INTRODUCTION

Heart failure (HF) is a rapidly growing public health issue affecting at least 2–4% of the global population according to the World Health Organization (WHO)¹. It is characterized by poor quality of life, recurrent hospitalization, and increased mortality risk, due to worsening HF symptoms². The main goal of HF treatment is to reduce symptoms, morbidity and mortality, and improve the quality of life^3,4. According to the European Society of Cardiology Guidelines for the diagnosis and treatment of HF⁵, self-management is integral to achieving best patient outcomes, to reduce mortality and improve quality of life. Such self-care management includes smoking cessation, blood pressure control, lipid management, weight reduction, and symptom monitoring⁶.

Several epidemiological studies have documented an association between lifestyle variables, such as smoking, obesity and dietary pattern, and HF risk in the general population^6-8. In particular, smoking is the most significant modifiable risk factor for cardiovascular disease (CVD) including HF⁹. The WHO reported that 10% of all deaths due to CVD can be attributed to smoking¹. Importantly, persistent smoking after HF diagnosis has been shown to worsen the long-term outcomes in HF and to reduce the efficacy of HF treatment¹⁰. Specifically, smoking not only increases the risk of coronary artery disease, a major cause of HF, but also increases systolic and diastolic blood pressure and oxidative stress. Furthermore, smoking can lead to vascular inflammation, worsening endothelial function and renal function, all of which have been implicated in the pathophysiology of HF¹¹.

Previous studies have reported that despite the recommendation to quit smoking after CVD diagnosis, many patients continued smoking, and their smoking behaviors varied extensively^10,12. A study by Lim et al.¹³ reported that approximately 50% of smokers continued smoking even after CVD events. Another study also reported that about 16% of participants with HF were currently smoking¹⁴. A meta-analysis of 14 studies showed that the smoking cessation rates after ischemic heart disease ranged from 7% to 63%¹⁵. Some studies have suggested that poor adherence to HF treatment, including smoking cessation, increases the risk of adverse health outcomes^15,16.

Despite this possible association between persistent smoking and adverse health outcomes among HF populations, to our knowledge, there has been no consensus on the magnitude of smoking-related morbidity and mortality in patients with HF^12,13. A recent review reported that more cardiovascular events in heavier smokers have failed to find a significant negative correlation between smoking habits and cardiovascular risk¹⁰. Furthermore, few studies have systematically reviewed the literature regarding this relationship. Therefore, the purpose of this study was to examine the association between persistent smoking and the risk of adverse health outcomes in patients with HF, through a critical appraisal of the literature.

METHODS

Study selection

Two independent reviewers screened the titles and abstracts of each included study. At this stage, the reviewers extracted possibly all relevant studies. To be included, studies and peer-reviewed articles needed to be: 1) published in English, and 2) original research using any observational study design (e.g. cross-sectional or cohort) that reported on smoking status in patients with HF as a primary study group. Studies excluded were: 1) intervention studies, reviews, conference abstracts, and letters; 2) studies with no outcome; and 3) studies with insufficient data. Disagreements on eligibility were reconciled by the reviewers. Because no primary data were to be collected, approval from an ethics committee was not required (Figure 1).

Figure 1

PRISMA flow diagram for study selection

RESULTS

The impact of smoking on adverse health outcomes

In the present study, persistent smoking showed association with mortality (n=5), readmission (n=3), and disease-specific health status (n=1), ventricular tachycardia (n=1), arterial stiffness (n=1), cognitive function (n=1), and length of stay (n=1) after the diagnosis of HF.

Mortality

Five studies provided data on mortality among those who continued smoking after HF diagnosis. Three of the studies used prospective cohort study design to assess mortality^21,22,26, one used a retrospective cohort study design27, while the other used a cross-sectional design¹⁴. One was conducted in Sweden, while four reported data from the same population in the USA.

Three of the included studies were metaanalyses14,21,27. The risk of mortality was greater in current smokers than in non-smokers (combined HR=1.384; 95% CI: 1.139–1.681; n=8958). The pooled results showed low heterogeneity (I2=5.5%; p=0.347) (Figure 2).

Figure 2

Mortality in current smokers vs non-smokers or never smokers

Subgroup analysis revealed a higher mortality risk ratio of never smokers versus current smokers of 1.820 (95% CI: 1.191–2.782; 1 study, n=1717) than that of non-smokers versus current smokers (1.402; 95% CI: 0.955–2.060; 2 studies, n=7241). Although two studies noted that continuous smoking after HF diagnosis was independently associated with mortality, they provided no statistical data^22,26.

