To assess the trends in overweight and obesity in pre-school children in urban areas of Ho Chi Minh City (HCMC), Vietnam, over the period 2002 to 2005.

Two cross-sectional studies were conducted in 2002 and 2005. Multistage cluster sampling was used in both surveys to select the subjects. Sociodemographic information was collected using a self-administered questionnaire given to parents in 2002 and using an interview-administered questionnaire to parents in 2005. Weight and height were measured using the same standard methods in both surveys. BMI (kg/m2) was calculated and overweight/obesity was defined using the age- and sex-specific BMI cut-off points proposed by the International Obesity Taskforce.

Children aged 4 to 5 years, attending pre-schools in urban areas of HCMC, Vietnam, in 2002 (n 492) and 2005 (n 670).

The prevalence of overweight and obesity almost doubled from 2002 to 2005 (21·4 % and 36·8 %, respectively). The increase was more evident in less wealthy districts than in wealthy districts. The proportion of boys classified as obese in 2005 (22·5 %) was three times that in 2002 (6·9 %).

The prevalence of overweight and obesity has increased rapidly in children aged 4 to 5 years in urban areas, and especially in less wealthy districts, over a 3-year period. These results signal an urgent need for prevention programmes to control and reverse this rapid upward trend in overweight and obesity in young children in HCMC.

To evaluate the effectiveness of a school nutrition programme on the weight gain and growth of Vietnamese schoolchildren.

A proximate cluster evaluation of children in seven schools, in which fortified milk and biscuits supplying 300 kcal of energy were being given on school days, compared with children in 14 nearby schools with no feeding. All children were dewormed.

Twenty-one primary schools in Dong Thap Province, Vietnam.

A cohort of 1080 children in grade 1 of 21 primary schools, and a cross-sectional interview of 400 children in grade 3.

The programme gave children the equivalent of 90 kcal day− 1 over 17 months. t-Tests showed a small but statistically significant difference between groups in their average gain in weight and height: 3.19 versus 2.95 kg (P < 0.001) and 8.15 versus 7.88 cm (P = 0.008). A multiple-level model showed that the programme was statistically significant after controlling for clustering of children in schools, sex, age and initial underweight (P = 0.024). A significant impact on height was also seen in a regression model, but not when controlling for school. The most undernourished children tended to gain the least weight. There was no evidence of substitution.

The programme had a small but significant effect on weight gain, but the most undernourished children benefited the least. Methods need to be developed to target them. This design may offer a means of estimating the impact of school feeding on growth in other programme settings.