Skip to main content
NCBI home page
PLOS ONE logoLink to PLOS ONE
. 2020 Jul 16;15(7):e0235276. doi: 10.1371/journal.pone.0235276

Protective effect of smoking cessation on subsequent myocardial infarction and ischemic stroke independent of weight gain: A nationwide cohort study

Jung-Hwan Cho 1,¤, Hye-Mi Kwon 1, Se-Eun Park 1, Jin-Hyung Jung 2, Kyung-Do Han 2, Yong-Gyu Park 2, Yang-Hyun Kim 3, Eun-Jung Rhee 1,*,#, Won-Young Lee 1,*,#
Editor: Michael Cummings4
PMCID: PMC7365437  PMID: 32673331

Abstract

Smoking cessation reduces the cardiovascular risk but increases body weight. We investigated the risk of subsequent myocardial infarction and ischemic stroke according to weight gain after smoking cessation, using a nationwide population based cohort. We enrolled 3,797,572 Korean adults aged over 40 years who participated in national health screenings between 2009 and 2010. Subjects who quit smoking were classified into three subgroups according to the weight change between baseline and 4 years prior. Myocardial infarctions and ischemic strokes were followed until the end of 2015. We compared the hazard ratios among smoking cessation subgroups, non-smokers, and current smokers. The mean changes in weight (1.5 ± 3.9 kg) of the smoking cessation group were higher than those of the other groups (p < 0.0001). A total of 31,277 and 46,811 subjects were newly diagnosed with myocardial infarction and ischemic stroke, respectively. Regardless of weight change, all subgroups of smoking cessation had significantly less risk than current smokers. The subgroup of smoking cessation with weight gain over 4kg showed the lowest risk for myocardial infarctions (hazard ratio 0.646, 95% confidence interval 0.583–0.714, p < 0.0001) and ischemic strokes (hazard ratio 0.648, 95% confidence interval 0.591–0.71, p < 0.0001) after multivariable adjustment. In conclusion, weight gain after smoking cessation did not adversely affect the cardiovascular protective effect.

Introduction

Smoking is an important, absolute risk factor in health behavior that can be corrected and significantly affects the incidence and mortality of cardiovascular disease (CVD) [1]. Since smoking is known to be related to the onset of inflammatory responses and the progression of thrombosis through increased oxidative stress and endothelial dysfunction, smoking cessation prevents the progression of atherothrombosis by rapidly reversing hemostatic and inflammatory markers [2]. However, people who quit smoking were exposed to the risk of weight gain resulting from smoking cessation [3]. Smoking cessation increases food intake and decreases energy expenditure via the effects of nicotine deficiency [4]. As the energy balance moves in a positive direction, obesity becomes worse and the metabolic profiles deteriorate, increasing the risk factors for cardiovascular disease including metabolic syndromes [5, 6].

Several studies investigated how weight gain after smoking cessation affect the relative risk of CVD. In the study of Japanese male [7], successful abstainers had an average weight gain of 2.4 kg over the 4 years after smoking cessation. Blood pressure, total cholesterol, triglyceride and fasting blood glucose of these subjects were also significantly worsened. Nevertheless, the estimated risk of CVD was reduced by 24% as compared to the baseline, conversely increased by 9% in continuing smokers. In the study using cohort data from the Framingham Offspring Study [8], weight gain following smoking cessation did not modify the association between quitting and lowering risk of CVD even though it did not yield enough statistical power in diabetic subjects. Subsequently, a study conducted in Korea with more participants demonstrated that quitters with change in BMI greater than 1.0 kg/m2 decreased the risk of myocardial infarction by 67%, and total stroke by 25% smokers with statistical significance compared to sustained smokers after adjusting for cardiovascular risk factors including blood glucose [9]. However, this study was only included males over 40 years of age because of the few number of female smokers. In addition, the body mass index (BMI) used to represent weight change in this study is affected by an individual's height. In a recent longitudinal cohort study performed in the United States [10], smoking cessation was associated with an increased short-term risk of type 2 diabetes with substantial weight gain over 5kg, but consistently reduced all-cause and cardiovascular mortality during extended follow-up durations regardless of the degree of weight gain in both men and women. However, recent quitters that did not gain weight had a higher risk of mortality relative to those who gained weight. Cardiovascular prognosis and mortality seems better in overweight or obese described by the “obesity paradox” [11], so the risk of actual cardiovascular events due to weight gain might be underestimated when determining mortality as an outcome.

Therefore, we aimed to investigate how the real world incidences and the risk of subsequent myocardial infarction (MI) and ischemic stroke (IS) were related to the weight gain after smoking cessation in nationwide Korean adults, through utilizing the national data of health screenings and insurance records.

Materials and methods

Study population

More than 95% of Koreans are covered by the National Health Insurance Service (NHIS). Nearly all Korean adults over 40 years of age undergo regular health checkups provided by the NHIS every one or two years. The NHIS organizes health checkup data together with national insurance claim records to provide comprehensive information for medical research from patient demographic information to hospital records. This information includes examinations, laboratory findings, diagnoses, and treatments. This database contains a representative population based cohort widely applicable to various clinical studies [12]. Our study was approved by the NHIS (Research Number: NHIS-2019-1-483).

A total of 12,724,418 Korean adults over 40 years of age participated in national health screenings between 2009 and 2010. Of these, 4,315,426 people who did not have missing data and whose health screening data including smoking history from 4 years prior could be accessed were included in this study. Individuals diagnosed with previous MIs or ISs (409,446) were excluded, as were 108,408 individuals previously diagnosed with cancer due to the possibility of unintentional weight change. Ultimately, we enrolled 3,797,572 subjects (Fig 1).

Fig 1. The selection process of study population.

Fig 1

Open in a new tab

Definition of smoking history and outcomes

We evaluated the subjects’ smoking history through self-questionnaires during health screenings including their current smoking status, duration, and amount of smoking. The participants responded to their current smoking status through one of three choices: never smoked, smoking in the past but now quitting, or continuing to smoke. Current smokers were defined as those who responded to smoke continuously from 4 years prior to baseline. Those who consistently never smoked were classified as non-smokers. The participants who quit smoking at baseline, but who were smokers 4 years prior were classified as the smoking cessation group.

Weight change was calculated as the difference of weight between baseline and 4 years prior. We reviewed all the participants’ international classification of disease (ICD) code claims through NHIS until the end of 2015 to verify the MI or IS diagnoses. The occurrence of a MI was defined as an ICD-10 I21 or I22 code claimed at least twice, or more than once with a hospitalization. The occurrence of an IS was defined as an ICD-10 I63 or I64 code claimed together with a hospitalization and a radiological examination (magnetic resonance imaging or computed tomography). The participants who had a history of MI or IS identified using these ICD code claims prior to baseline were excluded. Newly diagnosed MIs and ISs were followed from enrolment to the end of 2015. the mean follow-up periods for MI and IS were 5.9 ± 0.7 years and 5.8 ± 0.8 years, respectively.

