Abstract

Background

Smoking is a well-documented risk factor for numerous chronic diseases, and cessation is correlated with enhanced health outcomes. Nonetheless, the precise effects of smoking cessation on the health status of older adults with chronic conditions in China have not been thoroughly quantified.

Objective

This study aims to quantitatively assess the correlations between smoking cessation and enhancements in the health outcomes of elderly Chinese individuals with chronic diseases.

Method

This research drew upon data from the China Health and Retirement Longitudinal Study (CHARLS). A cohort of 9914 participants was ultimately included in our analysis. Group comparisons and linear regression analyses were utilized. The investigation delved into health status scores, hematological markers, and physiological parameters.

Result

With each additional year of smoking cessation, former smokers demonstrated improved self-rated health and reduced EQ-5D-3L scores. Regression analysis unveiled a positive correlation between smoking cessation and enhanced self-assessed health (β estimate = 0.198), while a notable adverse effect was observed in EQ-5D-3L scores (β estimate = -0.179) and grip strength (β estimate = -2.530). Blood biomarkers also displayed noteworthy relationships with smoking cessation, showcasing rehabilitation in LDL cholesterol, total cholesterol, glucose, cystatin C, creatinine, HbA1c, and uric acid levels.

Conclusion

This research provides evidence highlighting the favorable health ramifications associated with smoking cessation in elderly individuals with chronic illnesses. Noteworthy improvements in both subjective health assessments and blood-based markers were observed post-smoking cessation, with benefits becoming more prominent with prolonged abstinence. These results underscore the vital importance of smoking cessation in the holistic care of chronic conditions and broader health enhancement endeavors. Further validation of these findings through an extended follow-up period is anticipated to bolster these conclusions with increased confidence.

Introduction

Smoking is widely acknowledged as a primary behavioral risk factor in global public health. Over the past seven decades, a plethora of studies [1,2,3,4] has elucidated the connections between smoking and a range of chronic conditions, including but not limited to cancer [5], hypertension [6, 7], stroke [8], asthma [9], and myocardial infarction [10]. Undeniably, tobacco smoking emerges as a crucial determinant of health in individuals with chronic diseases. Recent research has underscored the possibility for smokers to prolong their life expectancy by up to a decade through smoking cessation [11]. Moreover, smoking cessation has been recognized as a fundamental element in the management of chronic diseases [12,13,14]. Wang Z et al [15] screened 13,460 literatures and included 11 studies of chronic obstructive pulmonary disease (COPD) patients. Using systematic review and meta-analysis, they analyzed forced expiratory volume in one second percentage predicted (FEV1% predicted), FEV1/forced vital capacity (FVC), 6-min walk test (6-MWT), and etc. They concluded that smoking cessation is beneficial to COPD patients, which can improve lung function, alleviate symptoms, enhance exercise tolerance, and may reduce mortality. Pan B et al [16] searched databases and included 14 studies with 303,134 subjects. Through meta-analysis of smoking and stroke factors like smoking status and gender, they found smoking increases stroke risk and cessation helps. Other reviews and meta-analyses have also reflected the benefits of smoking cessation on chronic patients, such as asthma [17], and ischemic heart disease [18].

Notably, China leads the world in tobacco consumption, with Chinese men representing approximately 40% of global cigarette usage [19]. Disturbing trends in cigarette consumption among Chinese youth over the last two decades indicate a potential rise in mortality risks linked to tobacco use for men in urban and rural regions alike [20]. As reported in other research [21], the smoking prevalence of the adults over 45 years old in a province of China is over 20%, much more higher than that in Europe and the United States of America.

Consequently, this study aims to quantify the correlations between smoking cessation and enhancements in the health status of elderly Chinese individuals with chronic conditions. All data utilized in this research were sourced from the China Health and Retirement Longitudinal Study (CHARLS) database. The most recent update of the CHARLS database in November 2023 enables us to access the most up-to-date information from China. The analysis encompassed health status assessments, blood biomarkers, and physical measurements.

