Association Between Body Mass Index and Suicidal Thoughts and Attempts Among US Adolescents

Abstract

s / Annals of Epidemiology 22 (2012) 661e680 674 than in children in other age group. There was no evidence of publication bias in our analysis; neither funnel plot nor Egger's test was significant. Conclusion: Short sleep duration is consistently associated with increased risk of development of obesity in children, but there is no consistent conclusion to identify if too long sleep duration do associate with increased risk of childhood obesity. P60-S. Association Between Body Mass Index and Suicidal Thoughts and Attempts Among US Adolescents D. Mowls, V.K. Cheruvu. College of Public Health, Kent State University, Kent, OH Purpose: The purpose of this study was to examine the association between feelings of sadness, suicidal thoughts, plans and attempts and body mass index (BMI) in a national sample of adolescents. Methods: Cross-sectional data for the current study were derived from the 2009 Youth Risk Behavior Surveillance System (n1⁄416410). Logistic regression models were used to model the likelihood of feelings of sadness, suicidal thoughts, plans and attempts in relation to BMI categories, adjusting for all potential confounders. Data were analyzed in 2012 and accounted for the complex sampling design of the YRBS to generalize the findings to the noninstitutionalized US population of adolescents in grades 9-12 in public and private schools. Results: The prevalence of underweight, healthy weight, overweight, and obese adolescents were 8.5%, 64.5%, 14.7%, and 11.1%, respectively. Underweight adolescents were more likely to have suicidal thoughts [Odds Ratio (OR): 1.3, 95% CI: 1.1 e 1.5]; suicidal plans [OR: 1.4, 95% CI: 1.2 e 1.6]; and suicidal attempts (two or more) [OR: 2.0, 95% CI: 1.6e2.6] when compared to healthy weight adolescents. In stratified analyses by gender, underweight males and obese females were more likely to have the outcomes when compared to healthy weights counterparts. Conclusion: Body mass index seems to be an important indicator of feelings of sadness, suicidal thoughts, plans and attempts among adolescents. These results provide new insights in understanding the association between categories of BMI and suicidal ideation. P61. Adolescent Fast Food Intake in Relation to Obesity and Chronic Disease Risk Factors M.A. Papas, K.J. Helzlsouer, L.E. Caulfield, A.J. Alberg, T.L. Gary-Webb. Department of Behavioral Health and Nutrition, University of Delaware, Newark, DE Purpose: Adolescent fast food consumption has been increasing rapidly and may contribute to obesity and chronic disease risk. The longitudinal association between adolescent fast food use, obesity, high blood pressure, and high cholesterol was investigated within a prospective cohort study. Methods: In 1989, 789 adolescents aged 12 to 18 years participated in CLUE II, an ongoing cohort study. Fast food restaurant use was collected by food frequency questionnaire; trained staff drew blood, took blood pressure, and measured weight and height. In 2003, 216 study participants aged 26 to 32 years completed follow-up questionnaires. Logistic regression was used to determine associations between adolescent fast food use and chronic disease risk factors at baseline and obesity in young adulthood. Results: Weekly adolescent fast food use was associated with adolescent obesity (OR1⁄41.2; 95% CI: 1.0, 1.3), high blood pressure (OR1⁄41.2; 95% CI: 1.0, 1.5), and high cholesterol (OR1⁄41.5; 95% CI: 1.0, 2.2). Increases in weekly fast food intake over the 2 time points were associated with obesity in young adulthood (OR1⁄41.2; 95% CI: 1.0, 1.5). Conclusion: Weekly consumption of fast food during adolescence was associated with obesity and chronic disease risks in adolescence and obesity in young adulthood. Adolescence may be a critical window for interventions that promote healthy diets to reduce the risk of chronic disease. P62. An Environment-Wide Association Study of Obesity M.A. Papas. Department of Behavioral Health and Nutrition, University of Delaware, Newark, DE Purpose: Obesity rates have been rising rapidly over the past twenty years. Poor diet and lack of exercise may not be the only factors responsible for this increase. Environmental exposures may alter metabolic processes and predispose some to gain weight. An environment-wide association study (EWAS) provides a useful tool for systematically evaluating multiple environmental factors potentially related to obesity. Methods: The association between more than 120 environmental factors and obesity was evaluated using data from the three most recent waves of the National Health and Nutrition Examination Survey (NHANES). Body mass index (BMI) was estimated for each adult participant with obesity defined as a BMI greater than 30. Environmental factors included all available clinical and laboratory measures in NHANES. Logistic regressionmodels with obesity as a dichotomous outcome were adjusted for age, sex, race, and socioeconomic status. Multiple comparisons were controlled using an a priori false discovery rate. Results: The study included over 6,500 adult participants and the prevalence of obesity was 35% across all three NHANES waves. Obesity was positively associated with certain heavy metals including barium (p < 0.01), cesium (p < 0.05), and thallium (p < 0.01). Conclusion: The EWAS method is a valuable initial strategy for evaluating multiple environmental factors potentially related to obesity. The analysis of three waves of NHANES participants confirm the results of recent investigations that identify potential associations between environmental exposure to heavy metals and obesity. P63. The Influence of Community Health Improvement Planning on Prioritizing Obesity Prevention in US Local Health Departments K.A. Stamatakis, R.C. Brownson. Division of Public Health Sciences, Washington University School of Medicine, St. Louis, MO Purpose: Obesity prevention may not be a priority programmatic area in local health departments (LHD) in many localities with high obesity prevalence. We examined whether developing integrated community health improvement plans (CHIP) moderated the likelihood that LHDs had obesity prevention programming in the highest prevalence localities. Methods: We conducted a descriptive, cross-sectional study by merging organizational data on LHDs from the 2005 National Profile of LHDs Study with county-level estimates of obesity prevalence from the Behavioral Risk Factor Surveillance System (n1⁄42,300). Multi-level logistic regression models were used to examine the moderating effect of CHIP on the relationship between obesity program implementation and local obesity prevalence; adjustment for other organizational characteristics and statelevel clustering was also assessed. Results: LHDs who developed CHIPs that were integrated with state health improvement plans were more likely to have obesity prevention programs in the highest prevalence counties (odds ratio [OR]1⁄42.0, 95% confidence interval [CI] 1.0-4.1); therewas no association among LHDs with no CHIP (p< .0001 for interaction). The relationship persisted after adjusting for organizational characteristics (e.g., size of service population). Conclusion: This study suggests that development of CHIPs may be a useful strategy for LHDs to improve prioritization of local obesity prevention efforts, particularly when integrated with state plans. P64-S. After MI: Perceived Ideal BMI andWeight Loss Behaviors in Women M.D. Zullo, S.M. Brady, J.T. Schaefer, V.K. Cheruvu. College of Public Health, Kent State University, Kent, OH Purpose: Research has demonstrated the importance of a healthy body mass index (BMI) to reduce risk after a cardiac event; however, research has not described what womenwith a history of myocardial infarction (MI) perceive as an ideal weight or behavioral methods used to reach that weight. Methods: Cross-sectional study using data from the 2001-2003 Behavioral Risk Factor Surveillance System (n1⁄41670). Body mass index (BMI) was categorized as normal, overweight, and obese. Logistic regression models with outcome “trying to lose weight” were run separately for white, nonHispanic (NH) and black-NH. Results: Mean BMI was 27.9 (standard deviation1⁄46.5) and 32.6(8.2), (p < 0.001) while reported ideal BMI was 24.0(3.2) and 26.7(4.2) (p < 0.001) for