Abstract
Blood pressure to height ratio (BPHR) has been suggested as a simple method for screening children with hypertension, but its discriminatory ability in young children is not as good as that in older children. Using data of 89,664 Chinese children aged 7 to 11 years, the authors assessed whether modified BPHR (BP:eHT13) was better than BPHR in identifying young children with hypertension. BP:eHT13 was estimated as BP/(height+7×(13−age in years)). Using Youden's index, the thresholds of systolic/diastolic BP:eHT13 for identifying prehypertension and hypertension were 0.67/0.44 and 0.69/0.45, respectively. These proposed thresholds revealed high sensitivity, specificity, negative predictive value, and area under the curve (AUC), ranging from 0.874 to 0.999. In addition, BP:eHT13 showed better AUCs and fewer cutoff points than, if not similar to, two existing BPHR references. BP:eHT13 generally performed better than BPHR in discriminating BP abnormalities in young children and may improve early hypertension recognition and control.
Hypertension remains one of the most important preventable contributors to disease and death.1 Elevated blood pressure (BP) in children is associated with detectable target organ damage and predicts hypertension in adulthood.2, 3 Therefore, prevention and treatment of childhood elevated BP can result in a life‐long reduction of BP and its associated conditions.4 However, the current approach to defining elevated BP, which incorporates age, sex, and height, as well as multiple cutoff points for prehypertension and hypertension diagnosis, is too cumbersome to be applied in pediatric clinics.5
In 2011, Lu and colleagues reported that BP to height ratio (BPHR) can identify hypertension easily and accurately in Chinese adolescents.6 Their observations have also been confirmed by studies conducted in other populations.7, 8, 9, 10 Using the nationally representative samples, there were two BPHR references for diagnosing prehypertension and hypertension in Chinese children to date.11, 12 Xi and colleagues11 established the optimal thresholds in children enrolled in the China Health and Nutrition Survey (CHNS) by age 6 to 11 years and 12 to 17 years, separately. However, the sensitivity and area under the curve (AUC) were lower in children aged 6 to 11 years than those in older children. Some researchers hypothesized that limitation was associated with the different growth rates of height and BP in young children.13 Using data from the Chinese National Survey on Students' Constitution and Health (CNSSCH), we observed that BPHR markedly declined with increasing age from 7 to 11 years and remained stable between 12 and 17 years.12 Therefore, we previously proposed the optimal BPHR thresholds by three age groups, 7 to 8 years, 9 to 11 years, and 12 to 17 years.12 However, by doing this, we increased the number of cutoff values.
Mourato and colleagues13 recently proposed a modified formula for calculating BPHR (BP:eHT13) and found that BP:eHT13 was better than BPHR for screening elevated BP in Brazilian children (5–12 years). Therefore, it would be interesting to explore whether BP:eHT13 could improve the ability to identify elevated BP in Chinese children. Using the same database as that for developing BPHR references,12 we established the optimal thresholds of BP:eHT13 for diagnosing prehypertension and hypertension, and further compared the performance of BP:eHT13 with two existing BPHR references in Chinese children aged 7 to 11 years.11, 12
Methods
Participants
The sampling procedures of CNSSCH 2010 have been published previously in detail.14 In brief, this survey used a stratified multistage sampling method to select Han ethnic students from primary and secondary schools from 30 of the 31 mainland provinces. Tibetan children were not included in this study because they had different BP levels than their Han counterparts.15 Participants in each province were classified into two area groups (urban and rural) according to their residential regions, which were further classified into sex‐ and age‐specific subgroups. In each subgroup, equal numbers of participants were selected by stratified cluster sampling from some classes, as clusters were randomly selected from each grade in the selected school. Of 89,758 children and adolescents aged 7 to 11 years, we excluded 94 participants (0.10%) with missing data or extreme height, sitting height, or BP (>6 sex‐ and age‐specific standard deviations [SDs] on either side of the mean). Thus, a total of 89,664 participants were included in the final analysis of this study.
This survey was conducted according to the guidelines from the Declaration of Helsinki and approved by six ministries of China, including the Ministry of Education, General Administration of Sport, Ministry of Health, State Ethnic Affairs Commission, Ministry of Science and Technology, and Ministry of Finance. Informed consent was obtained from both parents and students. The project of analyzing the survey data was approved by the medical research ethics committee of the University of Queensland (#2011001199).
Measurements
Anthropometric measurements were performed according to the same protocol at all survey sites.14 Students were asked to wear light clothes only and to stand straight without shoes. Height was measured using a wall‐mounted stadiometer to the nearest 0.1 cm, and weight was measured with a scale to the nearest 0.1 kg. Both height and weight were measured twice, and the mean value was recorded. Body mass index (BMI) was calculated with the values of weight divided by height squared (kg/m2) and categorized as thin, normal, overweight, or obese according to the sex‐ and age‐specific references for children suggested by Cole and colleagues.16, 17 Height was categorized into ˂10th (z‐score ˂−1.28), ≥10th and <90th (z‐score ≥−1.28 to <1.28), and ≥90th (z‐score ≥1.28) percentile groups based on the sex‐ and age‐specific SD scores (z‐score).
