Validation of Recalled Relative Adolescent Body Size with Photographs among a Sample of Young Non-Hispanic Black and White Women

Carlson, Nicole, Joseph J. Zilber School of Public Health, Lemke, Colleen, Mentor, Poster Presentation, Projects, School and College, Student, Velie, Ellen

Nicole Carlson and Colleen Lemke, “Validation of Recalled Relative Adolescent Body Size with Photographs among a Sample of Young Non-Hispanic Black and White Women”
Mentor: Ellen Velie, Public Health
Poster #32

Adolescence represents a critical period in obesity development, which is associated with negative health outcomes, including breast cancer. Adolescent body size is difficult to assess retrospectively, however, without historical health records. Data for this project is from the Young Women’s Health History Study (YWHHS), a population-based case-control study of breast cancer among non-Hispanic Black and White women under 50 years of age. Participants in YWHHS were interviewed and asked to rate their body size at age 12 and 18 years on a modified Stunkard scale, which consists of nine body figures ranging from underweight (1) to obese (9). Participants provided childhood photographs for those ages, which were rated by two evaluators using the same Stunkard scale. The objectives of this project are to evaluate: 1) the validity of self-reported relative adolescent body size at ages 12 and 18 years vs. rater’s assessment of body size based on photographs; 2) agreement between two evaluators of participant body size based on the childhood pictures; and 3) comparisons of rater and participant evaluations to recalled body mass index (BMI) at age 18 years. Agreement between participant’s and evaluators’ ratings, as well as comparison of the two evaluators’ ratings were examined using Cohen’s Kappa. Bland-Altman analysis was also used to assess bias. Results revealed that the comparison of the two evaluators’ ratings was slight at age 12 (kappa=0.16) and moderate at age 18 (kappa=0.44). Bland-Altman analysis revealed minimal bias. Additional analysis assessing agreement between the average of the two evaluators vs. participant self-reported body size and the average of the evaluator ratings vs. participant recalled BMI are in progress. Results of these analyses will inform our interpretation of studies evaluating the association between recalled adolescent body size and risk of breast cancer, as well as future studies of adolescent body size and adult health.