Corresponding author: Sunjoo BOO, PhD, RN College of Nursing, Research Institute of Nursing Science, Ajou University, 206 World cup-ro, Yeongtong-gu, Suwon 16499, Korea INTRODUCTION
National data showed that the self-reported smoking prevalence among Korean women was 5.9%, whereas the cotinine-verified prevalence was 13.9% [1], suggesting substantial underreporting in the context of persistent stigma surrounding female smoking in Korea [1,2]. Female workers employed in service and sales occupations have been reported to smoke at higher rates than female workers in other occupations, and emotional labor and job stress have been identified as correlates of smoking in these workers [3,4]. Call center work is a representative emotional labor occupation in which employees must regulate their emotions in response to customer demands and organizational display rules, and approximately one fifth of female call center workers in Korea were reported to be current smokers, with cigarettes suggested as a means of alleviating the emotional burden of work [3].
Smoking cessation is a behavioral process that involves changes in motivation and smoking behavior over time. The transtheoretical model describes this process as progression from no intention to quit through intention and preparation to action and maintenance [5]. Accordingly, the present study classified participants into three groups based on current smoking status and intention to quit: an abstinence group, a quit-intention group, and a no-plan group. Comparing the psychosocial characteristics of these groups may provide more useful information for developing tailored smoking cessation interventions than a binary comparison between current and former smokers.
Several psychosocial factors may be relevant to this process. Social nicotine dependence refers to cognitive and sociocultural dependence characterized by glamorizing, rationalizing, or minimizing the harms of smoking beyond physical nicotine dependence [6,7]. Smoking abstinence self-efficacy, defined as confidence in maintaining abstinence in tempting or stressful situations, is a consistent predictor of subsequent cessation [8]. Together, these constructs represent two opposing cognitive forces, one rationalizing smoking and the other supporting cessation. In addition, emotional labor [9,10], cognitive emotion regulation (CER) strategies [11], and perceived social support [12] may further shape how work-related emotional strain relates to smoking behavior.
A previous study from this survey population showed that female call center workers who currently smoked had higher social nicotine dependence, more permissive smoking attitudes, and higher emotional labor than ex-smokers and never smokers [13]. However, that study focused on smoking status and therefore could not determine whether psychosocial characteristics differed among female workers maintaining smoking abstinence, current smokers intending to quit, and those with no intention to quit. Understanding these group-specific differences may help identify appropriate intervention targets. Therefore, this study compared psychosocial factors—including social nicotine dependence, smoking attitude, smoking abstinence self-efficacy, emotional labor, CER strategies, and social support—across smoking cessation groups among female call center workers and identified factors independently associated with group membership. The findings may provide a basis for developing workplace interventions tailored to smoking status and quit intention among female emotional labor workers.
METHODS
Study design and participants
This cross-sectional study analyzed data from an anonymous online survey of female employees of three credit card call centers in Korea, the design of which has been described previously [13]. Participants in the original survey were recruited by convenience sampling from three credit card call centers; unit managers distributed research flyers containing a link to the voluntary anonymous survey. Eligibility—female employees with at least 6 months of work experience—was screened by self-report at the beginning of the survey. Of the 618 employees who completed the initial screening, 30 were ineligible, yielding 588 participants in the original survey.
The present analysis included 186 female workers with a current or past smoking history, defined as having smoked at least 100 cigarettes in their lifetime. Among current smokers, quit intention was reported as within 1 month (n=9), within 6 months (n=14), sometime in the future (n=50), or no intention to quit (n=42). Because the study focused on the presence versus absence of quit intention and the two near-term intention categories contained few participants, all current smokers reporting any quit intention were combined into the quit-intention group (n=73). Current smokers with no intention to quit formed the no-plan group (n=42), and former smokers who were not currently smoking formed the abstinence group (n=71); these categories were used as broader analytical groups.
A post-hoc sample size assessment indicated that the available sample exceeded the minimum number required for one-way ANOVA (n=159). In addition, the available sample was also considered acceptable for exploratory multinomial logistic regression based on the events-per-variable criterion [14]. This study was approved by the Institutional Review Board (IRB) of Ajou University (IRB No. AJIRB-SBR-SUR-20-561).
