Smoking cessation and mortality among middle-aged and elderly Chinese in Singapore: the Singapore Chinese Health Study

Abstract

Objective

This study determines if recent smoking cessation, compared with long-term cessation, can reduce mortality risk associated with smoking.

Methods

Data from the Singapore Chinese Health Study, a cohort study of middle-aged and elderly Chinese in Singapore, were analysed (n=48 251). Smoking status was evaluated at recruitment between 1993 and 1998 and reassessed between 1999 and 2004. Participants were classified as never-smokers, long-term quitters (quit before recruitment, mean 17.0 years), new quitters (quit between recruitment and second interview, mean 4.3 years) and current smokers. Mortality was ascertained by linkage with the nationwide death registry.

Results

After a mean follow-up of 8.1 years, 6003 deaths had occurred by 31 December 2009. Compared with current smokers, the adjusted HR (95% CI) for total mortality was 0.84 (0.76 to 0.94) for new quitters, 0.61 (0.56 to 0.67) for long-term quitters and 0.49 (0.46 to 0.53) for never-smokers. New quitters had 24% reduction in lung cancer mortality (HR: 0.76, 95% CI 0.57 to 1.00) and long-term quitters had 56% reduction (HR: 0.44, 95% CI 0.35 to 0.57). Risk for coronary heart disease mortality was reduced in new quitters (HR: 0.84, 95% CI 0.66 to 1.08) and long-term quitters (HR: 0.63, 95% CI 0.52 to 0.77), although the result for new quitters was of borderline significance due to relatively small number of cardiovascular deaths. Risk for chronic pulmonary disease mortality was reduced in long-term quitters but increased in new quitters.

Conclusion

Significant reduction in risk of total mortality, specifically for lung cancer mortality, can be achieved within 5 years of smoking cessation.

The evidence for the causal relationship between cigarette smoking and cardiovascular disease, cancer and respiratory disease is strong and universally accepted.¹ Since the 1960s, cigarette smoking has been consistently found to predict the incidence of lung cancer, cardiovascular disease, respiratory disease and death from these diseases in both the Western and Asian populations.^{2 3} In many developing countries in Asia, while cigarette smoking is becoming popular, the awareness and knowledge of the associated health risks remain low. It has been estimated that 30% of the global cigarettes are smoked in China alone.⁴ Smoking-related diseases have, in fact, replaced infectious disease as major health problem in many Asian countries.⁵

Because of vigorous tobacco control measures, the prevalence of smoking in Singapore is lower than that of its neighbouring countries.⁶ Even so, smoking-related diseases such as lung cancer, coronary heart disease (CHD) and chronic obstructive pulmonary disease (COPD) remain among the leading causes of death in Singapore. In 2007, an estimated 13.8% of Singaporeans aged 18 to 69 years smoked cigarettes daily, with a higher prevalence in men (23.7% of men vs 3.7% of women).⁷ One important public health strategy to reduce the mortality and morbidity associated with smoking is to encourage current smokers to stop smoking.

Smoking cessation in the Asian populations has yielded positive results in reduction of mortality; quitting smoking has been shown to reduce total mortality^8–11 as well as mortalities from specific causes such as cancer^11–13 and cardiovascular disease.^{11 12 14} These results concur with the finding from Western studies that quitting smoking reverses mortality risk.^15–17 However, it remains unknown how fast the benefits of smoking cessation can be observed after quitting smoking. In this study, we used data from a population-based cohort study in Singapore to determine whether cessation in smoking within about 5 years is associated with significant reduction in smoking-related mortality risk among middle-aged and elderly persons in an Asian population.

METHODS

Study population

The Singapore Chinese Health Study is a prospective cohort study of diet and health among middle-aged and elderly Chinese in Singapore. The design and methodology have been described elsewhere.¹⁸ Briefly, the cohort was drawn from permanent residents or citizens of Singapore who live in housing estates built by the government (about 86% of Singaporean Chinese lived in government apartments at the time of recruitment) and who were aged between 45 and 74 years at the time of recruitment between April 1993 and December 1998. Recruitment was limited to the Hokkiens who originated from the southern part of Fujian Province and the Cantonese who came from the central region of Guangdong Province. These two provinces are contiguous prefectures in southern China. Since marriage across dialect groups among the parents of our cohort participants was minimal, this cohort can be considered to be remarkably homogeneous in its genetic make-up. Approximately 85% of contacted people agreed and participated in the cohort study. All surviving cohort members were re-interviewed between 1999 and 2004. The Institutional Review Boards at the National University of Singapore and the University of Minnesota approved this study. Written informed consent was obtained from all participants.

