Development of age-specific and sex-specific blood pressure norms and their associations with lipid profile in children aged 6–16 years from urban Bengaluru: a cross-sectional study

Abstract

Introduction

Paediatric hypertension is a significant health problem, with elevated childhood blood pressure (BP) linked to adult hypertension and increased cardiovascular risk. Existing reference guidelines from the American Academy of Pediatrics (AAP) may not be appropriate for Indian children, emphasising the need to develop region-specific normative data.

Objective

(1) To derive population-specific BP percentiles for children from urban Bengaluru, aged 6–16 years, compare these to the AAP reference guidelines; (2) to estimate and compare the proportion of hypertension identified using the derived percentiles versus AAP reference guidelines and (3) to examine the association of elevated BP with lipid profiles.

Design

Cross-sectional observational study.

Setting

Urban school-based setting in Bangalore, India.

Participants

Apparently healthy school-going children (n=9051), aged 6–16 years were included. Children with known chronic illnesses or conditions affecting BP were excluded.

Interventions

None.

Main outcome measures

Sex-specific, age-specific and height-specific systolic BP (SBP) and diastolic BP (DBP); proportion of BP categories (normal, elevated, hypertensive) using AAP reference guidelines and the study-derived percentiles with Centers for Disease Control and Prevention (CDC) and Indian Academy of Pediatrics (IAP) height centiles; proportion of abnormal lipid profiles.

Results

The study-derived SBP and DBP percentiles increased steadily with age in both sexes, differing from the percentiles of AAP reference guidelines. Using AAP reference guidelines, 14.1% of children had elevated BP and 23.9% had hypertension, compared with 6.9% and 8.1%, using study-derived percentiles. Proportion of hypertension estimated from the derived BP percentiles using CDC and IAP height centiles were similar (p=0.178). Elevated BP was significantly associated with a higher proportion of abnormal total cholesterol (8.6% vs 3.2%, p=0.027) and higher low-density lipoprotein-cholesterol levels, predominantly among girls.

Conclusions

The study developed updated age-sex and height-specific BP percentiles for Indian children. Findings indicate that elevated BP may coexist with adverse lipid profiles, suggesting early clustering of cardiometabolic risk factors in this population.

WHAT IS ALREADY KNOWN ON THIS TOPIC

• Blood pressure (BP) percentiles used in India are largely based on American Academy of Pediatrics reference guidelines, which may not account for population-specific growth patterns and cardiometabolic risk profiles of Indian children.

WHAT THIS STUDY ADDS

• Updated sex-specific, age-specific and height-specific BP percentiles for children from urban Bengaluru aged 6–16 years, which can be used to improve the accuracy of hypertension diagnosis in Indian children.

• Evidence for significant association between elevated BP and adverse lipid profiles, suggesting the coexistence of cardiometabolic risk factors in Indian children.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• These percentiles can enhance the accuracy of paediatric hypertension screening in routine clinical practice and inform multicentre studies to derive nationally representative BP reference standards, thereby strengthening cardiovascular risk surveillance and guiding child health policy in India.

Introduction

Hypertension affects one in three adults globally, with prevalence rates stabilising over two decades; ~32% in women and 34% in men aged 30–79 years.^{1 2} In India, data from the National Family Health Survey indicate hypertension prevalence of 22.6% in adults.³ Though commonly associated with adults, hypertension has its origins in childhood; 4% in children and adolescents globally⁴ and ~7% of Indian school-going children are reported to be hypertensive, while among adolescents, the prevalence of hypertension is estimated at 7.6%, with significant heterogeneity across studies.^{5 6}

Paediatric hypertension has significant and prolonged negative health outcomes, both in childhood and later in life. Elevated blood pressure (BP) in childhood (aged 3–18 years) significantly increased the risk of hypertension in adulthood (ages 18–57 years),⁷ while children with elevated BP were observed to have 35% higher risk of developing elevated BP as adults, compared with those with normal BP.⁸ Longitudinal studies link elevated childhood BP to early signs of cardiovascular damage including increased pulse wave velocity, greater carotid intima–media thickness, left ventricular hypertrophy⁹ and increased rates of adult cardiovascular events by 40 years of age,¹⁰ highlighting the need for early screening, detection and timely intervention of elevated BP in children to mitigate the long-term burden of cardiovascular diseases.