Methods

Study design and sample source

The CHARLS represents a comprehensive initiative aimed at gathering high-quality microdata concerning Chinese households and individuals aged 45 and older. The inaugural national survey of CHARLS took place in 2011, encompassing 450 villages, 150 countries, and 28 provinces, involving approximately 17,000 individuals from 10,000 households. This survey captures fundamental demographic details about the participants and their families, intra-family financial transfers, the health status of respondents, healthcare access and insurance coverage, employment particulars, income, expenditures, assets, and more. Moreover, CHARLS involved 13 physical assessments and the collection of blood samples. To date, CHARLS has released five waves of data: the national baseline survey (Wave 1, 2011), and subsequent follow-up surveys (Wave 2, 2013; Wave 3, 2015; Wave 4, 2018; Wave 5, 2020). The CHARLS datasets are accessible for download on the CHARLS homepage at . Approval for the CHARLS survey project was obtained from the Biomedical Ethics Committee of Peking University, and all participants provided informed consent [22].

In this study, we selected data from all five waves of the CHARLS database. The inclusion criteria were as follows: (1) individuals diagnosed with any chronic disease in Wave 1; (2) individuals who underwent treatment for chronic diseases. To avoid bias stemming from treatment for chronic diseases and to ascertain the true impact of smoking cessation, we only selected individuals who received treatment. These chronic diseases encompassed hypertension, diabetes, cancer, chronic lung disease, heart attack, stroke, psychiatric problems, arthritis or rheumatism, dyslipidemia, any liver disease, any kidney disease, any stomach or other digestive disease, and asthma. Based on previous studies [5,6,7,8,9,10], hypertension, asthma, cancer, chronic lung disease, heart attack, and stroke were considered to be related to smoking. The exclusion criteria were as follows: (1) individuals aged under 45 years in Wave 1; (2) individuals quitting smoking outside the follow-up period. We used the STROBE cohort checklist when writing our report [23].

Variables

To demonstrate the association between smoking cessation and chronic disease status, the following variables were selected for analysis: (1) demographic factors including age, gender, marital status, educational background, public insurance, and residence area; (2) smoking status and duration since quitting; (3) chronic disease status and specific disease types. These participants were initially classified into three main groups based on their smoking status: non-smokers, current smokers, and former smokers. According some similar researches [24, 25], former smokers were defined as individuals who previously smoked but have since discontinued this habit.

To assess health status, we utilized various metrics, including self-rating, blood analyses, physical measurements, and the EuroQoL 5-Dimension 3-Level (EQ-5D-3L) questionnaire. In CHARLS, participants were requested to rate their health on a scale ranging from 1 to 5, with 1 indicating the worst and 5 indicating the best health. Blood analyses were conducted in Wave 1 and Wave 3, while physical measurements were conducted in Wave 1, Wave 2, and Wave 3 [26], although not all respondents underwent these assessments. The blood analyses encompassed parameters such as hemoglobin, hematocrit, white blood cell count (WBC), platelet count, mean corpuscular volume (MCV), C-reactive protein (CRP), glycosylated hemoglobin, type A1c (HbA1c), total cholesterol, high-density lipoprotein (HDL) cholesterol, low-density lipoprotein (LDL) cholesterol, triglycerides, glucose, blood urea nitrogen (BUN), creatinine, uric acid, and cystatin C. Physical measurements included walk measurements, blood pressure measurements, hand grip strength measurements, body mass index (BMI), lung function measurements, balance tests, and chair stand tests. Detailed units for these variables can be found in Table S-1.

The EQ-5D-3L questionnaire was employed to assess the health-related quality of life (HRQOL) of our participants. HRQOL was evaluated across five dimensions (5D): mobility (MO), self-care (SC), usual activities (UA), pain/discomfort (PD), and anxiety/depression (AD). Each dimension comprised three levels (3L): no problem, moderate problems, and severe problems [27]. Previous research has successfully developed Chinese utility values for EQ-5D-3L health states using the time trade-off method and established a scale and formula specific to Chinese residents [28,29,30](Table S-2). Based on other researchs [6], we selected five parts from CHARLS for each dimension: MO: Difficulty with Shopping for Groceries, SC: Difficulty with Preparing Hot Meals, UA: Difficulty with Household Chores, PD: Troubled with Body Pain, AD: Center for Epidemiologic Studies Depression Scale-10 (CESD-10) questionnaire in CHARLS, recognized as an effective mental health assessment tool for the elderly [31, 32].

