Childhood dietary habits are important because a food culture once adopted is apparently difficult to reverse(1). With obesity linked to an "obesogenic environment" in urban areas(2), knowledge of dietary patterns of our urban school children is important. Several cross-sectional studies in India indicate that the percentage of overweight children in cities is a matter of concern(3,4).

The aim of the study was to evaluate the dietary patterns and extent of overweight in urban school-going children in Bangalore.

Data was collected at baseline of a cross-sequential 3 year longitudinal study from a purposive sample of 307 children. Approximately 30 children were recruited at each year between 7 to 15 years from 3 schools representing different socio-economic status (SES) based on fee structure. Ethical approval was obtained from the institutional ethical review board. Permissions of the school principals, parental consent and participant assent were obtained.

General dietary habits were collected using an interviewer-administered questionnaire. Household assets were used as a surrogate measure of SES; derived by ranking each possession based on its value, and using the sum to obtain the final socio-economic ranking. The values obtained were divided into tertiles of SES for analysis.

A semi-quantitative food frequency questionnaire, for which validation is on-going, was used to assess food intakes. The macro-nutrient composition was calculated using standard portion sizes and a composite of the food composition tables of National Institute of Nutrition, ICMR, India and the United States Department of Agriculture, a method used earlier(5).

Height was measured to the nearest 0.1 cm and weight to the nearest 0.1 kg. The body mass index (BMI) was derived and overweight and obesity defined using Cole’s international cut-offs(6) (extrapolated to a BMI of 25 and 30 kg/m2 at age 18).

Data was analyzed using SPSS version 13.0. Nutrient and anthropometric data was analyzed using the general linear model adjusted for age and sex.

A total of 142 boys and 165 girls, (mean age 11.0 ± 2.5 years) were recruited; 73% were non-vegetarian. Of these, 7.2% of the children were overweight (7.7% boys and 6.7% girls) and 1% obese.

Daily energy and protein intakes were higher in boys than girls (1777 ± 452 kcal and 60 ± 15.6 g respectively for boys and 1588 ± 417 kcal and 52.7 ± 14.4 g respectively for girls). Daily fat intakes were significantly higher (P = 0.029) in boys than girls in the 7-9 yr (57.6 ± 17.2 g vs 50.7 ± 14.2 g) and 13-15 yr (54.8 ± 14.2 g vs 47.0 ± 18.7 g) age group while saturated fat intake was significantly higher (P = 0.016) in the 13-15 year old boys (20.4 g vs 16.8 g).

Fat and saturated fat expressed as a percentage of daily energy intake, increased with increasing level of SES (Table I), while percent energy from carbohydrates decreased with increase in SES, although overall energy intakes were not significantly different. The differences observed were significant between the 1st and 3rd tertile of SES. Children belonging to the 2nd tertile were significantly heavier than those in the first tertile of SES.

All values provided are means (± standard deviation).
*General linear model controlled for age and sex.

As age increased (especially in girls), step-wise regression indicated that the number of foods accounting for approximately 90% of variance in energy intake decreased, indicating that the variety of foods consumed decreased with increasing age. Contribution analysis for all age groups indicated that approximately 50% of energy intake was through milk, rice and dhal/ pulses. A little over 3% energy was from fruits and over 6% from sweets, chips and biscuits.

Further, increased frequency of eating out (Table II), resulted in increased daily energy intakes (P <0.001) while BMI did not (P = 0.573). with mean daily energy intakes lower in children in the lowest category of eating out compared to the other two groups. No significant correlation between the frequency of eating out and the SES was evident (r = 0.091, P = 0.113).

 *In a general linear model adjusting for age and sex, significant at p <0.001.

8.2% of children studied were overweight or obese, which is comparable to data obtained by Kapil, et al.(8). As SES increased the percent energy derived from fats and saturated fats increased, while that from carbohydrates decreased, similar to a study on children in Thailand(9). That frequency of eating out has an impact on energy intake of children is a major concern since the cumulative effects of an excess energy intake could in the long-term contribute to weight gain. Between 7 to 15 years of age, an excess of approximately 165 kcal (daily difference in energy intake between extreme groups of eating out) could lead to a theoretical 3 kg excess weight largely as fat per year. World-wide eating foods away from home have been implicated as a cause for increased energy intakes(10-13). In cross-sectional studies, increase in energy intake associated with increased frequency of eating out does not always translate into increases in BMI(11,12), although this could be more evident in longitudinal studies where the cumulative effects of small daily increases in energy intake can be documented(14).

With dietary patterns of urban children being influenced by their socio-economic status and the frequency of eating out, promoting healthy eating in schools in important.

We thank the principals of the three schools who permitted and cooperated with us to conduct the study, and all the children who participated in the study. We acknowledge Parimala R. for helping in data entry.

Contributors: SS was involved in conceptualization, data collection and analysis of the study and writing the first draft of the manuscript. She will act as the guarantor of the study. TT conducted the analysis and reviewed and approved the final manuscript. AVK interpreted the data, critically reviewed, and finally approved the manuscript. MV was involved in conceptualization of the study, interpretation of data and critically reviewed and edited the manuscript.

Funding: None.

Competing interests: None stated.