BP was measured according to the recommendation of the National High Blood Pressure Education Program (NHBPEP) Working Group in Children and Adolescents5 using an auscultation mercury sphygmomanometer. The cuff bladder width covered 50% to 75% of the circumference of the arm. BP measurements were taken at least 5 minutes after resting. Systolic BP (SBP) was defined as the first Korotkoff sound and diastolic BP (DBP) as the fifth Korotkoff sound. An average of three BP measurements at a single visit was calculated for each child. In addition, prehypertension was defined as either SBP/DBP greater than or equal to the referent sex‐, age‐, and height‐specific 90th percentile or SBP/DBP ≥120/80 mm Hg, and hypertension was defined as SBP/DBP ≥95th sex‐, age‐, and height‐specific percentile according to the references of the NHBPEP working group.5 Sex‐ and age‐specific percentiles of height were derived from the US Centers for Disease Control and Prevention growth chart as suggested by the NHBPEP working group.18
Systolic BPHR (SBPHR) was calculated as SBP/height (mm Hg/cm), and diastolic BPHR (DBPHR) as DBP/height (mm Hg/cm). According to previous reports,11, 12 two existing BPHR references based on national samples of Chinese children were used in this study. One of these references (BPHR‐D) was developed by our previous study based on the same CNSSCH data.12 That study proposed the SBPHR/DBPHR thresholds for defining prehypertension and hypertension were 0.84/0.55 and 0.87/0.57 for children aged 7 to 8 years, and 0.78/0.50 and 0.81/0.53 for children aged 9 to 11 years, respectively. The other reference for BPHR (BPHR‐X) was developed by Xi and colleagues11 and suggested 0.81/0.52 and 0.84/0.55 as the SBPHR/DBPHR thresholds for diagnosing prehypertension and hypertension, respectively, in children aged 7 to 11 years. BP:eHT13 was calculated as BP (mm Hg)/(height (cm)+7×(13−age in years)).13
Rigid quality control measures were enforced in this survey. All measurement instruments, including stadiometer, scale, and sphygmomanometer, were calibrated and compared with standard instruments before use. All technicians were required to pass the standard measurement test after a rigorous 1‐week training course, and the measurements were conducted by the same team of technicians in each region.
Statistical Analyses
Descriptive results are expressed as mean and SD by sex. The sex‐ and age‐specific optimal thresholds of BP:eHT13 for discriminating prehypertension and hypertension were determined by Youden's index (sensitivity+specificity−1), and the values corresponding to the maximum of Youden's index were selected using the references of the NHBPEP working group as the gold standard.5 Because those optimal thresholds changed little among different age and sex groups, we evaluated the optimal thresholds by combining sexes and ages together. To assess the performance of the determined optimal thresholds, the sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were estimated. In addition, AUC, estimated by receiver operating characteristic (ROC) curve analysis, was also used to evaluate the discrimination power of these thresholds, ranging from 0.5 (no discrimination) to 1 (perfect discrimination).19 Furthermore, the performance of BP:eHT13 was also compared with those of two existing BPHR references,11, 12 and χ 2 test was used for the comparison of AUCs between BP:eHT13 and BPHR.20 These analyses were repeated in the various BMI groups, as well as different height percentile categories. A P value <.05 was considered statistically significant. All analyses were performed with Stata 13 software (StataCorp, College Station, TX).
Results
A total of 44,830 boys and 44,834 girls aged 7 to 11 years were included in the current study. As shown in Table 1, the differences in SBP:eHT13, as well as DBP:eHT13, between the sexes were small. In addition, 6.1% and 7.8% of the participants were identified as having prehypertension and hypertension, and the prevalence of thinness, overweight, and obesity were 17.6%, 11.8%, and 3.3%, respectively.
Table 1.
Characteristics of Participants by Sex and Age (N=89,664)
| Age, y | No. | SBP, mm Hg | DBP, mm Hg | Height, cm | SBP:eHT13, mm Hg/cm | DBP:eHT13, mm Hg/cm |
|---|---|---|---|---|---|---|
| Boys | ||||||
| 7 | 8960 | 95.6 (11.1) | 59.3 (10.0) | 125.5 (6.0) | 0.57 (0.07) | 0.35 (0.06) |
| 8 | 8964 | 97.9 (11.0) | 61.0 (9.6) | 130.7 (6.2) | 0.59 (0.07) | 0.37 (0.06) |
| 9 | 8954 | 99.2 (11.2) | 62.0 (9.8) | 135.8 (6.6) | 0.61 (0.07) | 0.38 (0.06) |
| 10 | 8978 | 101.5 (11.1) | 63.6 (9.5) | 140.9 (6.9) | 0.63 (0.07) | 0.39 (0.06) |
| 11 | 8974 | 103.2 (11.2) | 64.4 (9.4) | 146.2 (7.9) | 0.64 (0.07) | 0.40 (0.06) |
| Girls | ||||||
| 7 | 8958 | 94.2 (11.0) | 58.6 (9.6) | 124.1 (5.9) | 0.57 (0.07) | 0.35 (0.06) |
| 8 | 8948 | 96.4 (10.9) | 60.3 (9.4) | 129.4 (6.2) | 0.59 (0.07) | 0.37 (0.06) |
| 9 | 8975 | 98.0 (10.9) | 61.8 (9.6) | 135.0 (6.8) | 0.60 (0.07) | 0.38 (0.06) |
| 10 | 8975 | 100.0 (10.9) | 63.2 (9.3) | 141.3 (7.4) | 0.62 (0.07) | 0.39 (0.06) |
| 11 | 8978 | 102.6 (11.0) | 64.7 (9.3) | 147.2 (7.7) | 0.64 (0.07) | 0.40 (0.06) |
Abbreviations: DBP, diastolic blood pressure; DBP:eHT13, modified diastolic blood pressure to height ratio; SBP, systolic blood pressure; SBP:eHT13, modified systolic blood pressure to height ratio. Values are expressed as means (standard deviations).
