The present study showed that malnutrition, whether undernutrition or micronutrient deficiencies, was highly prevalent at school age in urban areas. Almost 60% of the children examined had at least one sign of malnutrition and roughly 15% had at least two such indicators. Of particular concern, more than 40% of the subjects were anaemic and roughly the same percentage were vitamin A deficient. We verified that there was no significant difference in the prevalence of micronutrient malnutrition in the retained subjects compared to the 135 children excluded because their birth date was unknown (p = 0.133 and p = 0.183 for VAD and anaemia, respectively: data not shown). These high rates in children (mean age 11.5 ± 1.2 years) are similar to those of the national study of schoolchildren in Burkina Faso, whose mean age was 9.7 ± 5.8 years . In this study where rural schoolchildren were predominant, 40.5% were vitamin A deficient and 43.7% were anaemic. Similarly, in the baseline study of the red palm oil project in selected primary schools of two zones of Burkina Faso (out of the Central region, where Ouagadougou is located), more than 40% of the children were vitamin A deficient . A high prevalence of micronutrient malnutrition at school age is not uncommon in developing countries . In northern Ethiopia, the prevalence of VAD was 51.1% in a study conducted in 1997 in 824 pupils aged 6-9 years . In a report on six African and two Asian countries, 40.2% of children aged 7-11 years and 54.4% of those aged 12-14 years were anaemic . A similar increasing trend of anaemia with age is observed in the current study (table 3).
We also found that 13.7% of the children were thin, which is higher than the 8% prevalence previously reported for Burkina Faso schoolchildren outside the capital city of Ouagadougou . Thinness, or wasting, usually describes acute malnutrition. Our study was conducted between late 2008 and early 2009, that is, during the global economic and food crisis that hit developing countries so hard  and which was responsible for reduced access to food particularly among vulnerable populations . We observed that several schoolchildren stayed at school during lunchtime, but did not have pocket money to buy any street food, or did not have enough to eat an adequate meal. This may have played a role in the observed prevalence of thinness in schoolchildren of Ouagadougou. Notwithstanding, this level is far lower than that reported by the Partnership for Child Development (PCD) in schoolchildren of developing countries ten years ago .
The prevalence of stunting in our study was lower than in the recent national study of schoolchildren in Burkina Faso (8.8% vs 12%) , as well as in a nationwide survey in Chad (18.7%) in a sample of schoolchildren aged between 6 and 15 years . Stunting is an indicator of chronic malnutrition, and at school age, it may reflect malnutrition during the first years of life . Growth deficit tends to accumulate with age and particularly in boys, as observed in our study and in other studies of school-children in developing countries . The higher rate of stunting among older children, depicting an increasing vulnerability with age may also reflect some improvement of food and health conditions over recent years since most of the growth deficit or catch-up takes place before the age of 24 months . Except for overweight/obesity and anaemia, a higher proportion of boys than girls showed signs of malnutrition, as previously reported for stunting and wasting , and for VAD . A metanalysis of data from 16 demographic and health surveys conducted in 10 sub-Saharan countries  revealed that boys were more stunted than girls, and speculated on the role of cultural factors or natural selection .
Although malnutrition still appears as a priority problem, overweight/obesity should not be overlooked right at school age, as we detected a higher prevalence trend in the youngest group of children. At variance with our study, a much lower prevalence of obesity was reported in 2001 (0.26% vs 0.60%) in an adolescent population of Ouagadougou (mean age 13.8 y) . However, both studies are consistent in the significantly higher prevalence of overweight observed in girls compared with boys (Figure 2).
It was clearly apparent in our study that private school-children enjoyed a better nutritional status than those attending public schools, with anaemia and VAD significantly higher in the latter (30% vs 45% and 6% vs 53% respectively). However, it is of note that overweight/obesity was also significantly higher in private than public schools, which is in accordance with previous reports in other developing country schoolchildren . Socio-economic disparities likely underlie these differences . We did not examine the socio-economic conditions of the individual children, but mere differences in school registration fees are convincing: US $ 60 in private schools compared with only US $ 4 in public schools. Nevertheless, it is surprising that thinness was as common in private as in public schools in our study (13.0% and 14.0%, respectively). There is no obvious explanation for this high rate of thinness even in private school children.
As could be expected, stunting and thinness were significantly higher in peri-urban than urban schools, and VAD also tended to be higher in the former than latter schools (table 4). Poverty and low maternal education are among the determinants of child malnutrition . It is also known that the prevalence of malnutrition is higher in rural than urban areas, particularly stunting , which reflects poor socio-economic status as the community level. It is therefore not surprising to find a higher percentage of malnourished children in peri-urban areas, where people are poorer, and where schools also draw their pupils from the surrounding villages. Anaemia was observed in roughly the same proportion of urban and peri-urban schoolchildren (around 40%). This confirms that anaemia is the most widespread malnutrition problem in schoolchildren in developing countries .
While iron deficiency is the main factor of anaemia , it is not the only one, and infection plays a major role [15, 51], notably malaria and hookworms in African school-children . In Ouagadougou, for instance, the prevalence of malaria (41.4%) tended to be the highest in children aged 5-14 years, who were also at the highest risk of infection compared to infants and adults . Other micronutrient deficiencies may also be involved in the aetiology of anaemia . The "top three" micronutrient deficiencies are iron deficiency, VAD and Iodine Deficiency Disorders (IDD) . The high prevalence of anaemia could be a great threat for school-children, particularly since it was combined with VAD in one out of five children (20.2%) in our study. Indeed iron deficiency and VAD are interrelated [54–56]. In contrast, we detected no goitre using the palpation method recommended by WHO , which likely reflects the effectiveness of the salt iodization strategy of the past several decades .
To our knowledge, this nutrition study is the first of its kind among city schoolchildren of West Africa. Although the schools were not randomly selected, they represent a broad array of features: public and private, confessional and non-confessional, as well as urban and peri-urban schools. Furthermore, sample size was large enough and in a narrow age-range. However, because of these study features, the results cannot be extrapolated.
School nutrition and feeding programs are usually directed at rural areas . Furthermore, under-five children are the priority target group for strategies and actions to fight malnutrition. There is an urgent need to address nutrition problems among schoolchildren in developing countries, without neglecting urban areas, considering that malnutrition can impair their performance while in school and their productivity later on in life .