Article Text

Download PDFPDF
Correlates of hypertension among urban Asian Indian adolescents
  1. Ruchika Goel1,
  2. Anoop Misra2,
  3. Sunil K Agarwal1,3,
  4. Naval Vikram4
  1. 1University of Pittsburgh Medical Center, Children's Hospital of Pittsburgh, Pittsburgh, Pennsylvania, USA
  2. 2Department of Diabetes and Metabolic Disorders, Fortis Hospital, New Delhi, India
  3. 3Gillings School of Global Public Health, University of North Carolina, Chapel Hill, North Carolina, USA
  4. 4Department of Medicine, All India Institute of Medical Sciences, New Delhi, India
  1. Correspondence to Anoop Misra, Department of Diabetes and Metabolic Disorders, Fortis Hospital, Vasant Kunj, New Delhi 110070, India; anoopmisra{at}


Objective Primary hypertension has become increasingly common in children but remains largely understudied, underdiagnosed and undertreated. This study examines the relationship between hypertension in adolescents and various markers of obesity, serum lipid levels, fasting blood glucose (FBG), haemoglobin A1c (HbA1c) and family history of hypertension.

Design/settings A cross-sectional population-based study of 1022 students aged 14–19 years in New Delhi, India.

Main outcome/exposures Those with age, gender and height specific blood pressure >95th percentile (derived from study data) or >130/85 mm Hg were considered hypertensive. Various markers of obesity, FBG, HbA1c and serum lipid levels were divided into quartiles and the odds ratios of hypertension calculated for the highest quartiles with reference to the lowest.

Results Hypertension was seen in 65 (6.4%) adolescents (2.7% isolated systolic, 2.0% isolated diastolic and 1.7% both). The odds of having hypertension were higher for those in the highest versus lowest quartiles of various measures of obesity such as body mass index (OR 2.90; 95% CI 1.40 to 6.12) and waist circumference (OR 5.21; 95% CI 2.14 to 12.17). A parental history of hypertension was associated with diastolic hypertension in the child (OR 2.21; 95% CI 1.13 to 4.33); the odds ratio decreased after simultaneous adjustment for salt intake (OR 1.98; 95% CI 1.00 to 3.94). In a multivariable model with backward elimination, waist circumference and triglycerides were the strongest predictors of hypertension, further suggesting that the relationship is stronger with central than peripheral obesity.

Conclusion Hypertension in Asian Indian adolescents is associated with obesity, higher serum lipids and a family history of hypertension.

Statistics from

Request Permissions

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


Childhood obesity is a growing concern in developing countries.1 2 The prevalence of obesity and ensuing insulin resistance is increasing among children and adolescents in urban India.3 Concomitantly, primary hypertension has also become more common in children.4 Hypertension is a well recognised risk factor for cardiovascular disease.5 Further, hypertension in adults often has its origins in childhood.6 7 Nevertheless, essential paediatric hypertension remains largely understudied, underdiagnosed and undertreated.8 Few studies have reported the prevalence of hypertension and its correlates in children and adolescents of Asian Indian ethnicity.9,,12 Of note, we have recently reported blood pressure centile values for an adolescent population in New Delhi, India.13

What is already known on this topic

  • Hypertension among adolescents is often unrecognised, underdiagnosed and hence unmanaged.

  • Epidemiological studies among children, mostly in Western countries, suggest that obesity is associated with hypertension.

What this study adds

  • This is the first study in Asian Indian adolescents to report a strong association between hypertension and several measures of obesity, higher serum triglycerides and family history.

  • Waist circumference and triglycerides levels are the strongest correlates of hypertension among several studied markers of obesity in this population.

Despite comparable levels of adiposity, children of Asian Indian ancestry have more metabolic disorders and at a younger age compared with white Caucasian children.3 Inadequate maternal nutrition in pregnancy, low birth weight and childhood ‘catch-up’ obesity, dietary imbalance including a higher intake of fat particularly ω-6 polyunsaturated fatty acids, and lack of exercise are the most important modifiable risk factors of the observed metabolic deregulation.14 Subclinical inflammation as suggested by increased highly sensitive C-reactive protein levels, and hyperinsulinaemia are evident with dietary imbalance and higher adiposity in adolescent Asian Indians.3 15,,17 As in adults, hypertension has consistently been proposed as part of the definition of metabolic syndrome in the paediatric and adolescent age group.18 However, very little is known about hypertension in children in the overall context of obesity and the metabolic syndrome.19 Recent studies among Indian children have reported higher odds of prevalent hypertension among Indian adolescents with a higher body mass index (BMI).11 20 However, research on comprehensive measures of obesity including waist circumference and skinfold thickness is lacking in Asian Indian adolescents.