The sex‐ and age‐specific optimal thresholds of SBP:eHT13 for prehypertension (Figure a) ranged from 0.65 to 0.71, and thresholds for hypertension (Figure b) ranged from 0.66 to 0.73. Corresponding optimal thresholds of DBP:eHT13 for identifying prehypertension (Figure a) ranged from 0.42 to 0.46, and thresholds for hypertension (Figure b) ranged from 0.44 to 0.49. These cutoff values in boys were similar to those in girls. All of these cutoff values showed high discrimination, with AUCs above 0.95.
Figure 1.
Optimal thresholds of modified blood pressure to height ratio (BP: eHT13) to identify elevated blood pressure in Chinese children by sex and age. (a) Optimal thresholds for identifying prehypertension and (b) optimal thresholds for identifying hypertension. DBP:eHT13 indicates modified diastolic blood pressure to height ratio; SBP:eHT13, modified systolic blood pressure to height ratio.
As the optimal thresholds varied slightly among different sexes and ages, optimal cutoff values for age 7 to 11 were evaluated without being stratified by sex and age. Accordingly, the optimal cutoff values of SBP:eHT13 for SBP ≥90th percentile (or ≥120 mm Hg) and SBP ≥95th percentile were 0.67 and 0.69, respectively. The corresponding optimal cutoff values for DBP ≥90th percentile (or ≥80 mm Hg) and DBP ≥95th percentile were 0.44 and 0.45, respectively. Table 2 shows the AUCs of SBP:eHT13 and DBP:eHT13, as well as their comparison with those of two BPHR references. Generally, the differences of AUCs between BP:eHT13 and BPHR‐D were small. Meanwhile, BP:eHT13 revealed significantly larger AUCs than BPHR‐X.
Table 2.
Comparison of the AUCs Among Three Methods by Sex
| BP Percentile | Patients, No. | BP:eHT13 | BPHR‐D | BPHR‐X | |||
|---|---|---|---|---|---|---|---|
| Cutoff Value | AUC (95% CI) | AUC (95% CI) | P Valuea | AUC (95% CI) | P Valuea | ||
| Boys (n=44,830) | |||||||
| SBP ≥90th | 4097 | 0.67 | 0.932 (0.930–0.935) | 0.936 (0.933–0.938) | .031 | 0.904 (0.900–0.909) | ˂.001 |
| DBP ≥90th | 4091 | 0.44 | 0.953 (0.950–0.956) | 0.939 (0.937–0.942) | ˂.001 | 0.921 (0.917–0.925) | ˂.001 |
| SBP ≥95th | 2216 | 0.69 | 0.942 (0.938–0.946) | 0.952 (0.949–0.956) | ˂.001 | 0.934 (0.928–0.939) | .017 |
| DBP ≥95th | 1972 | 0.45 | 0.957 (0.955–0.960) | 0.964 (0.961–0.966) | ˂.001 | 0.961 (0.957–0.964) | .091 |
| Girls (n=44,834) | |||||||
| SBP ≥90th | 3845 | 0.67 | 0.931 (0.927–0.935) | 0.926 (0.922–0.930) | .055 | 0.872 (0.866–0.877) | ˂.001 |
| DBP ≥90th | 4165 | 0.44 | 0.952 (0.949–0.955) | 0.937 (0.934–0.940) | ˂.001 | 0.899 (0.894–0.904) | ˂.001 |
| SBP ≥95th | 2058 | 0.69 | 0.943 (0.938–0.947) | 0.947 (0.942–0.952) | .100 | 0.918 (0.911–0.925) | ˂.001 |
| DBP ≥95th | 2530 | 0.45 | 0.964 (0.961–0.966) | 0.959 (0.956–0.963) | .018 | 0.907 (0.900–0.914) | ˂.001 |
Abbreviations: AUC, area under the curve; BP, blood pressure; BP:eHT13, modified blood pressure to height ratio; BPHR‐D, thresholds of blood pressure to height ratio developed by Dong and colleagues12; BPHR‐X, thresholds of blood pressure to height ratio developed by Xi and colleagues11; CI, confidence interval; DBP, diastolic blood pressure; SBP, systolic blood pressure. a P value for AUC compared with BP:eHT13.
