## Physical Growth Standards for Urban Adolescents (10-15 Years) From South Gujarat

#### Author(s): Hitendra G. Thakor, Pradeep Kumar, Vikas K Desai, Ratan K Srivastava

#### Vol. 25, No. 2 (2000-04 - 2000-06)

Department of PSM, Government Medical College, Surat (India) 395 001

### Abstract:

**Research question:** What are the anthropometric measurements (height, weight and BMI) of school going adolescents of Surat city and how do they compare with ICMR and NCHS standards?

**Objectives:** (1) To record age, weight, height and BMI of children and to calculate the mean, standard deviation and the percentile values for the sample population.
(2) To compare these values with ICMR and NCHS standards.
(3) To calculate and study the correlates of weight with age, height, BMI and outdoor playing activities (hours spent per day).

**Study design:** Cross-sectional, school based study.

**Setting: **12 primary schools run by Surat Municipal Corporation and selected by stratified random sampling method.

**Sample size: **2250 children (1092 boys and 1158 girls), aged 10 years and above and attending the selected schools.

**Study variables:** Age, sex, weight, height and BMI.

**Outcome variable: **Mean, standard deviation and percentiles of observed anthropometric measurements.

**Statistical analysis:** Z-test, t test, correlation coefficient, coefficient of determination and simple and multiple regression equations.

**Results and conclusions:** Study provides the growth standards for this population (10-15 years) of Surat city. Appraisal of nutritional status, adjudged by the weight, height and BMI, revealed that the median parameters of the population are comparable to the ICMR standards but are far below the 50th percentiles of NCHS standards. Girls in the present study exhibited the better nutritional status in terms of "weight for age" and the BMI, than boys. Further, while comparing with the ICMR standards, a little slowing down of growth was observed in the present study in both the sexes (in terms of weight and height) after 13 or 14 years of age. Regression equations to calculate the weight and height for a given age and sex have also been constructed. Coefficient of determination calculated for weight and height shows the extent to which the variation in these variables could be explained by the variables (age, height, sex etc.) studied.

**Keywords:** Adolescents, Physical growth, BMI, Coefficient of determination.

### Introduction:

The period of adolescence is a crucial phase of growth and offers the last chance of catch-up growth. Achievement of optimum growth during this period is of utmost importance in maintaining good health thereafter^{1}. Growth monitoring by anthropometric measurement during this period, is not only an important health indica or but also a predictor of various morbidity in the community, though the anthropometry is universally applicable, simple, inexpensive and non-invasive technique, it is still an underutilized tool for guiding public health policy as well as individual clinical decision^{2}.

Simple measurement of height and weight serve as reliable means to evaluate the growth of a child and also to detect gross abnormalities even when no other clinical sign of illness is manifested^{3}.

In a country like India with wide variation in the growth determinants, it is essential that the normal values are developed region-wise and are redefined from time to time^{4,5}. Most of the attempts of generating anthropometric profile have so far focused on pre-school children and a very few have dealt with adolescents age groups^{6}. Non-availability of an anthropometric profile for adolescents of this region of south Gujarat especially for lower middle and lower social class urbanites prompted us to undertake this study. It has been carried out to develop the mean, standard deviation and percentile values (for weight, height and BMI) for studied population and to compare them with national (ICMR)^{7} and international (NCHS)^{8} standards. Attempt has also been made to study various correlates (age, sex, height and BMI) of body weight. Considering that the time spent on outdoor activities (hours per day) can be an important determinant of weight^{9,10}, it was also included to study its association with later.

### Material and Methods:

The study was carried out in 12 primary schools (up to the seventh standard), selected by stratified random sampling technique, from a total of 204 such schools run by Surat Municipal Corporation (SMC); catering for the children of lower middle to lower social class residents of the city. The city is divided into six administrative zones. One school each for boys and girls, was selected by random sampling to give due representation to all the areas and both the sexes. Selected schools were covered in toto and all the children above 10 years of age were included in the study.