Other baseline characteristics

This study included both male and female Koreans aged over 40 years. Socio-behavioral information such as drinking (more than 30 g of alcohol per day), regular physical activity (moderate exercise more than 3 days per week or vigorous exercise more than 3 days per week), and income (below the 20th percentile) were obtained through a standardized questionnaire. Weight (kg) was measured with an electronic scale and BMI was calculated using height (cm) and body weight. Waist circumference (cm) was measured by trained examiners at the midpoint between the rib cage and iliac crest. Abdominal obesity was defined in accordance with the standard criteria in Korean male and female (≥90 and ≥85 cm, respectively) [13]. Blood samples were collected during fasting and blood pressure was measured by a skilled examiner using a sphygmomanometer after a five minute rest. Hypertension, diabetes, and hyperlipidemia were diagnosed using previous claim record ICD codes (ICD-10 code I10 to I15; E11 to 14; E78) with medication or checking results from health screenings (systolic blood pressure ≥ 140 mmHg and diastolic blood pressure ≥ 90 mmHg; fasting blood glucose level ≥ 126 mg/dl; total cholesterol levels ≥ 240 mg/dl).

Statistical analysis

In order to demonstrate the effectiveness of smoking cessation, we performed primary analysis by identifying the crude incidence rate (IR) and the hazard ratios (HR) of MI or IS. HRs were compared among the smoking cessation, non-smoker, and current-smoker groups with current smokers as a reference group. We analyzed HR by performing multivariate adjustments with confounders from a non-adjusted model to a fully adjusted model (Model 1, non-adjusted; Model 2, age and sex; Model 3, age, sex, and baseline BMI; Model 4, age, sex, baseline BMI, alcohol drinking, low income, and regular exercise). To demonstrate the effects of weight change on the outcome, the smoking cessation group was divided into three subgroups based on cut-offs related to weight changes (≤ 0; 0–4; ≥ 4) and tertiles, and HRs of each divided subgroups were compared with current smokers and non-smokers. In addition, we performed secondary subgroup analyses to identify whether differences in sex, age, and the presence or absence of metabolic diseases such as hypertension, diabetes, hyperlipidemia, or abdominal obesity.

HRs in primary and secondary analyses were analyzed using a Cox proportional hazards model with a 95% confidence interval (CI). Continuous variables were analyzed by variance analysis, and categorical variables were analyzed using the chi-squared test. Statistical calculations and analyses were performed using SAS version 9.3 (SAS Institute Inc., Cary, NC, USA). Since the cohort data of NHIS was anonymous and adhered to confidentiality guidelines, we were exempt from securing informed consent. Our research complied with the Declaration of Helsinki through approved by the official review committee and the institutional review board of the Kangbuk Samsung Hospital (IRB Number: KBSMC 2017-06-004).

Results

Baseline characteristics

Of the total 3,797,572 subjects, approximately 4.5% quit smoking (n = 172,439) and 20% continued to smoke (n = 768,087) (Table 1). Smoking rates were higher in males than females and the amount of smoking in past smokers was about 21 pack-years. About 61% of the subjects who quit smoking gained weight. The mean weight change of the total smoking cessation group at baseline was 1.5 ± 3.9 kg, which was 0 ± 3.3 kg for non-smokers and 0 ± 3.6 kg for current smokers (p <0.0001). In addition to weight gain, systolic (1.0 ± 16.1 mmHg) and diastolic blood pressure (0.3 ± 11.5 mmHg), blood cholesterol (5.5 ± 34.7 mg/dl), and fasting blood sugar (5.1 ± 26.0 mg/dl) increased in the smoking cessation group at baseline compared to the 4 year prior.

Table 1. Baseline characteristics of the subjects according to smoking status.

Non-smoker Smoking cessation Current smoker
N (%) 2,857,046 (75.23) 172,439 (4.54) 768,087 (20.23)
Sex
    Male, N (%) 704,700 (24.67) 168,107 (97.49) 744,941 (96.99)
    Female, N (%) 2,152,346 (75.33)‬ 4,332 (2.51) 23,146 (3.01)
Amount of smoking, pack-years, Mean (SD) 0 21.2 ± 16.1 22.2 ± 13.7
    0–10, N (%) 35,657 (20.68) 103,939 (13.53)
    10–20, N (%) 49,385 (28.64) 221,667 (28.86)
    ≥20, N (%) 87,397 (50.68) 442,481 (57.61)
Age (group), Mean (SD) 55.8 ± 9.9 52.32 ± 9.2 51.5 ± 9.1
    ≥55, N (%) 1,369,268 (47.93) 59,023 (34.23) 234,855 (30.58)
Alcohol drinkinga, N (%) 55,318 (1.94) 25,732 (14.92) 136,818 (17.81)
Low incomeb, N (%) 559,664 (19.59) 22,481 (13.04) 111,680 (14.54)
Regular exercisec, N (%) 592,791 (20.75) 46,173 (26.78) 149,136 (19.42)
Hypertension, N (%) 902,467 (31.59) 56,810 (32.94) 219,192 (28.54)
Diabetes, N (%) 267,412 (9.36) 22,669 (13.15) 95,159 (12.39)
Hyperlipidemia, N (%) 685,405 (23.99) 40,664 (23.58) 151,655 (19.74)
Abdominal obesityd, N (%) 583,814 (20.43) 45,048 (26.12) 161,166 (20.98)
Waist circumference (cm), Mean (SD) 79.4 ± 8.4 85.0 ± 7.3 83.5 ± 7.6
Body mass index, Mean (SD) 23.8 ± 3.0 24.5 ± 2.8 23.8 ± 2.9
Height (cm), Mean (SD) 158.2 ± 8.0 168.7 ± 6.2 168.5 ± 6.4
Weight (kg), Mean (SD) 59.8 ± 9.5 69.9 ± 9.8 67.8 ± 10.2
Weight difference (subgroup), Mean (SD) 0 ± 3.3 1.5 ± 3.9 0 ± 3.6
    ≤ 0, N (%) 1,612,958 (56.46) 67,387 (39.08) 433,313 (56.41)
    0–4, N (%) 922,931 (32.3) 57,799 (33.52) 233,034 (30.34)
    ≥ 4, N (%) 321,157 (11.24) 47,253 (27.4) 101,740 (13.25)
Systolic blood pressure (mmHg), Mean (SD) 123.1 ± 15.3 125.6 ± 14.1 124.4 ± 14.3
    Difference, Mean (SD) -0.1 ± 16.5 1.0 ± 16.1 -0.4 ± 16.0
Diastolic blood pressure (mmHg), Mean (SD) 76.2 ± 9.9 78.9 ± 9.7 78.2 ± 9.8
    Difference, Mean (SD) -0.5 ± 11.3 0.3 ± 11.5 -0.6 ± 11.4
Fasting plasma glucose (mg/dl), Mean (SD) 97.6 ± 21.0 102.7 ± 26.3 101.6 ± 27.1
    Difference, Mean (SD) 3.1 ± 21.0 5.1 ± 26.0 3.9 ± 27.0
Total cholesterol (mg/dl), Mean (SD) 201.3 ± 36.6 202.0 ± 36.5 199.2 ± 36.2
    Difference, Mean (SD) 4.2 ± 35.3 5.5 ± 34.7 2.6 ± 33.2
Prevalence of myocardial infarction, N (%) 20,321 (0.71) 1,613 (0.94) 9,343 (1.22)
Prevalence of ischemic stroke, N (%) 33,087 (1.16) 2,005 (1.16) 11,719 (1.53)
Open in a new tab

All characteristics met p < 0.0001.

aDefined as drinking more than 30 g of alcohol per day.

bDefined as moderate exercise more than 3 days per week or vigorous exercise more than 3 days per week.

cDefined as income below the 20th percentile.

dDefined as waist circumference ≥ 90 cm in male and ≥ 85 cm in female.

