A Confirmatory Model for Substance Use Among Japanese American and Part-Japanese American Adolescents

Abstract

Few studies have examined the effect of ethnicity and cultural identity on substance use among Asian and Pacific Islander adolescents. A cross-sequential study conducted in Hawai'i with 144 Japanese and part-Japanese American adolescents assessed a model integrating Japanese ethnicity, cultural identity, substance use, major life events, and social support. Japanese American adolescents scored higher on the Japanese Culture Scale and on the Peers' Social Support than the part-Japanese American adolescents. Significant associations for substance use and impairment included culturally intensified events and Japanese cultural identity-behavior subset. Models had good overall fits and suggested that conflict surrounding cultural identity may contribute to substance use.

KEYWORDS:

• substance use
• Japanese American adolescents
• Japanese Culture Scale
• mixed ethnicity

Acknowledgments

This article was supported by the National Center for Indigenous Hawaiian Behavioral Health [National Institute of Mental Health (RCMI supplement NIH RR0361-06S1; NIMH; R24 MH5015-01, R24 MH57079-A1) and the Queen Emma Foundation], the Asian/Pacific Islander Youth Violence Prevention Center (Centers for Disease Control and Prevention; R49/CCR918619-05; 1 U49/CE000749-01), the Alcohol Research Center of Hawai'i (National Institute on Alcohol Abuse and Alcoholism; UO1AA014289-01), and National Center on Minority Health and Health Disparities. The contents of this article are solely the responsibility of the authors and do not necessarily represent the official views of the funding agencies. The authors would like to thank the researchers and administrators of the National Center on Indigenous Hawaiian Behavioral Health, Asian/Pacific Islander Youth Violence Prevention Center, Alcohol Research Center of Hawai'i, and Suzanne Baker, MA.

Notes

Note: F tests involved analysis of variance (ANOVA). Wald χ² involved univariate logistic regressions; R ² based on maximum value.

Note: F tests involved analysis of variance (ANOVA). Wald χ² involved univariate logistic regressions; R ² based on maximum value.

*t > 1.96.

Item 1: N = 144, R ² = 9.05%, χ² = 99.92, df = 37, RMSEA =.1090, RMR =.0477, GFI =. 9144, p < .0001.

Item 2: N = 144, R ² = 4.01%, χ² = 100.37, df = 37, RMSEA =.1094, RMR =.0476, GFI =. 9140, p < .0001.

Item 3: N = 144, R ² = 7.79%, χ² = 100.10, df = 37, RMSEA =.1092, RMR =.0476, GFI =. 9142, p < .0001.

Item 4: N = 144, R ² = 5.33%, χ² = 99.58, df = 37, RMSEA =.1088, RMR =.0476, GFI =. 9147, p < .0001.

Item 5: N = 144, R ² = 7.25%, χ² = 99.55, df = 37, RMSEA =.1087, RMR =.0476, GFI =. 9147, p < .0001.

Item 6: N = 144, R ² = 12.89%, χ² = 99.47, df = 37, RMSEA =.1087, RMR =.0476, GFI =. 9148, p < .0001.

Overall: N = 144, R ² = 9.17%, χ² = 100.15, df = 37, RMSEA =.1092, RMR =.0476, GFI =. 9142, p < .0001.

Factor 1: N = 144, R ² = 8.40%, χ² = 100.52, df = 37, RMSEA =.1096, RMR =.0476, GFI =. 9139, p < .0001.

Factor 2: N = 144, R ² = 8.96%, χ² = 99.52, df = 37, RMSEA =.1087, RMR =.0476, GFI =. 9147, p < .0001.

Note: RMSEA = root mean square error of approximation; RMR = root mean-square residual; GFI = Goodness-of-Fit Index.