Prior permission was obtained from the authority. The purpose and the process of the study was explained to all the participants and their teachers. Age was verified from the school records and recorded in completed years. Weight and height for each child were recorded by using standaridized equipment. Mean, standard deviation and percentile values (5th, 50th and 95th) were calculated for weight, height and BMI for all ages and both the sexes. Values thus obtained, were compared with 50th percentiles of ICMR and NCHS standards. In order to test the significance of the differences, various statistical tests such as Z test and unpaired t test were used, wherever found necessary. Simple and multiple regression equations for weight and height were also constructed. Apart from calculating correlation coefficients (r) for weight with age, height, BMI and outdoor playing, coefficient of determination were also calculated. While correlation coefficient describes the degree to which one variable is linearly related to the another, the coefficient of determination (r^{2}) measures the extent or strength of the association that exists between the two variables^{11}. The two are interrelated as shown below:

Correlation coefficient (r) = square root of coefficient of determination (r^{2})

(r^{2}) can directly also be calculated once the linear regression equation of two variables has been constructed (shown below)

y = a+b * x

where y = dependent variable,

x = independent variable,

a = y intercept and

b = slope of the line

Now

(r^{2}) = a*Ey + b *Ey - n*(^{-}y)^{2}
Ey^{2} -n*(^{-}y)^{2}

In short, r^{2} measures how well the independent variable (x) explains the dependent variable (y). Coefficient of determination for an independent variable further helps in finding out the fraction of total variation of the dependent variable that is explained by the regression line^{11}.

### Results:

A total of 2250 children (1092 boys and 1158 girls) from 12 schools of Surat city constituted the study sample. Adolescents as per WHO^{2} are children between 10-19 years of age. Our study was done in the schools up to the 7th standard only, therefore, the age variation observed in the studied children was between 10-15 years. Out of 2250 children, most belonged to 11 and 12 years of age (30.8% each), followed by 13 (15.8%), 10(10.1%) and 14(8.9%) years of age. Very few (3.5%) belonged to 15 years age. Overall sex ratio was 1060.4. It showed a progressive decline from 1270.0 (10 years) to 645.8 (15 years).

#### Table I: Anthropometric profile of study population - height in centimeters.

Boys | Percentile values | Girls | Percentile values | |||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|

Age (yrs.) | Total examined | Mean | SD | 5th | 50th | 95th | Total examined | Mean | SD | 5th | 50th | 95th |

10 | 100 | 133.7 | 6.3 | 124 | 133 | 144 | 127 | 132.8 | 10.9 | 124 | 132 | 148 |

11 | 305 | 135.4 | 7.2 | 125 | 135 | 146 | 388 | 136.7 | 9.7 | 125 | 137 | 150 |

12 | 324 | 139.5 | 9.7 | 128 | 139 | 153 | 370 | 141.2 | 20.6 | 128 | 141 | 153 |

13 | 190 | 142.8 | 10.9 | 128 | 142 | 157 | 166 | 144.7 | 8.3 | 130 | 145 | 158 |

14 | 125 | 146.7 | 9.1 | 133 | 145 | 161 | 76 | 147.4 | 7.1 | 136 | 148 | 160 |

15 | 48 | 153.6 | 11.1 | 135 | 154 | 166 | 31 | 150.0 | 20.6 | 134 | 146 | 168 |

#### Table II: Anthropometric profile of study population - weight in Kgs.

Boys | Percentile values | Girls | Percentile values | |||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|

Age (yrs.) | Total examined | Mean | SD | 5th | 50th | 95th | Total examined | Mean | SD | 5th | 50th | 95th |

10 | 100 | 25.0 | 4.3 | 19 | 25 | 34 | 127 | 26.1 | 4.8 | 21 | 25 | 35 |

11 | 305 | 25.9 | 4.0 | 20 | 25 | 33 | 388 | 28.4 | 5.9 | 21 | 28 | 40 |

12 | 324 | 27.7 | 4.5 | 21 | 27 | 36 | 370 | 29.4 | 6.3 | 22 | 29 | 39 |

13 | 190 | 30.2 | 5.5 | 23 | 30 | 40 | 166 | 33.4 | 5.9 | 25 | 33 | 44 |

14 | 125 | 31.7 | 5.8 | 21 | 31 | 44 | 76 | 35.9 | 6.9 | 25 | 35 | 50 |

15 | 48 | 36.2 | 7.4 | 25 | 37 | 48 | 31 | 38.0 | 7.1 | 31 | 36 | 54 |

Anthropometric profile of the studied sample in terms of mean, standard deviation and percentile values for height and weight for both the sexes are shown in Table I and II. The height correlated positively with the age and increased with its increase (Table I). In boys, mean height from its minimum (133.7±6.3 cms) in 10 years age rose to the maximum (153.6±11.1 cms) in 15 years age. Similarly, in girls, mean height increased from 132.8±10.9 cms in 10 years to 150±20.6 cms in 15 years age. Median (50th percentile) values largely coincided with the mean values in various age groups and both the sexes. Mean height was slightly more in girls than boys at various ages except at the extremes (10 and 15 years), however, the difference everywhere was statistically not significant.

