Using Tobit Regression Analysis to Further Understand the Association of Youth Alcohol Problems with Depression and Parental Factors among Korean Adolescent Females.
10.3961/jpmph.2007.40.2.145
- Author:
Jorge DELVA
1
;
Andrew Grogan KAYLOR
;
Emily STEINHOFF
;
Dong Eok SHIN
;
Kristine SIEFERT
Author Information
1. Young Again Adult Day Health Care Center, University of Michigan, USA. jadelva@umich.edu
- Publication Type:Original Article
- Keywords:
Adolescent;
Female;
Alcoholism;
Depression;
Regression analysis
- MeSH:
Regression Analysis;
Questionnaires;
*Parent-Child Relations;
Korea;
Humans;
Female;
*Depression;
*Alcoholism;
Adolescent
- From:Journal of Preventive Medicine and Public Health
2007;40(2):145-149
- CountryRepublic of Korea
- Language:English
-
Abstract:
OBJECTIVES: This study characterized the extent to which youth depressive symptoms, parental alcohol problems, and parental drinking account for differences in alcoholrelated problems among a large sample of adolescent females. METHODS: The stratified sample consists of 2077 adolescent females from twelve female-only high schools located in a large metropolitan city in the Republic of Korea. Students completed a questionnaire about alcohol use nd alcohol problems, their parents' alcohol problems, and a number of risk and protective factors. Data were analyzed using tobit regression analyses to better characterize the associations among variables. RESULTS: Almost two-thirds of students who consume alcohol had experienced at least one to two alcohol-related problems in their lives and 54.6% reported at least one current symptom of depression, with nearly one-third reporting two depressive symptoms. Two-thirds of the students indicated that at least one parent had an alcoholrelated problem, and that approximately 29% had experie nced several problems. Results of tobit regression analyses indicate that youth alcohol-related problems are positively associated with depressive symptoms (p<0.01) and parent drinking problems (p<0.05). Parental drinking is no longer significant when the variable parental attention is added to the model. Decomposition of the tobit parameters shows that for every unit of increase in depressive symptoms and in parent drinking problems, the probability of a youth experiencing alcohol problems increases by 6% and 1%, respectively. For every unit of increase in parental attention, the probability of youth experiencing drinking problems decreases by 5%. CONCLUSIONS: This study presents evidence that alcoholrelated problems and depressive symptoms are highly prevalent among adolescent females. Although a comprehensive public health approach is needed to address drinking and mental health problems, different interventions are needed to target factors associated with initiation of alcohol problems and those associated with increased alcohol problems among those who already began experiencing such problems.