Data analysis

Descriptive analyses of demographics and health status were conducted using frequency counts, proportionate ratios, means, and standard deviations. Group comparisons for continuous variables were performed using the Wilcoxon rank-sum test. For former smokers, the parameters before and after the cessation were calculated. For example, if a respondent quit smoking in 2014, the record in Wave 3 (2015), Wave 4 (2018), and Wave 5 (2020) would be described as 1-year cessation, 3-year cessation, and 6-year cessation respectively. The average of Wave 1 and Wave 2 was considered a pre-cessation value, and the average of the other three waves was considered a post-cessation value. All blood analysis outcomes were natural log-transformed to improve normality, and thus, results were presented as percentage differences: (exp (β-coefficient) -1) × 100%. To investigate the relationships between smoking status and the parameters included in the study, univariable and multivariable linear regression analyses were employed. Model 1 (M1) featured a binary indicator for smoking cessation status, while Model 2 (M2) adjusted for demographic variables. Moreover, the duration since quitting smoking was taken into account. The statistical analyses were executed using R (version 4.3.3) with a significance level of α = 0.05. Most of the results were rounded up to two decimals places.

Result

Descriptive statistics

Finally, a cohort of 9,914 individuals was identified for subsequent investigations (Fig.Ìý1). Table 1 details the descriptive analysis of demographic characteristics. The mean age of the former group was slightly higher than the overall. Around 80% of male participants enrolled in the study had a history of smoking, aligning with findings from prior research studies.

Fig.Ìý1

Flowchart for inclusion and exclusion in our study

TableÌý1 Characteristics of included respondents in the study

Table 2 presented the distribution of respondents with various chronic diseases, revealing that hypertension and arthritis or rheumatism emerged as the predominant chronic conditions among the participants. Each chronic ailment displayed a relatively consistent prevalence across all groups. However, concerning respiratory system disorders, the prevalence rate among current smokers was approximately double that observed among non-smokers.

TableÌý2 Number of respondents with each type of chronic disease

Increasing self-rate and decreasing EQ-5D-3L scores

FigureÌý2A showed the trend of self-rate and EQ-5D-3L scores among former smokers. Upon analysis, we observed a progressive rise in self-rating scores over time, contrasting with a declining trend in EQ-5D-3L scores. FigureÌý2B illustrates a comparison between pre- and post-cessation periods. It revealed that individuals who quit smoking exhibited higher self-rating scores but lower EQ-5D-3L scores. Therefore, a detailed examination of EQ-5D-3L data indicated that, in comparison to smokers, quitters demonstrated an increase in the AD coefficient by 0.117, along with incremental changes in the SC, UA, and PD coefficients by 0.017, 0.018, and 0.015, respectively. This analysis suggested that the reduction in EQ-5D-3L scores primarily stemmed from the elevated AD coefficient (Table S-3).

Fig.Ìý2

A The trend of self-rate and EQ-5D-3L scores for former smokers. The red and blue numbers indicate the number of respondents for each corresponding quitting year.ÌýB Self-rate and EQ-5D-3L scores for former smokers before and after cessation. *: p < 0.05, **: p < 0.01, ***: p < 0.001, ****: p &±ô³Ù; 0.0001

Health status comparisons between all subgroups

As shown in TableÌý3, almost all parameters in the current smoker and pre-cessation groups displayed similar levels, except for self-rate, EQ-5D-3L, HbA1c, grip strength, and lung function. In blood analyses, compared with the pre-cessation group, the post-cessation group demonstrated a decline in hemoglobin, total cholesterol, LDL cholesterol, and cystatin C, alongside an increase in creatinine, HbA1c, and uric acid. Non-smokers exhibited lower values in hemoglobin, hematocrit, MCV, BUN, creatinine, uric acid, cystatin C, limb strength, and lung function, while a higher level in HbA1c compared to the pre-cessation group.