Inclusion/exclusion criteria

A total of 63 257 participants were interviewed at baseline. At the follow-up interview conducted between 1999 and 2004, about 17% were lost to follow-up due to death, physical disability or because they were no longer contactable. A total of 52 322 participants were successfully re-contacted, and the mean period between interviews was 5.8 (SD 1.5) years. Participants who started smoking after the baseline interview (n=2564) and those whose responses were inconsistent (n=1507) (eg, those who said that they were current smokers at baseline but never-smokers at the second interview) were excluded, leaving a final sample of 48 251 in the analyses for this study.

Exposure assessment

At the time of recruitment (1993–1998), face-to-face interviews were conducted by trained interviewers at each participant’s home. A structured questionnaire was used to obtain information on lifestyle factors such as current diet, lifetime use of tobacco, alcohol intake, physical activity as well as medical history and reproductive history (for women only). At the follow-up interview (1999–2004), cigarette smoking, other health behaviours and medical histories were assessed, either in-person or by phone, using a standardised questionnaire.

In the baseline interview, participants were grouped into three groups according to the answers given to the following question, “Have you ever smoked at least 1 cigarette a day for 1 year or longer?” Subjects who answered ‘no’ were classified as ‘never-smokers’, those who answered ‘yes, but I quit smoking’ were classified as ‘former smokers’ and those who answered ‘yes, and I currently smoke’ were classified as ‘current smokers’. At the follow-up interview, participants were asked the following question, “Have you ever smoked more than 100 cigarettes in your lifetime?” Those who answered ‘no’ were classified as ‘never-smokers’ and those who answered ‘yes’ were asked a second question, “Have you smoked at least one cigarette in the past 30 days?” Those who answered ‘no’ were classified as ‘former smokers’ and those who answered ‘yes’ were classified as ‘current smokers’. In both surveys, we also obtained information on number of cigarettes smoked per day and age of smoking initiation from ever-smokers, and if they had quit, the number of years since quitting.

Smoking and smoking cessation categories

Based on smoking status ascertained from the baseline and follow-up interviews, participants were classified into four groups: current smokers reported smoking at both interviews, long-term quitters reported having quit smoking at baseline and remained so at follow-up interview, new quitters were smoking at baseline but had quit before follow-up interview and never-smokers were those who had never smoked.

Ascertainment of mortality

Deaths were identified through record linkage with the Singapore Registry of Births and Deaths. For the current analysis, we updated mortality data through 31 December 2009. As of June 2011, only 47 subjects were known to be lost to follow-up due to migration out of Singapore or for other reasons. This suggests that emigration among these subjects was negligible and that vital statistics at follow-up was virtually complete.

Underlying causes of death were coded according to the International Classification of Diseases, Ninth Revision; codes 140–208 for cancer deaths, code 162 for lung cancer death, codes 410–414 for CHD deaths, codes 430–438 for stroke deaths and codes 490–496 for COPD deaths.

Statistical analysis

Demographic and health characteristics of the four smoking categories were compared using χ² test for categorical variables and analysis of variance for continuous variables. For each participant, person-years of follow-up were counted from the date of the follow-up interview to the date of death or 31 December 2009, whichever occurred first. The Cox proportional hazards model was used to examine the association between smoking cessation and the risk of all-cause mortality. Covariates in the model were age at follow-up interview, interval between baseline and follow-up interview (year), year of follow-up interview, body mass index at follow-up interview (<20, 20 to <24, 24 to <28, 28+), sex, dialect (Hokkien, Cantonese), education (no formal education, primary school, secondary school or higher), total ethanol consumption at follow-up interview (grams per day), physical activity at baseline interview (no, 0.5–3, 4+ h/week), self-reported history of physician-diagnosed hypertension, diabetes mellitus, stroke and cardiovascular disease, and diagnosis of cancer through linkage with the national cancer registry (none, reported at baseline interview, not reported at baseline but at follow-up interview).