Currently, paediatric hypertension is classified using percentile-based criteria adjusted for age, sex and height^{11 12}; systolic BP (SBP) or diastolic BP (DBP) ≥95th percentile or 130/80 mm Hg for age >13 years is classified as hypertension, while values between the 90th and 95th percentiles are considered elevated BP. Direct application of the American Academy of Pediatrics (AAP) reference guidelines, in which BP was measured in American children, using the Centers for Disease Control and Prevention (CDC) height centiles,¹² to Indian children may have challenges due to significant physiological, socioeconomic and epidemiological differences, making region-specific norms essential for accurate diagnosis and management. While few Indian studies have developed BP percentiles for children aged 5–16 years, these were conducted over two decades ago and some did not account for sex-based differences,^{13 14} highlighting the need to establish updated age-sex and height-specific BP percentiles tailored for Indian children to enable accurate diagnosis and effective clinical management. Indian Academy of Pediatrics (IAP) published height centiles,¹⁵ being routinely used by paediatricians to monitor growth.

Indian children demonstrate a high prevalence of elevated BP, and those with high BP were more likely to exhibit coexisting risk of overweight/obesity, high fasting blood glucose, high triglycerides and high low-density lipoprotein cholesterol (LDL-C).¹⁶ The early coexistence of these factors may increase the risk of cardiovascular morbidity and mortality in adulthood. Unique body composition, genetic and environmental factors render Indian children more susceptible, necessitating early screening to plan targeted prevention strategies and address India’s unique early onset and high burden of cardiovascular disease.

Hence, the objectives of the present study were (1) to derive sex, age and height specific BP percentiles for children aged 6–16 years from urban Bengaluru using both CDC and IAP height centiles, (2) to estimate and compare the proportion of elevated BP and hypertension using the newly derived percentiles with currently used AAP reference guidelines and (3) to examine the associations of elevated BP with lipid profile of the study population.

Methods

Study design and participants

The present work was nested within the larger Pediatric Epidemiology and Child Health (PEACH I–III) cohort, conducted by the Division of Nutrition, St. John’s Research Institute, Bengaluru, India from 2008 to 2018, which recruited apparently healthy school-aged children (5–18 years), born at term and residing in urban Bengaluru using convenience sampling technique.^{17 18} For this analysis, cross-sectional data on 9051 children aged 6–16 years was included from the PEACH II and III phases (2012–2018). To achieve the objective of the study, a minimum sample of 200 children per age-sex group was required based on the recommendation for generating normative values and to provide adequate precision for estimating location, scale and shape parameters of BP in the Generalised Additive Model for Location, Scale and Shape (GAMLSS) framework.^{19 20} Children with any history of long-term medical conditions were excluded from the study. Information on comorbidities was first obtained through a structured questionnaire completed by parents/guardians. Subsequently, a paediatrician assigned to the study cohort reviewed the medical history of each child and screened specifically for chronic illnesses or long-term medical conditions. Sociodemographic details including parental education and monthly income were recorded through a self-administered questionnaire, filled by the parents/guardians. Dietary patterns were assessed by recording the frequency of intake of milk, vegetables, fruits, sweets, salty snacks, fast-foods, baked items and sugar-sweetened beverages. Physical activity was evaluated using a questionnaire capturing the frequency and duration of activities on weekdays and weekends, as well as time spent in sedentary behaviours. Activities were categorised as light, moderate or vigorous using metabolic equivalent tasks.²¹ All the measurements were conducted at the school premises.

Patient and public involvement

Patients and the public were not involved in the development of the research questions, study design or outcome measures. Participants were recruited through schools and had no role in recruitment or study conduct. The study procedures and time commitment were clearly explained to school authorities, parents and children prior to participation. Study findings will be shared with participants following the publication through school-based awareness sessions and disseminated more widely via a press release.

Anthropometric measurements

Anthropometry was measured by trained nutritionists following standard protocols; height was measured to the nearest 0.1 cm (SECA 213, USA), and weight was measured in light clothing, on empty bladder, without footwear to the nearest 0.1 kg using a calibrated digital scale (Salter, Germany). Waist circumference was measured to the nearest 0.1 cm with a non-stretchable measuring tape (ADC 396, USA).²² The % technical error of measurement for anthropometric measurements was <0.5% intraobserver and 1.3% for interobserver variability. Body mass index (BMI)-for-age Z-scores were calculated using the WHO’s AnthroPlus software (V.1.0.4; Geneva, Switzerland). Children with >+1 SD Z scores were categorised as overweight, >+2 SD as obese and values between −1 and +1 SD were considered normal.²³

