Assessment of Obesity among Adolescents from an Urban Area of Western India

Introduction

India is witnessing a double epidemic of communicable as well non-communicable diseases in the current century. Youth represents a substantial proportion of Indian population and the eating habits of the youth today will determine the health of the nation in the future. Obesity and sedentary habits are on the rise in urban areas because of a huge penetration of modern technology that makes people more and more inert and inactive.

The National Family Health Survey (2015-16) district data showed that prevalence of obesity was 26.8% and 39.8% among men and women aged 15-49 years, respectively.⁽¹⁾ The currently available Indian growth reference curves are based on the data collected by Agarwal, et al.⁽²⁾⁽³⁾ in 1989 which were published in 1992 and 1994 and were then adopted by the Indian Academy of Pediatrics for growth monitoring in 2007⁽⁴⁾. The IAP standards were revised by Khadilkar et al⁽⁵⁾ in his 2007 Multi-centric Study among affluent Indian children residing in all 5 zones of India. The WHO has been persuading paediatricians and governments all over the world to use the WHO growth charts for identifying underweight and overweight.⁽⁶⁾In February 2007, the Ministry of Women and Child Development (nodal ministry for the Integrated Child Development Services [ICDS] program) and the Ministry of Health and Family Welfare in India changed over from the IAP growth curve to the WHO child growth curves⁽⁷⁾.Hence, the current study used the WHO and CDC reference charts for comparison, rather than the IAP charts.

Studies on adolescent obesity are generally not comparable because of different standards used for categorization. National Surveys like NFHS, Census etc. generally do not focus on adolescents and throw light on the burden only in the adult population. Hence, a need was felt to use internationally validated WHO 2007 BMI-for-age standards on a sample of urban adolescents residing in Western India.

Further, it was noted that previous studies for comparing growth standards are lacking in an Indian sample.

The current study is an effort to document the prevalence of obesity in a sample of urban adolescents residing in a city in Western India and to compare their BMI for age with WHO 2007 and CDC 2000 growth charts for adolescents.

Results

Out of 300 study participants, 157 (52.3%) were males, 199 (66.3%) were from age groups 18-19 years, 283 (94.3%) were Hindus and around 50% were living with biological parents (Table 1).

Table 1- Socio-Demographic Profile of the study participants n=300

Gender	Total (n=300)
Male	157 (52.3)
Female	143 (47.6)
Age (in years)
15	1 (0.3)
16	27 (9.0)
17	73 (24.3)
18	127 (42.3)
19	72 (24.0)
Religion
Hindu	283 (94.3)
Others	17(5.6)
Language
Gujarati	250 (83.3)
Hindi	30 (10.0)
Others	20 (6.6)
Parents' Highest Educational Qualification
Primary	109 (36.3)
Secon dary	109 (36.3)
Pre-University	37 (12.3)
Gradu ate	24 (8.0)
Post Graduate	21 (7.0)
Current living status
With biological parents	146 (48.6)
Residential facility (Rent/Hostel etc.)	141 (47.0)
With biological m other/father	7(2.3)
With adoptive parents	6(2.0)

*Percentages may not add to 100% because of rounding.

Table 2 shows that the mean BMI of males was higher than their female counterparts. Similarly, BMI was found to increase with age and rising parental education. The mean BMI of adolescents having Post-Graduate parents was 21.78 kg/m², whereas the mean BMI was 20.46 kg/m² for adolescents with Primary educated parents. Parental education is an important proxy indicator of socio-economic status (SES), thereby indicating that BMI increases with rising SES. The mean BMI in the different categories of parental education were compared using one-way Analysis of Variance (ANOVA), but the difference was not statistically significant (p value > 0.05).

Table 2- Comparison of mean BMI on the basis of socio-demographic variables

Socio-demographic Variables	BMI (in kg/m2)		tstat	p value
	Mean	SD
Gender
Male	21.71	3.54	1.94	.053
Female	20.83	4.33
Age (in years)
16	20.39	3.71	1.825	.143
17	20.91	3.95
18	21.89	4.23
19	20.98	3.42
Language
Guja rati	21.21	3.93	-1.079	.281
Others	22.3	3.96
Parental Education*
Primary	20.46	2.89	1.076	.368
Secondary	20.31	3.37
Pre-University	21.48	3.40
Graduate	21.12	4.04
Post Graduate	21.78	4.30

* This was categorised into five categories; Primary (upto class 8^th), Secondary (upto class 10^th), Pre-university (upto class 12^th), Graduate and Post-graduate.

The BMI-for-age of the study participants was compared with the WHO 2007 BMI-for-age growth charts (Table 3). It was found that 10.3% of the adolescents were overweight, 4.3% were obese and 9.0% were underweight. No significant association was found between gender and any of the BMI- for-age categories according to WHO 2007 criteria. This indicates that nearly equal proportions of males and females were present in each of the BMI categories.

