Cigarette smoking is a widespread unhealthy behavior that leads to several diseases and premature death (1). Although the tobacco control policy reduced the worldwide prevalence of smoking, 9.5% of women were found to be active smokers in 2015 (2) indicating that women's smoking is an important public issue. Men and women smokers vary significantly in their physiological, cultural, and behavioral factors (3). For example, most men smoke for the pleasure and enjoyment of smoking, while women smoke to relieve stress and control their body weight (BW) (4, 5). Smoking activates dopamine in the ventral striatum of men, and the discrete subregion of the dorsal putamen of women (6). Plasma and saliva cortisol levels have predicted smoking relapse in men but not in women (7). Although the prevalence of women smokers is much lesser than that of men smokers (2), the former are more likely to get smoking attributable diseases, even under controlled tobacco exposure (8, 9) In addition, smoking during pregnancy is associated with poor birth outcomes (10) and early onset of smoking in children (11).

Abstinence is the most important way to reduce the harm caused by smoking. Studies have shown that women face greater difficulty in quitting smoking as compared with men (12). One plausible reason is that women are more likely to continue smoking for controlling their BW. Body-image challenges motivate young women to smoke (13), and smoking cessation is often associated with an increase in BW. A previous meta-analysis inferred that after 12 months of abstinence from smoking, 84% of individuals gained 4 to 5 kg mean BW (14). However, during the cessation program smoking abstinence-related negative mood and greater motivation to relieve distress from withdrawal symptoms were reported more in women than in men (15), suggesting that the barriers to smoking cessation differ between the sexes.

To the best of our knowledge, studies on the predictive factors of cessation failure between different sexes are scarce. In this study, we hypothesized that the predicted factor of failed smoking-cessation differs between the sexes and that the BW gained by women during the cessation program is associated with failed abstinence. Therefore, we performed a retrospective cohort study by using a well-registered, real-world in-hospital cessation program data from a Medical Center in the southern Taiwan to verify the hypothesis.

Among the 2,635 individuals who participated in the smoking-cessation program during the study period, 303 participants were excluded as they were lost to follow-up when assessing the 6-month prolonged smoking abstinence, 804 participants were excluded because of their missing information, and 53 participants were excluded because of duplicated data. The remaining 1,475 participants were included in the analysis (Figure 1). Table 1 shows the demographic and clinical characteristics of 1,339 (90.8%) men and 136 (9.2%) women among a total of 1,475 participants. Compared with men, a larger proportion of the women were younger than 65 years, unmarried, had fewer comorbidities, PHQ-2 ≥ 2, and were initiated through outpatient service. The FTND score, BW gain during the program, and usage of types of cessation medication were similar for both the sexes. The 6-month prolonged abstinence rate of women was lower than that of men (21.3 vs. 36.7%, P < 0.001).

Previous studies have found that smoking behavior differed between the sexes (3), and women smokers have more difficulty in quitting smoking than men smokers (13). To lower the prevalence of smoking among women, it is important to understand the factors associated with the failure of cessation in women smokers. However, studies investigating the predictors of failed smoking abstinence in hospital-based cessation programs are paucity. In this retrospective cohort study, data from 1,475 individuals were analyzed who had participated in a hospital-based smoking-cessation program. The collected data were considered credible because the factors, including age, sex, comorbidities, FTND scores, and cessation medications were monitored by the Taiwan Bureau of Health Promotion. In the multivariable logistic regression model adjusted for age, FTND score, PHQ-2, a cessation medication, other factors, and the BW gain was found to be associated with a failed 6-month prolonged smoking abstinence in women (OR: 3.10, 95% CI: 1.10 to 9.04) and with a successful 6-month prolonged smoking abstinence in men (OR: 0.77, 95% CI: 0.61 to 0.98). The sensitivity analyses showed that BW loss during the cessation program was associated with a successful 6-month prolonged smoking abstinence in women, but not in men. Thus, it was concluded that BW gain might be a sex-specific predictor of failed smoking abstinence in smoking-cessation programs.

Body-image dissatisfaction differs between the sexes. Women choose a smaller size whereas men choose a similar or larger size than their ideal body sizes (29, 30). In comparison with men, women have a higher risk of dissatisfaction with their body weight, report more disordered eating, and are less likely to recover their ideal BW from eating disorders (31, 32). Weight control is one of the main reasons leading to tobacco abuse in women (4, 5). On the other hand, weight gain might be one of the barriers in quitting smoking for women (33, 34). Cropsey et al. reported that women prisoners, who arguably have a stable energy intake and output, had average 10 pounds weight gain on the 6th month of smoking cessation, although the gain of body weight was decreased on the 12th month of cessation (35). The effect of smoking cessation on body mass index was higher for older women than for younger (36). In a previous smoking cessation study for postmenopausal women, the average weight gained was 6.5 kg (14 pounds) after 16 months of quitting smoking (37). To avoid failure of smoking abstinence, multidisciplinary weight control strategies (including education, calorie restriction, increasing physical activity, and cessation medication) should be introduced to the participants (38), especially women. Metanalysis concluded that pharmacological interventions limited the short-term weight gain during smoking cessation. However, there are no clinically effective interventions to restrict long-term weight gain currently (39).

Our study results showed that women participants with a PHQ-2 score ≥ 2 were associated with successful abstinence from smoking. However, previous research showed that depressive symptoms predicted a lower smoking abstinence rate, and there was no evidence of sex-specific effects of depression on smoking cessation (40, 41). PHQ-2 is a depression screening questionnaire rather than a diagnostic tool (17). Further investigation regarding PHQ-2 and smoking cessation in women is warranted.

Several limitations required to be acknowledged in this study. First, this study was conducted in Taiwan and body dissatisfaction differs based on cultural background, race, and economic status, (42) thus, the result of this study should be generalized with caution. Second, as this study was conducted retrospectively and the smoking history was dependent on the report of the participants, recall bias could not be excluded. However, other data including comorbidities, FTND, and cessation medication were regulated by the Taiwan Bureau of Health Promotion, so the robustness and accuracy of these data are supposed to be reliable. Third, apart from the information regarding the marital status of participants, information regarding active pregnancy, or child-bearing was unavailable. Although several pregnant women try to quit smoking due to the health concerns surrounding them and the fetus (3). A previous study had reported that among the few Taiwanese women who quit smoking during pregnancy, most relapsed in the 1st year after childbirth (43). Thus, the absence of pregnancy and child-bearing data should not affect the results of this study. Fourth, the women comprise only 9.2% of the participants. Given the large size of our study cohort, the sample size of women participants were sufficient. At last, the doses of NRT are different among the participants. However, the dose of NRT was determined by the participants' daily cigarette consumption in our study. This dosing strategy is widely used in the clinical trials and real-world conditions, and this limitation should not affect the results of our study.

BW gain during a hospital-based smoking-cessation program is a sex-specific predictor of failed abstinence. Furthermore, multicenter studies including participants of different races and cultural backgrounds are warranted.

The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author.

The studies involving human participants were reviewed and approved by Institutional Review Board of NCKUH (ID number: B-ER-110-126). The Ethics Committee waived the requirement of written informed consent for participation.

C-WK, C-FL, and TY designed the study. C-WK, C-FL, C-YChe, C-WC, and C-CS searched the literature. C-WK, C-YChe, R-HW, C-YCho, H-JC, and J-SW collected the data. C-WK, C-FL, and TY did the statistical analysis. C-WK, C-FL, and C-YChe wrote the manuscript. All the authors contributed to and approved the final manuscript.

This study was funded by grants from the National Cheng Kung University Hospital (NCKUH-11106011).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.