Measurements
Social nicotine dependence was measured using the Korean version of the 10-item Kano Test for Social Nicotine Dependence (KTSND) [6,15], scored from 0 to 30, with higher scores indicating stronger tendencies to rationalize and glamorize smoking (Cronbach’s α=0.88). Smoking attitude was measured using the 7-item smoking attitude scale derived from the Teenage Attitudes and Practices Survey and translated into Korean, as described previously [13], with higher scores indicating a more permissive attitude toward smoking (Cronbach’s α=0.82). Smoking abstinence self-efficacy was measured using a 9-item scale based on the relapse self-efficacy model [16,17] (range, 9–45), with higher scores indicating greater confidence in remaining abstinent from smoking (Cronbach’s α=0.90). Emotional labor was measured using the 14-item Emotional Labor Scale developed by Brotheridge and Lee [18] and validated in Korean [19], rated on a 5-point scale, with higher mean scores indicating a greater degree of emotional labor (Cronbach’s α=0.84). CER strategies were measured using the Korean version [20] of the Cognitive Emotion Regulation Questionnaire [11], which comprises adaptive strategies and maladaptive strategies (Cronbach’s α=0.89 and 0.80, respectively). Social support was measured using the 12-item Multidimensional Scale of Perceived Social Support [21], which assesses support from family, friends, and significant others, with higher scores indicating greater perceived support (Cronbach’s α=0.93). General and occupational characteristics included age, marital status, education, household income, perceived health, sedentary time, drinking days per month, task type (inbound or outbound), job satisfaction, working period, and number of customers handled per day.
Statistical analysis
General characteristics were summarized as frequencies with percentages or means with standard deviations, and their homogeneity across the three groups was tested using the chi-square test and one-way ANOVA. Differences in the major study variables across groups were tested using one-way ANOVA with Scheffé post-hoc tests. Relationships among the study variables were examined using Pearson correlation coefficients. To identify factors independently associated with smoking cessation group, multinomial logistic regression was performed with the abstinence group as the reference category; social nicotine dependence, abstinence self-efficacy, emotional labor, adaptive and maladaptive CER strategies, social support, and drinking days per month were entered as covariates. Smoking attitude was excluded from the model because of its strong correlation with social nicotine dependence (r=0.711), which raised concern about multicollinearity [22]. Statistical analyses were performed using IBM SPSS Statistics for Windows, version 26.0 (IBM Corp.), and a two-sided P<0.05 was considered statistically significant.
RESULTS
General and occupational characteristics
The mean age of the participants was 37.64±8.40 years; 41.4% were living with a spouse, 48.9% had a college education or higher, and 42.5% had a monthly household income of 4 million Korean won or more. Half of the participants (50.0%) rated their health as fair, and the mean sedentary time was 8.16±2.61 hours per day. Inbound tasks accounted for 62.9% and outbound tasks for 37.1% of the participants, 50.0% were satisfied with their job, and the mean working period was 63.45±54.55 months with an average of 94.78±66.48 customers handled per day. None of the general or occupational characteristics, including perceived health (P=0.778), differed significantly among the three groups, except for drinking days per month, which was significantly higher in the no-plan group (7.70±8.07 days) than in the abstinence group (4.11±5.86 days) (F=3.45, P=0.034) (Table 1).
Differences in major study variables by smoking cessation stage
Social nicotine dependence increased stepwise and significantly across the abstinence (13.66±5.97), quit-intention (19.26±5.67), and no-plan (22.86±4.72) groups (F=39.05, P<0.001; all pairwise comparisons significant). Smoking attitude was significantly less permissive in the abstinence group (7.46±4.41) than in the quit-intention (11.14±3.64) and no-plan (12.76±3.38) groups (F=28.59, P<0.001), and smoking abstinence self-efficacy was significantly higher in the abstinence group (35.31±8.16) than in the two current-smoker groups (25.74±8.17 and 22.81±8.68) (F=37.91, P<0.001). Adaptive CER strategies were significantly lower in the no-plan group (62.93±10.88) than in the abstinence group (68.30±11.09) (F=3.67, P=0.027), whereas emotional labor (P=0.070), maladaptive CER strategies (P=0.872), and social support (P=0.549) did not differ among the groups (Table 2).