In the analysis of cause-specific mortality, deaths due to other causes were regarded as competing risks.¹⁹ The effect of smoking on cause-specific mortality was modelled using the competing risks regression²⁰ and quantified based on the subdistribution HR and its corresponding 95% CI. To reduce potential bias due to pre-existing disease or smoking cessation related to chronic illness, all analyses were repeated after excluding participants with reported prevalent diseases (diabetes mellitus, stroke, cardiovascular disease and cancer). Finally, the analysis was repeated only among ever-smokers in order to investigate whether the relationship between smoking cessation and mortality persisted after controlling for the dosage of smoking (number of cigarettes smoked per day). All reported p values are two sided; p<0.05 was considered statistically significant. The statistical software STATA V.11 was used for the analysis.

RESULTS

Among 48 251 participants in the analysis, 15.0% (n=7223) were current smokers, 3.6% (n=1737) were new quitters, 8.9% (n=4283) were long-term quitters and 72.5% (n=35 008) were never-smokers. Overall, there were more ever-smokers among the 19 423 men in this study (30.9% current smokers, 7.4% new quitters and 19.9% long-term quitters) compared with the 28 828 women (4.2% current smokers, 1.0% new quitters and 1.5% long-term quitters). New quitters and long-term quitters were of comparable age at follow-up interview and older than current and never-smokers. New quitters also had similar proportion of regular alcohol drinkers compared with long-term quitters. For self-reported physician-diagnosed medical conditions, new quitters had the highest prevalence of diabetes mellitus, CHD and stroke compared with the other groups. New quitters also had the highest prevalence of cancer at the time of follow-up interview. New quitters had smoked for a longer duration than long-term quitters since new quitters had quit only recently, but the mean number of cigarettes smoked per day was comparable. The mean duration of quitting among new quitters was 4.3 years (IQR 1–5 years) and that for long-term quitters was 17.0 years (IQR 10–23 years) (table 1).

Table 1.

Demographic and health characteristics of 48 251 participants (mean and SD or number and %) by smoking category in the Singapore Chinese Health Study (1993–2009)

		Smoking category*
	Total (n=48 251)	Current smokers (n=7223)	New quitters (n=1737)	Long-term quitters (n=4283)	Never-smokers (n=35 008)
Men	19 423 (40.3%)	6006 (83.2%)	1436 (82.7%)	3860 (90.1%)	8121 (23.2%)
Age at follow-up interview (years)	61.5 (7.9)	62.3 (7.6)	64.3 (7.9)	64.8 (8.0)	60.7 (7.8)
Time between two interviews (years)	5.8 (1.5)	6.0 (1.4)	6.3 (1.5)	5.7 (1.4)	5.7 (1.4)
Follow-up time (years)	8.1 (1.9)	7.6 (2.2)	7.4 (2.5)	7.7 (2.2)	8.3 (1.7)
Body mass index (kg/m2)	23.1 (3.5)	22.3 (3.5)	23.0 (3.6)	23.4 (3.4)	23.3 (3.5)
Dialect group, Cantonese	23 404 (48.5%)	2820 (39.0%)	725 (41.7%)	2164 (50.5%)	17 695 (50.5%)
No formal education	12 468 (25.8%)	1550 (21.5%)	357 (20.6%)	578 (13.5%)	9983 (28.5%)
Alcohol intake	1.6 (7.3)	5.2 (13.5)	3.1 (10.3)	3.0 (10.5)	0.7 (3.6)
Weekly drinkers	3811 (7.9%)	1117 (15.5%)	213 (12.3%)	583 (13.6%)	1898 (5.4%)
Daily drinkers	1538 (3.2%)	729 (10.1%)	110 (6.3%)	221 (5.2%)	478 (1.4%)
Prevalence of hypertension	18 794 (38.9%)	1973 (27.3%)	736 (42.2%)	2009 (46.9%)	14 076 (40.2%)
Prevalence of diabetes	6854 (14.2%)	835 (11.6%)	335 (19.3%)	826 (19.3%)	4858 (13.9%)
Prevalence of coronary heart disease	3409 (7.1%)	518 (7.2%)	264 (15.2%)	600 (14.0%)	2027 (5.8%)
Prevalence of stroke	1708 (3.5%)	257 (3.6%)	157 (9.0%)	251 (5.9%)	1043 (3.0%)
Prevalence of cancer	2347 (4.9%)	266 (3.7%)	125 (7.2%)	241 (5.6%)	1715 (4.9%)
Cigarettes smoked per day		17.3 (10.9)	18.7 (13.7)	18.7 (14.7)	–
Length of smoking (years)		41.9 (11.1)	39.4 (11.9)	27.3 (12.8)	–
Length of quitting (years)		–	4.3 (5.8)	17.1 (10.5)