Body weight also increased in boys as well as girls with the increase in age (Table II). In boys, mean weight increased from 25.0±4.3 kg (10 years) to 36.2±7.4 kg (15 years). Weight gain was more in later age groups. In girls also, similar trend was seen, whereby the weight increased from 26.1±4.8 kg at 10 years to 38.0±7.1 at 15 years. Mean as well as median weight were higher for girls (than boys) at all ages and the differences were statistically significant at different ages except 10 and 15 years.

#### Table III: Anthropometric profile of study population - BMI.

Boys | Percentile values | Girls | Percentile values | |||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|

Age (yrs.) | Total examined | Mean | SD | 5th | 50th | 95th | Total examined | Mean | SD | 5th | 50th | 95th |

10 | 100 | 13.9 | 1.8 | 11.5 | 13.7 | 17.5 | 127 | 15.0 | 3.8 | 12.0 | 14.3 | 18.3 |

11 | 305 | 14.1 | 1.5 | 11.8 | 14.1 | 16.6 | 388 | 15.3 | 3.2 | 12.2 | 14.8 | 19.3 |

12 | 324 | 14.3 | 1.7 | 12.1 | 14.2 | 17.4 | 370 | 15.1 | 2.8 | 12.5 | 14.8 | 18.9 |

13 | 190 | 14.8 | 2.2 | 12.7 | 14.6 | 18.3 | 166 | 15.9 | 2.0 | 13.2 | 15.6 | 19.8 |

14 | 125 | 14.7 | 1.8 | 12.1 | 14.6 | 17.5 | 76 | 16.4 | 2.3 | 12.7 | 16.2 | 20.3 |

15 | 48 | 15.3 | 2.4 | 12.0 | 15.2 | 18.7 | 31 | 17.2 | 3.1 | 14.5 | 17.0 | 22.6 |

BMI exhibited similar trends in both the sexes. Median values for BMI, like height and weight, coincided with the corresponding mean values of BMI.

#### Table IV: Correlation (r) and determination (r^{2}) coefficients between weight and different variables.

Variable | Age | Height | BMI | Outdoor play | ||||
---|---|---|---|---|---|---|---|---|

Group | r | r^{2} |
r | r^{2} |
r | r^{2} |
r | r^{2} |

All | 0.4204** | 0.1767 | .6267** | 0.3879 | .5796** | 0.3359 | -0.1396** | 0.0195 |

Male | 0.4778** | 0.2283 | .6956** | 0.4839 | .6922** | 0.4791 | -0.0550 | 0.0030 |

Female | 0.4233** | 0.1792 | .5980** | 0.3576 | .5152** | 0.2654 | -0.0896* | 0.0080 |

10 | - | - | .4747** | 0.2253 | .4661** | 0.2172 | -0.1481 | 0.0219 |

11 | - | - | .6284** | 0.3949 | .5526** | 0.3054 | -0.2074** | 0.0430 |

12 | - | - | .5971** | 0.3565 | .5309** | 0.2819 | -0.1306** | 0.0171 |

13 | - | - | .5974** | 0.3569 | .7317** | 0.5354 | -0.2307** | 0.0532 |

14 | - | - | .6633** | 0.4400 | .8338** | 0.6952 | -0.2328** | 0.0542 |

15 | - | - | .3286** | 0.1080 | .6867** | 0.4716 | -0.0316 | 0.0010 |

One tailed significance: p<0.01, **p>0.01

The values of simple correlation coefficient for weight and their statistical interpretation in relation of sex, age, height, BMI and hours spent per day on the outdoor playing are shown in Table IV. The coefficient of determination for weight in relation of sex, age, height, BMI and hours spent per day on the outdoor playing are shown in Table IV. The coefficient of determination for weight in relation to same variables are also shown there. Weight correlated positively and significantly with age, height and BMI in total population as well as in various sub-groups. However, the time spent per day on outdoor games exhibited a weak and negative correlation, which was also statistically significant in total population, girls and various age groups excluding 10 and 15 years of age. Coefficient of determination for different variables showed the amount of variation in weight, to the extent explained by the variation of that individual variable. Height and BMI showed maximum influence on weight and could explain 48.4 to 10.8 percent and 69.5 to 21.7 percent of weight variations respectively. BMI explained the maximum weight variation during 13-15 years. Even though the time spent on outdoor playing correlated significantly, the variation in weight explained by it was minimal (0.1-5.4%).