As presented in Table 3, the optimal thresholds of BP:eHT13 for identifying prehypertension and hypertension revealed a high sensitivity, specificity, and NPV, ranging from 87.4% to 99.9% depending on sex. These performances were better, if not similar, than those of BPHR‐D and BPHR‐X. Furthermore, BP:eHT13 mostly demonstrated a significantly larger AUC than both BPHRs for discriminating prehypertension, and similar or greater than BPHRs for discriminating hypertension. These performances of three methods were also presented by BMI category (Table 4). BP:eHT13 showed significantly larger AUCs than both BPHRs for identifying prehypertension, and generally demonstrated a similar, if not greater, AUC for identifying hypertension, across various BMI levels. These patterns were also true among different height percentile groups, particularly in comparison with BPHR‐X (Table 5).
Table 3.
Performance of BH:eHT13 and BPHR for Identifying Prehypertension and Hypertension in Chinese Children Aged 7 to 11 Years by Sex
| Test Results | Prehypertension and Higher | Hypertension | ||
|---|---|---|---|---|
| Boys | Girls | Boys | Girls | |
| Patients, No. | 6279 | 6119 | 3320 | 3641 |
| BP:eHT13 | ||||
| Cutoff value | 0.67/0.44 | 0.69/0.45 | ||
| Sensitivity, % | 97.4 (97.0–97.8) | 95.8 (95.2–96.3) | 98.6 (98.2–99.0) | 97.4 (96.8–97.9) |
| Specificity, % | 87.4 (87.0–87.7) | 89.3 (89.0–89.6) | 88.1 (87.8–88.4) | 90.7 (90.4–91.0) |
| PPV, % | 55.6 (54.7–56.6) | 58.5 (57.6–59.5) | 39.9 (38.8–41.0) | 48.1 (47.0–49.3) |
| NPV, % | 99.5 (99.4–99.6) | 99.3 (99.2–99.3) | 99.9 (99.8–99.9) | 99.7 (99.7–99.8) |
| AUC | 0.924 (0.921–0.926) | 0.925 (0.922–0.928) | 0.934 (0.931–0.936) | 0.940 (0.937–0.943) |
| BPHR‐D | ||||
| Cutoff valuea | 0.84/0.55 (7–8 y); 0.78/0.50 (9–11 y) | 0.87/0.57 (7–8 y); 0.81/0.53 (9–11 y) | ||
| Sensitivity, % | 98.4 (98.1–98.7) | 96.7 (96.3–97.2) | 98.3 (97.7–98.7) | 96.3 (95.7–96.9) |
| Specificity, % | 84.4 (84.1–84.8) | 85.7 (85.3–86.0) | 89.9 (89.6–90.2) | 91.8 (91.5–92.0) |
| PPV, % | 50.7 (49.8–51.6) | 51.7 (50.7–52.6) | 43.8 (42.7–45.0) | 50.9 (49.7–52.1) |
| NPV, % | 99.7 (99.6–99.8) | 99.4 (99.3–99.5) | 99.8 (99.8–99.9) | 99.6 (99.6–99.7) |
| AUC | 0.914 (0.912–0.917)c | 0.912 (0.909–0.915)c | 0.941 (0.938–0.944)c | 0.941 (0.937–0.944) |
| BPHR‐X | ||||
| Cutoff valueb | 0.81/0.52 | 0.84/0.55 | ||
| Sensitivity, % | 93.3 (92.7–93.9) | 88.5 (87.7–89.3) | 95.2 (94.4–95.9) | 87.5 (86.4–88.5) |
| Specificity, % | 85.4 (85.1–85.8) | 86.0 (85.6–86.3) | 91.2 (90.9–91.5) | 92.2 (91.9–92.4) |
| PPV, % | 51.0 (50.1–52.0) | 49.9 (49.0–50.9) | 46.4 (45.2–47.6) | 49.7 (48.5–50.9) |
| NPV, % | 98.7 (98.6–98.9) | 97.9 (97.8–98.1) | 99.6 (99.5–99.6) | 98.8 (98.7–98.9) |
| AUC | 0.894 (0.890–0.897)c | 0.872 (0.868–0.877)c | 0.932 (0.928–0.936) | 0.898 (0.893–0.904)c |
Abbreviations: AUC, area under the curve; BP, blood pressure; BP:eHT13, modified blood pressure to height ratio; BPHR‐D, thresholds of blood pressure to height ratio developed by Dong; BPHR‐X, thresholds of blood pressure to height ratio developed by Xi; PPV, positive predictive value; NPV, negative predicted value. Values are expressed as percentages or means (95% confidence interval). aDeveloped by Dong and colleagues.12 bDeveloped by Xi and colleagues.11 c P<.01 for AUC compared with BP:eHT13.
Table 4.
