Study design and sample

Data for this study came from China Health and Retirement Longitudinal Study (CHARLS) [11]. CHARLS is a longitudinal study that assessed the social, economic, and health status of a nationally representative sample of Chinese adults aged ≥45 years. The study covers 28 provinces, 150 county-level units and 450 village-level units and uses probability proportional to size (PPS) sampling. Regular follow-up with strict quality control maintained the representativeness of the middle-aged and elderly population, and formed a large-scale cohort data on aging problems in China [11].

The Biomedical Ethics Review Committee of Peking University approved the CHARLS study in January 2011. All participants were informed and provided informed consents before interviews. For this study, we applied to the CHARLS team and obtained anonymous data of participants.

The 2011 CHARLS survey provided baseline data. The follow-up rate of participants recruited in 2011 was 88% in 2013 and 87% in 2015. The present study included 3708 participants who were current smokers in 2011, and were followed up in both 2013 and 2015. We checked the data integrity, and excluded 109 participants with missing or abnormal data. Our study was a secondary analysis of data from the CHARLS.

Measurements

Ever smokers in the CHARLS study were defined as participants who smoked more than 100 cigarettes in their lifetime, according to answers to the following question: “have you ever chewed tobacco, smoked a pipe, smoked self-rolled cigarettes, or smoked cigarettes/cigars?”. Ever smokers were classified as current or former smokers (quitters) through the question “do you still have the habit or have you totally quit?”. Current smokers in 2011 were re-interviewed in 2013 and classified into two categories: continued smokers or quitters in 2013. Current smokers in 2013 were classified as remained smokers and quitters in the 2015 survey, and quitters in 2013 were classified as quitters or relapsers in 2015 (Fig 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Flow diagram for the process of participant selection and follow-up.

https://doi.org/10.1371/journal.pone.0240806.g001

The CHARLS study measured the level of nicotine dependence based on cigarettes per day, age at starting smoking, and time to the first cigarette after waking up. The CHARLS data also included the following information on participants: sociodemographic variables including age, sex, rural/urban location, educational level, marital status, region and self-perceived health condition. The perception of health by participants was measured by asking the following question: "Would you say your health is very good, good, fair, poor or very poor?"

Statistical analysis

The quit rate was estimated by dividing the number of quitters at the end of the follow-ups in 2013 and 2015 by the number of current smokers before the follow-up in 2011 and 2013 (Fig 1). Therefore, quit rate in this study was self-reported point quit prevalence. To facilitate comparisons with different studies, we estimated the annualized quit rate (q) based on an assumption of constant quit rate during the follow-up period: q = 1−exp {ln (1−nq/ns)/yr}, in which ns refers to smokers before the follow-ups, nq refers to the number of quitters at the end of follow-ups, and yr refers to the number of follow-up years. The annualized quit rate had been used in previous studies with different years of follow-ups [12].

Between smokers and quitters, the differences in characteristics in the frequency of categorical variables were analyzed using the chi-squared test. Logistic regressions modeled the association between smoking cessation (dependent variable) and participant characteristics (independent variables). The selection of independent variables was restricted by data availability, and according to our understanding of factors that may be possibly associated with the dependent variable. We used the following independent variables: age, sex, rural/urban location, marital status, geographical region, educational level, self-perceived health status, cigarettes per day, age at starting smoking, and time to the first cigarette after waking up. Data were analyzed using Stata/MP 16 (Statacorp, TX). Two sided p ≤0.05 was considered as statistically significant.

Findings and comparison with other studies

The quit rate of current smokers in 2011 was 16.6% at the 4-year follow up in 2015, which corresponded to an annualized quit rate of 4.4%. This estimate of annualized quit rate was surprisingly similar to findings from previous large-scale longitudinal studies in China and in other countries, given the considerable differences in smokers’ characteristics, socioeconomic and health care circumstances, and definition of smoking cessation outcomes. For example, a study in six cities in China found an annualized quit rate of 4.1% [10]. A study in the United States reported that 30.2% of the 5127 current smokers in 1993 stopped smoking by 2001, with an annualized quit rate of 4.4% [12]. A 1-year follow-up study in 1996 in the UK reported a quit rate of 5% in smokers [13]. In a longitudinal study of 4636 smokers from seven centers in Northern Europe, the crude quit rate was 4.5 per 100 person-years [14]. In a Japanese study, the quit rate in 1358 smokers after 1-year follow-up was 4.6% (≥3-months’ abstinence) or 3.1% (≥6 months’ abstinence) [15].

The current study found that smoking cessation was associated with age, region, perception of health status, and nicotine dependence, which are similar to previous studies [10, 16]. Older smokers are more likely to quit smoking, which may be partially due to increased concerns about health [10], and the inverse U-shaped relationship of nicotine dependence with age [15]. Smokers in northeast China had a lower smoking cessation rate, which is similar to the findings of the ITC study and likely due to differences in tobacco culture across regions in China [10].

The smoking cessation rate was relatively high in smokers with self-perceived poor health. According to the previous studies, concern about personal health was a common reason for smokers to quit smoking [17]. When smokers realize that smoking is affecting their health, they may be more motivated to quit smoking [18]. Therefore, doctors and family members should take the opportunity to promote smoking cessation in patients with smoking related diseases.

Our results confirm findings from previous studies that smoking cessation was associated with tobacco dependence, including the number of cigarettes smoked daily, age of smoking initiation and time to the first cigarette in the morning [7]. Compared with participants who were less addicted, smokers with more severe nicotine dependence were less likely to quit [19, 20]. Adult smokers with more health problems may be more motivated to quit smoking than smokers with fewer health problems [21].

There are various interventions to help smokers quit smoking, as specified by the World Health Organization Framework Convention on Tobacco Control (FCTC) and MPOWER (Monitor, Protect, Offer, Warn, Enforce, Raise) [13]. Strong evidence revealed that policies of tobacco control intervention reduced the smoking rate of general population [22]. However, the implementation of the FCTC in China must be further strengthened, including political leadership, government oversight of the tobacco industry, and advocacy and support [23]. Smoking cessation support in China is rarely available, and only 5.6% smokers in China used smoking cessation medications [24]. Therefore, smoking cessation in China was predominantly unassisted compared to quitters in other countries [25]. Therefore, cessation treatments were unlikely to be an important factor that directly affected the quit rate or motivation to quit in the study participants.

We also found that smoking relapse was lower in older adults, smokers with less severe tobacco dependence, and who had perceived poor health. In addition to the self-perception of poor health, previous studies found that smoking relapse was associated with perceived stress, anxiety or depression [9, 26]. Further research is required to improve our understanding of smoking relapse and related factors.

Strengths and limitations

The present study was based on data from a large, nationally representative longitudinal study that included urban and rural areas. Another strength of the study is the longitudinal data from two-waves of follow-ups, which provided rare opportunity for us to estimate quit rates and related factors in current smokers in China. However, this study has some limitations. First, the study relied on self-reported data and excluded participants with missing follow-up data, which may result in recall bias and errors. Second, we measured point prevalence of smoking status at the two follow-ups in 2013 and 2015, and did not assess possible changes in smoking status between the assessments. Third, we conducted multiple statistical comparisons, and did not correct for possibly inflated type I errors. Furthermore, we did not consider the impacts of smoking cessation interventions (e.g., tobacco control policies, smoking cessation therapy) on smoking and quitting. These limitations should be appropriately addressed in further studies.