Effect of smoking cessation on the outcome

A total of 31,277 subjects (0.82%) were diagnosed with MIs and 46,811 subjects (1.23%) were diagnosed with ISs. The crude incidence of MI (IR 1.603 per 1,000 person-years) and IS (IR 1.994 per 1,000 person-years) were higher in the smoking cessation group than in non-smokers, but lower than that in current smokers. After multivariable adjustment, smoking cessation significantly lowered the risk of MIs (Model 4; HR 0.703, 95% CI 0.666–0.741, p < 0.0001) compared to current smokers. In the analysis of IS, smoking cessation consistently lowered the risk (Model 4; HR 0.7, 95% CI 0.667–0.734, p < 0.0001) (Table 2).

Table 2. Incidence Rate (IR) and multivariate-adjusted Hazard Ratios (HRs) (95% confidence intervals) of myocardial infarction and ischemic stroke.

Myocardial infarction
HR (95% Cl)
Smoking status Number of subjects Events IR (per 1000 person years) Model 1a Model 2b Model 3c Model 4d
Non-smoker 2,857,046 20,321 1.215 0.582 (0.567–0.596) 0.485 (0.47–0.501) 0.473 (0.458–0.488) 0.47 (0.455–0.485)
Smoking cessation 172,439 1,613 1.603 0.768 (0.728–0.809) 0.724 (0.686–0.763) 0.7 (0.663–0.738) 0.703 (0.666–0.741)
Current smoker 768,087 9,343 2.088 1 (Ref.) 1 (Ref.) 1 (Ref.) 1 (Ref.)
Ischemic stroke
HR (95% Cl)
Smoking status Number of subjects Events IR (per 1000 person years) Model 1a Model 2b Model 3c Model 4d
Non-smoker 2,857,046 33,087 1.982 0.755 (0.739–0.771) 0.563 (0.548–0.578) 0.55 (0.536–0.565) 0.564 (0.55–0.58)
Smoking cessation 172,439 2,005 1.994 0.76 (0.725–0.797) 0.706 (0.673–0.74) 0.689 (0.656–0.722) 0.7 (0.667–0.734)
Current smoker 768,087 11,719 2.623 1 (Ref.) 1 (Ref.) 1 (Ref.) 1 (Ref.)
Open in a new tab

aNon-adjusted.

bAdjusted for age and sex.

cAdjusted for age, sex, and body mass index.

dAdjusted for age, sex, body mass index, alcohol drinking, low income, and regular exercise.

Difference in the effect according to the change in weight

Regardless of changes in weight, all smoking cessation subgroups showed significantly less risk than current smokers when spanning from the non-adjusted model to the fully adjusted model (Table 3). The subgroup of smoking cessation with weight gain over 4kg (mean increase in weight: 6.0 ± 2.4 kg) showed the lowest risk for MIs (Model 4; HR 0.646, 95% CI 0.583–0.714, p < 0.0001) and ISs (Model 4; HR 0.648, 95% CI 0.591–0.71, p < 0.0001).

Table 3. Incidence Rate (IR) and multivariate-adjusted Hazard Ratios (HRs) (95% confidence intervals) of myocardial infarction and ischemic stroke according to the change in weight.

Myocardial infarction
HR (95% Cl)
Smoking status (weight change) Number of subjects Events IR (per 1000 person years) Model 1a Model 2b Model 3c Model 4d
Non-smoker 2,857,046 20,321 1.215 0.582 (0.567–0.596) 0.485 (0.47–0.501) 0.473 (0.458–0.488) 0.47 (0.455–0.485)
Smoking cessation
    ≤ 0 67,387 722 1.845 0.884 (0.819–0.953) 0.78 (0.723–0.841) 0.778 (0.721–0.839) 0.782 (0.725–0.843)
    0–4 57,799 499 1.477 0.707 (0.646–0.773) 0.674 (0.615–0.737) 0.648 (0.592–0.708) 0.651 (0.594–0.712)
    ≥ 4 47,253 392 1.416 0.678 (0.612–0.749) 0.696 (0.628–0.769) 0.645 (0.582–0.712) 0.646 (0.583–0.714)
Current smoker 768,087 9,343 2.088 1 (Ref.) 1 (Ref.) 1 (Ref.) 1 (Ref.)
Ischemic stroke
HR (95% Cl)
Smoking status (weight change) Number of subjects Events IR (per 1000 person years) Model 1a Model 2b Model 3c Model 4d
Non-smoker 2,857,046 33,087 1.982 0.755 (0.739–0.771) 0.563 (0.548–0.578) 0.55 (0.536–0.565) 0.565 (0.55–0.58)
Smoking cessation
    ≤ 0 67,387 915 2.340 0.893 (0.834–0.954) 0.758 (0.709–0.811) 0.757 (0.707–0.809) 0.772 (0.721–0.825)
    0–4 57,799 617 1.828 0.697 (0.642–0.755) 0.658 (0.607–0.713) 0.639 (0.589–0.692) 0.649 (0.598–0.704)
    ≥ 4 47,253 473 1.710 0.651 (0.594–0.713) 0.68 (0.619–0.744) 0.642 (0.584–0.703) 0.648 (0.591–0.71)
Current smoker 768,087 11,719 2.623 1 (Ref.) 1 (Ref.) 1 (Ref.) 1 (Ref.)
Open in a new tab

The change in weight was calculated as the weight difference between baseline and four years prior.

The mean ± standard deviation increase in weight (kg) was -2.1 ± 2.5 at ≤ 0; 2.0 ± 0.8 at 0–4; 6.0 ± 2.4 at ≥ 4.

aNon-adjusted.

bAdjusted for age and sex,.

cAdjusted for age, sex, and body mass index.

dAdjusted for age, sex, body mass index, alcohol drinking, low income, and regular exercise.

When the subgroups were divided into tertiles, the 3rd tertile (mean increase in weight: 5.2 ± 2.4 kg) showed the lowest risk for MIs (HR 0.641, 95% CI 0.587–0.699, p < 0.0001), and the 2nd tertile (mean increase in weight: 1.0 ± 0.8 kg) showed the lowest risk for ISs (HR 0.657, 95% CI 0.606–0.711, p < 0.0001) (S1 Table). In secondary subgroup analyses, the reduction of risk after smoking cessation was consistent across the categories of sex, age, hypertension, diabetes, hyperlipidemia, and abdominal obesity (S2 Table).

Discussion

According to our retrospective cohort study of about 3.8 million nationwide Korean adults, weight gained on average and metabolic profiles were worsened after smoking cessation. Before our study, numerous studies have confirmed increases in weight gain after quitting [7, 14, 15]. Smoking increases insulin resistance and central fat accumulation, raising the risk of metabolic syndromes and diabetes [16]. In the early stages of smoking cessation, continuous increased β-cell secretion in response to glucose and fasting insulin resistance is responsible for weight gain, along with an increase in energy balance due to nicotine withdrawal [17]. Several studies have shown an increase in visceral adipose after quitting, which over time gradually dropped compared to that of non-smokers [6, 18]. Changes in abdominal obesity, insulin resistance, and the incidence of type 2 diabetes mellitus showed a similar pattern that increases after smoking cessation and then decreases [10, 19].

