The major aim of the current study was to investigate thinness in a representative sample of Australian school children and our findings suggest that approximately 4% of the sample of 8550 Australian schoolchildren was thin. This finding is identical to a similar national study  conducted in 2007 confirming the results reported in this paper. The thin students were distributed among the three SES levels. In contrast, the obese children were predominately from low SES school communities, as shown previously [18–21].
In European countries, the prevalence of thinness has been reported to range from 4.8% to 11.9% in girls, and 3.1% to 9% in boys [1, 3, 22–25] but the prevalence of thinness in these countries differs by gender, age of children and region. The prevalence of thinness has decreased in some countries where under-nutrition has been a problem [24, 25], but has increased in regions where there is a general desire for thinness. In some regions the rate of thinness in girls is concerning because it exceeds the rate of obesity [1, 3, 25]. Our study differs from those reported above in that the children were sampled from a wide diversity of regions in Australia and included children aged from 6 to 18 years.
Thin children displayed smaller stature than normal weight children. Other studies have shown that height is related to body weight and that obese children may be taller than normal weight children [4, 5, 8, 26–28]. Thin children tend to be characterized by not only a lower percentage of body fat, but also by lower muscle mass and lower lean body mass . Thinness due to under nutrition may result from food fussiness, increased satiety responsiveness  or fear of fatness leading to disordered eating. Bulik and Allison  present arguments for the genetic basis of thinness in humans and animals.
The results of the current study highlight several areas of concern regarding thin children. First, 12% of thin girls and 7% of thin boys said that they wanted to be a little or a lot lighter and these students were also dieting to lose weight. Some thin children were told that they "should lose weight" or that they "eat too much." The role of parental comments is important for influencing disordered eating and body dissatisfaction [31–33] and this may be an explanation for our current findings. Interestingly, whilst the questionnaire used in the study was not designed to diagnose clinical eating disorders, the findings may serve this purpose.
In one study of 14-year-olds, Allen et al.  estimated that 6% of the sample met full or partial criteria for a DSM-IV eating disorder and an additional 3% were at risk for developing an eating disorder. Being female and being perceived as overweight by the child's parents were the strongest predictors for eating problems in this study. This is of particular concern because dieting, body image dissatisfaction and teasing are all known predictors of eating disorders . On the other hand, 41.3% of thin students in this survey reported that parents and peers told them that they "did not eat enough." Thin students in the current study were much less concerned than their peers about the need to "get fitter" or to "do more exercise," with only 20.2% being advised to do so. The findings of the current study warrant further investigation using a longitudinal study design in order to further clarify if some thin children deliberately engage in weight loss behaviours that may be triggered by parental advice.
The health benefits of physical activity for children are supported by many studies [35, 36]. There is no evidence that thin children exercise more or are more fit than other children, and a recent study  showed that thin adolescents are less active than their normal weight peers. In contrast, participation in physical activity would be particularly important for thin children to overcome potential problems related to their low muscle mass and decreased muscle strength  and sport and physical activity are known to be important in the development of child and adolescent self-esteem  and positive adult body image . Parental encouragement of physical activity is necessary for thin children, as children are more likely to have higher levels of physical activity when their parents and others encourage and support their participation [34, 39, 40].
Exercise in childhood is predictive of adult levels of exercise . Low physical activity and poor cardiovascular fitness are independent predictors of mortality related to type 2 diabetes and chronic disease in general . In fact, thinness is related to increased mortality and morbidity compared to normal and overweight (but not obese) adults [12, 13]. The negative attitudes of thin students concerning exercise identified in the current study, if continued into adolescence and adulthood, do not bode well for a healthy population.
Finally, the findings of this study suggest that thin children may be at risk of body image problems, as only 57.4% of thin students considered themselves to be "about right" compared with 83.1% of normal weight students and 63.7% of overweight students. Children can develop body image concerns at very young ages . Underweight as well as overweight youth reported higher levels of teasing than normal weight youth, and both groups face stereotypical attitudes and prejudices . Weight teasing has been significantly associated with disordered eating behaviours in a large study from the USA  and thin children were found in another US study to dislike school and consider themselves to be poor students . Thus, thin children (as well as children of all sizes) need to be considered and included in health promotion initiatives for self-esteem and positive body image.
This study has strengths in its large, generalized sample size, high participation rate, and accurately measured height and weight data, but has limitations in that there are no direct measures of caloric consumption or physical activity. Further research using qualitative interviews may better explain some of the findings of the current study. Of particular interest would be the attitudes of children and parents from various cultural and ethnic groups. Longitudinal studies could also further explore these issues over time, encompassing puberty and such studies could also incorporate actual health outcomes.