Linear regression equations constructed for predicting weight and height are as follows :

#### In Boys:

Height (cms) = 96.06+3.59* Age (yrs) (r^{2} = 0.27, F = 412.6)

Weight (kg.) = 3.54+2.05* Age (yrs) (r^{2} = 0.23, F = 321.9)

#### In Girls:

Height (Cms) = 100.28+3.34* Age (yrs) (r^{2} = 0.20, F = 292.45)

Weight(kg.) = 1.79+2.4* Age (yrs) ((r^{2} = 0.19, F = 266.9)

Multiple regression equations to predict variation in the weight on the basis of variations in age and height (for two sexes separately) were also constructed and presented below :

#### In Boys:

Weight (kg.) = -39.88 + 0.43* Age (years) + 0.45*
Height(cms) (r^{2} = 0.61, F = 837.3)

#### In Girls:

Weight (kg.) = -49.0+0.7* Age (yrs) + 0.51* Height (cms) (r^{2} = 0.56, F = 727.7)

#### Table V: Statistical significance of the difference in mean values between two sexes for different variables.

Age(yrs.) | Height | Weight | BMI | |||
---|---|---|---|---|---|---|

Z value | Interpretation | Z value | Interpretation | Z value | Interpretation | |

10 | 0.84 | NS | 1.71 | NS | 2.78 | HS |

11 | 1.90 | NS | 6.63 | HS | 6.26 | HS |

12 | 1.42 | NS | 4.13 | HS | 4.94 | HS |

13 | 1.81 | NS | 5.23 | HS | 4.91 | HS |

14 | 0.56 | NS | 4.34 | HS | 5.77 | HS |

15 | 0.89 | NS | 1.09 | NS | 2.90 | HS |

HS = Highly significant (Z>2.58; p<0.01), NS = Not significant (Z<1.96; p>0.05)

The mean values of BMI were significantly higher for girls than boys, in all age groups (Table V).

### Discussion:

Adolescents are persons aged 10-19 years^{2}. However, the present study covered only the young adolescents (10-15 years), as it focused on primary schools. Progressively declining female ration with age signifies the increasing drop-out of females from their study. The median height in boys as well as in girls in present study were largely comparable with the corresponding ICMR values^{7}. In fact, the median height in both the sexes was slightly higher in the present study in the 10 to 12 years of age, but, thereafter, it was less than the ICMR standards. The median heights in the present study (in different ages and sexes) were far lower than the median values of NCHS standards^{8} and were comparable only with the 5th percentile of the later.

Similar to the height, median body weights in both the sexes were marginally higher than the ICMR^{7} standards up to 11 years of age, but were lower thereafter. Median NCHS standards^{8} were very high when compared with the present study and here also median values were comparable only with the 5th percentile of NCHS standards.

The relationship between weight and height is best expressed as BMI (weight (Kg.)/height^{2} (meter). Mean BMI in the present study were higher in females than males in all age groups. Median BMI were far less than the 50th percentiles of NCHS standards and were even less than the 5th percentiles of NCHS standards in boys. To some extent BMI is age independent in the sub group of adolescents. It is worth comparing the overall BMI observed in the present study with BMIs from other countries.

Country | Boys | Girls | ||||
---|---|---|---|---|---|---|

Mean age | Mean BMI | Sample size | Mean age | Mean BMI | Sample size | |

Singapore | 13.7 | 16.5 | 1606 | 13.9 | 17.2 | 1499 |

Thailand | 13.4 | 16.6 | 1377 | 13.5 | 17.6 | 1394 |

Somalia | 14.3 | 16.6 | 389 | 14.2 | 18.3 | 304 |

USA | 14.5 | 20.6 | 1506 | 14.5 | 20.8 | 1381 |

India* | 12.1 | 14.4 | 1092 | 11.8 | 15.4 | 1158 |

Source: WHO^{12}. *Present study

It is evident from the above that the mean BMI in the present study are lowest in both sexes when compared to other countries including the country like Somalia which is lagging far behind us in development.