Despite weight gain, we found that the risk of MIs and ISs in quitters was lower than current smokers. Furthermore, weight gain after quitting was not associated with a relative increase in the risk. Several mechanisms could explain the prevention of cardiovascular disease despite weight gain and the deteriorated metabolic profile after smoking cessation. Post-cessation-related obesity could contribute to insulin resistance, but the benefits of stopping smoking due to reverse the worsening of insulin resistance caused by nicotine outweigh the risks [20]. Insulin modulates lipoprotein lipase (LPL) activity, and this enzyme expressed by the adipose tissue has an anti-atherogenic effect through improving circulating lipoprotein profiles [21]. The paradoxical response of adipose LPL to glucose in smokers was improved by stopping smoking, contributing to weight gain and increasing adipose tissue LPL activity [22, 23]. Researchers from the REGRESS study group demonstrated that LPL activity was inversely associated with severity of angina pectoris [24]. A clinical study investigating the preheparin LPL mass in patients with coronary atherosclerosis showed that the amount of preheparin serum LPL was significantly lower in patients with coronary atherosclerosis, which was negatively correlated with triglycerides and positively correlated with HDL cholesterol [25]. Smoking is a more potent mediator for normalization of cardio-protective high density lipoprotein cholesterol (HDL-C) than weight gain [26]. In aforementioned study of Japanese male, it was observed that HDL-C increased steadily despite weight gain after smoking cessation [7]. Adiponectin, which is a fundamental factor of lipid metabolism and plays an important role in the development of cardiovascular disease, increases after smoking cessation and does not decrease in spite of increased body weight and abdominal obesity due to quit smoking [27].

One interesting result which emerged from our study was that the group with weight loss after smoking cessation had a significantly lower risk of MIs or ISs than current smokers, but seemed to have a higher incidence and relative risk compared to the group with weight gain after quitting. In the previous study of Koreans with similar designs to ours, no significant risk reduction of MI and total stroke were identified in quitters with BMI loss compared to sustained smokers [9]. This tendency to increase the relative risk of the weight loss group might have been influenced by other causes accompanying unintentional weight loss and/or sarcopenia, despite we excluded patients previously diagnosed with cancer. Subsequent unintentional weight loss due to aging, health problems or chronic disease is closely related to sarcopenia [28]. Several studies in Koreans found the association between sarcopenia and the risk of cardiovascular disease [29, 30]. In addition, we did not exactly know why the subjects of our study quit smoking and there is a possibility that those who attempted to quitting were influenced by other health problems. Therefore, it would be necessary to conduct further research that including the causes and intentions of weight loss after smoking cessation, and whether accompanied by sarcopenia or not.

Our study including both sexes and all adults over 40 years of age who had undergone periodic medical examinations across the country, represented the entire nation of Korean adults. We used the measured weight rather than self-checked weight to reduce error. In order to ensure statistical significance, large number of individuals in the smoking cessation group were included, and various models were calibrated to minimize confounding outcomes between smoking cessation and weight change. Despite these efforts, our study had several limitations. First, our study only analyzed one-time point for smoking cessation and we could not confirm the exact starting point and duration of smoking cessation. This ‘point prevalence’ design which could not identify the changes of exposure over time could lead to an error in weight gain measurement and result interpretation [10, 31]. Secondly, a selection bias cannot be ruled out because only the participants with access to the medical examination record and smoking history of the 4 years prior were allowed to the study. Thirdly, there is a possibility of bias in the confirmation of characteristics through self-questionnaires and the diagnosis of outcomes through insurance claims. Lastly, since we included only Koreans in our study, the effect of smoking cessation on weight gain and its impact on the outcome may be different in other countries and ethnicities.

In conclusion, smoking cessation was effective in reducing the risk of MIs and ISs. Although weight gain after smoking cessation had a temporary adverse effect on the metabolic profile, it did not affect the protection that smoking cessation provided to the cardiovascular system.

Supporting information

S1 Table. Incidence Rate (IR) and multivariate adjusted Hazard Ratios (HRs) (95% confidence intervals) of myocardial infarctions and ischemic strokes according to the tertiles of weight change.

(DOCX)

Click here for additional data file. (29.9KB, docx)
S2 Table. Secondary subgroup analyses according to sex, age, and the presence or absence of hypertension, diabetes, hyperlipidemia, or abdominal obesity.

(DOCX)

Click here for additional data file. (32.4KB, docx)

Acknowledgments

We would like to express our sincere gratitude to NHIS and its employees for creating and providing excellent cohort data for our research.

Data Availability

The data underlying our study are third party data. Data are available from the NHIS Institutional Data Access/Ethics Committee for researchers who meet the criteria for access to confidential data. For information on request data from the NHIS Institutional Data Access/Ethics Committee, please see: https://nhiss.nhis.or.kr/bd/ab/bdaba032eng.do.

Funding Statement

The authors received no specific funding for this work.

