Association of smoking status, cumulative smoking, duration of smoking cessation, age of starting smoking, and depression in Korean adults

Epidemiologic [1–4] and clinical [5, 6] studies have reported a strong association between smoking and depression. Several explanations for this association have been proposed.

The first is that underlying environmental or genetic factors predispose individuals to both smoking and depression [2, 7–9]. The second is that depression leads to smoking. Several studies reported that smoking may increase as a result of self-medication for depression [10–12]. The third is that smoking itself increases the risk of depression [13–18]. Some studies found that smoking among adolescents increased the risk of depression, and another reported that smoking was highly associated with depression in adults. However, in these studies, many confounding factors, such as socioeconomic status, which may affect the association between smoking and depression, were not sufficiently controlled. Furthermore, to our knowledge, no previous study has simultaneously evaluated several aspects of smoking status, including the amount of smoking, age of starting smoking, and duration of smoking cessation.

The objective of this study was to investigate the cross-sectional association between smoking and depression after controlling for socioeconomic factors and many aspects of smoking in population-based subjects.

The results of this study demonstrated a stronger association between smoking status and depression in women than in men. However, the dose–response relationship for smoking and depression among current-smokers was only significant in females. No significant associations of the age of starting smoking and duration of smoking cessation with depression were observed among former-smokers.

In this study, the smoking rate in women was lower than that in men. The gender difference in smoking might be due to the effects of socio-cultural factors. Sex role norms and general expectations concerning gender-appropriate behavior have had a variety of effects on gender differences in smoking [21]. Traditional sex roles lead to social pressure against women smoking [21]. Additionally, differences in the demographic and socioeconomic characteristics of smokers and non-smokers have been reported [22–24]. The relationship between smoking and depression could be affected by these factors. However, it seems unlikely that these factors affected our results, as we adjusted for them in our analysis.

The CES-D is an adequate screening instrument for depressive disorder in the population [25–28]. We used a CES-D cutoff score of 21 as the definition of depression. Although the CES-D has low sensitivity and specificity for minor depression, it is excellent for use as a screening instrument for major depression, and the optimum cutoff point for the CES-D was found to be 21 [25]. Cho et al. also reported that a CES-D score of 21 most effectively detected a range of depressive symptoms during screening. They suggested that the higher cutoff point in Korea than in Western countries might be due to differences in the expression of affect, especially the suppression of positive affect, in cultures based on Confucian ethics [20].

Previous reports have proposed mechanisms for the association of smoking with depression. One pharmacological explanation for the influence of smoking on depression is that a reduction in monoamine oxidase B (MAO B) activity might synergize with nicotine to produce the diverse behavioral and epidemiological effects of smoking. The brains of smokers showed a 40% decrease in MAO B level relative to nonsmokers or former-smokers [29]. Another explanation is that vulnerability to depression is related to certain factors. Nicotine is known to affect the neurotransmitters involved in major depression [30]. Nicotine use might increase vulnerability to depression [31], and dysregulation of the dopaminergic system in an addictive state is a plausible mechanistic pathway to depressive vulnerability [32]. A further explanation is that smoking generates free radicals, causing lipid peroxidation, oxidation of proteins, and other tissue damage in smokers [33]. Depression has been associated with elevated reactive oxidative species such as anti-oxidative enzyme activities and lipid peroxidation [34].

Our data suggest that associations between smoking status and depression were more likely in females. Some studies reported a strong relationship between smoking and depression in females, rather than in males. Others reported a gender-specific effect, and that females showed a much greater smoking-depression relationship than males in both adolescents [14, 35] and adults [15, 16]. However, some studies have reported conflicting results. Nakata et al. reported that gender did not modify the effects of smoking on depression [36], and in some studies, an increased risk of depression was observed in male, but not female, smokers [37–39]. One possible explanation for this gender difference is that female smokers may have a lower threshold for depression than male smokers. McKee et al. reported that stressful life events were more strongly related to smoking in females than in males [40]. However, gender-specific effects in the smoking–depression relationship remain incompletely understood.

In this study, current-smokers had a higher risk of depression than never-smokers among both males and females. Although no statistically significant association was seen in males, former-smokers had a lower risk of depression than current-smokers. Wisebeck et al. reported that former-smokers carried a risk of depression that was significantly lower than that of current-smokers, but higher than in never-smokers [41]. They suggested that their results are in accordance with the hypothesis that smoking cessation was an effective way to reduce the risk of depression, and if so, depression should be less frequent in former-smokers who stopped smoking a long time ago compared to those who stopped recently [41]. However, in their study, the difference between the association of distant (first 3 years) smoking cessation and recent (within 1 year) smoking cessation with depression was not significant [41]. Mykletun et al. also reported no associations between depression and time since cessation in former-smokers [38]. They explained that although a trend toward a lower prevalence of depression with time since cessation was detected, it was not statistically significant [38]. Our results are similar; the risk of depression did not significantly decrease with increasing duration of smoking cessation in both males and females. We also evaluated the relationship between the age of starting smoking and depression, and no significant increased risk of depression was observed according to decreasing age of starting smoking. Some studies have reported that smoking during adolescence predicted depression later in life compared with never-smokers [14, 42, 43]. However, to our knowledge, no study had evaluated the relationship between the age of starting smoking and depression among current-smokers. We suggest that current smoking status is a more important factor for risk of depression.

If a physiologic link indeed exists between smoking and the subsequent development of depression, then it would be reasonable to expect that evidence of this link would became more pronounced as the number of cigarettes consumed increases. Klungsøyr et al. reported a dose–response relationship with an increasing risk for former-smokers and for increasing numbers of cigarettes smoked per day for current-smokers, and also for an increasing number of smoking-years [18]. Pasco et al. reported that compared with nonsmokers, the odds for major depressive disorder increases by 1.47-fold for females who smoked 11–20 cigarettes per day and more than doubled for those who smoked >20 cigarettes per day [44]. However, in their studies, dose–response relationships of smoking were evaluated by comparing them with never-smokers. Duncan and Reese reported very little evidence that smoking intensity was related to depression in smokers [45]. In their study, smoking an extra pack of cigarettes per month was associated with a statistically nonsignificant 0.02 increase in male CES-D scores and a 0.01 increase in female CES-D scores after controlling for smoking status [45]. In the present study, no dose–response relationship of smoking with depression was observed among current-smokers in males, but female heavy smokers had a higher risk of depression. We suggest that this difference might be related to a gender-specific effect of smoking on depression.

Our data demonstrate that smoking is strongly associated with depression, particularly among females. However, a dose–response relationship was observed only between smoking and depression in females, and not in males, and duration of smoking cessation and age of starting smoking were not significantly associated with depression. Further large population-based prospective studies are needed to confirm the causal effects of smoking on depression. These findings suggest that depression prevention may need to be combined with smoking prevention and that different strategies in males and females may be necessary.