The present study analyzed the construct validity for the figure rating scale put forward by Childress et al.  for Brazilian adolescents from ten to nineteen years old. In this study, the nutritional status was assessed only in terms of normal or overweight because there were no underweight adolescents. The data suggest construct validity of the scale because the following hypotheses are confirmed: 1) positive correlation between Current Body Size and BMI z-score in both sexes; 2) greater degree of body dissatisfaction (in module) by BMI z-score value in girls.
The last hypothesis was not expected to be proven in boys. Data from scientific research show that underweight or overweight boys are at greater risk of body dissatisfaction [11, 12, 15]. The hypothesis was expected to be proven for girls since the trend to the slim ideal begins in girlhood and accompanies women into adult life. Several studies show that the ideal of the female body image, both for adolescents and for adults, is predominantly geared toward weight loss, associated with the ideal female esthetic of slimness [27–29].
The use of figure rating scales is recurrent among researchers studying issues of body image and body dissatisfaction in populations of children and adolescents. Data from studies show that from the age of eight children have reliable responses in identifying body shape. However, there is still a lack of data about the validity of figure rating scales , since those that exist suggest correlations with variables such as social comparison, social and family support, the presence of eating disorders and BMI.
Wertheim and coworkers  found a correlation of 0.69 between the body size identified by female adolescents and BMI. The scale used in that study is different from that used in the present study. However, both are similar with regard to the larger number of figures (9 and 8 respectively) and subtle gradations from one figure to the next, which provides a greater range of choice. The similar correlations also suggest that the scales probably measure the same phenomenon. Other figure rating scales deemed valid in the scientific literature have an 0.41  and 0.53  correlation with BMI. In Wal & Thelen's study  the correlation of BMI with body dissatisfaction in adolescent females was 0.42, similar to the present study (0.37).
The figure rating scale used in the present study was constructed by Childress et al , adapted from the figure rating scales produced by Stunkard & Sorensen . The adaptation aimed to enable the Stunkard scale to be used in evaluating body image and body dissatisfaction in children and adolescents. Interest in the Childress et al. scale  is due to the fact that the Stunkard and Sorensen figures  have been validated for the Brazilian adult population , and it was therefore important to validate an instrument with similar psychometric properties, at least regarding the construction of the instrument. In the validation study by Scagliusi et al.  the Spearman correlation between Current Body Size and BMI was 0.76, which indicates that the scale used for the Brazilian adult population and the scale adapted for use in Brazilian adolescents are possibly psychometrically equivalent.
The scientific community has been interested in the discussion on body image and body dimensions, above all body dissatisfaction, since the mid-twentieth century. Special importance has been given in this context to the issue of changes in body image over the several stages of a person's life, from childhood, through puberty and adolescence, to adult life. Aspects such as predictors, the consequences of body dissatisfaction, and an understanding of the processes that lead to changes in body image-related phenomena therefore lie within the scope of researchers interested in the topic .
The present study is therefore important in the discussion of changes in body image and body dissatisfaction over several life cycles in the Brazilian population, since it enables researchers to use an instrument for Brazilian adolescents that is psychometrically similar to the instrument that has already been validated for the adult population. The results of this study suggest that information obtained by the use of the Childress et al. scale  to evaluate body image in adolescents has considerable validity, and thus enables inferences to be drawn about the latent construct in question.
Although we described the reasons for choosing the scale adapted by adapted by Childress et al , it is also important to mention the existence of other relevant available tools well recognized in the literature such as the Veron-Guidry and Williamson's tool [34, 35]. These findings indicate construct validity for the Childress figure rating scale  when used in 10-19-year-old Brazilian adolescents.
The present study thus describes the first evidence of a figure rating scale for Brazilian adolescents. We may consider our findings as a tool able to perform an association between body image dissatisfaction and BMI by means of novel statistical approaches such as spline function . Further studies should be carried out to verify the validity of criterion and reliability of the Childress et al. figure rating scale  for Brazilian adolescents.