References

  • 1.Mons U, Müezzinler A, Gellert C, Schöttker B, Abnet CC, Bobak M, et al. Impact of smoking and smoking cessation on cardiovascular events and mortality among older adults: meta-analysis of individual participant data from prospective cohort studies of the CHANCES consortium. BMJ: British Medical Journal. 2015;350:h1551 10.1136/bmj.h1551 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Wannamethee SG, Lowe GDO, Shaper AG, Rumley A, Lennon L, Whincup PH. Associations between cigarette smoking, pipe/cigar smoking, and smoking cessation, and haemostatic and inflammatory markers for cardiovascular disease. European Heart Journal. 2005;26(17):1765–73. 10.1093/eurheartj/ehi183 [DOI] [PubMed] [Google Scholar]
  • 3.Filozof C, Fernández Pinilla MC, Fernández-Cruz A. Smoking cessation and weight gain. Obesity Reviews. 2004;5(2):95–103. 10.1111/j.1467-789X.2004.00131.x [DOI] [PubMed] [Google Scholar]
  • 4.Audrain-McGovern J, Benowitz N. Cigarette Smoking, Nicotine, and Body Weight. Clinical Pharmacology & Therapeutics. 2011;90(1):164–8. 10.1038/clpt.2011.105 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Yoon C, Goh E, Park SM, Cho B. Effects of smoking cessation and weight gain on cardiovascular disease risk factors in Asian male population. Atherosclerosis. 2010;208(1):275–9. 10.1016/j.atherosclerosis.2009.07.024 [DOI] [PubMed] [Google Scholar]
  • 6.Matsushita Y, Nakagawa T, Yamamoto S, Takahashi Y, Noda M, Mizoue T. Associations of Smoking Cessation With Visceral Fat Area and Prevalence of Metabolic Syndrome in Men: The Hitachi Health Study. Obesity. 2011;19(3):647–51. 10.1038/oby.2010.237 [DOI] [PubMed] [Google Scholar]
  • 7.Tamura U, Tanaka T, Okamura T, Kadowaki T, Yamato H, Tanaka H, et al. Changes in Weight, cardiovascular risk factors and estimated risk of coronary heart disease following smoking cessation in Japanese male workers: HIPOP-OHP study. J Atheroscler Thromb. 2010;17(1):12–20. Epub 2010/01/19. 10.5551/jat.1800 . [DOI] [PubMed] [Google Scholar]
  • 8.Clair C, Rigotti NA, Porneala B, Fox CS, D’Agostino RB, Pencina MJ, et al. Association of Smoking Cessation and Weight Change With Cardiovascular Disease Among Adults With and Without DiabetesSmoking Cessation, Weight Change, and CVD. JAMA. 2013;309(10):1014–21. 10.1001/jama.2013.1644 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Kim K, Park SM, Lee K. Weight gain after smoking cessation does not modify its protective effect on myocardial infarction and stroke: evidence from a cohort study of men. European Heart Journal. 2018;39(17):1523–31. 10.1093/eurheartj/ehx761 %J European Heart Journal. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Hu Y, Zong G, Liu G, Wang M, Rosner B, Pan A, et al. Smoking Cessation, Weight Change, Type 2 Diabetes, and Mortality. New England Journal of Medicine. 2018;379(7):623–32. 10.1056/NEJMoa1803626 . [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Hainer V, Aldhoon-Hainerová I. Obesity Paradox Does Exist. Diabetes Care. 2013;36(Supplement 2):S276–S81. 10.2337/dcS13-2023 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Lee J, Lee JS, Park S-H, Shin SA, Kim K. Cohort Profile: The National Health Insurance Service–National Sample Cohort (NHIS-NSC), South Korea. International Journal of Epidemiology. 2016;46(2):e15–e. 10.1093/ije/dyv319 [DOI] [PubMed] [Google Scholar]
  • 13.Yoon YS, Oh SW. Optimal Waist Circumference Cutoff Values for the Diagnosis of Abdominal Obesity in Korean Adults. Endocrinol Metab. 2014;29(4):418–26. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Williamson DF, Madans J, Anda RF, Kleinman JC, Giovino GA, Byers T. Smoking Cessation and Severity of Weight Gain in a National Cohort. New England Journal of Medicine. 1991;324(11):739–45. 10.1056/NEJM199103143241106 . [DOI] [PubMed] [Google Scholar]
  • 15.Aubin H-J, Farley A, Lycett D, Lahmek P, Aveyard P. Weight gain in smokers after quitting cigarettes: meta-analysis. BMJ: British Medical Journal. 2012;345:e4439 10.1136/bmj.e4439 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Faeh D, Paccaud F, Cornuz J, Chiolero A. Consequences of smoking for body weight, body fat distribution, and insulin resistance. The American Journal of Clinical Nutrition. 2008;87(4):801–9. 10.1093/ajcn/87.4.801 [DOI] [PubMed] [Google Scholar]
  • 17.Marietta S, Larissa T, Angela S, Michael W, Slobodan P, Christian B, et al. Effects of smoking cessation on β-cell function, insulin sensitivity, body weight, and appetite. European Journal of Endocrinology. 2014;170(2):219–27. 10.1530/EJE-13-0590 [DOI] [PubMed] [Google Scholar]
  • 18.Lee K, Lee CM, Kwon HT, Oh S-W, Choi H, Park JH, et al. Associations of smoking and smoking cessation with CT-measured visceral obesity in 4656 Korean men. Preventive Medicine. 2012;55(3):183–7. 10.1016/j.ypmed.2012.06.009 [DOI] [PubMed] [Google Scholar]
  • 19.Kim K-W, Kang S-G, Song S-W, Kim N-R, Rho J-S, Lee Y-A. Association between the Time of Length since Smoking Cessation and Insulin Resistance in Asymptomatic Korean Male Ex-Smokers. Journal of Diabetes Research. 2017;2017:7 10.1155/2017/6074760 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Harris KK, Zopey M, Friedman TC. Metabolic effects of smoking cessation. Nature reviews Endocrinology. 2016;12(5):299–308. Epub 2016/03/05. 10.1038/nrendo.2016.32 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Mead JR, Ramji DP. The pivotal role of lipoprotein lipase in atherosclerosis. Cardiovascular research. 2002;55(2):261–9. Epub 2002/07/19. 10.1016/s0008-6363(02)00405-4 . [DOI] [PubMed] [Google Scholar]
  • 22.Chajek-Shaul T, Berry EM, Ziv E, Friedman G, Stein O, Scherer G, et al. Smoking depresses adipose lipoprotein lipase response to oral glucose. European journal of clinical investigation. 1990;20(3):299–304. Epub 1990/06/01. 10.1111/j.1365-2362.1990.tb01859.x . [DOI] [PubMed] [Google Scholar]
  • 23.Ferrara CM, Kumar M, Nicklas B, McCrone S, Goldberg AP. Weight gain and adipose tissue metabolism after smoking cessation in women. International journal of obesity and related metabolic disorders: journal of the International Association for the Study of Obesity. 2001;25(9):1322–6. Epub 2001/09/26. 10.1038/sj.ijo.0801716 . [DOI] [PubMed] [Google Scholar]
  • 24.Kastelein JJ, Jukema JW, Zwinderman AH, Clee S, van Boven AJ, Jansen H, et al. Lipoprotein lipase activity is associated with severity of angina pectoris. REGRESS Study Group. Circulation. 2000;102(14):1629–33. Epub 2000/10/04. 10.1161/01.cir.102.14.1629 . [DOI] [PubMed] [Google Scholar]
  • 25.Hitsumoto T, Ohsawa H, Uchi T, Noike H, Kanai M, Yoshinuma M, et al. Preheparin serum lipoprotein lipase mass is negatively related to coronary atherosclerosis. Atherosclerosis. 2000;153(2):391–6. Epub 2001/02/13. 10.1016/s0021-9150(00)00413-5 . [DOI] [PubMed] [Google Scholar]
  • 26.Chelland Campbell S, Moffatt RJ, Stamford BA. Smoking and smoking cessation—the relationship between cardiovascular disease and lipoprotein metabolism: a review. Atherosclerosis. 2008;201(2):225–35. Epub 2008/06/21. 10.1016/j.atherosclerosis.2008.04.046 . [DOI] [PubMed] [Google Scholar]
  • 27.Komiyama M, Wada H, Yamakage H, Satoh-Asahara N, Sunagawa Y, Morimoto T, et al. Analysis of changes on adiponectin levels and abdominal obesity after smoking cessation. PLOS ONE. 2018;13(8):e0201244 10.1371/journal.pone.0201244 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Bales CW, Ritchie CS. Sarcopenia, weight loss, and nutritional frailty in the elderly. Annu Rev Nutr. 2002;22:309–23. Epub 2002/06/11. 10.1146/annurev.nutr.22.010402.102715 . [DOI] [PubMed] [Google Scholar]
  • 29.Chin SO, Rhee SY, Chon S, Hwang YC, Jeong IK, Oh S, et al. Sarcopenia is independently associated with cardiovascular disease in older Korean adults: the Korea National Health and Nutrition Examination Survey (KNHANES) from 2009. PLoS One. 2013;8(3):e60119 Epub 2013/03/28. 10.1371/journal.pone.0060119 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Kim JH, Cho JJ, Park YS. Relationship between sarcopenic obesity and cardiovascular disease risk as estimated by the Framingham risk score. J Korean Med Sci. 2015;30(3):264–71. Epub 2015/03/03. 10.3346/jkms.2015.30.3.264 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Klesges RC, Winders SE, Meyers AW, Eck LH, Ward KD, Hultquist CM, et al. How much weight gain occurs following smoking cessation: A comparison of weight gain using both continuous and point prevalence abstinence. Journal of Consulting and Clinical Psychology. 1997;65(2):286–91. 10.1037//0022-006x.65.2.286 [DOI] [PubMed] [Google Scholar]

Decision Letter 0

Michael Cummings

18 Mar 2020

PONE-D-20-01312

Protective effect of smoking cessation on subsequent myocardial infarction and ischemic stroke independent of weight gain: A nationwide cohort study

PLOS ONE

Dear Dr Lee,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Both reviewers commented on the large sample size and potential inherent in the analyses of these data.  That said, both reviewers felt that analyses as presented were inadequate and need to be substantially revised before this paper can be considred for publication.  The authors should heed the recommended changes in defining outcomes and presenting results if they are interested in revising this paper for consideration in PLOS ONE.  Also, we encourage the authors to carefully check their paper for english grammar to improve communication of their findings.   

