were 490%.
Consistent with previous research,15 stage 1
hypertension was much more common in our
population with an approximate 7:1 ratio compared
with stage 2 hypertension. However, adolescents
with stage 2 hypertension may need more prompt
evaluation and pharmacologic therapy. BPHR was
really useful, especially for stage 2 hypertension. In
identifying stage 2 hypertension, when the sensitivities were 100%, the specificities were 98.6% for
boys and 99.1% for girls.
The main limitation of our study was that it only
included adolescents of the Han ethnicity, limiting
Diagnostic criteria of hypertension in adolescents
Q Lu et al
5
the ability to apply the study results to the general
Chinese population. However, many studies have
shown that height is independently related to blood
pressure in different ethnic groups.16–20 The effect of
height on blood pressures did not differ among
different ethnic groups.19 Height was independently
related to blood pressure for all ages.20 Thus, we
speculate that this method could be applied to other
ethnic groups. Another limitation was that we did
not have the data regarding children. In a further
study, we will analysis the feasibility of using this
method in children.
After the simple calculations, only eight threshold
values need to be remembered. The number of
threshold values was significantly reduced through
the calculation. It will reduce the trouble of
clinicians and non-professionals. When identifying
adolescents with hypertension, they do not have to
lookup the complicated table. In addition, this
method is easy to understand and master. Therefore,
we believe that this method will be accepted by the
majority of clinicians and non-professionals.
In summary, we conclude that BPHR is a simple,
easy, inexpensive and accurate index for identifying
hypertension in Han adolescents, especially for
stage 2 hypertension. This method might be applicable to other ethnic groups, but should be validated
in those groups.
5
6
7
8
9
10
11
What is known about this topic
K The prevalence of hypertension in adolescents seems to be
increasing. Diagnosis of hypertension in adolescents is
complicated because blood pressure values vary with age,
gender and height.
12
What this study adds
K Blood pressure-to-height ratio is a simple, easy, inexpensive
and accurate index for identifying hypertension in Han
adolescents, especially for stage 2 hypertension.
13
14
15
Conflict of interest
16
The authors declare no conflict of interest.
References
1 Sorof JM, Lai D, Turner J, Poffenbarger T, Portman RJ.
Overweight, ethnicity, and the prevalence of hypertension in school-aged children. Pediatrics 2004; 113:
475–482.
2 Lauer RM, Clarke WR. Childhood risk factors for high
adult blood pressure: the Muscatine Study. Pediatrics
1989; 84: 633–641.
3 Kiessling SG, McClanahan KK, Omar HA. Obesity,
hypertension, and mental health evaluation in adolescents: a comprehensive approach. Int J Adolesc Med
Health 2008; 20: 5–15.
4 Berenson GS, Srinivasan SR, Bao W, Newman III WP,
Tracy RE, Wattigney WA. Association between multi-
17
18
19
20
ple cardiovascular risk factors and atherosclerosis in
children and young adults. The Bogalusa Heart Study.
N Engl J Med 1998; 338: 1650–1656.
Berenson GS. Childhood risk factors predict adult risk
associated with subclinical cardiovascular disease.
The Bogalusa Heart Study. Am J Cardiol 2002; 90:
3L–7L.
Hansen ML, Gunn PW, Kaelber DC. Underdiagnosis of
hypertension in children and adolescents. JAMA 2007;
298: 874–879.
Weili Y, He B, Yao H, Dai J, Cui J, Ge D et al. Waist-toheight ratio is an accurate and easier index for
evaluating obesity in children and adolescents. Obesity
(Silver Spring) 2007; 15: 748–752.
McCarthy HD, Ashwell M. A study of central fatness
using waist-to-height ratios in UK children and
adolescents over two decades supports the simple
message–‘keep your waist circumference to less than
half your height’. Int J Obes (Lond) 2006; 30: 988–992.
Group of China Obesity Task Force. Body mass index
reference norm for screening overweight and obesity in
Chinese children and adolescents. Chin J Epidemiol
2004; 25: 97–102.
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:
555–576.
Zou KH, O’Malley AJ, Mauri L. Receiver-operating
characteristic analysis for evaluating diagnostic
tests and predictive models. Circulation 2007; 115:
654–657.
Rozanov VB, Aleksandov AA, Shugaeva EN, Perova
NV, Maslennikova G, Smirnova SG et al. Primary
prevention of cardiovascular diseases: long term
results of five year long preventive intervention in
12-year old boys (ten year prospective study). Kardiologiia 2007; 47: 60–68.
Luma GB, Spiotta RT. Hypertension in children and
adolescents. Am Fam Physician 2006; 73: 1558–1568.
Sinaiko AR. Hypertension in children. N Engl J Med
1996; 335: 1968–1973.
McNiece KL, Poffenbarger TS, Turner JL, Franco KD,
Sorof JM, Portman RJ. Prevalence of hypertension and
pre-hypertension among adolescents. J Pediatr 2007;
150: 640–644, e1.
Ejike CE, Ugwu CE, Ezeanyika LU, Olayemi AT. Blood
pressure patterns in relation to geographic area of
residence: a cross-sectional study of adolescents in
Kogi state, Nigeria. BMC Public Health 2008; 8: 411.
Gutgesell M, Terrell G, Labarthe D. Pediatric blood
pressure: ethnic comparisons in a primary care center.
Hypertension 1981; 3: 39–47.
Harrabi I, Belarbia A, Gaha R, Essoussi AS, Ghannem
H. Epidemiology of hypertension among a population
of school children in Sousse, Tunisia. Can J Cardiol
2006; 22: 212–216.
Daniels SR, McMahon RP, Obarzanek E, Waclawiw
MA, Similo SL, Biro FM et al. Longitudinal correlates
of change in blood pressure in adolescent girls.
Hypertension 1998; 31: 97–103.
Rosner B, Prineas RJ, Loggie JM, Daniels SR. Blood
pressure nomograms for children and adolescents, by
height, sex, and age, in the United States. J Pediatr
1993; 123: 871–886.
Journal of Human Hypertension