**3. Results**

The sociodemographic characteristics of the study sample are compared by BMI status in Table 1. Of the 633 participants, 54% are 11–14 years of age, 46% are between 15 and 18 years old, 52% are female, 22% are African American, 25% are Caucasian, 24% are US Latinos, and 29% are youth who live in Mexico. Thirty percent of youth have a healthy BMI, 30% are overweight, and 40% are obese. Thirty-seven percent of participants are from Seattle, WA, 35% are from Los Angeles, CA, and 29% are from Cuernavaca, Mexico. Chi-square tests were used to assess differences by weight status for each of the study variables, separately by country. There are no significant differences by country of residence in terms of sociodemographic characteristics for each of the three BMI categories, except for education level among the US participants. (Table 1)


**Table 1.** Sample characteristics by body mass index (BMI) categories and country (*n* = 633).

Sample sizes may not add up to marginal totals due to missing values. 1 Differences between proportions were performed using chi-square tests of homogeneity by weight status for each of the study variables, separately by country. Statistically significant results are in bold.

Tables 2 and 3 compare the mean scores for various psychosocial, behavioral, and socio-environmental variables, by country of residence and BMI status. Within the domain of psychosocial factors, overweight or obese youth in Mexico and the US are more likely to report being dissatisfied with their body image, to perceive themselves as overweight or obese, and to have lower weight-specific QOL scores than healthy-weight youth. However, overweight or obese youths in Mexico are not more likely to report more depressive symptoms than healthy-weight youths, unlike obese youths in the US, who are more likely to report depressive symptoms than healthy-weight youths (3.2 vs. 2.3, respectively). The presence of depressive symptoms is greater among healthy-weight, overweight, and obese youth in Mexico (3.1, 3.7, 3.9, respectively) than those in the US (2.1, 2.8, 3.2, respectively); and the weight-related QOL reported by overweight or obese youths in Mexico is lower than those in the US (65.1 and 52.9 vs. 78.1 and 67.0, respectively).

In terms of behavioral factors, obese youth in the US have lower healthy lifestyle priorities (3.0 vs. 3.1, respectively), are less physically active (1.6 vs. 1.8, respectively), and are less likely to eat breakfast (1.5, 1.6, respectively) than healthy-weight youth in the US. Overweight and obese participants in both countries are also significantly more likely to engage in weight control behaviors, such as exercise and restrained eating, as compared to healthy-weight youth. The only statistically significant socio-environmental factors reported by obese youths in both countries include being more likely to think that their parents are concerned about their weight and that their parents have a larger body size than healthy-weight youth. However, overweight youths in the US are not more likely to report that their parents are concerned about their weight, as compared to healthy weight youths (Tables 2 and 3). All the aforementioned results had a *p*-value of <0.05, which was considered statistically significant.



\* A higher score indicates a greater frequency or agreement; ¶ Reference category for comparisons between BMI groups; **ഝ** Differences between means were performed using t-tests; ᐫ Cuzick's trend test; statistically significant results are in bold.


**Table 3.** Comparison of various psychosocial, behavioral, and socio-environmental factors by BMI category among youths in the Unites States (*n* = 452).

\* A higher score indicates a greater frequency or agreement; ¶ Reference category for comparisons between BMI groups; **ഝ** Differences between means were performed using t-tests; ᐫ Cuzick's trend test; statistically significant results are in bold.

The standardized and adjusted odds ratios for various psychosocial, behavioral, and socio-environmental factors, by BMI status, among youths in Mexico and the US (controlling for age, sex, race/ethnicity, and country of residence) are reported in Table 4. Overweight and obese youth have significantly greater odds of reporting body image dissatisfaction (OR = 1.67, OR = 2.95), having depressive symptoms (OR = 1.08, OR = 1.12), perceiving themselves as overweight or obese, and having a lower weight-specific QOL (OR = 0.97, OR = 0.95), than healthy-weight youth. Obese youth in both countries also have significantly lower odds of having healthy lifestyle priorities (OR = 0.75), being physically active (OR = 0.79), consuming fast food (OR = 0.68), and eating breakfast (OR = 0.47), than healthy-weight youth. Overweight and obese youth are significantly more likely to engage in weight control behaviors (OR = 5.19, OR = 8.88), such as exercise (OR = 1.99, OR = 2.12), as well as eating less, fewer calories, and lower-fat food (OR = 2.15, OR = 2.32) than healthy-weight youth. In addition, overweight and obese youth have significantly greater odds of restrained eating behaviors (OR = 1.86, OR = 2.35) than healthy-weight youth. Both groups are also significantly more likely to perceive their parent as overweight or obese (OR = 1.49, OR = 1.71), and obese youth have significantly greater odds of reporting that their parents are very concerned about their weight (OR = 1.56), compared to healthy-weight youth. The standardized odds ratio results indicate that the following psychosocial, behavioral, and socio-environmental factors are most significantly associated with overweight and obesity: Perceived body shape (OR = 6.31, OR = 25.89), restrained eating behaviors (OR = 1.7, OR = 2.08), and perceived parent body shape (OR = 1.88, OR = 2.34), respectively. The standardized odds ratios of measures with scales that have a wider range, such as the CDI-S (0–20) and the YQOL-W (0–100), show a stronger association with overweight and obesity, than the non-standardized odds ratios. (Table 4) All the aforementioned results had a *p*-value of <0.05, which was considered statistically significant.


**Table 4.** Standardized and adjusted odds ratios for psychosocial, behavioral, and socio-environmental factors by BMI status, among youth in Mexico and the US (*n* = 633).

¶ Standardized odds ratios; **ഝ** Healthy is reference category for comparison between BMI groups; ∞ Adjusted for age, gender, race/ethnicity, and country; \* *p*-value ≤ 0.001; \*\* *p*-value < 0.05; significant results are in bold.

Table 5 presents the logistic regression results for the psychosocial, behavioral, and socio-environmental factors, stratified by sex. Some important differences are observed by sex. For example, overweight or obese boys are more likely to report dissatisfaction with their body image (OR = 1.81 and OR = 3.21, respectively) than girls (OR = 1.59 and OR = 2.78, respectively). However, the presence of depressive symptoms is significantly greater among overweight and obese females (OR = 1.14 and OR = 1.16, respectively) but not among males. Girls are also more likely to perceive themselves as overweight or obese and "feel fat" (OR = 8.91 and OR = 34.28, respectively) than boys (OR = 7.14 and OR = 32.28, respectively). Obese females are significantly less likely to be physically active (OR = 0.72) and eat breakfast than healthy-weight females (OR = 0.40), but this association was not found to be significant among males. Overweight or obese males are more likely to engage in weight control behaviors (OR = 13.77 and OR = 12.69, respectively) than obese females (OR = 8.02), especially exercise (OR = 2.67 and OR = 2.59, respectively) and eating less/few calories/low-fat foods (OR = 3.40 and OR = 2.93, respectively). However, obese girls are significantly more likely to consume diet pills, powders or liquids (OR = 9.59) than boys (Table 5). All the aforementioned results had a *p*-value of <0.05, which was considered statistically significant.


**Table 5.** Association between psychosocial, behavioral, and socio-environmental factors and overweight or obesity, by sex (*n* = 633).

**ഝ** Healthy is reference category for comparison between BMI groups; ∞ Adjusted for age, race/ethnicity, and country; \**p*-value≤0.001;\*\**p*-value<0.05;statisticallysignificant resultsareinbold.