^*Demographic and health characteristics of the four smoking categories were compared using χ² test for categorical variables and analysis of variance for continuous variables. All comparisons were statistically significant, p<0.0001.

After a mean of 8.1 (range, 0.01–10.5) years of follow-up, there were 6003 deaths among current smokers (n=1590), new quitters (n=437), long-term quitters (n=850) and never-smokers (n=3126). Specifically, 609 were deaths due to lung cancer (10.1%), 1566 were due to other cancers (26.1%), 1168 due to CHD (19.5%), 488 due to stroke (8.1%) and 249 due to COPD (4.1%).

Table 2 shows mortality risk by smoking category in our study. In model 1, we examined the association between smoking status and mortality with inclusion of age, lifestyle factors and co-morbidities as potential confounders. Compared with current smokers, never-smokers had the lowest risk of all-cause mortality (HR: 0.49, 95% CI 0.46 to 0.53) and also of mortality due to cancer, CHD, stroke and COPD. While long-term quitters had a 39% reduction in total mortality (HR: 0.61, 95% CI 0.56 to 0.67), there was a 16% reduction in mortality risk (HR: 0.84, 95% CI 0.76 to 0.94) among new quitters. This reduction in mortality risk among new quitters remained essentially the same after additional adjustment for dosage of smoking. Similar trends were noted for death due to lung cancer; new quitters showed a 24% reduction (HR: 0.76, 95% CI 0.57 to 1.00) in risk compared with a 56% reduction (HR: 0.44, 95% CI 0.35 to 0.57) among long-term quitters. For death due to cancers other than lung cancer, while there was moderate and significant risk reduction in long-term quitters, there was a non-statistically significant risk reduction among new quitters.

Table 2.

Mortality risk by smoking category among 48 251 participants in the Singapore Chinese Health Study (1993–2009)

		Model 1 (n=48 251)		Model 2 (n=36 387)		Model 3 (n=13 243)
Cause of death	Mortality	Adjusted HR	95% CI	Adjusted HR	95% CI	Adjusted HR	95% CI
All-cause
Current smokers	1590	1.00		1.00		1.00
New quitters	437	0.84	0.76 to 0.94	0.83	0.71 to 0.97	0.85	0.76 to 0.95
Long-term quitters	850	0.61	0.56 to 0.67	0.56	0.49 to 0.63	0.62	0.56 to 0.68
Never-smokers	3126	0.49	0.46 to 0.53	0.46	0.41 to 0.50	–
Lung cancer
Current smokers	289	1.00		1.00		1.00
New quitters	64	0.76	0.57 to 1.00	0.61	0.41 to 0.90	0.78	0.59 to 1.02
Long-term quitters	89	0.44	0.35 to 0.57	0.40	0.29 to 0.55	0.45	0.35 to 0.57
Never-smokers	167	0.16	0.13 to 0.20	0.15	0.11 to 0.20	–
Cancers other than lung cancer
Current smokers	347	1.00		1.00		1.00
New quitters	98	0.90	0.71 to 1.14	1.06	0.77 to 1.44	0.92	0.72 to 1.17
Long-term quitters	203	0.81	0.67 to 0.97	0.74	0.57 to 0.95	0.83	0.69 to 1.00
Never-smokers	918	0.71	0.61 to 0.82	0.70	0.58 to 0.85	–
Coronary heart disease
Current smokers	280	1.00		1.00		1.00
New quitters	86	0.84	0.66 to 1.08	0.74	0.49 to 1.11	0.85	0.66 to 1.09
Long-term quitters	181	0.63	0.52 to 0.77	0.56	0.41 to 0.76	0.63	0.51 to 0.78
Never-smokers	621	0.58	0.49 to 0.69	0.56	0.43 to 0.73	–
Stroke
Current smokers	99	1.00		1.00		1.00
New quitters	25	0.73	0.47 to 1.14	0.66	0.35 to 1.26	0.72	0.46 to 1.12
Long-term quitters	62	0.74	0.53 to 1.02	0.68	0.44 to 1.07	0.75	0.53 to 1.04
Never-smokers	302	0.78	0.59 to 1.03	0.67	0.46 to 0.97	–
Chronic obstructive pulmonary disease
Current smokers	120	1.00		1.00		1.00
New quitters	38	1.23	0.84 to 1.79	1.22	0.78 to 1.91	1.19	0.82 to 1.74
Long-term quitters	52	0.68	0.48 to 0.95	0.59	0.39 to 0.90	0.66	0.46 to 0.94
Never-smokers	39	0.15	0.10 to 0.24	0.15	0.09 to 0.26	–