Performance of BH:eHT13 and BPHR for Identifying Prehypertension and Hypertension in Chinese Children Aged 7 to 11 Years by BMI Group
| Test Results | BMI Group | |||
|---|---|---|---|---|
| Thinness (n=15,786) | Normal (n=60,347) | Overweight (n=10,547) | Obesity (n=2984) | |
| Prehypertension and higher | ||||
| Patients, No. | 1494 | 7492 | 2394 | 1018 |
| BP:eHT13 | ||||
| Sensitivity, % | 96.9 (95.9–97.7) | 96.7 (96.2–97.1) | 97.0 (96.2–97.6) | 94.7 (93.1–96.0) |
| Specificity, % | 88.0 (87.4–88.5) | 88.5 (88.2–88.7) | 87.5 (86.8–88.2) | 89.8 (88.4–91.1) |
| PPV, % | 45.7 (44.0–47.5) | 54.3 (53.5–55.2) | 69.6 (68.0–71.1) | 82.7 (80.5–84.9) |
| NPV, % | 99.6 (99.5–99.7) | 99.5 (99.4–99.5) | 99.0 (98.7–99.2) | 97.0 (96.1–97.8) |
| AUC | 0.925 (0.919–0.930) | 0.926 (0.923–0.928) | 0.922 (0.917–0.927) | 0.922 (0.913–0.932) |
| BPHR‐D | ||||
| Sensitivity, % | 99.1 (98.4–99.5) | 98.0 (97.6–98.3) | 96.5 (95.7–97.2) | 95.0 (93.5–96.2) |
| Specificity, % | 84.1 (83.5–84.7) | 85.0 (84.7–85.3) | 86.7 (85.9–87.4) | 86.4 (84.8–87.9) |
| PPV, % | 39.5 (37.9–41.1) | 48.1 (47.3–48.9) | 68.0 (66.4–69.6) | 78.4 (76.0–80.6) |
| NPV, % | 99.9 (99.8–99.9) | 99.7 (99.6–99.7) | 98.8 (98.6–99.1) | 97.1 (96.2–97.8) |
| AUC | 0.916 (0.912–0.920)a | 0.915 (0.913–0.917)a | 0.916 (0.911–0.921)b | 0.907 (0.897–0.917)a |
| BPHR‐X | ||||
| Sensitivity, % | 96.6 (95.5–97.4) | 91.5 (90.8–92.1) | 86.9 (85.5–88.3) | 87.9 (85.8–89.9) |
| Specificity, % | 84.9 (84.3–85.5) | 85.6 (85.3–85.9) | 88.3 (87.6–89.0) | 83.7 (82.0–85.3) |
| PPV, % | 40.0 (38.4–41.7) | 47.4 (46.6–48.2) | 68.6 (66.9–70.2) | 73.6 (71.0–76.1) |
| NPV, % | 99.6 (99.4–99.7) | 98.6 (98.5–98.7) | 95.8 (95.4–96.3) | 93.0 (91.8–94.2) |
| AUC | 0.907 (0.902–0.913)a | 0.885 (0.882–0.889)a | 0.876 (0.869–0.884)a | 0.858 (0.845–0.871)a |
| Hypertension | ||||
| Patients, No. | 889 | 4145 | 1307 | 620 |
| BP:eHT13 | ||||
| Sensitivity, % | 99.0 (98.1–99.5) | 98.4 (97.9–98.7) | 96.9 (95.8–97.7) | 96.1 (94.3–97.5) |
| Specificity, % | 90.6 (90.1–91.1) | 89.9 (89.7–90.2) | 85.6 (84.9–86.3) | 84.3 (82.7–85.7) |
| PPV, % | 38.6 (36.6–40.6) | 41.9 (40.9–42.9) | 48.7 (46.8–50.7) | 61.6 (58.4–64.6) |
| NPV, % | 99.9 (99.9–100.0) | 99.9 (99.8–99.9) | 99.5 (99.3–99.6) | 98.8 (98.2–99.2) |
| AUC | 0.948 (0.944–0.952) | 0.942 (0.939–0.944) | 0.912 (0.906–0.918) | 0.902 (0.891–0.913) |
| BPHR‐D | ||||
| Sensitivity, % | 98.9 (97.9–99.5) | 97.8 (97.3–98.3) | 96.1 (94.9–97.1) | 93.5 (91.3–95.4) |
| Specificity, % | 90.7 (90.3–91.2) | 91.1 (90.8–91.3) | 90.4 (89.8–91.0) | 88.5 (87.1–89.8) |
| PPV, % | 38.9 (36.9–41.0) | 44.6 (43.6–45.7) | 58.6 (56.5–60.7) | 68.1 (64.8–71.2) |
| NPV, % | 99.9 (99.9–100.0) | 99.8 (99.8–99.9) | 99.4 (99.2–99.5) | 98.1 (97.5–98.7) |
| AUC | 0.948 (0.944–0.952) | 0.944 (0.942–0.947) | 0.933 (0.926–0.939)a | 0.910 (0.899–0.922) |
| BPHR‐X | ||||
| Sensitivity, % | 95.5 (93.9–96.8) | 91.9 (91.0–92.7) | 87.7 (85.8–89.4) | 87.3 (84.4–89.8) |
| Specificity, % | 91.5 (91.0–91.9) | 91.6 (91.4–91.9) | 93.0 (92.4–93.5) | 89.3 (88.0–90.5) |
| PPV, % | 40.0 (37.9–42.1) | 44.8 (43.7–45.8) | 63.8 (61.6–66.1) | 68.1 (64.8–71.4) |
| NPV, % | 99.7 (99.6–99.8) | 99.3 (99.3–99.4) | 98.2 (97.9–98.4) | 96.4 (95.5–97.1) |
| AUC | 0.935 (0.928–0.942)a | 0.918 (0.913–0.922)a | 0.903 (0.894–0.913) | 0.883 (0.868–0.897)b |
Abbreviations: AUC, area under the curve; BMI, body mass index; BP, blood pressure; BP:eHT13, modified blood pressure to height ratio; BPHR‐D, threshold of blood pressure to height ratio developed by Dong and colleagues12; BPHR‐X, threshold of blood pressure to height ratio developed by Xi and colleagues11; PPV, positive predictive value; NPV, negative predicted value. Values are expressed as percentages or means (95% confidence interval). a P<.01 and b P<.05 for AUC compared with BP:eHT13.