Body composition measurements

The body composition was estimated by measuring the body volume using the principle of air displacement plethysmography,²⁴ using BOD POD (Cosmed, Italy, software V.5.2.0). Details regarding the calibration of BOD POD and methodology of assessing body composition using BOD POD have been published previously.²⁵ Briefly, body volume was measured with the child seated inside the chamber wearing appropriate clothing. An electronic scale was used to measure the body weight of the child, and the body density (kg/L) was determined as the ratio of body weight (kg) to body volume (L), from which the proportions of fat mass (%FM) and fat-free mass (%FFM) were determined using age-sex specific equations.²⁶ Fat mass index (FMI) and fat-free mass index (FFMI) were obtained by dividing total FM and total FFM, respectively, in kilograms (kg) by height (m²). Daily calibrations were performed, and the coefficient of variation (CV) for %FM was 2.3%.²⁵

BP measurement

BP was assessed at a single study visit, measured using a mercury sphygmomanometer (Diamond Regular 112 MAR, India) with an appropriate paediatric cuff, following recommended guidelines by two trained technicians. The technicians were trained following the standard protocol recommended by the American Heart Association. Measurements were conducted after the children remained calm in the familiar school environment with minimal distractions. Prior to the measurement, each child was made to sit quietly for 5 min on a chair with back support and the right arm was positioned at a 90° angle. Two BP readings were recorded to the nearest 2 mm Hg, with a minimum interval of 5 min between measurements.¹¹ To ensure accuracy, BP readings were measured in a pilot sample where technicians remained blinded to each other’s reading. The interobserver variability was minimal, with differences in SBP and DBP between each pair of readings being <5 mm Hg in 95% of measurements.

Blood biochemistry

On the day of the measurement, the children were advised to report to school following an overnight fast. About 4 mL of blood was collected by a trained technician. Lipid profile including total cholesterol (TC), triglycerides (TG), high-density lipoprotein cholesterol (HDL-C) and LDL-C were analysed using a point-of-care device—Cholestech LDX analyser (Alere San Diego, San Diego, California, USA).^{27 28} The inter-assay CV by the internal quality controls was <5% for all assays. Cut-offs to define abnormality were: hypercholesterolaemia—serum cholesterol ≥200 mg/dL; high LDL-C—serum LDL-C ≥130 mg/dL; low HDL-C—serum HDL-C <40 mg/dL; hyper-triglyceridaemia—serum TG 5–9 years: ≥100 mg/dL and 10–19 years: ≥130 mg/dL.²⁹

Statistical analysis

Descriptive statistics were reported as mean±SD for continuous variables and frequencies with percentages for categorical variables. Anthropometry, BP and body composition were compared between the sexes using independent sample t-test at each age group. GAMLSS method was performed to derive the age-sex and height specific percentile curve for SBP and DBP for the children with normal BMI.¹⁹ GAMLSS method enabled the modelling of SBP and DBP distributions based on multiple covariates, accounting for variability in dispersion, skewness and kurtosis. Box-Cox power exponential (BCPE/BCPEo), Box-Cox Cole and Green (BCCG/BCCGo) and Box-Cox t (BCT) distribution were examined and best fit was chosen for SBP and DBP. The BCPE distribution was identified as the ‘best fit’ for both sexes for SBP. For DBP, the BCCG distribution was selected for boys and the BCPE distribution for girls. The effects of age and height on distribution parameters were modelled as constant, linear or smoothed (cubic/penalised) splines depending on the best model fit. Worm plots were used as diagnostic tools to assess the need for kurtosis adjustments. Goodness of fit, assessed by the Akaike Information Criterion, global deviance and Q–Q plots were used to choose the final model based on optimal fit.

Height centiles were derived using both CDC 2000 growth charts via Epi Info V.7³⁰ and IAP growth charts.¹⁵ Age-sex-specific 50th, 90th and 95th percentiles for SBP and DBP were computed across the 5th, 10th, 25th, 50th, 75th and 95th centiles of height derived using CDC growth charts. Similarly, age-sex-specific 50th, 90th and 95th percentiles for SBP and DBP were computed across the 3rd, 10th, 25th, 50th, 75th and 97th centiles of height, derived using growth charts of IAP.¹⁵ Comparison between the study-derived age-sex specific 90th, and 95th percentiles of SBP and DBP at 50th height centile with the corresponding percentiles from AAP reference guidelines were reported.