Model 1: adjusted for age at follow-up interview, year of interview, years between baseline and follow-up interview, body mass index group (<20, 20 to <24, 24 to <28, 28+), gender, father’s dialect (Hokkien, Cantonese), education (non-formal education, primary school, secondary school or higher), total ethanol intake (grams per day), moderate activity (no, 0.5–3, 4+ h/week), hypertension, diabetes mellitus, stroke, coronary heart disease (CHD) and cancer.

Model 2: excluding participants with history of diabetes mellitus, stroke, CHD and cancer.

Model 3: excluding never-smokers and further adjusted for number of cigarette smoked per day in addition to covariates in model 1.

For CHD mortality, compared with current smokers, never-smokers and long-term quitters had a similar risk reduction of about 40%, which was statistically significant. Comparatively, there was a 16% reduction in risk among new quitters (HR: 0.84, 95% CI 0.66 to 1.08). For stroke death, compared with current smokers, the other three groups had non-statistically significant reduction in risk of similar magnitude (HRs: 0.73, 0.74 and 0.78 for new quitters, long-term quitters and never-smokers, respectively). For death due to COPD, while the risk was reduced by 32% in long-term quitters (HR: 0.68, 95% CI 0.48 to 0.95), there was a non-statistically significant increase in risk among new quitters.

In model 2, after excluding subjects with history of diabetes mellitus, hypertension, CHD, stroke and cancer, there was a further reduction in risk of total and cause-specific mortalities for long-term quitters and new quitters. In model 3, after adjusting for number of cigarettes smoked per day, most of the risk estimates for new quitters were slightly attenuated. However, there was still a statistically significant 15% reduction in total mortality risk (HR: 0.85, 95% CI 0.76 to 0.95). In a subanalysis, we re-analysed our data stratified by age at enrolment (<60, 60+ years at enrolment). For all-cause mortality, the HR (95% CI) for new quitters of age <60 years at recruitment was 0.76 (0.63 to 0.93) and that for new quitters of age 60 years and above was 0.88 (0.77 to 1.00). The risk estimates for long-term quitters in these two age groups were 0.62 (0.53 to 0.73) and 0.62 (0.56 to 0.68), respectively. This suggests that with time, the risk reduction in mortality was stable across ages, although this should be further verified in other studies.

DISCUSSION

In this study, we compared all-cause and cause-specific mortalities among never-smokers, new quitters, long-term quitters and current smokers and found risk reduction in all-cause, lung cancer and cardiovascular mortalities in the new quitters who quit for a mean of only 4 years. The most impressive reduction was in lung cancer mortality. The results show that the excess risk of mortality associated with smoking can be reduced rapidly upon cessation.

