ESPE Abstracts

Objective: To evaluate the severity and the correlates of depressive symptoms in adolescents with obesity comparing to normal adolescents in Hong Kong using a validated screening tools: Patient Health Questionnaire-9 (PHQ-9) for age 11-17 in Hong Kong.

Methods: In this prospective and retrospective cross-sectional study. We recruited 840 adolescents, including 306 adolescents with obesity from clinic and local schools. We also randomly selected age- and sex-matched controls for 1:1 comparison.

Results: Adolescent with obesity and normal control both had a mean age of 13.57 years, 69.6% male and 30.4% female participants. The obese group had a significantly higher mean BMI z-score of 2.15 compared to the normal control with a mean BMI z-score of 0.01 (P-value <0.001). The mean socioeconomic scores were 13.41 for the obese group and 12.13 for the normal control (P-value= 0.005). A positive correlation was found between BMI z-score and PHQ-9 score (r = 0.177, P-value= 0.002) in all obese subjects. ROC analysis on adolescent with obesity attending clinic showed an optimal cut-off of BMI z-score of ≥2.677 (equivalent to BMI percentile ≥99.6%) with a fairly good AUC value (0.79) and exhibiting a sensitivity of 57.1% and a specificity of 95.7%. The segmented regression analysis indicated no significant trend between PHQ-9 score and BMI z-score before the cut-off, but an increasing trend was observed after the cut-off (slope = 12.89, P-value= 0.008). Subgroup analysis by univariable logistic regression revealed that BMI percentile ≥99.6% and being female were significant risk factors for a higher likelihood of having a PHQ-9 score ≥10. (BMI≥99.6% group: OR 1.518, p-value 0.000; female: OR 1.156, p-value 0.001). Age, socioeconomic status and the total number of complications did not show a significant increase in risk. In the multiple logistic regression assessing the combined effects of the significant risk factors, only the group with BMI≥99.6% remained significant (OR: 1.459, P-value= 0.000), and no collinearity was detected between the risk factors based on the variance inflation.

Conclusion: Using the BMI ≥ 95 ^th percentile definition of obesity, our study did not demonstrate any increased risk of major depression compared to normal population. However, for adolescents with BMI ≥ 99.6^th percentile, the risk of major depression is statistically significantly higher. Therefore, we recommend routine PHQ-9 screening for all adolescents with BMI ≥ 99.6^th percentile.