Children were classified as normotensive if both SBP and DBP were <90th percentile, elevated BP if either SBP or DBP was between the 90th and 95th percentiles, and hypertension if either was >95th percentile based on BP percentiles derived from CDC and IAP height centiles. Agreement between the AAP reference guidelines and the study-derived BP classification (using the CDC height centiles) was assessed using Cohen’s Kappa statistic. The proportion of hypertension was compared between study-derived BP percentiles using CDC and IAP height centiles using McNemar χ² test. For further analysis, children with elevated BP and hypertension were combined into a single category of elevated BP. Association of elevated BP with the presence of abnormal lipid markers was analysed using the χ² test among children with available data without excluding them based on weight status. A p<0.05 was considered statistically significant. All statistical analyses were performed using R (V.4.5.0),³¹ with {package: ‘gamlss’ used for percentile derivation^{19 32}} and IBM SPSS V.29.0.

This study was reported according to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.

Results

Of the total 11 196 children assessed, 0.9% were underweight, 16.2% were overweight, 2.0% were obese (figure 1) and were excluded from the data analysis. The final analysis included 9051 children: 48% boys and 52% girls. There was no significant difference in the age-sex distribution between the included and excluded children except for socioeconomic status (SES) of the school to which the child belongs (online supplemental table 1).

Figure 1. Details of selection of participants. PEACH, Pediatric Epidemiology and Child Health.

Descriptive statistics for anthropometry, BP and body composition parameters by age and sex are presented in table 1. Up to age 10 years, boys and girls had similar mean height and weight. Between 10 and 13 years, girls were significantly taller, heavier and had higher FFMI (p<0.05). From 14 years, except for FFMI, boys exceeded girls in all the above measures, while girls had significantly higher FMI at most ages, except between 9 and 10 years, where boys had significantly higher values. The mean SBP showed a trend of being higher in boys compared with girls across all age groups, except for ages 12 and 13 years, while girls showed a higher trend of DBP across all ages.

Table 1. Age-specific and sex-specific characteristics of study children.

Age (years)	n	Weight (kg)	Height (cm)	Systolic BP (mm Hg)	Diastolic BP (mm Hg)	Fat mass index (kg/m2)	Fat free mass index (kg/m2)
Boys (n=4329)
6	266	18.3±2.0	114.9±5.3	105.2±7.7	60.7±6.6	1.9±0.7	12.0±0.8*
7	338	20.0±2.4	119.7±6.0	107.3±8.1*	62.2±7.1	2.0±0.7	11.9±0.7
8	366	22.1±2.7	125.3±6.0	108.8±8.4	63.2±7.0	2.0±0.7	12.0±0.8
9	391	24.5±2.9	130.3±5.7	108.7±8.6	63.2±7.5	2.3±0.9	12.1±0.9
10	444	27.0±3.4	135.1±6.3	109.3±8.6	63.9±8.4	2.5±1.0*	12.2±0.9
11	454	29.6±4.0	139.8±6.6	110.2±9.7	64.2±8.5	2.7±1.0	12.3±1.0
12	481	32.8±4.8	145.7±7.4	109.5±9.5	63.6±7.9	2.8±1.1	12.6±1.0
13	480	36.8±5.9	151.4±8.5	110.6±10.2	64.5±8.8	2.9±1.2	13.0±1.3
14	494	41.8±6.6	158.2±8.8*	112.7±10.0	65.5±9.5	2.8±1.1	13.8±1.4
15	415	45.7±6.4*	163.1±7.8*	115.7±10.4*	65.7±8.2	2.6±1.1	14.5±1.5
16	200	47.0±5.4*	164.4±6.2*	116.9±9.9*	65.0±7.6	2.5±1.1	14.9±1.3*
Girls (n=4722)
6	256	18.1±2.1	114.3±5.1	104.4±8.0	61.8±7.2	2.0±0.8	11.8±0.9
7	361	20.1±2.6	119.5±5.9	105.7±8.2	62.6±7.3	2.1±0.9	12.0±0.9
8	373	22.0±2.9	124.5±5.9	107.7±8.7	63.6±7.4	2.2±0.9*	12.0±0.9
9	370	24.4±3.2	130.4±6.2	108.2±8.8	64.5±8.1*	2.2±0.9*	12.3±0.9
10	404	27.7±3.8*	136.0±6.3*	108.4±9.4	63.9±9.0	2.3±1.0	12.7±1.0*
11	398	31.9±5.0*	142.7±7.1*	110.1±10.6	64.6±9.0	2.7±1.1	13.0±1.1*
12	434	36.2±5.5*	148.5±7.2*	110.7±10.4	65.0±8.6*	2.8±1.1	13.6±1.2*
13	519	39.8±5.1*	152.4±5.9*	110.9±9.5	65.8±8.2*	3.2±1.2*	13.9±1.1*
14	655	41.8±4.8	154.2±5.5	111.7±9.6	65.6±8.3	3.4±1.3*	14.2±1.1*
15	681	43.0±4.7	154.5±5.6	112.0±9.5	66.3±8.3	3.6±1.2*	14.5±1.1
16	271	43.6±5.1	154.3±6.0	111.4±9.1	66.1±8.5	3.8±1.4*	14.5±1.1

