Smoking contributes extensively to cardiovascular diseases (CVD), which are the leading causes of death and the second largest source of healthcare costs in Western countries [1, 2]. Continuing to smoke after a coronary event is associated with mortality levels that are significantly higher than the mortality levels among those who quit. In addition, smoking cessation rapidly reduces the risk of a coronary event once smoking stops, with the risk of such an event approaching that of non-smokers by 3 years after cessation [3‐5]. A meta-analysis of cohort studies assessing the effects of smoking cessation on mortality after a myocardial infarction estimated that the combined odds ratio of death for smoking cessation compared with smoking continuation was 0.54 (95% confidence interval [CI], 0.46–0.62) [6]. Furthermore, quitting smoking has been associated with a substantial reduction (36%) in the risk of all-cause mortality among patients with CVD [7].

Nevertheless, there is evidence that a substantial number of patients who were smokers prior to hospitalisation for coronary heart disease continue to smoke after discharge, despite advice to stop [7‐9]. Although it could be presumed that hospitalisation serves as an important event—a “teachable moment”—that could motivate smokers to quit smoking [10], this effect seems to be short term in many cases: most smokers resume smoking after hospitalisation [11].

Many factors are associated with the continuation of smoking among cardiac patients: younger patient age, higher level of nicotine dependency, lack of future intentions to quit [1], lack of will or confidence, absence of social support [12], belief that “the damage is done,” history of smoking relapse, hostility [13], and absence of smoking cessation advice [10, 14]. Furthermore, Berndt et al. [1] distinguished three types of hospitalised cardiac patients according to their smoking characteristics and predictors of smoking abstinence 1 month after discharge. The first and largest cluster (38.4% of patients) comprises patients with high scores on self-efficacy toward non-smoking, high scores on perceived social influence toward non-smoking, and the highest likelihood of successfully abstaining from smoking. The second and smallest cluster (26.4%) differs from the first one because its members reported the presence of many smokers in their social environments, but they showed high levels of smoking cessation. The third cluster—the hard-core smokers (comprising 35.2% of all patients)—had the highest risk of continuing or resuming smoking after hospital discharge and were characterised as smokers with low future intentions to quit [1].

Although the socio-demographic and psychological factors associated with smoking cessation or relapse have been studied in detail, the socio-economic factors that might affect smoking outcomes (e.g., educational level, income, occupational status) have been under evaluated. However, such factors have been shown to independently affect these outcomes [15]. Therefore, our objective was to examine the influence of income differences on smoking cessation success among patients who were smokers when they were admitted for a cardiac event at the National Institute for Cardiac Surgery and Interventional Cardiology (INCCI) in Luxembourg.

In total, 1289 patients participated in the follow-up study, of which 250 were ever smokers. The follow-up group of respondents who were smokers was not significantly different from the non-respondents who were smokers for all characteristics, except for physical activity (Table 1). Compared with the smokers in the initial study (n = 1001), the smokers in the follow-up study were more likely to have a Luxembourg nationality, to be married or living with a partner, and to have regular physical activity (Table 1). Of the 250 participants who were current smokers at the time of the initial study, 43.2% indicated that they still smoked, while 56.8% had stopped smoking by the follow-up (Table 2). Smoking cessation was not significantly associated with gender, age, nationality, marital status, or awareness of tobacco as a CV risk factor, but was significantly associated with income, initial diagnosis, and health-related behaviour (practicing regular physical activity and BMI) by the follow-up.

In Table 3, awareness of tobacco as a CV risk factor is shown in relation to patient characteristics and smoking status. The awareness of tobacco as a CV risk factor was associated with age, household income, and BMI for all smokers. For former smokers, awareness of tobacco as a CV risk factor was significantly associated with age, household income, and physical activity. In current smokers, this awareness was only associated with nationality and household income.

In Table 4, the full results from all three models without the interaction term are presented. In the first multivariate logistic model, quitting smoking was associated with income after controlling for socio-demographic indicators (sex, age-group, nationality, marital status). The odds of smoking cessation were higher among patients belonging to the two highest income groups (odds ratio [OR] = 2.0; 95% CI, 1.1–3.8 and OR = 2.0; 95% CI, 1.0–4.0, respectively) than among those belonging to the lowest income group.