We would appreciate receiving your revised manuscript by May 02 2020 11:59PM. When you are ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

Please include the following items when submitting your revised manuscript:

  • A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.

  • A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.

  • An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Michael Cummings, PhD

Academic Editor

PLOS ONE

Journal Requirements:

1.  When submitting your revision, we need you to address these additional requirements.

 

Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

http://www.journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and http://www.journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: No

Reviewer #2: Partly

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: No

Reviewer #2: No

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

Reviewer #2: No

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: No

Reviewer #2: No

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: 1. While the scale of this analysis which includes 5.2 million Korean adults is impressive neither the topic nor the results are novel.

2. The analytic approach is confusing and potentially flawed. The dependt variable of a 1-2 point change in BMI appears to be differential depending on an individual's height. For example, a 6 foot 1 inch individual demonstrating an 8 pound weight gain from 200 to 208 pounds would show a 1.0 change in BMI (from 26.4 to 27.4). Conversely a 5 foot 6 inch individual at a baseline weight of 120 pounds which show a 4.5 pound weight gain in order to achieve a 1.0 change in BMI (from 19.1 to 20.1). A much more straightforward approach to the analyses would have simply reported an overall change in weight rather than BMI.

3. Introduction: The text is very general and seems to overlook relative versus absolute effects of cardiac risk factors. The text does not include any information about affect sizes. Among all the risk factors identified smoking far and away is the most significant.

4. Study population: The exclusion of 17,000 individuals diagnosed with an MI or ischemic stroke within 1 year of follow-up is confusing.

5. Classification of the population into smokers and "those who had not smoked during this time" as non-smoker is confusing and inaccurate. Moreover, this approach is inconsistent with the common classification system noting individuals as current, former or never smokers.

6. table 1. What is "difference" under systolic blood pressure, diastolic blood pressure, fasting plasma glucose, total cholesterol?

7. Table 2. Body mass index is not weight gain. Adding additional variables into their multivariate models parens alcohol, income, exercise, hypertension, diabetes, hyperlipidemia) does not change risk estimates beyond more parsimonious models. Authors are urged to simplify their models and descriptions.

8. Table 3. The results and conclusions would be much easier to understand if they were presented as actual weight gains rather than changes in BMI.

8. Supporting information is confusing. Figure 1 raises additional questions about the study design since it suggests that BMI was assessed over an interval of 4 years prior to baseline while MI and ischemic strokes were assessed for a period of 5 years after baseline. Content of figure 2 is also unclear.

9. As a minor point the manuscript would benefit from careful editing to clarify wording and grammar.

Reviewer #2: Introduction

1. The first sentence of the second paragraph is confusing. Consider rewording.

2. Consider rewording the second sentence to "Clair et al. [8] showed that body weight change after smoking cessation....."

3. Explain what you mean by "mortality seems better in higher BMI groups"

Methods

Study population:

4. Correct the following sentence"Our study approved by the NHIS" to "Our study WAS approved by the NHIS"

5. Correct "4 years ago" to "4 years PRIOR" in line 77/78

6. Correct "Individuals diagnosed with previous MIs or ISs" to "Individuals PREVIOUSLY DIAGNOSED with MIs or ISs"

7. Correct "We further excluded 17,192 patients who diagnosed with MI or IS..." to "We further excluded 17,192 patients who WERE diagnosed with MI or IS..."

8. Explain why 17,192 who were diagnosed with MI or MS were excluded. What was unclear about the causal relationship?

9. Excluding patients with the outcome without further explanation contributes to selection bias.

Definition of smoking history and outcomes:

10. Replace "years ago" with "years prior"

11. Explain what you mean by "self-questionnaire". Do you mean self-report?

12. How did you define "steadily smoking"? Consider adding the actual questions that were asked to this section. Same for those who quit.

13. State the follow-up period in this section.

14. This sentence in unclear "The exclusion criteria for previous MIs or ISs were the same as above." State clearly what the exclusion criteria were.

Other baseline characteristics:

15. Include sex and age in this section.

Statistical analysis:

16. It is unclear whether this was a primary or secondary data analysis.

17. Incidence rates are obtained by a Binomial/Poisson regression, Risk Differences are obtained by a normal/log-normal regression, and hazard ratios are obtained by time-to-even regression. These are all different models assessing different measures. Please explain which procedure was used to obtain which measure and what the exposure and outcome were for each procedure. Also, explain whether or not models were adjusted, and if so, the variables that were adjusted for.

Table 1:

18. Indicate which variables are shown as N(%) and which are shown as Mean (SD).

19. Add a column fourth for the total and percentages for smoking status.

20. Replace "Number" by "Characteristic" and "No. (%)" by "N (%)".

21. Footnote: use lower case "p" indicating p-value.

22. What do you mean by "All characteristics met P < 0.0001"? What is being being tested here? Please specify.

23. Add the proportion of daily and non-daily smokers to Table 1. Pack years cannot be calculated for non-daily smokers.

Table 2:

24. Consider changing the reference group to current smokers.

25. Was the incidence rate adjusted or not?

Table 3:

26. Consider changing the reference group to current smokers.

27. Was the incidence rate adjusted or not?

28. It would be worthwhile to perform the regression of the outcomes on BMI (4 categories) stratified by smoking status adjusting for all identified confounders except BMI.

Results

29. Please do not use risk ratio and hazard ratio interchangeably, RR does not take into effect the time to event while HR does.

Discussion

30. Be aware of overarching statements or statements that are not supported by evidence such as the first 2 sentences of the discussion.

31. This paper did not assess the effect of BMI on the risk of MS and IS as mentioned in the second paragraph of the discussion, however, this would be possible with the analyses that I previously suggested under table 3.

32. Discussion will likely have to be revised according to the suggested analyses.

General:

33. Consider restricting the analysis to those who are 40 years of age or older as incidence of cardiovascular outcomes are likely to be low in the younger age groups.

34. Specify the novelty of this article and what value it adds to the existing literature.

35. Figure 2 does not add any more information than Table 3, consider removing.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org. Please note that Supporting Information files do not need this step.

PLoS One. 2020 Jul 16;15(7):e0235276. doi: 10.1371/journal.pone.0235276.r002

Author response to Decision Letter 0

1 May 2020

Here are the responses to the comments of the reviewers:

To reviewer #1

1. While the scale of this analysis which includes 5.2 million Korean adults is impressive neither the topic nor the results are novel.

: Thank you for your comment. Although there have been various studies before our study on the relationship between weight gain after smoking cessation and cardiovascular disease, we demonstrated significant results by evaluating the occurrence of real cardiovascular events as outcome in millions of people including both sexes across the country, complementing the shortcomings of these studies.

2. The analytic approach is confusing and potentially flawed. The dependt variable of a 1-2 point change in BMI appears to be differential depending on an individual's height. For example, a 6 foot 1 inch individual demonstrating an 8 pound weight gain from 200 to 208 pounds would show a 1.0 change in BMI (from 26.4 to 27.4). Conversely a 5 foot 6 inch individual at a baseline weight of 120 pounds which show a 4.5 pound weight gain in order to achieve a 1.0 change in BMI (from 19.1 to 20.1). A much more straightforward approach to the analyses would have simply reported an overall change in weight rather than BMI.

