The results of this study showed significant differences in social contextual factors between non- and former vs. current smokers and different patterns by gender. Inconsistent with other studies on the older population and general population [9, 12, 30, 31], the stress level of current smokers was not higher than that of non- and past smokers, and a moderate social network level significantly influenced smoking behaviors among women only. In addition, a high level of social support was positively related to smoking.
The smoking prevalence in this study reported 17.4%. This prevalence was quite lower compared to that among Koreans aged over 19 years who was about 25% . However, mean age of this study participant was 60 years old and lived in rural area. And other study on the relations of social support to the health behaviors and health status in elderly was surveyed the smoking behavior for 8,688 elderly people, and that smoking rate was 18.6% . Therefore, once considering the demographic characteristics of study participants, this smoking prevalence seemed not quite low.
There were few studies on why is the difference of smoking prevalence between urban and rural population in Korea. However, recently reported study showed that the inequality of smoking prevalence by age and education level have been increased, but the inequality of smoking prevalence by residence area using location quotient was not a difference after controlling the socio-economic characteristics at individual level . Results of the inequality of smoking prevalence considered socio-economic characteristic, but not smoking related other psychosocial and health status. And other daily living pattern, which is different between urban and rural area, could be considered such as leisure time activity, working content & time, social relationship, and so on. Therefore, regarding lack of studies in these issues, further studies are recommended for evident explanation.
Previous studies have been reported that smoking behavior is affected by stress as the response to interact with a social-environmental context even though perceived subjective stress [34, 35, 12], and social support regulates the effect of stress on smoking behaviors . However, the result of this study was inconsistent with previous studies. The finding showed that the level of stress in non- and former smoker was significantly higher than that in current smoker, but that the relationship of smoking behavior and stress was no longer significant once demographic, psychosocial, and other factors had been adjusted statistically, including gender, age, education level, family income, BMI, alcohol intake, social support and social network. The high level of social support had likely to be more smoke, contrary to expectation. These findings suggest the possibility that smoking as a means for reducing stress could be used, which might be encouraged by families or friends who smoke [36, 37, 32].
Social support as a function of social relation has been introduced and recognized as a facilitator for improving health behavior . The size of the social network as a resource of social support positively affects health behavior, especially among older people [15, 38]. However, other studies have focused on the closeness within social network members and have pointed out that high homogeneity of a social network could reinforce smoking . Furthermore, social networks change over time, and older people have a tendency to construct a social network with close social partners . The result of this study that a high intimacy level of social network did not serve a protective role for smoking but, rather, that a moderate social network level played a protective role for smoking is in partial agreement with studies on Korean older populations [32, 40].
Social support has been defined as a facilitator for reaching goals with others and then induced the change of situation and as a qualitative aspect of social relations [18, 41]. Generally, social support has been recognized as a moderator of stress and has been used in interventions such as smoking cessation and alcohol abuse [18, 42]. In fact, group intervention for smoking cessation using social support group has been shown to be more effective . However, the current study showed that those in the high-level social support group were likely to be current smokers. This result is in disagreement with previous studies, but consistent with some studies in South Korea [32, 40]. Smokers have a tendency to continue their smoking behavior as a result of rationalizations about the benefit of smoking filtered cigarettes [44, 45]. Social norms about smoking and attitudes toward smoking as a product of interaction with members within the social network may reinforce smoking behavior . In addition, smokers have a high degree of knowledge about the health risks of smoking, but a low will to quit smoking . Positive social norm about smoking may be influenced by self-appraisal of benefit-harm on smoking, smoking acceptability within social network, delivering the misinformation by significant others, and so on [47, 27]. Also, relatively low price of tobacco in Korea could be contributed to smoke in that make to improve the accessibility of social resource availability as part of social support . The negative effect of social support on smoking behaviors in this study can be explained by the older age of the participants and the results of interactions between social support and social networks in rural areas . These results suggest that in order to improve the effect of smoking-related interventions or policies, especially in rural areas, it is necessary to identify the features of social support and social networks.