Adding diagnostic factors (acute myocardial infarction, ischaemic heart disease, and angina pectoris) in the second model did not alter the odds of smoking cessation among patients belonging to the two highest income groups (OR = 2.1; 95% CI, 1.1–4.0 and OR = 2.6; 95% CI, 1.3–5.3, respectively). In the third and full model, risk behaviour elements (BMI, physical activity) were added as cardiovascular risk factors. The odds of quitting smoking among patients belonging to the two highest income categories remained significantly higher compared with patients belonging to the lowest income group (OR = 2.8; 95% CI, 1.3–6.1 and OR = 2.8; 95% CI, 1.2–6.5). Awareness of tobacco as a CV risk factor was not associated with the probability of smoking cessation.

In Table 5, an interaction term was added to model 3. The odds of quitting smoking were only associated with income when patients were aware of tobacco as a CV risk factor. Indeed, among patients who were aware that tobacco is a CV risk factor, the odds of smoking cessation were 5.62 (95% CI: 2.13–14.86) and 3.65 (95% CI: 1.51–8.86) times in patients with household incomes of 36,000–53,999 euros and ≥54,000 euros, respectively, compared to patients with an household income of <36,000 euros.

This study showed an independent association between smoking cessation and income status among adult patients with CVD 5 years after an angiography in Luxembourg. Patients belonging to the higher income groups were more likely to quit smoking compared to patients belonging to the lowest income group, and this association held even after accounting for diagnostic factors and cardiovascular risk factors or awareness of tobacco as a CV risk factor. These results are similar to those found in other studies examining educational level or occupational status as a proxy for socio-economic status [15, 18‐20]. People with a low socio-economic status are thus confronted with simultaneous problems: they have higher rates of coronary heart diseases and a higher prevalence of smoking, as well as a higher risk of relapse and a decreased likelihood of successfully quitting smoking [15, 20]. It is also important to stress that the association between socio-economic status—regardless of the measure (i.e. income, educational level, or occupational status)—and smoking cessation in coronary patients is independent of other well-known predictors of successful cessation or relapse.

However, for this analysis, household income was selected as a proxy for socio-economic status instead of educational level because a majority (around 80%) of the participating patients (in both the initial and follow-up study) were over 55 years old making educational level less relevant as a socio-economic indicator when compared with income. Furthermore, although smoking initiation has been shown to be strongly related to educational level, smoking cessation has been shown to be strongly related to income [15, 21]. Indeed, the association found between smoking cessation and income status in patients with CVD in this study supports the findings in the general population. For example, studies examining socio-economic status and smoking cessation in Western countries generally indicate that smokers from lower income groups are less likely to be successful in quitting smoking [22]. Furthermore, in hospitals, income-related inequalities in smoking cessation may be due to the degree of patient adherence to advice or recommendations regarding health behaviour change. Indeed, patients with chronic conditions belonging to a low-income category were less likely to adhere to the recommendations to reduce or stop smoking [23], and our results suggest that having been admitted to a hospital for a coronary event is less effective in serving as a “wake-up call” for patients with a low income.

Although awareness of tobacco as a CV risk factor is necessary for a behavioural change, it is not a sufficient reason to quit smoking, particularly for individuals in lower income groups. In our study, awareness of tobacco as a CV risk factor was not significantly associated with the probability of smoking cessation, even when we only adjusted for age and sex. However, by assuming an interaction between the income level of patients and their awareness of the effects of smoking on CVD, we obtained different results, which allowed us to understand one of the mechanisms that may explain the association between smoking cessation and income in patients with CVD. Indeed, the awareness of tobacco as a CV risk factor appears to play a mediating role (or has a modifying effect) between income and smoking cessation. Income was only associated with smoking cessation among patients who were aware of the health risks of tobacco. In addition, smokers with higher or medium incomes who were aware of tobacco as a CV risk factor were more likely to quit smoking.

Our study showed evidence of socio-economic inequalities in smoking cessation 5 years after a coronary angiography, where quitting smoking is an important measure of secondary prevention in patients with CVD. Indeed, quitting smoking reduces the risk of recurrent events, improves patient quality of life, and is cost-effective [30, 31]. This study also highlighted the influence of income on adopting positive behaviours regarding CVD risk factors (such as smoking) and showed that patients in the highest income groups were more likely to quit smoking. This influence may be explained by the existence of a significant interaction between awareness of tobacco as a CV risk factor and the level of income. Given the persistence of a relatively high level of smoking among people at high risk, significant efforts should be made to encourage and support smokers to quit smoking in order to improve social welfare and substantially reduce morbidity and premature death associated with tobacco. Thus, it is important that practitioners develop intervention strategies that enable disadvantaged patients to stop smoking. In light of the benefits of smoking cessation on health and society, intervention strategies combining behavioural and pharmacological therapies should be implemented in major health facilities and should target lower income groups.