Values are presented as mean±SD.

^*Indicates that the marked sex has a significantly higher mean than the other sex; p<0.05 indicates a statistically significant difference between the sexes based on independent samples t-test.

BP, blood pressure.

Age-sex and height-specific percentile values based on CDC height centiles from 5th to 95th percentiles for SBP and DBP are presented in online supplemental tables 2 and 3, respectively, and an example of their application provided in online supplemental appendix S1. In both boys and girls, SBP and DBP percentiles showed a gradual rise from ages 6 to 16 years. At a given age and sex, BP percentiles increased with height centiles. For example, at 95th percentile of SBP, BP values increased by 3–9 mm Hg across 5th–95th height centiles in boys and approximately 1–5 mm Hg in girls. For DBP, the 95th percentile value differed by 1–5 mm Hg in boys and 1–2 mm Hg in girls across the same height range.

Comparison of study-derived 90th and 95th SBP and DBP percentiles at 50th CDC height centile across ages, with the AAP reference guidelines is depicted in figure 2. For boys, the study-derived percentiles and the percentiles of AAP reference guidelines showed an increasing trend in SBP from ages 6 to 16 years, although it increased rapidly after 13 years. While in girls, the percentiles of AAP reference guidelines showed a marked increase starting at age 11 year, in contrast to the study-derived percentiles, which showed a slower and uniform increase up to the age 12, followed by a plateau for both 90th and 95th percentiles. Among boys and girls, the 90th and 95th percentiles of SBP based on AAP reference guidelines were lower than the 90th percentile of study-derived SBP values up to the age of 14 years. However, the 95th DBP percentile of AAP reference guidelines overlapped with the study-derived 90th DBP percentile in boys up to 8 years of age and in girls up to 11 years. Beyond these ages, the percentiles of AAP reference guidelines increased but remained lower than the study-derived percentiles, except in boys older than 15 years, where both the 90th and 95th percentile of AAP reference guidelines were higher.

Figure 2. Comparison of study-derived 90th and 95th systolic and diastolic BP percentiles at 50th height centile across ages with the percentiles of AAP reference guidelines in boys and girls. (A) Boys and (B) girls represent the 90th and 95th percentiles of systolic and diastolic BP at the 50th height centile. Black lines represent study-derived BP percentiles based on CDC height centiles; red lines represent BP percentiles of AAP reference guidelines. The solid line (—) and dashed line (----) represents the 95th and 90th percentile, respectively. AAP, American Academy of Pediatrics; BP, blood pressure; CDC, Centers for Disease Control and Prevention.

When we directly applied the AAP reference guidelines to our study children, 14.1% and 23.9% were classified as having elevated BP and hypertension, respectively. However, using the study-derived BP percentile (by CDC height centiles), 6.9% of the children were classified as having elevated BP and 8.1% as hypertensive. There was a significant difference in the proportion of children identified with hypertension when using the study-derived BP percentiles compared with the AAP reference guidelines (8.1% vs 23.9%; p<0.001).

Figure 3 depicts the agreement in classifying BP categories using the AAP reference guidelines and the present study-derived percentiles based on CDC height centiles. The classification of hypertension using both the methods had an agreement of 100%. Among children classified as having elevated BP using the study-derived percentiles, only 19% were identified as having elevated BP using AAP reference guidelines, while the remaining were classified as hypertensives. Among children with normal BP levels by the study-derived percentiles, only 73% were correctly classified as normotensive by AAP reference guidelines, while 12% and 15% were classified as hypertensives and elevated BP, respectively. Overall agreement between the two methods across the BP categories as assessed by Kappa statistic was 0.36 (95% CI 0.34 to 0.37).