An understanding of the correlates of early hypertension will help identify risk factors occurring at a young age, guide preventive programs focussing on risk factors for hypertension, and also help define ethnic-specific metabolic syndrome in children.

This community based cross-sectional study aims to explore the relationship between hypertension in Asian Indian adolescents and various markers of obesity, serum lipid levels, fasting blood glucose (FBG), haemoglobin A1c (HbA1c) and a family history of hypertension.



For the present study, data on 1022 apparently healthy school and college students (14–19 years) in New Delhi, India were obtained from the Epidemiological Study of Adolescents and Young Adults.21 For recruitment, multi-stage cluster sampling, based on the modified World Health Organization Expanded Program of Immunization Sampling Plan22 was used as described previously.21 23

The study was approved by the Director of Education, Delhi Government and an institutional review board at the All India Institute of Medical Sciences. Written informed consent was given by all study participants or their parents (when younger than 18 years).

Blood pressure measurements

Blood pressure was measured using a mercury sphygmomanometer (Industrial Electronic and Allied Products, Pune, India) after the subject had rested for 5 min in the sitting position, using an appropriate cuff size. In case of an abnormal blood pressure recording, another reading was obtained after a 5 min rest and the mean of the two values calculated. The blood pressure readings were taken by the same physician for all subjects using the same instrument. The instrument was periodically validated against a Hawksley random zero sphygmomanometer (Hawksley, Lancing, Sussex, UK).

Definition of hypertension

Using the study data, linear regression models were fitted with systolic (and separately with diastolic) blood pressure as outcome and with gender and second-order polynomials for both age and height as explanatory variables. From these models, age, gender and height specific predicted 95th percentiles of systolic and diastolic blood pressure were derived. Those with systolic blood pressure, diastolic blood pressure or either above the 95th percentile of the predicted value for their respective gender, age and height were considered to have systolic hypertension, diastolic hypertension or overall hypertension, respectively. Additionally, those with systolic blood pressure over 130 mm Hg, diastolic blood pressure over 85 mm Hg or either were considered to have systolic, diastolic or overall hypertension, respectively.

Explanatory variables

Measures of obesity

BMI, waist circumference, hip circumference, skinfolds (biceps, triceps, suprailiac, subscapular), percentage body fat and fat mass were measured as described previously.21 Briefly, we measured height (to the nearest 0.005 m), weight (to the nearest 0.1 kg), waist (at the mid-abdomen level to nearest 0.001 m) and hip circumferences (at the largest posterior extension of the buttocks, to the nearest 0.001 m) using a heavy duty inelastic tape. Biceps, triceps, subscapular and suprailiac skinfold thickness were measured using Lange skinfold calipers (Beta Technology, Santa Cruz, California, USA). The mean of three measurements was taken for waist circumference, hip circumference and skinfold thickness. We defined Σ4SF as the sum of two central (suprailiac and subscapular) and two peripheral (biceps and triceps) skinfolds. Percentage body fat was measured using a two-point leg-to-leg bioelectrical impedance method (Tanita TBF 300; TANITA, Tokyo, Japan). Recent studies have shown a strong correlation (r=0.85–0.95) for estimates of percentage body fat, fat mass and fat-free mass as measured by bioelectrical impedance analysis and dual-energy x-ray absorptiometry.24 25 The bioelectrical impedance method has also been validated for Asian children and adolescents26 and also specifically for Asian Indian men and women.27 We also used this method in our previous studies.16 21


FBG and lipid profile (total cholesterol, high-density lipoprotein cholesterol (HDL-C) and low-density lipoprotein cholesterol (LDL-C), and serum triglycerides) were determined as described previously.16 17 A commercially available radioimmunoassay kit (Medicorp, Montreal, Canada) was used to determine serum insulin levels as previously.16 The intra-assay and interassay percentage variations were 2.4% and 3.2%, respectively. A faculty member of the Department of Biochemistry supervised the quality control of all biochemical measurements and insulin assays.

BMI, waist circumference, waist–hip circumference ratio, body fat mass and the sum of four skinfold thickness were used as markers of obesity.

Family history

A family history was defined as a history of hypertension in the father or mother diagnosed by a physician or intake of antihypertensive medication.