Correlations among study variables
Social nicotine dependence was strongly and positively correlated with smoking attitude (r=0.711, P<0.01) and negatively correlated with abstinence self-efficacy (r=–0.599, P<0.01); smoking attitude was also negatively correlated with abstinence self-efficacy (r=–0.499, P<0.01). Adaptive CER strategies were positively correlated with abstinence self-efficacy (r=0.174, P<0.05), emotional labor (r=0.389, P<0.01), and social support (r=0.393, P<0.01), and maladaptive CER strategies were positively correlated with emotional labor (r=0.318, P<0.01). Emotional labor and social support were not significantly correlated with the smoking-related variables in the total sample (Table 3).
Factors associated with smoking cessation stage
In the multinomial logistic regression with the abstinence group as the reference (likelihood ratio chi-square=102.03, P<0.001; McFadden pseudo R-squared=0.256), higher social nicotine dependence was independently associated with membership in the quit-intention group (odds ratio [OR]=1.10; 95% confidence interval [CI], 1.02–1.19; P=0.012) and, more strongly, in the no-plan group (OR=1.26; 95% CI, 1.13–1.40; P<0.001). Lower abstinence self-efficacy was associated with membership in both current-smoker groups (OR=0.89; 95% CI, 0.85–0.94 and OR=0.89; 95% CI, 0.84–0.96, respectively). Emotional labor was not significantly associated with membership in the quit-intention group, although the point estimate was elevated and the CI was wide (OR=2.18; 95% CI, 0.88–5.41; P=0.093). CER strategies, social support, and drinking days were not independently associated with group membership after adjustment (Table 4).
DISCUSSION
This study found that cessation stage among female call center workers was mainly distinguished by social nicotine dependence and abstinence self-efficacy rather than by emotional labor or perceived social support alone. Social nicotine dependence increased stepwise from the abstinence group to the quit-intention and no-plan groups, whereas abstinence self-efficacy showed the opposite pattern, and both factors independently distinguished the two current-smoker groups from the abstinence group in multivariable analysis.
The first major finding was the strong role of social nicotine dependence. Each 1-point increase was associated with higher odds of belonging to the current-smoker groups, particularly the no-plan group (OR=1.26). This is consistent with previous KTSND studies showing that scores differ by smoking status and reflect cognitive rationalization of smoking [6,7,15], and it extends the earlier status-based comparison in this population, in which social nicotine dependence rose from never smokers through ex-smokers to current smokers [13], by showing that the gradient extends across the three cessation groups. Because the KTSND can be assessed in former smokers, it may be useful both for identifying smokers resistant to cessation and for monitoring residual smoking-rationalizing beliefs among female workers maintaining abstinence as a marker of relapse risk. Reducing social nicotine dependence may require more than information about the harms of tobacco; cognitive restructuring of beliefs such as “smoking relieves stress” or “smoking is part of workplace breaks” is likely needed.
Abstinence self-efficacy was the second key factor. The abstinence group reported the highest self-efficacy, and lower self-efficacy independently predicted membership in both current-smoker groups even after accounting for social nicotine dependence, emotional labor, CER strategies, social support, and drinking, consistent with social cognitive theory and meta-analytic evidence [8,23]. The negative correlation between social nicotine dependence and self-efficacy further suggests that rationalizing beliefs may undermine confidence in quitting. Interventions for smokers with no intention to quit may therefore need to weaken smoking-rationalizing cognitions first and then build mastery through small, achievable abstinence tasks.
