The purpose of the proposed research is to carry out a study on a probability sample of midlife adults (N = 1,000 divided equally by gender; age range:30-79) from the Tokyo metropolitan area. Respondents will complete a 50-page questionnaire that obtains information about sociodemographic characteristics (age, gender, marital status, educational status, income), psychosocial characteristics (e.g., personality traits, sense of control, independence/interdependence, social support, family obligation, etc.), and health (e.g., health behaviors, mental health: depression, anxiety, well-being; physical health: chronic conditions, health symptoms, functional limitations). These measures parallel those obtained in a national longitudinal sample of midlife Americans. Thus, a central objective is to compare the Japanese study (MIDJA) with the U.S. study (MIDUS) to test the hypothesis that well-being and health in Japan are predicted by an interdependent model of self, whereas in the U.S. health and well-being are predicted by an independent model of self. Nonetheless, based on pilot findings that compare Japanese and U.S respondents, we also predict cultural similarities in various age profiles of well-being (e.g., declining scores with age on purpose in life). A second major objective of the proposed study is to collect biomarkers on approximately half of the Japanese sample. These will include assessments of neurendocrine regulation, immune function, and cardiovascular risk. Parallel biological data collection is in progress via Project 4 of the MIDUS II P01. Thus, in both cultures we will examine linkages between psychosocial factors and biology to test the hypothesis that interdependent measures of self are more strongly linked with biomarkers in Japan, while independent measures of self are more strongly linked with biomarkers in the U.S. We also predict cultural similarities in age and gender profiles on the various biomarkers. The overarching goal is to combine these various domains of assessment to carry out preliminary cross-sectional analyses of integrative pathways (combining sociodemographic, psychosocial, and biological information) to health. Recursive partitioning is elaborated a methodology for testing the integrative hypotheses detailed in the proposal.