Readmission rates

Three studies reported on the association between readmission and persistent smoking after HF diagnosis. Two of such studies used retrospective cohort study design to assess readmission^20,27 while the other was a prospective cohort study21. All studies were conducted in the US. The risk of readmission was greater in current smokers than in non-smokers (combined HR=1.448; 95% CI: 1.086–1.930; n=7457). The pooled results showed moderate heterogeneity (I²=57.14%; p=0.097) (Figure 3). Subgroup analysis revealed a higher readmission risk ratio of never smokers versus current smokers of 1.680 (95% CI: 1.081–2.612; one study, n=1717) than non-smokers versus current smokers (1.402; 95% CI: 0.955–2.060; 2 studies, n=7457).

Figure 3

Readmissions for heart failure in current smokers vs non-smokers or never smokers

Other adverse health outcomes

One cross-sectional study conducted in the US examined the disease-specific health status depending on the smoking status after HF diagnosis¹⁴. Statistically significant differences in the diseasespecific health status were observed between the smoker and non-smoker groups in terms of physical function, symptoms, social function, self-efficacy, and quality of life (F=3.47; p=0.03; n=537). In particular, the disease-specific health status was lower in current smokers than in never smokers (p=0.01) or former smokers (p=0.02). One prospective cohort study conducted in the US examined the incidence of ventricular tachycardia (VT) during sleep depending on the smoking status after HF diagnosis²³. The incidence of VT was greater in current smokers than in non-smokers (OR=9.96; 95% CI: 1.93–51.48; n=87). One cross-sectional study conducted in China examined the incidence of arterial stiffness depending on the smoking status after HF diagnosis²⁵. Current smoking was an independent determinant of arterial stiffness (β=0.121; p=0.013; n=354). One retrospective cohort study conducted in the US examined the cognitive function depending on the smoking status after HF diagnosis²⁴. Both current (β=0.321; p=0.016; n=2043) and former smokers (β=0.204; p=0.041) had higher cognitive function than never smokers. One prospective cohort study conducted in the US examined the length of stay in the hospital, depending on the smoking status after HF diagnosis²². Smokers had similar length of stay in the hospital compared to non-smokers (relative risk, RR=0.37; 95% CI: 0.94–0.99; n=48612).

DISCUSSION

Despite the strong association of smoking with cardiovascular disease and adverse events, little is known about the impact of persistent smoking after a HF diagnosis on patient outcomes. In this review, we found that persistent smoking after HF diagnosis, compared with non-smoking (former or never smoking), increased the risk of adverse health outcomes. In particular, our main finding was that current smokers, who continued smoking after HF diagnosis, were more susceptible to mortality and readmissions than non-smokers.

Our meta-analysis demonstrated that mortality was the most frequently examined variable out of the seven types of adverse health outcomes associated with smoking status. Furthermore, persistent smokers showed higher mortality than non-smokers. However, we could not verify whether the higher mortality among persistent smokers was primarily ascribable to the functional deterioration of the cardiovascular system or to other causes, given that most data used in the studies indicated ‘all-cause mortality’ as the cause of death. Specifically, chemical constituents of smoke have high oxidant and inflammatory power that can directly induce endothelial damage and potentiate inflammatory response^28-30. Namely, cigarette smoking can promote atherosclerosis by its effects on lipid profile^25,31. Moreover, according to one study³², nicotine along with carbon monoxide and oxidative stress induce fibrosis at different cardiac sites; thus, generating a structural remodeling that may favor cardiac arrhythmia. For these reasons, the damaged cardiovascular system induces myocardial systolic and diastolic dysfunction through smoking, thus aggravating HF symptoms²¹, which can result in death. Consequently, further studies are necessary to completely understand the exact mechanisms underlying the vascular damage induced by smoking. Such knowledge can help healthcare providers to educate active or heavy smokers after a HF diagnosis against smoking.