: Thank you for your comment. Based on your feedback, we’ve conducted a new analysis with a straightforward approach using weight, not BMI. As a result, it is consistently confirmed that weight gain after smoking cessation does not adversely affect the cardiovascular protection effect.

3. Introduction: The text is very general and seems to overlook relative versus absolute effects of cardiac risk factors. The text does not include any information about affect sizes. Among all the risk factors identified smoking far and away is the most significant.

: Thank you for your comment. We also agree that smoking cessation is the most effective and powerful factor for reducing the risk of cardiovascular disease. We’ve revised the text so that the relative risk is not overestimate relative to the absolute risk of smoking.

4. Study population: The exclusion of 17,000 individuals diagnosed with an MI or ischemic stroke within 1 year of follow-up is confusing.

: Thank you for your comment. We initially thought it was unclear whether a casual relationship existed between the incidence of myocardial infarctions or strokes within one year from baseline and exposure to weight change. However, we agree that you were confused with this approach. We’ve included all MIs and ISs that occurred after the baseline without a one-year washout period in the analyses.

5. Classification of the population into smokers and "those who had not smoked during this time" as non-smoker is confusing and inaccurate. Moreover, this approach is inconsistent with the common classification system noting individuals as current, former or never smokers.

: Thank you for your comment. We evaluated the subjects’ smoking history through self-questionnaires during health screenings including their current smoking status, duration, and amount of smoking. The participants responded to their current smoking status through one of three choices: never smoked, smoking in the past but now quitting, or continuing to smoke. Current smokers were defined as those who responded to smoke continuously from 4 years prior to baseline. Those who consistently never smoked were classified as non-smokers. The participants who quit smoking at baseline, but who were smokers 4 years prior were classified as the smoking cessation group. We’ve added these to the text to avoid confusion of communication.

6. table 1. What is "difference" under systolic blood pressure, diastolic blood pressure, fasting plasma glucose, total cholesterol?

: Thank you for your comment. The "difference" of these characteristics refers to the numerical change of the baseline and the 4 year prior. For example, systolic (1.0 ± 16.1 mmHg) and diastolic blood pressure (0.3 ± 11.5 mmHg), blood cholesterol (5.5 ± 34.7 mg/dl), and fasting blood sugar (5.1 ± 26.0 mg/dl) increased in the smoking cessation group at baseline compared to the 4 year prior.

7. Table 2. Body mass index is not weight gain. Adding additional variables into their multivariate models parens alcohol, income, exercise, hypertension, diabetes, hyperlipidemia) does not change risk estimates beyond more parsimonious models. Authors are urged to simplify their models and descriptions.

: Thank you for your comment. In the selection of confounders in our study, we selected variables that influence the cardiovascular outcomes. We agree that chronic disease such as hypertension, diabetes, and hyperlipidemia are not appropriate to be included as variables under the influence of weight gain. We’ve subtracted these mediators from variables, and instead presented a subgroup analysis with or without mediators.

8. Table 3. The results and conclusions would be much easier to understand if they were presented as actual weight gains rather than changes in BMI.

: Thank you for your comment. As we answered in comment #2, we’ve conducted a new analysis using actual weight gains.

9. Supporting information is confusing. Figure 1 raises additional questions about the study design since it suggests that BMI was assessed over an interval of 4 years prior to baseline while MI and ischemic strokes were assessed for a period of 5 years after baseline. Content of figure 2 is also unclear.

: Thank you for your comment. As you said, weight change was assessed over an interval of 4 years prior (2005~2006) to baseline (2009 ~ 2010). MI and ischemic strokes were followed up from baseline (2009 ~ 2010) to the end of 2015, and the mean follow-up periods for MI and ischemic strokes were 5.9 ± 0.7 years and 5.8 ± 0.8 years, respectively. We’ve modified figure 1 and deleted figure 2.

10. As a minor point the manuscript would benefit from careful editing to clarify wording and grammar.

: We thank you for pointing out this and have tried to clarify wording and correct the English grammar to improve communication.

To reviewer #2

1. The first sentence of the second paragraph is confusing. Consider rewording.

: Thank you for your comment. We’ve reworded the first sentence of the second paragraph which can be confusing.

2. Consider rewording the second sentence to "Clair et al. [8] showed that body weight change after smoking cessation....."

: Thank you for your comment. We’ve reworded the second sentence to “In the study using cohort data from the Framingham Offspring Study...”

3. Explain what you mean by "mortality seems better in higher BMI groups"

: Thank you for your comment. Cardiovascular prognosis and mortality seems better in overweight or obese described by the “obesity paradox”. This means that the risk of actual cardiovascular events due to weight gain might be underestimated when determining mortality as an outcome. As a result of the study by Hu Y et al. mentioned in that section, recent quitters that did not gain weight had a higher risk of mortality relative to those who gained weight.

Methods

Study population:

4. Correct the following sentence"Our study approved by the NHIS" to "Our study WAS approved by the NHIS"

: Thank you for your comment. We’ve corrected it as your comment.

5. Correct "4 years ago" to "4 years PRIOR" in line 77/78

: Thank you for your comment. We’ve corrected it as your comment.

6. Correct "Individuals diagnosed with previous MIs or ISs" to "Individuals PREVIOUSLY DIAGNOSED with MIs or ISs"

: Thank you for your comment. We’ve corrected it as your comment.

7. Correct "We further excluded 17,192 patients who diagnosed with MI or IS..." to "We further excluded 17,192 patients who WERE diagnosed with MI or IS..."

: Thank you for your comment. We’ve removed this sentence from the text.

8. Explain why 17,192 who were diagnosed with MI or MS were excluded. What was unclear about the causal relationship?

: Thank you for your comment. We initially thought it was unclear whether a casual relationship existed between the incidence of myocardial infarctions or strokes within one year from baseline and exposure to weight change. However, we agree that you and reviewer #1 were confused with this approach. We’ve included all MIs and ISs that occurred after the baseline without a one-year washout period in the analyses.

9. Excluding patients with the outcome without further explanation contributes to selection bias.

: Thank you for your comment. Since our study made the first diagnosed cardiovascular disease as an outcome, the participants who were previously diagnosed with cardiovascular disease could not be included in the study. In addition, people with a history of cardiovascular disease could have unintentional weight loss and/or sarcopenia. This can be a bias to the weight change after smoking cessation, an exposure of our study.

Definition of smoking history and outcomes:

10. Replace "years ago" with "years prior"

: Thank you for your comment. We replaced it with "years prior".

11. Explain what you mean by "self-questionnaire". Do you mean self-report?

: Thank you for your comment. “Self-questionnaire" are questions that must be answered in a medical examination. Questions on current smoking status, duration, and amount of smoking are written in the questionnaire. The participants responded to their current smoking status through one of three choices: never smoked, smoking in the past but now quitting, or continuing to smoke.

12. How did you define "steadily smoking"? Consider adding the actual questions that were asked to this section. Same for those who quit.

: Thank you for your comment. The participants responded to their current smoking status through one of three choices: never smoked, smoking in the past but now quitting, or continuing to smoke. Current smokers were defined as those who responded to smoke continuously from 4 years prior to baseline. Those who consistently never smoked were classified as non-smokers. The participants who quit smoking at baseline, but who were smokers 4 years prior were classified as the smoking cessation group. We’ve added these to the text to avoid confusion of communication.

13. State the follow-up period in this section.

: Thank you for your comment. We’ve written a follow-up period in this section. The mean follow-up periods for MI and ischemic strokes were 5.9 ± 0.7 years and 5.8 ± 0.8 years, respectively.