Our study has several strengths. Almost all previous studies could only ascertain the smoking status of the participants at a single time point in which ‘new quitters’ and ‘long-term quitters’ were commonly grouped as ‘former smokers’. Our study assessed smoking status at two separate time points and could therefore ascertain smoking status more reliably. The unique strength of this study lies in its ability to separate individuals who had quit smoking in the shorter term (mean 4.3 years) from those who had quit smoking for a much longer period (mean 17.0 years). The survival experiences between these two ex-smoker groups were clearly different, with the long-term quitters having more favourable survival than new quitters after about 8 years of follow-up.

The second strength of the study is the control of chronic diseases as potential confounders since these diseases are related to both smoking status and mortality risk. The new quitters in our study reported greater prevalence of diabetes, CHD, stroke and cancers, which rendered them more likely to die from these diseases. In fact, the diagnosis of these chronic diseases is often a great motivator to quit smoking. Our study attempted to reduce the confounding in two ways. First, we included participants’ co-morbidities (hypertension, diabetes, stroke, cancer and CHD) at the time of self-report (baseline or follow-up interview) into the statistical model. Second, we excluded all participants who had these diseases from the analysis and showed that the risk estimates in new quitters were further reduced for total mortality and also for mortalities due to lung cancer, CHD and stroke. In the final model, we included dosage of smoking as a covariate in order to reduce the potential bias that successful quitters may have smoked less heavily than current smokers. Our final analysis still showed a statistically significant reduction in risk for all-cause mortality for new quitters compared with current smokers. Other strengths of the study include its large sample size, its representativeness of Chinese Singaporeans and the fact that causes of death were ascertained from the nationwide registry, where mortality assessment is complete and reliable.

There are, however, several limitations in the current study. First, the smoking status was assumed to be static in the follow-up period of the study. It is possible that new quitters and long-term quitters may start smoking again or that current smokers may quit smoking during the follow-up period. Such non-differential misclassification of smoking status could lead to an underestimation of the risk reduction associated with smoking cessation. Second, the proportion of ever-smokers in the cohort study was very low among women compared with men and precluded us from detecting a difference in effect of smoking cessation by sex. Third, validation of the self-reported smoking status by bioassay (eg, cotinine or thiocyanate) was not done in this study. We have, however, tried to overcome this limitation by using this study design that included two time point assessments of smoking habit in order to be more certain about the cessation of smoking.

The causal relationship of cigarette smoking to cause-specific mortality is likely to be strongest for lung cancer compared with other cancers. Hence, it is not surprising that smoking cessation also has the greatest impact on lung cancer mortality compared with other cancers, and while recent quitters were able to see significant reduction in lung cancer mortality, only long-term quitters benefited from a reduction in mortality risk from other cancers. This finding confirmed a previous study, which found a greater reduction in lung cancer mortality compared with mortalities due to all cancer for former smokers compared with current smokers.¹¹

In the present study, there was 16% reduction in CHD mortality risk among new quitters. The rapid decline of CHD mortality following smoking cessation has been documented in the Western studies.^15–17 The US Surgeon General’s Report in 1990 concluded that excess risk of CHD caused by smoking was reduced to about half after 1 year of cessation.¹⁵ In the Multiple Risk Factor Intervention Trial, quitters had a lower RR of CHD mortality compared with non-quitters (RR: 0.63) after only 1 year of smoking cessation.²¹ In our study, long-term quitters reached the risk level of never-smokers for CHD mortality. This finding is supported by two Japanese cohort studies^{22 23} and one Taiwanese cohort study,²⁴ which found mortality risk reduced to the level of never-smokers after ≥10 years of quitting.

With respect to mortality risk from stroke, our study showed that mortality risk reduced markedly, albeit not reaching statistical significance, in new quitters in comparison with current smokers. Furthermore, the mortality risk of new quitters was similar to that of long-term quitters and never-smokers. In the Nurses’ Health Study,²⁵ mortality risk for vascular diseases declined more rapidly compared with lung cancer and COPD in the first 5 years after quitting. In a meta-analysis of cohort studies from the Asia Pacific Cohort Studies Collaboration,¹⁴ significant reduction of risk in stroke mortality with quitting (HR=0.84, 95% CI 0.76 to 0.92) was reported. Within only 4 years of follow-up, quitting smoking has been shown to reduce 33% mortality risk for stroke among elderly Chinese in Hong Kong.¹¹ In these Asian studies, however, the number of years of quitting among the former smokers was not indicated. It is unclear whether the reduction in stroke mortality was different between short-term and long-term quitters. To our knowledge, only one other Asian study analysed the smoking cessation on stroke mortality as a function of years of quitting.²³ Similar to our finding, this Japanese study found that total stroke mortality was reduced by 27% between 2 and 4 years of quitting. This mortality risk was further reduced to about half between 10 and 14 years of quitting (RR=0.48, 95% CI 0.31 to 0.74). The lack of significant association of smoking cessation and mortality risk from stroke in our study may be due to small number of stroke death in the study population.