In previous studies, tobacco smoking was associated with higher risk of cognitive decline. These associations are thought to be caused mainly by smoking-related damage, which develops during cardiovascular and respiratory processes in the brain^33,34. However, in a study in which cognitive function was reported depending on the smoking status, current smokers were found to have higher cognitive function than never smokers²⁴. This may be explained by the fact that a cognitive function decline is associated not only with smoking but also with other factors, such as age, hemoglobin level, and body mass index^35,36. These factors act as confounding variables, and there is a need to adjust for them. Moreover, only one of the studies analyzed in this systematic review examined the association between cognitive function and smoking status after HF diagnosis, and the validity of its findings will have to be verified by additional research. In one study that examined the length of stay in the hospital in relation to smoking status, persistent smokers and non-smokers showed similar lengths of stay²². This study noted that the sudden smoking cessation after hospitalization might have resulted in the improved health status. Although the average length of stay in the hospital reported in this study was 4 days, too short a period of smoking cessation to improve health. Taken together, HF patients who are smoking at diagnosis may benefit from counseling regarding the harms of continued smoking after a diagnosis of HF, as one way to improve quit rates. Secondhand smoke exposure is related to increased risk of CVD and ischemic heart disease^37,38. Thus, smoke-free policies within hospitals can help persistent smokers to attempt to quit smoking.

In our review, of 70461 participants, 11252 (16%) from nine studies continued smoking after HF diagnosis. Similar percentages were found in previous studies for diseases associated with smoking: 14.1% of 1700 patients with CVD¹³; 17.6% of 5185 patients diagnosed with cancer³⁹; and 17% of 893 patients with lung cancer⁴⁰. Most of the patients diagnosed with HF undergo interventions designed to prevent the aggravation of HF. Such interventions include smoking cessation, low-salt diet, abstention from drinking, regular exercise, and management of risk factors such as hypertension, diabetes, and hyperlipidemia⁴¹. The finding of this study showing that 16% of patients with HF failed to quit smoking highlights the need of customized health education for the patients by identifying the causes of persistent smoking and negative attitudes towards health-promoting behaviors.

Our review of the studies included, found that the follow-up period (12–41 months) was not sufficiently long to verify the adverse health outcomes for persistent smoking after a HF diagnosis. A cohort study should be performed to clearly indicate the temporal sequence between exposure and outcome, or with a long follow-up period in case of a low incidence rate⁴². In particular, a lifestyle variable like smoking requires a study design providing sufficient follow-up for the target outcome, given its long-term negative health consequences. Furthermore, the smoking intensity, such as cumulative smoking in pack-years, number of cigarettes smoked per day, and duration of smoking, was not sufficiently examined. In previous studies, dose effect association between pack-years of exposure and HF risk was observed⁴³. Furthermore, current smokers’ risk of HF increased as their daily tobacco consumption increased⁴⁴. That is, since smoking affects the cardiovascular system to a greater or lesser extent depending on smoking intensity, strategies for smoking reduction will have to be developed by investigating the threshold smoking intensity for exacerbating HF. For HF diagnosis, medical records (objective data) were used in all studies, while NYHA functional class, LVEF, and BNP, were used to verify the severity of HF. NYHA functional class is a subjective measure useful for determining the motor skills and symptoms of patients with HF and widely used for monitoring patients with HF⁴⁵. NYHA functional class and LVEF were well documented in 77.8% of the included studies, to describe the subjective and objective characteristics of the patients with HF. When the level of LVEF is ≥40%, the ventricle volume increases, and pressure overload must be monitored by additionally checking the increase in BNP^46,47. Of the two studies reporting LVEF ≥40% in this systematic review, only one investigated BNP. To ensure a clear explication of the association between smoking status after HF diagnosis and adverse health outcomes, NYHA, BNP and LVEF will have to be clearly determined as the diagnostic criteria for HF.

CONCLUSIONS

Our study showed that approximately 16% of smokers continued smoking even after HF diagnosis. This supports the need to assess patients’ smoking status during outpatient follow-up visits. Importantly, our review identified that adverse health outcomes were more common in persistent smokers after HF diagnosis than in non-smoker patients. Healthcare professionals should create awareness regarding the harm of continued smoking in HF patients. Particularly, patients willing to try to reduce or stop smoking should be encouraged. Further studies are needed to explore underlying reasons why HF patients continue to smoke after a HF diagnosis, in order to develop patient-centered smoking cessation strategies. Furthermore, the causal relationships between persistent smoking and various aspects of adverse health outcomes can be determined with a large and culturally diverse HF population.