Statistical analysis

Unpaired two-sample t tests were carried out to compare the means of the anthropometric and biochemical variables between normotensive and hypertensive subjects. All markers of obesity, FBG, HbA1c and lipid levels were divided into quartiles. The odds ratios and 95% CI of hypertension (systolic, diastolic and overall) for the highest quartiles with reference to the lowest were calculated for the various explanatory variables discussed above. As age, gender, and height may confound these associations, and so that average estimates independent of the above variables could be obtained, they were adjusted for in the estimation of the odds ratios using multivariable logistic models. Multiple linear regression was used to model the continuous blood pressure measurements. Logistic regression using stepwise elimination was carried out to identify the variables most predictive of systolic, diastolic and overall hypertension. A p value of less than 0.05 on a two sided null hypothesis test was considered statistically significant.


A total of 1022 adolescents (54.3% males, mean age 16.7 years, range 13–19 years) were included in the analyses. Sixty-five (6.4%) met the study definition of hypertension. Twenty-eight (2.7%) had isolated systolic, 20 (2.0%) isolated diastolic and 17 (1.7%) both systolic and diastolic hypertension. Either a systolic blood pressure >130 mm Hg or a diastolic blood pressure >85 mm Hg was seen in 41 (4%) subjects and both systolic blood pressure >130 mm Hg and diastolic blood pressure >85 mm Hg was seen in 10 (1%) children.

Table 1 shows the anthropometric and biochemical variables of subjects with and without hypertension. Notably, subjects with hypertension had a faster heart rate and higher values for all measures of obesity. The odds of being hypertensive were higher for those in the highest quartiles of various measures of obesity as compared to those in lowest quartiles of the same measure (table 2). The association of waist circumference, waist–hip ratio and the sum of four skinfold thicknesses seemed to be stronger with diastolic pressure than with systolic pressure. For the odds ratios of systolic, diastolic, and overall hypertension per standard deviation increment, see online supplementary table 1.

Table 1

Comparison of the characteristics of Asian Indian adolescents in Delhi by hypertension* status (n=1022)

Table 2

Odds ratios of systolic, diastolic and overall hypertension* in Asian Indian adolescents in Delhi comparing the highest with the lowest quartile of various characteristics (n=1022)

Systolic (statistically not significant) and diastolic hypertension were associated with a history of clinical hypertension in either of the parents (table 2). The odds of having diastolic hypertension in the presence of a family history attenuated and was statistically non-significant (OR 1.98; 95% CI 1.00 to 3.94) after additional adjustment for BMI and daily salt intake.

Logistic regression with all the variables listed in table 2 with stepwise selection showed that waist circumference and serum triglyceride level were the strongest predictors of diastolic, systolic and overall hypertension. For hypertension, the odds estimates from a model with waist circumference and serum triglyceride only were 1.62 (95% CI 1.28 to 2.05) and 1.08 (95% CI 1.06 to 1.16) for every 10 cm and 10 mg/dl increase, respectively. Further, waist circumference explained a higher variance (r2=0.09, p<0.001) as compared to BMI (r2=0.00, p=0.9) in age, gender and height adjusted systolic blood pressure (figures 1 and 2). For instance, a 3 cm increase in waist circumference increased the age, gender and height adjusted systolic blood pressure by 1 mm Hg.

Figure 1

Scatter plot with linear fit line between mean systolic blood pressure (age, gender and height adjusted) and body mass index (r2=0.00, p=0.9).

Figure 2

Scatter plot with linear fit line between mean systolic blood pressure (age, gender and height adjusted) and waist circumference (r2=0.09, p<0.001).


This is the first study among Asian Indian adolescents to show that various measures of obesity including BMI, waist circumference, percentage body fat and various skinfold thicknesses and a family history of hypertension are associated with higher blood pressure. The relationship of hypertension with waist circumference (central adiposity) appears to be stronger than with peripheral skinfold thickness (peripheral adiposity) or BMI (overall obesity) in adolescents and this relationship is stronger with diastolic than systolic pressure. Also, those in the highest quartiles of FBG, serum triglycerides and serum LDL-C had higher odds of having hypertension, although not reaching statistical significance. The results for the above measures did not reach statistical significance due to the small number of hypertensive children in the lowest quartiles.