Figure 3. Agreement between study-derived percentiles and the AAP reference guidelines in classifying the BP categories. Three-dimensional clustered column chart depicting the percentage of agreement between BP classifications based on study-derived percentiles (x-axis) and the AAP reference guidelines (y-axis). The percentage of agreement is shown on the z-axis. Columns are colour-coded to represent BP categories: black for hypertensive, dark grey for elevated BP and light grey for normotensive. AAP, American Academy of Pediatrics; BP, blood pressure.

Online supplemental tables 4 and 5 represent the sex, age and height specific percentile values based on IAP height centiles from 5th to 95th percentiles for SBP and DBP. The proportion of hypertension and elevated BP classified using the study BP percentiles derived using IAP height centiles were 8.2% and 7.0% respectively. There was no significant difference in the proportion of hypertension between the study-derived BP percentiles using CDC and IAP height centiles (p=0.178).

In the overall study sample (n=2855), an association between the elevated BP derived using the study-derived percentiles and the abnormal lipid profiles was noted where a significantly higher proportion of children with elevated BP had abnormal TC levels (8.6%), compared with those with normal TC levels (3.2%) (p=0.027). Similarly, a higher proportion of children with elevated BP had abnormal LDL-C levels, though not statistically significant (8.0% vs 3.3%, p=0.088). This significance was primarily noted among girls for TC (13.6% vs 4.2%, p=0.012) and LDL-C (13.3% vs 4.4%, p=0.044). This association was also noted with the elevated BP derived using the AAP reference guidelines.

From the analysis of dietary habits (n=7855), none of the processed food categories; sweets (p=0.572), baked items (p=0.408), namkeens (p=0.140), western fast-foods (p=0.909), Indian fast-foods (p=0.653), fruit juices (p=0.976) or SSB (p=0.100) were significantly associated with the BP categories. Although more than 50% of the children did not meet the recommended hours for physical activity, the median moderate-to-vigorous physical activity was 4 (0, 9) hours/week with no significant difference across the BP categories classified using the study-derived percentiles (p=0.237).

Discussion

The present study established age-sex and CDC height-specific BP percentiles for South Indian children using GAMLSS modelling and compared them to the AAP reference guidelines. Since the majority of the paediatricians in India use the IAP height centiles,¹⁵ the study also generated BP percentiles using the IAP height centiles. The study-derived percentiles using CDC height centiles classified ~15% of the children as having elevated BP/hypertensive, vs ~38% as having elevated BP/hypertensive by the AAP reference guidelines, suggesting possible overestimation using international references. The study-derived BP percentiles using IAP height centiles have also classified about 8.2% of the children as hypertensive and 7.0% as having elevated BP. Findings highlight the need for region-specific BP standards considering Indian children’s unique body composition, genetic predispositions and cardiometabolic risk patterns. Additionally, children with elevated BP showed higher prevalence of TC and LDL-C, highlighting clustering of cardiometabolic risk and the potential for early onset of cardiovascular diseases.

Sex-specific differences were observed in the present study for the BP values; SBP was higher in boys at most ages, except for 12–13 years, while DBP was higher in females across the ages. The higher SBP in boys may be due to increased testosterone during puberty, linked to higher muscle mass and cardiac output.^{33 34} In girls during puberty, increased FM, hormonal fluctuations and endothelial reactivity may contribute to higher DBP in late adolescence, as observed in our study where the girls had higher FMI and higher DBP after 12 years. Additionally, the age-sex and height specific percentile derived from the study showed a steady rise in BP from ages 6 to 9 years, which aligned with consistent somatic growth, where height and weight increase proportionally, while the BP percentiles plateaued in both sexes, between 10 and 13 years, likely due to hormonal and vascular tone changes stabilising BP during puberty.^{33 34} After 13 years, BP increased sharply in boys, may be due to greater lean body mass, cardiac output, sympathetic activity, and activation of the renin–angiotensin–aldosterone system.^{33 34} In girls, BP rise was less steep, likely influenced by oestrogen-mediated vasodilation and adipose tissue accumulation.^{33 34} Previous multicentric study involving non-overweight children and adolescents aged 6–19 years from seven countries also reported similar sex-specific patterns in BP distribution, which showed comparable BP levels between boys and girls until around 13 years, followed by higher BP values in males thereafter.³⁴