Significant positive correlation coefficients were seen for body weight with age, height and BMI separately. While BMI correlated more strongly with body weight in later age groups, the correlation between height and weight decreased in 15 years of age. It can be due to the fact that growth spurt is achieved by this age and any increase in height afterwards, is either minimal or nil. Various coefficients of determination showed the proportion of variation in weight which can be explained by the variation in the other variable. For example if coefficient of determination (r^{2}) of BMI for weight in total population is 0.34 (Table IV), it means that one third of variation in weight in the population can be attributed to the variation in the BMI. Outdoor playing activities were negatively correlated with the weight but could explain only a small 0.3-5.4%) variation of weight.

Regression equations for height and weight in boys and girls, according to the age, constructed in this study will be useful in predicting the same for a given age. Height and weight both are more age dependent in boys than girls as the r^{2} were more (0.23 and 0.27) in boys than in girls (0.19 and 0.20). Multiple regression equations to calculate weight for a given age, height and sex were more precise as the r^{2} were 0.61 and 0.56 for boys and girls respectively. In other words the variation in age and height in the present study could explain 61 and 56 percent variations in body weight in boys and girls respectively.

### Conclusions:

Present study provides the growth standards for the school based population (10-15 years age) of Surat city. Appraisal of nutritional status adjudged by the weight, height and BMI revealed that the median parameters of the population were comparable to the ICMR standards but were far below the 50th percentiles of NCHS standards. Girls in the present study exhibited better nutritional status in terms of "weight for age" and the BMI, than boys. Further, a little slowing down of growth was observed in the present study in both sexes (in terms of weight and height) after 13 or 14 years of age. While up to the age of 12 or 13 years our parameters were little higher than those of ICMR, thereafter, the ICMR standards were higher than the parameters of the present study.

Being a school based study, it has all the limitations of such study; i.e. children not attending these schools or absent on the day of examination were not covered.

#### Acknowledgment:

Authors are grateful to Shri KF Patel, then education officer at SMC, Principals, teachers and participants of the concerned schools for their co-operation.

### References:

- Adolescent growth in girls - The Indian perspective Editorial. Indian Pediatrics 1990; 27: 1 49-5.
- World Health Organization: Physical status - the use and interpretation of anthropometry. Report of the WHO Expert Committee. Technical Report Series No. 854, Geneva, World Health Organisation, 1995; 263-311.
- Khadgavat R, Dabadghao P, Mehrotra RN, Bhatia V: Growth
^{1}charts suitable for evaluation of India children. Indian Pediatrics 1998; 35: 859-65. - Singh BM, Vashisht S, Bachani D: Utilization of anthropometric parameters to monitor growth pattern of school going children. Indian J of Prev. & Soc. Med. 1992; 23(4): 129-32.
- Mandowara SL: IAP criteria for grading the nutritional status of Indian children need modification (letter to the editor). Indian Journal Pediatrics 1986; 53: 127-8.
- Qumra SR, Mehta S, Deodhar SD: Physical growth in school girls: Relationship to socio-economic status and dietary intake-II. Indian Pediatrics 1990; 27: 1051-65.
- Indian Council of Medical Research: Growth and physical development of Indian infants and children. TRS. No. 18; 1989. Delhi, Indian Council of Medical Research.
- Hamill PVV, Drizd TA, Johnson CL et al: Physical growth; National Centre for Health Statistics percentiles. American Journal of Clin. Nurt. 1979; 32: 609-10.
- Stone EJ, Mckenzie TL, Welk GJ, Booth ML: Effects of physical activity interventions in youth - Review and synthesis. American Journal of Prev. Medicine 1998; 15(4): 298-315.
- Dwyer T, Coonan WE, Leitch DR, Hetzel BS, Baghurst RA: An investigation of the effects of daily physical activity on the health of primary school students in South Australia. Int. J. Epidemiology 1983; 12: 308-13.
- Levin RI, Rubin DS: Simple Regression and Correlation in Statistics for Management (fifth edition). Editors: same, New Delhi, Printice Hall of India Pvt. Ltd. 1992: 478-593.
- World Health Organization: The health of youth - facts for action. Youth and Nutrition. A-42/Technical Discussions/8. World Health Organization, Geneva, 1989.