The 130th Annual Meeting of APHA

Abstract #48352

Howard Andrews, PhD¹, Carlos M. Jusino, MS², Robin Garfinkel, PhD³, Lori Hoepner, MPH³, Frederica P Perera, PhD⁴, and Virginia Rauh, ScD⁵. (1) Center for Child Environmental Health, Columbia University, 1051 Riverside Drive, Box 47, New York, NY 10032, 212-543-5585, hfa1@columbia.edu, (2) West Harlem Environmental Action, Inc., 271 West 125th Street, Suite 308, New York, NY 10027, (3) Data Coordinating Center, Columbia University Psychiatry Department, 1051 Riverside Drive, New York, NY 10032, (4) Center for Child Environmental Health, Columbia University, Mailman School of Public Health, 100 Haven Ave., New York, NY 10032, (5) Columbia Center for Children's Environmental Health (CCCEH) at the Joseph L. Mailman School of Public Health, 60 Haven Avenue, B-116, New York, NY 10032

Quantitative analyses of relationships among child health outcomes and community-level social and environmental indicators were first carried out over forty years ago. However, technical advances within the past 5 years have revolutionized the potential for assessing the impact of community social and environmental factors on individual health outcomes. In this presentation, we review these methodological advances, using research conducted at the Columbia Center for Child Environmental Health (CCCEH) to illustrate state-of-the-art multi-level modeling of child health and developmental outcomes. New technologies and information systems now readily available to the research community include: 1) GIS applications that can warehouse, analyze and display a variety of geographically-related information; 2) Accessible, computerized data characterizing the child’s social and physical environment from the time of conception through adulthood; 3) Computerized record linkage techniques that can greatly expand the scope of project data while ensuring confidentiality of research subjects; 4) Statistical software that makes it possible to: a) determine the relationship between higher-level environmental factors and individual-level outcomes and to assess interactions between environmental factors and individual-level risk, as well as b) appropriately address concerns about statistical independence of observations that plagued earlier attempts at multi-level analysis. In city-wide analyses, we found strong associations at the community level between childhood asthma rates, point sources of pollution (bus terminals), poverty and other negative indicators of community burden. Multi-level modeling of low birth weight indicates that community poverty is a significant predictor even when the effects of individual poverty and other individual-level risks are controlled.

Learning Objectives: At the conclusion of the session, the participant will be able to

• Understand 4 technological advances that facilitate multi-level modeling of asthma and other child health outcomes.
• Design a research plan and a data management model for conducting multi-level research.
• Understand basic concepts in multi-level statistical modeling, including main community-level effects, cross-level effects, 'slopes as outcomes' and independence of observations.

Keywords: Environmental Health, Asthma

Presenting author's disclosure statement:
Organization/institution whose products or services will be discussed: Columbia University School of Public Health Center for Child Environmental Health
I do not have any significant financial interest/arrangement or affiliation with any organization/institution whose products or services are being discussed in this session.