The present study found an increased but non-statistically significant risk for COPD death in new quitters, which has also been reported in other Chinese studies of smoking cessation and mortality.^{9 11} The apparent increased mortality due to COPD (‘reverse causality’) could be explained by the fact that many Asian smokers quit because they were ill or they already experienced respiratory symptoms.^{9–11 26} The finding of our study contributes to the knowledge that reduction in risk of COPD can be observed after >10 years of cessation in the Asian populations. Nevertheless, even with long-term quitting, former smokers were still at higher risk of lung cancer and COPD deaths compared with never-smokers, suggesting that ever-smokers may never completely recover from the damage to the lungs. This finding is consistent with the results from the study on insured U.S. veterans²⁷ and the Nurses’ Health Study,²⁵ which found that >20 years might be needed before the full benefits of smoking cessation on lung cancer or COPD could be observed.

The benefits of smoking cessation for CHD and stroke were realised earlier than that of lung cancer and COPD, which can be explained by the differential biological effects of smoking on cardiovascular death, cancer death and COPD death. Within days or weeks, former smokers may experience the benefits of quitting such as increase in platelet activation and change in fibrinolytic system.¹⁵ However, carcinogenic transformation of individual lung cells and the metastatic growth of the cancer cells is a long process, and reversal of risk of cigarette on lung cancer risk will likely to take longer time before full benefits can be observed.¹⁵ Similarly, long latency period to develop COPD before its clinical manifestations emerge may explain the persistent elevation of COPD mortality risk after many years of cessation.

In conclusion, this study demonstrates that benefit in terms of reduction in all-cause and lung cancer mortalities can be realised even after short-term quitting. While Western countries, such as USA, have reached the peak of the tobacco epidemic,²⁸ the smoking prevalence is rapidly increasing in Asian countries such as China. Mortality attributed to smoking in Asian populations is expected to rise with the disturbing observations that the mean cigarette consumption by Chinese is increasing rapidly²⁹ and that smoking is starting at an earlier age.³⁰ Hence, it is important to document the benefit of smoking cessation on mortality in Chinese populations to reinforce the public health importance of smoking cessation. This is especially relevant as many Asian populations, including China, are also ageing. Hence, there is a need to quantify the benefit of smoking cessation among the elderly Chinese population.

It is now recognised that in order to effectively curb the smoking epidemics, multilevel and multilateral efforts are needed.³¹ In addition to increasing taxes on tobacco, enacting bans on tobacco advertising and promotion, government and public health officials should provide enabling environment (eg, smoke-free environment) and tools (eg, quit line and nicotine replacement therapy) to help smokers quit smoking. In a national health survey in Singapore, the intention to quit smoking has been found to be lower among smokers over the age of 45 compared with younger smokers.⁷ Many older adults may be unaware of the benefits of smoking cessation and the resources available to help them quit smoking. Further studies should identify the barriers and successful strategies to quitting among elderly smokers. This study found that even among smokers over the age of 60 years, mortality risk can be reduced within a few years after quitting. Hence, physicians should encourage patients of all ages to quit smoking.

What this paper adds.

In this study, using two interviews 6 years apart, we accurately identified new quitters and long-term quitters in a cohort of middle-aged and elderly Chinese in Singapore and found a significant reduction in all-cause and cause-specific mortalities in the new quitters who quit for a mean of only 4 years. The most impressive reduction was in lung cancer mortality. The results show that even among smokers over the age of 60, mortality risks can be reduced within a few years after quitting. Hence, elderly smokers should be encouraged to quit smoking.