The distribution of BP in Indian children of the present study differed significantly from AAP reference guidelines, being consistent with previous studies from India.^{13 14 35} These discrepancies are likely due to differences in growth patterns, body composition, pubertal timing, nutrition, SES, genetic influences, and environmental exposures between Indian and U.S. American children.¹³ The AAP reference guidelines, developed primarily on taller American children (typically 8–10 cm taller than their Indian counterparts), may not be appropriate for Indian children, and could result in lower BP cut-offs and overestimation of elevated BP.^{5 36} In the present study, height was closely associated with BP, with both SBP and DBP increasing steadily across the 5th to 95th percentiles in both sexes. Taller children are likely to have higher BP due to greater blood volume, cardiac output and earlier pubertal progression.^{32 33} Paediatricians in India routinely use the revised IAP height centiles¹⁵ for assessing growth and monitoring children aged 5 to 18 years, which have been specifically developed based on nationally representative Indian data. The proportion of hypertension and elevated BP using IAP height centile-based BP percentiles did not differ significantly from those derived using CDC height centile-based BP percentiles.

In the present study, 8.1% of the urban school-aged children were classified as hypertensive by the study-derived percentiles, compared with 23.9% identified using AAP reference guidelines (p<0.001). Similarly, the proportion of children having elevated BP was also higher with AAP reference guidelines (14.1%) when compared with study-derived percentiles (6.9%), with limited concordance between methods. Nationally, hypertension is estimated at ~7% and elevated BP at 10% among Indian children,^{5 6} but secondary analysis of the Comprehensive National Nutrition Survey (CNNS) dataset revealed even higher rates among Indian adolescents (35.1% in children aged 10–12 years and 25.1% in ≥13 years) using AAP reference guidelines.¹⁶ Thus overestimation of elevated BP and hypertension rates as previously observed¹⁶ can occur with the direct use of AAP reference guidelines in Indian children. Therefore, region-specific BP percentiles derived on Indian children are needed to ensure accurate identification and appropriate clinical decision-making for paediatric hypertension.

The children of the present study with elevated BP had a significantly higher prevalence of abnormal TC (8.6% vs 3.2%) and a non-significant higher prevalence for LDL-C, consistent with the national CNNS data which reports the coexistence and clustering of risk factors among urban and older children.¹⁶ Studies have previously reported that elevated BP in children often coexists with other risk markers such as high TG, elevated fasting glucose, increased body fat and low cardiorespiratory fitness. Children aged 6–15 years from Brazil who were followed for 2 years demonstrated that those who had one risk factor (hypertension) at baseline increased the likelihood of showing additional risk factors over time, demonstrating moderate stability and tracking of clustered risks.³⁷ Similarly, hypertensive/prehypertensive children from Brazil had high rates of overweight and lipid abnormalities.^{38 39} In Chinese children aged 8–12 years, significant differences in lipid levels were observed across BP categories, with the LDL-C/HDL-C ratio being higher in hypertensive children.⁴⁰ Early clustering of cardiometabolic risk factors may predispose Indian children to earlier onset and higher burden of cardiovascular disease, highlighting the need for timely identification and intervention.⁴¹

This present study had certain limitations; cross-sectional study design limited the ability to assess causality or track BP trajectories over time. The clinical diagnosis of hypertension could not be confirmed due to the absence of repeated BP measurements among children with elevated readings. Pubertal staging, which could have further clarified sex-specific trends in BP due to pubertal onset, was not performed due to lack of consent from the parents. The socio-economic differences observed between the included and excluded participants may have introduced selection bias in the findings. The reference values derived from urban school children may not be directly applicable to rural or diverse Indian populations and should not be extrapolated without further validation. Future multicentre studies across the country are needed to enhance representation.

Conclusions

This study established updated, age-sex and height-specific BP percentiles for Indian children using rigorous statistical modelling. Compared with AAP reference guidelines, the derived BP percentiles identified a lower proportion of children at risk for elevated BP and were associated with abnormal lipid profiles. These findings highlight the need for population-specific paediatric BP references to improve early detection and intervention. Future research should validate these percentiles in larger and more diverse Indian populations and integrate them into national paediatric screening guidelines.

Data availability statement

Data are available on reasonable request.