The prevalence of hypertension among children is estimated to be about 2–5 % in the USA.4 It was 6.4% in our school- and college-based study and is comparable to the prevalence in an outpatient clinic in Mumbai (West India) in 2001,28 and to the 6.2% observed in a recent study of school children in Mysore (South India).29 However, prevalence depends on the definition used and usually regresses to lower values on repeated examination,29 30 emphasising the need for repeated examination using a standard protocol for clinical diagnosis.31 Further, systemic underestimation of blood pressure in younger children may result from use of cuff covering two thirds of arm length, but this is of lesser concern in teenagers.32

It is often difficult to diagnose hypertension in children. About three-quarters of those with hypertension were not diagnosed in the large outpatient clinics of an urban medical centre in the USA.8 The authors gave two explanations for this under-diagnosis: (1) a lack of awareness of normal blood pressure in children owing to numerous age, height and gender specific cut-offs, and (2) a lack of awareness of previous blood pressure readings in a patient. It is also important to note that although most of these children had essential hypertension, secondary causes of hypertension are more common in children than in adults, necessitating careful examination.31 A study of children with sustained hypertension at our institution has shown that 80% had an underlying cause with reno-vascular and endocrine disorders being the most common.33

To our knowledge, no investigator has examined the diagnosis and management of childhood hypertension in India. Given its strong association with obesity and family history, it should be investigated in both paediatric and adult clinics and its presence could serve as a strong motivation for family lifestyle changes. Identification of hypertension and obesity in parents could prompt screening of other members of the family, including children, for hypertension. Also, lifestyle management could lessen hypertension, the metabolic syndrome and other cardiovascular risk factors often seen in Asian Indian children.

About one-third of overweight or obese children and adolescents exhibit features of the metabolic syndrome.17 Because the metabolic syndrome continues into adulthood, its manifestations need to be recognised early for the prevention of diabetes and coronary heart disease.19 Adult blood pressure has been show to be highly correlated with childhood blood pressure, body size and increasing adiposity in adulthood.6 Adult adiposity was related to childhood adiposity, and the most obese as adults showed the greatest increase in weight since childhood.6 Among adults, about 22% of men and 27% of women in the highest quintile of systolic blood pressure were also in the highest quintiles for three out of the five other risk factors for cardiovascular disease (high cholesterol, low HDL-C, high BMI, high triglycerides and high FBG).34 It appears that such clustering of these common risk factors begins early in life.23 35 Additionally, the effects of early hypertension and the clustering of other risk factors on physical and mental growth and development remain unexplored and need to be studied.

Although, genotypic differences explain much variability in BMI,36 other than rare gene disorders for secondary hypertension, no genetic abnormality has been found to be responsible for essential hypertension.37 It is evident that the association of high blood pressure with family history is partly mediated by a shared and modifiable environment including higher fat and salt intake,38 lower fruit and vegetable intake,29 and lower physical activity.37

Among the characteristics reported in table 2, waist circumference and triglycerides were the strongest predictors of hypertension. Kahn and Valdez39 have suggested that a combination of these two variables could be used to identify individuals with a high risk metabolic profile.

Among adults, a reduction in systolic blood pressure of 2 mm Hg resulted in 6%, 4% and 3% reductions in 1-year stroke related mortality, coronary heart disease related mortality and overall mortality, respectively.40 It is highly likely that controlling blood pressure at an early age will result in greater benefits. Thus, community interventions including the promotion of a healthy lifestyle, dietary education and increasing physical activity through nationwide regulatory changes in schools may help to reduce obesity, improve diet and combat sedentary lifestyles among adolescent urban Asian Indians. We have launched a large scale educational program ‘MARG’ to help control childhood obesity. In this program, children, parents and teachers are taught the beneficial effects of a healthy diet and increased physical activity.2


The authors are thankful to the Ministry of Education, Government of New Delhi for their assistance in conducting the study. Mr Ramesh Giri assisted in anthropometry and body fat measurement, Mr Inder Taneja, Mr Gian Chand and Mrs Alice Jacob performed biochemical investigations and insulin assays, and Mr R L Taneja supervised the quality control of biochemical tests. The co-operation of the children who took part in the study, and the help extended by the principals, teachers and staff of the various schools and colleges where the study was conducted are greatly appreciated.



  • Funding This study was funded by a grant from the Department of Science and Technology, Ministry of Science and Technology, Government of India, New Delhi.

  • Competing interests None.

  • Ethics approval This study was conducted with the approval of the Ministry of Education, Government of India, New Delhi, and the All India Institute of Medical Sciences, New Delhi, India.

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