Table 5.
Performance of BH:eHT13 and BPHR for Identifying Prehypertension and Hypertension in Chinese Children Aged 7 to 11 Years by Height Percentile Group
| Test Results | Height Percentile Group | ||
|---|---|---|---|
| <P10th (n=9291) | ≥P10th, <P90th (n=70,946) | >P90th (n=9427) | |
| Prehypertension and higher | |||
| Patients, No. | 1102 | 9661 | 1635 |
| BP:eHT13 | |||
| Sensitivity, % | 100.0 (99.7–100.0) | 97.4 (97.0–97.7) | 89.7 (88.2–91.2) |
| Specificity, % | 76.4 (75.5–77.3) | 88.9 (88.7–89.2) | 95.9 (95.5–96.4) |
| PPV, % | 36.3 (34.6–38.1) | 58.1 (57.4–58.9) | 82.2 (80.4–84.0) |
| NPV, % | 100.0 (99.9–100.0) | 99.5 (99.5–99.6) | 97.8 (97.4–98.1) |
| AUC | 0.882 (0.878–0.887) | 0.932 (0.930–0.934) | 0.928 (0.921–0.936) |
| BPHR‐D | |||
| Sensitivity, % | 100.0 (99.7–100.0) | 98.9 (98.7–99.1) | 88.1 (86.5–89.7) |
| Specificity, % | 69.2 (68.2–70.2) | 85.5 (85.3–85.8) | 98.0 (97.6–98.3) |
| PPV, % | 30.4 (28.9–32.0) | 51.9 (51.2–52.6) | 90.1 (88.6–91.5) |
| NPV, % | 100.0 (99.9–100.0) | 99.8 (99.8–99.8) | 97.5 (97.2–97.9) |
| AUC | 0.846 (0.841–0.851)a | 0.922 (0.921–0.924)a | 0.931 (0.923–0.939) |
| BPHR‐X | |||
| Sensitivity, % | 100.0 (99.7–100.0) | 93.2 (92.7–93.7) | 71.1 (68.9–73.3) |
| Specificity, % | 71.0 (70.1–72.0) | 86.3 (86.1–86.6) | 96.2 (95.7–96.6) |
| PPV, % | 31.7 (30.2–33.3) | 51.8 (51.1–52.6) | 79.7 (77.5–81.7) |
| NPV, % | 100.0 (99.9–100.0) | 98.8 (98.7–98.9) | 94.1 (93.5–94.6) |
| AUC | 0.855 (0.850–0.860)a | 0.898 (0.895–0.901)a | 0.837 (0.825–0.848)a |
| Hypertension | |||
| Patients, No. | 663 | 5553 | 745 |
| BP:eHT13 | |||
| Sensitivity, % | 100.0 (99.4–100.0) | 99.1 (98.8–99.3) | 87.9 (85.4–90.2) |
| Specificity, % | 79.2 (78.3–80.1) | 90.3 (90.1–90.5) | 93.0 (92.4–93.5) |
| PPV, % | 27.0 (25.2–28.8) | 46.4 (45.5–47.3) | 51.9 (49.1–54.6) |
| NPV, % | 100.0 (99.9–100.0) | 99.9 (99.9–99.9) | 98.9 (98.6–99.1) |
| AUC | 0.896 (0.892–0.900) | 0.947 (0.945–0.948) | 0.905 (0.893–0.917) |
| BPHR‐D | |||
| Sensitivity, % | 100.0 (99.4–100.0) | 98.8 (98.5–99.1) | 83.4 (80.5–86.0) |
| Specificity, % | 79.1 (78.2–79.9) | 91.5 (91.3–91.7) | 97.6 (97.2–97.9) |
| PPV, % | 26.9 (25.1–28.7) | 49.7 (48.8–50.6) | 74.5 (71.4–77.5) |
| NPV, % | 100.0 (99.9–100.0) | 99.9 (99.9–99.9) | 98.6 (98.3–98.8) |
| AUC | 0.896 (0.891–0.900) | 0.952 (0.950–0.953)a | 0.905 (0.891–0.918) |
| BPHR‐X | |||
| Sensitivity, % | 100.0 (99.4–100.0) | 92.1 (91.3–92.8) | 76.2 (73.0–79.3) |
| Specificity, % | 82.3 (81.5–83.1) | 92.2 (92.0–92.4) | 97.1 (96.7–97.4) |
| PPV, % | 30.3 (28.4–32.3) | 50.1 (49.1–51.1) | 69.2 (65.9–72.3) |
| NPV, % | 100.0 (99.9–100.0) | 99.3 (99.2–99.3) | 97.9 (97.6–98.2) |
| AUC | 0.912 (0.908–0.916)a | 0.921 (0.918–0.925)a | 0.867 (0.851–0.882)a |
Abbreviations: AUC, area under the curve; BP, blood pressure; BP:eHT13, modified blood pressure to height ratio; BPHR‐D, threshold of blood pressure to height ratio developed by Dong and colleagues12; BPHR‐X, threshold of blood pressure to height ratio developed by Xi and colleagues11; CI, confidence interval; PPV, positive predictive value; NPV, negative predicted value. Values are expressed as percentages/means (95% confidence interval). a P<.01 for AUC compared with BP:eHT13.