Emotional labor did not differ among the groups and was not an independent predictor; nevertheless, its OR for the quit-intention group was of moderate magnitude (OR=2.18). This does not imply that emotional labor is irrelevant; rather, it may operate as a contextual stressor whose effect depends on cognitive appraisal and coping resources. Emotional labor was positively correlated with both adaptive and maladaptive CER strategies, indicating that workers under higher emotional demands mobilize more cognitive coping overall, and prior research has linked surface acting and emotional dissonance to impaired well-being [9,10]. For cessation, the critical question may be not whether emotional labor is high but whether workers respond to it through adaptive regulation or through smoking-rationalizing beliefs and cigarette use; mediation or moderation models should be tested in future studies. The elevated but imprecise OR for the quit-intention group should therefore be interpreted as a possible trend rather than evidence of an independent effect; limited precision and statistical power may have contributed to the non-significant result.
Adaptive CER strategies were lower in the no-plan group in univariable analysis but not independently associated with group membership after adjustment, suggesting that adaptive cognitive coping is a supportive resource rather than a direct determinant of group membership. The attenuation may partly reflect conceptual overlap with abstinence self-efficacy, because cognitive reappraisal and coping processes can support confidence in managing high-risk situations [8,11,20,23]. CER training may still be clinically useful when explicitly linked to self-efficacy-building exercises. Such training could include identifying automatic smoking-related thoughts after difficult calls, practicing positive reappraisal and refocusing on an action plan, rehearsing brief urge-management statements, and recording successful smoke-free coping episodes to build mastery.
Perceived social support did not differ by cessation stage and was not independently associated with group membership. This likely indicates that general support is less sensitive than cessation-specific support, whose type, source, and specificity determine its effect [12]. General support may not translate into quitting unless it specifically helps workers manage urges after difficult calls or maintain smoke-free routines; interventions may need to assess and strengthen such behavior-specific support. Similarly, drinking frequency differed among groups in univariable comparison but not in the adjusted model, suggesting shared variance with the cognitive determinants; given the well-documented co-use of alcohol and tobacco and its contribution to relapse [24], brief alcohol-risk counseling may still be a useful component of workplace programs.
These findings support workplace interventions tailored to smoking status and quit intention. For smokers with no intention to quit, the first target should be smoking-rationalizing beliefs, followed by gradual enhancement of abstinence self-efficacy. For smokers intending to quit, preparation for high-risk situations, strengthening of abstinence confidence, and practice of adaptive CER are indicated. For female workers maintaining abstinence, relapse prevention should include monitoring residual social nicotine dependence and reinforcing alternatives to smoking during emotionally demanding work situations. In practical terms, these components could be delivered within the natural rhythm of call center work rather than as separate clinic visits. For example, brief (5–10 minute) post-call debriefs led by supervisors or trained peer facilitators could help workers reframe smoking-rationalizing beliefs (e.g., challenging the notion that a cigarette is the only way to recover after a difficult call); short micro-sessions during scheduled work breaks could offer alternative stress-relief routines—paced breathing, light stretching, or peer conversation—to substitute for smoking breaks; and posted or mobile app–delivered prompts could reinforce messages that lower social nicotine dependence and acknowledge small abstinence milestones. Embedding cessation support into existing work breaks and debriefs in this way may improve its feasibility and reach in this time-constrained workforce. Cessation should be framed not as an issue of individual willpower alone but as a behavior shaped by the cognitive and emotional conditions of work.
This study has several limitations. First, participants were recruited through convenience sampling from a limited occupational context, and response rates could not be calculated because the number of employees who received or viewed the invitation was unavailable; therefore, self-selection bias cannot be excluded and generalizability is limited. Second, the cross-sectional design precludes causal inference; whether cognitive factors led to cessation or cessation reshaped cognition remains unclear. Third, smoking status was self-reported and may be underreported given the stigma against female smoking in Korea [1]. Fourth, the relatively small no-plan group may have limited power to detect weaker associations. Nevertheless, this study is meaningful in that it simultaneously examined cognitive, emotional, and social correlates of cessation stage in an understudied group of female emotional labor workers. In conclusion, social nicotine dependence and abstinence self-efficacy were the most robust correlates of smoking cessation group membership, and workplace cessation interventions for female call center workers should prioritize cognitive restructuring of smoking-rationalizing beliefs and enhancement of abstinence self-efficacy, incorporating adaptive emotion regulation tailored to emotionally demanding work.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation
Print
