AIM: The study examined the association between perceived body size (through self-selection of Stunkard's body shape silhouettes) and measured body mass index (BMI) in Mexican Mestiso adult men and women, and determined the BMI values that corresponded to each silhouette. SUBJECTS: A sample of 1092 men and 1247 women, apparently healthy, aged 20-69, was studied. METHODS: Participants were asked to identify the silhouette that most accurately represented his/her body size. Weight and height were measured in duplicate to calculate BMI. A simple linear regression analysis was used to adjust BMI values to silhouette categories by gender. Areas under the receiver operating characteristic curves (AUC) were calculated to assess accuracy. RESULTS: Gender-specific BMI values for each silhouette were obtained. Correlation coefficients between silhouette ratings and BMI were 0.702 in males and 0.766 in females. Overweight ranged from silhouette 4 to 6 and obesity from silhouette 6 to 9, in both men and women. In the regression models, 49.3% of BMI variance in males and 58.7% in females were explained statistically by silhouette self-selection. AUC were higher than 0.8 for overweight and obesity in males and females. CONCLUSIONS: Both male and female models had a good fit, indicating that BMI can be associated with body shape silhouettes. Silhouettes can be useful for defining overweight and obesity in settings such as schools or worksites, or where scales and stadiometers might not be available. They can also be used to provide health advice to the public.
AIM: The study examined the association between perceived body size (through self-selection of Stunkard's body shape silhouettes) and measured body mass index (BMI) in Mexican Mestiso adult men and women, and determined the BMI values that corresponded to each silhouette. SUBJECTS: A sample of 1092 men and 1247 women, apparently healthy, aged 20-69, was studied. METHODS:Participants were asked to identify the silhouette that most accurately represented his/her body size. Weight and height were measured in duplicate to calculate BMI. A simple linear regression analysis was used to adjust BMI values to silhouette categories by gender. Areas under the receiver operating characteristic curves (AUC) were calculated to assess accuracy. RESULTS: Gender-specific BMI values for each silhouette were obtained. Correlation coefficients between silhouette ratings and BMI were 0.702 in males and 0.766 in females. Overweight ranged from silhouette 4 to 6 and obesity from silhouette 6 to 9, in both men and women. In the regression models, 49.3% of BMI variance in males and 58.7% in females were explained statistically by silhouette self-selection. AUC were higher than 0.8 for overweight and obesity in males and females. CONCLUSIONS: Both male and female models had a good fit, indicating that BMI can be associated with body shape silhouettes. Silhouettes can be useful for defining overweight and obesity in settings such as schools or worksites, or where scales and stadiometers might not be available. They can also be used to provide health advice to the public.
Authors: Jennifer S Mindell; Alison Moody; Andres I Vecino-Ortiz; Tania Alfaro; Patricia Frenz; Shaun Scholes; Silvia A Gonzalez; Paula Margozzini; Cesar de Oliveira; Luz Maria Sanchez Romero; Andres Alvarado; Sebastián Cabrera; Olga L Sarmiento; Camilo A Triana; Simón Barquera Journal: Am J Epidemiol Date: 2017-09-15 Impact factor: 4.897
Authors: Catherine G Sutcliffe; Kathleen Schultz; Julitta M Brannock; Francis M Giardiello; Elizabeth A Platz Journal: Cancer Causes Control Date: 2014-11-07 Impact factor: 2.506