TableÌý3 Group comparisons between different smoking status

Regression analysis revealed health benefits

Table S-4 presented the results of linear regression analyses examining the relationship between smoking status and all parameters, with all significant outcomes depicted in Fig.Ìý3. A similar trend was observed in Model 1 (M1) and Model 2 (M2), except for several parameters. Notably, the analysis under M2, which accounted for demographic variables, revealed several noteworthy associations. A significant reduction in EQ-5D-3L (β estimate = -0.18) and grip strength (β estimate = -2.53) was observed. However, a significant positive correlation was found between self-rate and smoking cessation (β estimate = 0.20). In the realm of blood analyses, smoking cessation exhibited significant associations with various blood biomarkers after adjusting for demographic factors: 12.03% lower LDL cholesterol, 7.70% lower total cholesterol, 7.26% lower glucose, 5.50% lower cystatin C, 4.67% lower hemoglobin, 3.39% higher creatinine, 11.40% higher HbA1c, and 12.03% higher uric acid.

Fig.Ìý3

Forest plot of linear regression. Model 1 was not adjusted. Model 2 was adjusted for demographic factors. Only significant outcomes were plotted. Blood analysis outcomes were natural log-transformed and presented as percentage differences (A). Other outcomes were not transformed (B). *: p < 0.05, **: p < 0.01, ***: p < 0.001, ****: p &±ô³Ù; 0.0001. Abbreviation: EQ-5D-3L: EuroQoL 5-Dimension 3-Level; LDL: low-density lipoprotein; HbA1c: glycosylated hemoglobin, type A1c

Considering cessation duration, trend analysis was performed and several significant associations were identified. With each additional year of smoking cessation, self-rate increases by 0.036, while EQ-5D-3L decreases by 0.031. Additionally, glucose decreases by 2.30%, HbA1c increases by 3.29%, uric acid increases by 3.79%, and walk time decreases by 0.474 (TableÌý4).

TableÌý4 Trend analysis of smoking cessation duration and parameters

Furthermore, based on the chronic disease classification mentioned above, we divided the respondents into two subgroups: those with smoke-related diseases and those without (Table S-5, Table S-6). However, a similar trend was shown in the two groups and no distinctive outcomes were observed.

Additionally, significant associations were confirmed between smoking behavior and certain parameters. Current smokers demonstrated a worse health status in comparison to non-smokers, which is suitable for prevailing perceptions (Table S-7).

Discussion

Our discoveries underscored the manifold advantages of abstaining from smoking, consistent with the prevailing scientific consensus regarding the adverse health consequences of tobacco consumption. We successfully quantify the enhancements in health status and blood indicators.

Our findings revealed noteworthy enhancements in specific health markers. The rise in hemoglobin levels can be attributed to the formation of carboxyhemoglobin, a product of the combination of carbon monoxide released from smoke and hemoglobin, leading to tissue hypoxia, heightened erythropoietin secretion, and increased erythropoiesis [33]. Elevated serum cholesterol and LDL levels, previously noted in smokers and considered a risk factor for cardiovascular disease [34], may result from modification of apolipoproteins [35]. Our findings demonstrate an improving trend after smoking cessation. In a prior meta-analysis, smokers were found to have higher HbA1c levels than non-smokers [36], but we found HbA1c increased after cessation, which was similar to other researchers’ findings [37, 38]. It was pointed out that the level of HbA1c would not immediately decrease in the short term after cessation. Furthermore, a decreased glucose level was observed after cessation, possibly linked to nicotine’s stimulation of insulin-antagonizing hormones [39]. Prior studies demonstrated a substantial decrease in cystatin C levels after smoking cessation, possibly indicating a reduction in cardiovascular risks [40]. The outcome of our study supported a recovery in blood parameters after smoking cessation.

