Overweight status of adolescent girls with polycystic ovary syndrome: body mass index as mediator of quality of life.
Trent M, Austin SB, Rich M, Gordon CM.
Ambul Pediatr. 2005 Mar-Apr;5(2):107-11
Division of General Pediatrics and Adolescent Medicine, Johns Hopkins School of Medicine, Baltimore, MD 21287, USA.
PURPOSE: Adolescent girls with polycystic ovary syndrome (PCOS) have significant disruption in health-related quality of life (HRQL) compared to their healthy peers, but it is unclear which aspects of the disorder have the greatest impact on these quality-of-life changes. Studies of adult women have indicated that weight status has a significant impact on subjective symptomatology; however, a similar effect of weight status on HRQL in adolescents with PCOS has not been established. This study evaluated the effect of body mass index (BMI) on quality-of-life disturbances in adolescent girls with PCOS.
CONCLUSION: This study demonstrates that BMI is a primary mediator in the relationship between PCOS and the health-related quality of life (HRQL) reductions experienced by girls with the disorder. Clinical interventions that effectively address weight issues in adolescents with PCOS may improve overall HRQL and other obesity-related clinical outcomes.
METHODS: Data were derived from the Adolescent PCOS Quality of Life Study, a cross-sectional study of healthy (n = 186) and PCOS-affected adolescent girls (n = 97). The study was conducted at an urban, hospital-based adolescent medicine clinical practice. Medical staff obtained anthropomorphic measurements. Participants in both groups completed the Child Health Questionnaire-Version CF-87 as a measure of HRQL. BMI was tested as a mediator of HRQL using bivariate analyses and multivariate linear regression models.
RESULTS: The average BMI was 31.7 kg/m(2) (standard deviation [SD] = 8.4) for adolescents with PCOS and 23.5 kg/m(2) (SD = 4.2) for healthy adolescents. BMI was associated with PCOS status and HRQL. Girls with PCOS scored lower on the general health perceptions scale, physical functioning scale, family activities scale, and the general behavior scale and higher on the change in health in the last year subscale. When BMI was added to the multivariate linear regression models, coefficients were reduced on average by 3 points and became nonsignificant, indicating that the HRQL differences are mediated by the higher BMI.