Discussion
Our study estimated the optimal thresholds of BP:eHT13 for diagnosing prehypertension and hypertension in Chinese children aged 7 to 11 years. These proposed thresholds revealed high sensitivities, specificities, NPVs, and AUCs, and showed a better ability for discriminating prehypertension and similar, if not greater, performance for identifying hypertension than two existing BPHR references. Our findings suggest that BP:eHT13 can be a simple measure for identifying abnormal BP in children aged 7 to 11 years.
Although BPHR was suggested as a simple and accurate index for detecting elevated BP in childhood, some studies showed that its performance in young children was not as good as that in older children.10, 21 Mourato and colleagues13 assumed that the difference in performance may relate to the faster growth in height compared with BP in that period of life. Consistent with their assumption, a significant decrease in age‐specific mean BPHRs, as well as the optimal thresholds of BPHR for discriminating elevated BP, was observed in our previous study in boys and girls aged 7 to 11 years, followed by a relative constant tendency between age 12 and 17 years.12 In this study, due to the fact that BPHR‐D was developed for two age groups, it performed better than BPHR‐X, which was for children of 6 to 11 years as a single group.
To attenuate the impact of distinct rates of height and BP, Mourato and colleagues adjusted the height by (7×(13−age in years)) in young children to overcome the disadvantage of BPHR.13 They found that modified BPHR, BP:eHT13, increased the ability to accurately detect BP abnormalities in children aged 5 to 12 years in comparison with the original BPHR. In the present study, BP:eHT13 revealed a slight change in age‐specific optimal thresholds, which allowed the elevated BP to be evaluated without stratifying by age. This might also contribute to the better performance of our proposed BP:eHT13 thresholds for discriminating BP abnormalities than existing BPHR references. In addition, the cutoff points obtained using BP:eHT13 were very similar between the sexes, as suggested by the previous study.13 Therefore, we developed the BP:eHT13 cutoff values for the combined age and sex group, and this resulted in fewer cutoff points for prehypertension and hypertension. This is particularly important because a more practical screening method for elevated BP should include as few cutoff points as possible.
Study Strengths and Limitations
Although there was only a slight improvement in detecting hypertension using BP:eHT13, a significant development was observed for discriminating prehypertension regardless of various BMI and height levels. Since children with prehypertension also demonstrate evidence of target organ damage, including increased left ventricular mass, elevated carotid intima‐media thickness, and diastolic dysfunction,3 accurate diagnosis and intervention of prehypertension in early life could reduce the risk of greater target organ disease in the future. Moreover, our proposed thresholds presented NPVs of nearly 100%, suggesting that all individuals without BP abnormalities could be identified.22 Our proposed BP:eHT13 thresholds have potential to replace more complicated and time‐consuming sex‐, age‐, and height‐specific BP references to confirm negative cases.5, 13
According to current guidelines, high BP level tends to fall on subsequent measurements as a result of the accommodation effect and regression to the mean. Consequently, both prehypertension and hypertension should be confirmed on repeated visits.5 In addition, only Han nationality was included in the present study. This limits its generalization to other ethnicities. However, the cutoff values of our study were similar with those conducted in Brazilian children.13
Conclusions
In general, BP:eHT13 performed better than, if not similar to, BPHR in discriminating prehypertension and hypertension in Chinese children aged 7 to 11 years with a small number of cutoff points. Our proposed thresholds are simple for hypertension screening in Chinese young children.
Acknowledgments
The authors would like to thank Wen‐Ke. Liao, Wen‐Hua Xing, Xin Zhang, and the members of the Chinese National Survey on Students' Constitution and Health for providing access to the survey data. We appreciate the helpful comments from Tania Pan and Luke Arnold, and we are grateful to the students who participated in the survey for their cooperation.
Author contributions
BD conceptualized and designed the study and drafted the initial manuscript. ZW designed the study, carried out the initial analyses, and reviewed and revised the manuscript. JM and HW designed the data collection instruments, coordinated and supervised data collection, carried out the analyses, and critically reviewed the manuscript. All authors approved the final manuscript as submitted.