14. This sentence in unclear "The exclusion criteria for previous MIs or ISs were the same as above." State clearly what the exclusion criteria were.

: Thank you for your comment. The occurrence of a MI was defined as an ICD-10 I21 or I22 code claimed at least twice, or more than once with a hospitalization. The occurrence of an IS was defined as an ICD-10 I63 or I64 code claimed together with a hospitalization and a radiological examination (magnetic resonance imaging or computed tomography). The participants who had a history of MI or IS identified using these ICD code claims prior to baseline were excluded. We've updated this in the text.

Other baseline characteristics:

15. Include sex and age in this section.

: Thank you for your comment. We’ve written statements about age and sex in this section.

Statistical analysis:

16. It is unclear whether this was a primary or secondary data analysis.

: Thank you for your comment. We’ve corrected the paragraph so that the primary and secondary analyses are not confused.

17. Incidence rates are obtained by a Binomial/Poisson regression, Risk Differences are obtained by a normal/log-normal regression, and hazard ratios are obtained by time-to-even regression. These are all different models assessing different measures. Please explain which procedure was used to obtain which measure and what the exposure and outcome were for each procedure. Also, explain whether or not models were adjusted, and if so, the variables that were adjusted for.

: Thank you for your comment. The incidence rates were expressed as crude rates without statistical test analysis. HRs in primary and secondary analyses were analyzed using a Cox proportional hazards model with a 95% confidence interval (CI). We’ve added a description of variables used for adjustment in each model to the text.

Table 1:

18. Indicate which variables are shown as N(%) and which are shown as Mean (SD).

: Thank you for your comment. We’ve inserted N(%) and Mean(SD) in Table 1, which are appropriate for each characteristics.

19. Add a column fourth for the total and percentages for smoking status.

: Thank you for your comment. We’ve added the number of subjects in each group, along with the percentages of the total.

20. Replace "Number" by "Characteristic" and "No. (%)" by "N (%)".

: Thank you for your comment. The "number" we entered was the number of subjects in each group in the row. We’ve modified "No. (%)" to "N (%)".

21. Footnote: use lower case "p" indicating p-value.

: Thank you for your comment. We replaced uppercase P by lowercase.

22. What do you mean by "All characteristics met P < 0.0001"? What is being being tested here? Please specify.

: Thank you for your comment. This indicates that there are significant differences in characteristics of the three groups according to smoking history. The difference between the three groups is significant, even if only one comparison of the three groups is significant.

23. Add the proportion of daily and non-daily smokers to Table 1. Pack years cannot be calculated for non-daily smokers.

: Thank you for your comment. Unfortunately, the daily and non-daily smokers cannot be distinguished from the questionnaire used in the medical examination, so this may be a limitation of our study. We describe this 'point prevalence' design in the limitation section of the discussion.

Table 2:

24. Consider changing the reference group to current smokers.

: Thank you for your comment. HRs were compared among the smoking cessation, non-smoker, and current-smoker groups with current smokers as a reference group. If the meaning of this comment needs to be modified "1 (Ref.)", we will revise it as soon as you respond.

25. Was the incidence rate adjusted or not?

: Thank you for your comment. The incidence rates were expressed as crude rates without adjustment.

Table 3:

26. Consider changing the reference group to current smokers.

: Thank you for your comment. HRs were compared among the smoking cessation, non-smoker, and current-smoker groups with current smokers as a reference group. If the meaning of this comment needs to be modified "1 (Ref.)", we will revise it as soon as you respond.

27. Was the incidence rate adjusted or not?

: Thank you for your comment. The incidence rates were expressed as crude rates without adjustment.

28. It would be worthwhile to perform the regression of the outcomes on BMI (4 categories) stratified by smoking status adjusting for all identified confounders except BMI.

: Thank you for your comment. We agreed with reviewer # 1's comment #2, so we've conducted a new analysis with a straightforward approach using weight, not BMI. Therefore, baseline BMI has been included in variables.

Results

29. Please do not use risk ratio and hazard ratio interchangeably, RR does not take into effect the time to event while HR does.

: Thank you for your comment. We have fixed the risk ratio to the hazard ratio as your opinion.

Discussion

30. Be aware of overarching statements or statements that are not supported by evidence such as the first 2 sentences of the discussion.

: Thank you for your comment. We’ve tried to reduce exaggerated expressions that are not supported by evidence.

31. This paper did not assess the effect of BMI on the risk of MS and IS as mentioned in the second paragraph of the discussion, however, this would be possible with the analyses that I previously suggested under table 3.

: Thank you for your comment. As we responded to your comment #28, we’ve conducted a new analysis using weight.

32. Discussion will likely have to be revised according to the suggested analyses.

: Thank you for your comment. We’ve added related contents to the discussion.

General:

33. Consider restricting the analysis to those who are 40 years of age or older as incidence of cardiovascular outcomes are likely to be low in the younger age groups.

: Thank you for your comment. We agree with your comment and have restricted the research group to those over 40 years of age. The results are consistent after restriction.

34. Specify the novelty of this article and what value it adds to the existing literature.

: Thank you for your comment. Although there have been various studies before our study on the relationship between weight gain after smoking cessation and cardiovascular disease, we demonstrated significant results by evaluating the occurrence of real cardiovascular events as outcome in millions of people including both sexes across the country, complementing the shortcomings of these studies.

35. Figure 2 does not add any more information than Table 3, consider removing.

: Thank you for your comment. We’ve retained Table 3 and removed figure 2.

Attachment

Submitted filename: Response to Reviewers.docx

Click here for additional data file. (25.8KB, docx)

Decision Letter 1

Michael Cummings

12 Jun 2020

Protective effect of smoking cessation on subsequent myocardial infarction and ischemic stroke independent of weight gain: A nationwide cohort study

PONE-D-20-01312R1

Dear Dr. Lee,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Michael Cummings, PhD

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #2: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #2: No

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #2: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #2: (No Response)

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #2: No

Acceptance letter

Michael Cummings

6 Jul 2020

PONE-D-20-01312R1

Protective effect of smoking cessation on subsequent myocardial infarction and ischemic stroke independent of weight gain: A nationwide cohort study

Dear Dr. Lee:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Michael Cummings

Academic Editor

PLOS ONE

Associated Data

    This section collects any data citations, data availability statements, or supplementary materials included in this article.

    Supplementary Materials

    S1 Table. Incidence Rate (IR) and multivariate adjusted Hazard Ratios (HRs) (95% confidence intervals) of myocardial infarctions and ischemic strokes according to the tertiles of weight change.

    (DOCX)

    Click here for additional data file. (29.9KB, docx)
    S2 Table. Secondary subgroup analyses according to sex, age, and the presence or absence of hypertension, diabetes, hyperlipidemia, or abdominal obesity.

    (DOCX)

    Click here for additional data file. (32.4KB, docx)
    Attachment

    Submitted filename: Response to Reviewers.docx

    Click here for additional data file. (25.8KB, docx)

    Data Availability Statement

    The data underlying our study are third party data. Data are available from the NHIS Institutional Data Access/Ethics Committee for researchers who meet the criteria for access to confidential data. For information on request data from the NHIS Institutional Data Access/Ethics Committee, please see: https://nhiss.nhis.or.kr/bd/ab/bdaba032eng.do.

    Articles from PLoS ONE are provided here courtesy of PLOS

    RESOURCES