Various smoking cessation interventions have been performed. Among them, group behavioral intervention and group counseling have been taken up as effective programs except population-based mass media campaigns [50, 51]. The advantage of group intervention or counseling is that it strengthens the will to quit and induces the motivation to quit by changing misconceptions, beliefs, and attitudes about smoking through participant interaction . Therefore, simple stay in grouping intervention or counseling is hard to increase the quitting rates. To improve the quitting rates of smokers, especially older smokers, interaction among participants should be activated. Social contextual factors such as social support and social networks should be used as a pathway to activate interaction among smokers. However, social contextual factors may differ by age, gender, education level, and living area. A previous study showed that women with a low level of social network had a tendency to be smokers . In fact, this study found a difference in psychosocial factors and social contextual factors by gender and smoking status, with men in the high positive interaction support group being more likely to smoke, but not women. These results provide evidence that a specific approach for quitting by gender using the pathway of social support and social networks is needed.
A study on the functions of the social networks of rural elders in Korea pointed out that although elders recognize their sons and daughters as important members of their social network, friends and neighbors function as the most influential aspects of the social network in daily living . In fact, the duration of acquaintance with neighbors was more than ten years for half of the elders in rural areas . This indicates that the characteristics of rural elders' social networks could transition from heterogeneity and peripheral partners to more homogeneity and closer partners over time . Smoking pattern by the level of social network can be explained by aspect of social activity, partially. This suggest the possibility that women smoker with high social network could have or make many opportunity to be smokers as intimated networkers, whereas women smoker with low social network may rather go out for smoking with intimated networkers than regulate their stress with smoking alone. In Korea, these patterns have been formed by lesser reluctant environment for smoking of elderly women than that of younger women. Therefore, a high level of social network ties may not serve a protective role in smoking among women. The result of this study is consistent with that of previous studies that have reported that poor social networks among women are associated with smoking behavior [47, 49]. In addition, this finding provides evidence that it is necessary to sustain the optimal social network level not to smoke in women .
Social networks can be defined as a resource of social support, a type of function of social relations, and objective informal social support [39, 55, 56]. A study on HIV risk behaviors described the characteristics of social networks and showed that social networks that consisted of persons with risky health behavior had a tendency to decrease in size and that localized social networks play a central role in unknowingly spreading health risk behaviors . This feature of social networks can be applied to delivering a smoking intervention program by identifying high-risk groups . As tobacco policies such as clean air ordinances and home restriction have increased, smokers are likely to be isolated from public places . Therefore, smoking social networks may be localized and stronger. In this study, the findings of poor social networks and high social support among female smokers suggest that this feature of social networks in rural areas in Korea may be similar. Therefore, in rural areas, identification of smoking social networks including smoking-related risk groups should precede anti-smoking interventions.
A limitation of this study is that social norms, attitudes, and beliefs about smoking behavior were not assessed. This study could not identify the social norms and attitudes toward smoking behavior as a product of interaction among social network members. A comprehensive understanding of the relationship of social support and social networks is difficult to achieve due to a lack of information on them. Therefore, future studies should be conducted to identify the relationship between social contextual factors and high-risk behavior among elderly rural persons.
Other limitation is the small number of women smoker to make a comparison with the influence of psychosocial factors. Although the smoking prevalence of women in this study was not low, the number of women smoker included in each group, which was divided by the level of stress, social support, and social network, was not enough to be compared with non- and former-women smoker. Therefore, the effect of social network in women may be biased result. To get a more information about the interaction among smoking behavior, stress, social support, and social network in detail should be considered the low smoking rate in women in Korea.
For the limitations on representativeness of the study subjects and generalization of our findings exist. We recruited voluntary participants in study areas without any randomization and/or stratification considering age and gender distribution of eligible population. Therefore, some selection bias and difficulties to generalize our finding could be. Comparing the age and gender distribution between study subjects and eligible population, there is not much difference, even though proportion of female and proportion of aged under 40 years in study subjects is relatively higher and lower than those in eligible population, respectively. However, these issues are usually happened in many other community based survey, and the proportion of old aged subjects in study population was similar with that in eligible population. In addition, many of middle aged residents who were registered in rural areas might not reside in their registered habitat, because they usually live in other areas for job seeking or convenience of daily living. These situations could affect to lessen the actual number of middle aged subject we can contact in the study areas.