Individuals who quit smoking demonstrated a notable decline in EQ-5D-3L scores, indicating a potential initial reduction in the quality of life. In previous studies [41,42,43], the benefits from smoking cessation have been proved in certain chronic diseases, especially in respiratory system disease. Respondents’ self-rate is often based on daily-life feelings like reduced coughing or easier breathing, which may make them more satisfied with their quality of life. However, the improvement in blood analyses may have less impact on the quality of life. The inverse correlation between smoking cessation and EQ-5D-3L scores necessitates cautious interpretation, possibly also attributable to withdrawal symptoms or lifestyle adjustments, which was proved by a decreased AD coefficient. Smokers may encounter irritability, anxiety, and depression upon abstaining from smoking, which is reliably alleviated by smoking, despite the fact that smoking initially induced these psychological disturbances [44]. Since we combined and analyzed the effects of multiple chronic diseases, the impacts of smoking cessation on different chronic diseases may mask each other. In previous researches, the outcome of smoking cessation is still uncertain. There is still debate about the benefits of smoking cessation for patients with different diseases. In COPD patients [42] and psychiatric problems patients [44], smoking cessation would lead to an improvement in quality of life, however, in cardiovascular disease patients [43, 45, 46] and arthritis patients [47], the benefits are still controversial. Meanwhile, it is worth mentioning that few research focused on older adults as we did. In general, the health status deteriorates as people get old. We must pay attention to the impacts brought about by advanced age.

The regression models provide a nuanced comprehension of the correlation between smoking cessation and health outcomes. Each model was meticulously crafted to sequentially account for potential confounding variables, thus isolating the impact of smoking cessation. M1, which specifically adjusts for smoking cessation status, revealed significant associations that were predominantly consistent across the adjusted models (M2) and the trend analysis, underscoring the robustness of our findings. The consistent results across models, particularly the notable associations observed in blood analyses, underscored the potential advantages of smoking cessation. However, the adverse association with EQ-5D-3L highlights the necessity for supportive interventions to address the immediate challenges quitters face.

The CHARLS dataset serves as a robust foundation for analysis due to its extensive sample size and longitudinal framework. However, it is important to acknowledge limitations such as potential selection bias and the generalizability of our findings to non-Chinese populations. Additionally, there remain other confounders that might not have been controlled for, even though we have corrected for a variety of possible confounding factors. The detailed treatment for chronic diseases was hard to classify. For example, we cannot be sure whether Chinese traditional medicine is suitable for patients. Besides, the use of self-made cigarettes by many individuals in rural areas presented difficulties in quantifying smoking amounts and detailing smoking behaviors. Moreover, the absence of access to medical records in the CHARLS dataset meant that respondents relied on memory to report physician-diagnosed diseases, potentially introducing recall bias. Finally, the cross-sectional nature of the data limits our ability to establish causality. Moving forward, it is imperative to conduct future longitudinal studies with larger sample sizes and extended follow-up periods to validate these findings and explore the temporal dynamics of smoking cessation's effects on health outcomes more comprehensively.

These findings advocate for the integration of smoking cessation interventions into routine care for older adults with chronic diseases, suggesting significant health benefits that warrant further exploration in clinical settings. Future studies should examine the specific effects of smoking cessation on various chronic disease populations and investigate its long-term impacts on healthcare utilization and costs.

Conclusion

The study's findings underscore the significant health benefits of smoking cessation among older Chinese adults with chronic diseases. Despite an initial decline in perceived quality of life, as indicated by EQ-5D-3L, long-term health markers demonstrate substantial improvements. The favorable trends in blood parameters highlight the physiological advantages of abstaining from smoking. These results reinforce the critical need to incorporate smoking cessation assistance into healthcare protocols, especially for elderly individuals dealing with chronic conditions. The study calls for additional research to investigate the precise effects of smoking cessation on different chronic disease populations and the factors driving smokers to quit.