The Society for Social Work and Research

2014 Annual Conference

January 15-19, 2014 I Grand Hyatt San Antonio I San Antonio, TX

Latent Class Growth Modeling of Korean Women's Longitudinal Smoking Patterns

Saturday, January 18, 2014
HBG Convention Center, Bridge Hall Street Level (San Antonio, TX)
* noted as presenting author
Sunju Sohn, PhD, Assistant Professor, Cheongju University, Cheongju, South Korea
Jina Jun, PhD, Research Fellow, Korean Institute of Health and Social Affairs, Seoul, South Korea
Hyunyong Park, MSSW, Doctoral Research Assistant, Case Western Reserve University, Cleveland, OH
Purpose: Female smokers tend to underreport their smoking behaviors, hence the discrepancy between the self-reported smoking rates and the actual prevalence rates. A recent study showed that about half of Korean women tend to falsely report their true smoking and six times more likely to conceal their smoking than men, indicating that more women than the estimated population are exposed to risks involved in smoking. Currently, limited research is available to enhance our understanding of female smokers, as research focus has been on young, unmarried, or male smokers. This presentation provides longitudinal findings on distinct sub-groups of Korean female smokers and their smoking patterns using latent class growth modeling (LCGM).

Method: This study used LCGM and secondary data from the Korean Welfare Panel Study on 329 Korean female smokers to identify longitudinal smoking trends developed across three years based on the total numbers of cigarettes smoked per day. This study also explored possible linkages between the identified smoking trajectories and levels of depressive symptoms (11-item CES-D, alpha=.88-.89), smoking cessation attempt, and sociodemographic factors (i. e., average age, marital status, education attainment, and household income). For the purpose of the study, only the current female smokers and those who reported smoking at least one cigarette per day throughout the three year period were included in the analyses. A binary logistic regression was performed to examine any differences between the identified smoking trajectories by levels of depressive symptoms, smoking cessation experiences, and sociodemographic factors.

Results: Two distinctive smoking trajectories emerged from the analysis: "decliners" whose smoking have early established at a relatively lower level and decreased over time (n=255, 77.5%); and "incliners" whose smoking started at a comparably higher level and steady increased across time (n=74, 22.5%). The initial average smoking of incliners was twice as their counterparts with an average of 14.35 cigarettes smoked per day compared to 6.57 cigarettes. Each year, the incliners’ average daily consumed cigarettes significantly increased by .62 cigarettes, compared to 1.37 decrease in the decliners. These findings translate into one in four Korean female smokers who are likely to engage in smoking behaviors heavier and persistently than the remaining three-quarters of the sample. Binary logistic regression results indicate that a higher percentage of unmarried women composed the incliners (68.9% vs. 48.2%) and relatively more married among the decliners (36.4% vs. 24.3%). No differences were observed in levels of depressive symptoms, cessation attempt, and other sociodemographic factors.

Implications: Findings suggest that heterogeneity in smoking patterns exists within female smokers. The findings also provide insight into understanding longitudinal aspects of how smokers can develop into continuous and heavy smoking and the role of initial average daily smoking in predicting long-term smoking behaviors. Given varying trajectories, future research need to incorporate various mental health factors related to smoking (e.g., stress or anxiety) to expand our knowledge on why many female smokers are unwilling or cannot quit smoking despite diverse negative health consequences and/or comorbid issues and to develop antismoking campaigns and treatment programs for increasing female smoking population.