Disclosure
Specific funding for this analysis was obtained from the National Health and Medical Research Council of Australia (number APP1045000) awarded to Zhiqiang Wang. The authors declare that they have no conflicts of interest.
J Clin Hypertens (Greenwich). 2016;18 557–564. DOI: 10.1111/jch.12712. © 2015 Wiley Periodicals, Inc.
References
- 1. James PA, Oparil S, Carter BL, et al. 2014 evidence‐based guideline for the management of high blood pressure in adults: report from the panel members appointed to the Eighth Joint National Committee (JNC 8). JAMA. 2014;311:507–520. [DOI] [PubMed] [Google Scholar]
- 2. Chen X, Wang Y. Tracking of blood pressure from childhood to adulthood: a systematic review and meta‐regression analysis. Circulation. 2008;117:3171–3180. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Urbina EM, Khoury PR, McCoy C, et al. Cardiac and vascular consequences of pre‐hypertension in youth. J Clin Hypertens (Greenwich). 2011;13:332–342. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Chiolero A, Bovet P, Paradis G. Screening for elevated blood pressure in children and adolescents: a critical appraisal. JAMA Pediatr. 2013;167:266–273. [DOI] [PubMed] [Google Scholar]
- 5. 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):555–576. [PubMed] [Google Scholar]
- 6. Lu Q, Ma CM, Yin FZ, et al. How to simplify the diagnostic criteria of hypertension in adolescents. J Hum Hypertens. 2011;25:159–163. [DOI] [PubMed] [Google Scholar]
- 7. Outdili Z, Marti‐Soler H, Simonetti GD, et al. Performance of blood pressure‐to‐height ratio at a single screening visit for the identification of hypertension in children. J Hypertens. 2014;32:1068–1074. [DOI] [PubMed] [Google Scholar]
- 8. Galescu O, George M, Basetty S, et al. Blood pressure over height ratios: simple and accurate method of detecting elevated blood pressure in children. Int J Pediatr. 2012;2012:253497. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Kelishadi R, Heshmat R, Ardalan G, et al. First report on simplified diagnostic criteria for pre‐hypertensionand hypertension in a national sample of adolescents from the Middle East and North Africa: the CASPIAN‐III study. J Pediatr (Rio J). 2014;90:85–91. [DOI] [PubMed] [Google Scholar]
- 10. Xi B, Zhang M, Zhang T, et al. Simplification of childhood hypertension definition using blood pressure to height ratio among US youths aged 8–17 years, NHANES 1999–2012. Int J Cardiol. 2015;180:210–213. [DOI] [PubMed] [Google Scholar]
- 11. Xi B, Zhang M, Zhang T, et al. Hypertension screening using blood pressure to height ratio. Pediatrics. 2014;134:e106–e111. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Dong B, Wang Z, Wang HJ, Ma J. Blood pressure‐to‐height ratio for screening prehypertension and hypertension in Chinese children. J Hum Hypertens 2015;29:618–622. [DOI] [PubMed] [Google Scholar]
- 13. Mourato FA, Nadruz W Jr, Moser LR, et al. A modified blood pressure to height ratio improves accuracy for hypertension in childhood. Am J Hypertens. 2015;28:409–413. [DOI] [PubMed] [Google Scholar]
- 14. Ji CY, Chen TJ. Empirical changes in the prevalence of overweight and obesity among Chinese students from 1985 to 2010 and corresponding preventive strategies. Biomed Environ Sci. 2013;26:1–12. [DOI] [PubMed] [Google Scholar]
- 15. Xu T, Zhu G, Liu J, et al. Gender‐specific prevalence and associated risk factors of high normal blood pressure and hypertension among multi‐ethnic Chinese adolescents aged 8–18 years old. Blood Press. 2015;24:189–195. [DOI] [PubMed] [Google Scholar]
- 16. Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. Establishing a standard definition for child overweight and obesity worldwide: international survey. BMJ. 2000;320:1240–1243. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Cole TJ, Flegal KM, Nicholls D, et al. Body mass index cut offs to define thinness in children and adolescents: international survey. BMJ. 2007;335:194. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Kuczmarski RJ, Ogden CL, Guo SS, et al. 2000 CDC Growth Charts for the United States: methods and development. Vital Health Stat 11 2002;246:1–190. [PubMed] [Google Scholar]
- 19. Sedgwick P. Receiver operating characteristic curves. BMJ. 2013;346:f2493. [DOI] [PubMed] [Google Scholar]
- 20. Zou KH, O'Malley AJ, Mauri L. Receiver‐operating characteristic analysis for evaluating diagnostic tests and predictive models. Circulation. 2007;115:654–657. [DOI] [PubMed] [Google Scholar]
- 21. Guo X, Zheng L, Li Y, et al. Blood pressure to height ratio: a novel method for detecting hypertension in paediatric age groups. Paediatr Child health. 2013;18:65–69. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Sedgwick P. Measuring the performance of screening tests. BMJ. 2014;348:g4438. [DOI] [PubMed] [Google Scholar]