Table 3- BMI of study participants compared to the WHO Growth Chart BMI-for-Age 2007

Interpretation	Males	Females	Total	OR	χ2 value	p value
Normal	125 (79.6)	104 (72.7)	229 ( 76.3)	1.463	2.335	.505
Overweight: >+1SD	15 (9.5)	16 (11.2)	31 (10.3)	0.839
Obesity: >+2SD	5 (3.2)	8 (5.6)	13 (4.3)	0.556
Thinness: <-2SD	12 (7.6)	15 (10.5)	27 (9.0)	0.707
Total	157	143	300

The BMI-for-age of the study participants was also compared with the CDC 2000 BMI-for-age growth charts (Table 4). Using WHO classification above, the prevalence of underweight was 9% which increased to 22.6% using CDC 2000 growth charts. The finding of no significant association between gender and any of the BMI-for-age categories demonstrated earlier (Table 3) was also replicated by using CDC 2000 criteria.

Table 4- BMI of participants compared to the CDC Growth Chart BMI-for-Age 2000

Interpretation of cut-offs	Males	Females	Total	OR	χ2 value	p value
Normal-5th to <85th	109 (69.4)	90 (62.9)	199 (66.3)	1.411	1.956	.581
Overweight: 85th to 95th	10 (6.4)	13 (9.1)	23 (7.6)	0.783
Obesity: =95th	4 (2.5)	6 (4.2)	10 (3.3)	0.223
Thinness: <5th	34 (21.7)	34 (23.7)	68 (22.6)	.1919
Total	157	143	300

As shown in Figure 1 the proportion of normal BMI identified by WHO standards was higher than that identified by CDC standards and this difference was found to be statistically significant. Similarly, the proportion of underweight identified by the CDC charts was significantly higher than that of WHO charts.

Close

Figure 1-

Comparison of CDC 2000 and WHO 2007 BMI-for-age charts in study participants

Export to PPT

Discussion

Obesity and its determinants are well identified by a plethora of scientific literature, beginning from the findings of the Framingham Heart Study in the 1960s. A systematic review done in 2016 in India reported that overweight and obesity rates in children and adolescents are increasing not just among the higher socio-economic groups but also in the lower income groups where underweight still remains a major concern⁽⁹⁾. A similar study done by Alok et al from Western India in 2012 reported the prevalence of overweight 25.8% and obesity 12.8%⁽⁸⁾. This is much higher than the 10.3% overweight and 4.3% obesity prevalence in the current study. The difference might be because of difference in standards used: IOTF (International Obesity Task Force)- Cole 2000⁽¹⁰⁾ versus WHO 2007. A study done by Gupta et al⁽¹¹⁾ in 2013 from Bankura, West Bengal using WHO standards reported prevalence of overweight 7.7% and obesity 4.0%. The findings are comparable to the current study. However, Gupta et al reported higher prevalence in males than females, whereas the current study identified higher prevalence in females. In the 2013 study done by Moitra et al ⁽¹²⁾ in the same setting, overall prevalence of overweight and obesity was 10.3% and 5.9%, respectively and the prevalence was higher in boys in comparison to the girls. Although this sex differential was not found by current study, the prevalence estimates are somewhat similar as both the studies used WHO standards.

Another study from South India⁽¹³⁾ reported that the prevalence of overweight and obesity in the study population was 9.7% and 4.3%, respectively but there was no difference in prevalence between males and females. Kotian et al (2010) reported that the prevalence of overweight among adolescents varies between 10 and 30%⁽¹⁴⁾. Regional variations in patterns of obesity are also documented with Northern and Eastern India reporting higher prevalence in comparison to Western and Southern India.⁽⁹⁾

Current study shows that BMI increases as the parental education rises. This finding is similar to previous studies done in developing countries, but is reverse to the findings from developed countries. Goyal et al in their study from Ahmedabad⁽¹⁵⁾used family income as a proxy of SES and reported that prevalence of overweight among children was higher in middle SES as compared to higher SES group in both the gender. Hence, the relation between SES and obesity is not linear as conventionally understood. In line with previous research, socio-economic status was assessed by using the proxy of parental education level⁽¹⁶⁾. We could not demonstrate a significant relation between BMI and parental education in our study, as education might not be perfect proxy of socio-economic status in Indian set-up.

For any given population, WHO-BMI z-scores associates with increased percentage of obese and reduced percentage of under-weight when compared to CDC-BMI z-scores⁽¹⁷⁾. This finding was also demonstrated by the current study. It was also found that the prevalence of underweight was 5.9% by WHO standards which increased to 19.8% using CDC 2000 growth charts. This is because both criteria have different base populations⁽¹⁸⁾. WHO charts are more representative of growth patterns in our study population because it is based on a much more diverse and representative sample.

Conclusion and Recommendations:

As demonstrated by the current study, overweight and obesity is equally prevalent in both the genders and in all SES groups. Early Interventions need to be directed at adolescent and younger population irrespective of gender or SES in order to prevent the epidemic of non-communicable diseases from taking a serious magnitude in the future. Further studies should be done to collect more nationally comparable data on overweight and obesity using uniform criteria and standards. In the absence of a nationally validated growth standard, the WHO 2007 standards can be used as the common yardstick for comparing obesity statistics throughout the country.