2023 Volume 30 Issue 4 Pages 348-363
Aims: Active cigarette smoking was intensively reported to increase the risk of aortic mortality while research on the association between smoking cessation and aortic mortality remains scarce. This study aimed to reconfirm the associations of exposure to cigarettes and smoking cessation associated with aortic mortality in a large Japanese population.
Methods: In the Japan Collaborative Cohort (JACC) Study, 91,141 residents (57±10 years; men, 43%) who were free of stroke, coronary heart disease, and cancer were followed up from 1989–90 until 2009 during which 110 deaths from aortic dissection and 112 deaths from aneurysm were identified. Cox proportional hazard model was used to estimate multivariable hazard ratios (95%CI) for total and specific aortic mortality.
Results: Compared to never smoking, HRs for total aortic mortality were 2.39 (1.40-4.08) for ＜20, 3.57 (2.19-5.83) for 20-39, and 3.92 (2.37-6.48) for ≥ 40 pack-years exposure. Compared to current smoking, HRs for total aortic mortality were 0.42 (0.18-0.97) for 10-15 years, 0.27 (0.11-0.66) for ＞15 years of cessation, and 0.24 (0.13-0.44) for never smoking. Similar inverse dose-response pattern was observed between smoking cessation duration and risk of mortality from aortic aneurysm (p for trend=0.001), but the association with aortic dissection mortality did not reach statistical significance.
Conclusions: Cigarette smoking was associated with an increased risk of aortic mortality while smoking cessation was so with a reduced risk among the Japanese population.
Aortic diseases (dissection and aneurysm) are large arterial diseases with vascular degeneration. Aortic dissection (AD), which mainly occurs in the proximal portion of the aorta, is a lesion that allows blood to flow between the layers of the aortic wall and thus force the aortic wall apart1). Aortic aneurysm (AA) is a focal dilatation of the aorta that exceeds the normal diameter by 50% or 3cm, mostly occurs in the abdominal aorta2). Despite the low incidence compared to those of stroke and coronary heart disease (CHD), aortic diseases are potentially fatal. Approximately 20%-50% of patients with acute AD and a ruptured abdominal AA died before reaching any hospitals3, 4). Therefore, it is of importance to confirm risk factors for aortic diseases for identifying people at high risk and prevention campaigns.
Cigarette smoking is an established risk factor for cardiovascular disease (CVD) and smoking cessation is emphasized as a crucial public health measure that reached the international consensus through the World Health Organization’s Framework Convention on Tobacco Control5). Substantial studies have revealed a positive association between smoking and the risk of most cardiovascular diseases6-8) as well as aortic diseases9-13). Meanwhile, smoking cessation was reported to reduce the risk of CHD and stroke5, 6) and aortic aneurysm11-13). However, only a few of these studies investigated the impact of active smoking or smoking cessation on aortic dissection and are mainly focused on the Western population9, 10). The reason why current research was preoccupied with AA may be attributed to the fact that the incidence of AA is 2-3 times higher than that of AD10, 14). This lack of attention on AD magnified among Asian research may be further related with the lower death rate of general aortic diseases among Asian population compared to that among Western European and North American populations (4.36 vs. 7.68 and 6.81 per 100,000 population)15). Therefore, this study aims to investigate the association of active smoking and smoking cessation and mortality from aortic dissection and aneurysm in a large Japanese community-based population.
Details of the Japan Collaborative Cohort (JACC) Study for Evaluation of Cancer Risk have been described elsewhere16). Briefly, the JACC study, sponsored by the Japanese Ministry of Education, Science, Sports, and Culture, is a large community-based cohort study established in 1988-90 and followed up until 2009. 110,585 individuals aged 40 to 79 years (46,395 men and 64,190 women) from 45 areas throughout Japan were enrolled at baseline. They were asked to complete a valid self-administered questionnaire about individual medical history (including major cardiovascular diseases, diabetes mellitus, and cancer), socio-demographics, and lifestyle (including smoking behavior). Informed consent was obtained from individuals before participation in the forms of written, oral, and group consent where appropriate.
Among the baseline participants, we excluded 5,850 individuals who had a history of stroke, CHD, or cancer; 10,207 individuals who did not report their smoking status; 1,531 current smokers who did not report their daily cigarette consumption and smoking duration; and 1,856 former smokers who did not report the daily cigarette consumption, the age when they started smoking, and the age when they quit. Ultimately, a total of 91,141 men and women were eligible for the analysis (Supplemental Fig.1). This study was approved by the institutional review boards of Hokkaido University (approval number 14-044) and Osaka University (approval number 14285).
Flowchart of participant selection
Qualitative and quantitative smoking information were acquired from the baseline assessment. Smoking status was derived from the question of “Do you smoke cigarette?” with answers to “Never”, “Used to”, or “Currently”. Daily cigarette consumption and duration of smoking for current and former smokers were further acquired from answers to the question “How many cigarettes do you smoke/did you smoke per day?” and “How long have/had you smoked? (recorded in years)”. Smoking cessation duration for former smokers were acquired from answer to the question “How many years since you have quit smoking?”. Smoking status was accordingly categorized into groups of never, former smokers, and current smokers with consumption of ＜15, 15-24, ≥ 25 sticks per day, respectively. Cumulative exposure to cigarettes was measured by the index of pack-year, which is a unit calculated using the following formula: pack-years=number of cigarettes smoked per day/20 (1 pack)×years one has smoked17). We defined cumulative exposure being mild with ＜20 pack-years, intermediate with 20-39 pack-years, and severe with ≥ 40 pack-years for former and current smokers. We also classified smoking cessation for former smokers into categories of recent quitting (0-4 years), quitted for 5-9, 10-15, and long-term quitting (＞15 years). The validity of self-reported smoking behavior was not tested in JACC study. However, a study compared self-reported current smoking with salivary cotinine among 273 Japanese workers showed the receiver operating curve with an AUC of 0.983, which indicated that self-reported smoking was highly valid18).
Mortality SurveillanceWe conducted an annual follow-up survey and tracked mortality and loss to follow-up in the cohort. We confirmed the date and cause of death with the official death certificates acquired from public health centers in every study area, where the underlying cause of death was coded according to the 10th Revision of the International Classification of Diseases on Mortality Statistics (ICD-10). Deaths from aortic dissection (code I710) and aortic aneurysm (codes I711-I719) were defined as the outcome. Despite no access to diagnostic imagining or autopsy data, a previous hospital-based autopsy study revealed that little misclassification was found between clinical autopsy based and death certificate derived diagnoses of aortic disease death19). We also verified the data on moving communities with population registry sheets and regarded participants moved out of the study areas or died of other causes were censored ones. The follow-up continued until the end of 1999 in 4 areas, 2003 in 4 areas, 2008 in 2 areas, and otherwise the end of 2009.
Covariate MeasurementCovariates information was extracted from valid answers to the self-reported questionnaire on histories and medications of hypertension and diabetes mellitus, education and employment status, lifestyle including drinking, walking, exercise behavior, and perceived mental stress at baseline. Body mass index (BMI) was calculated using self-reported height and weight (BMI=weight (kg) / [height (m)]2)20).
Statistical AnalysisThe sex- and age-adjusted mean values and proportions of selected cardiovascular risk characteristics according to smoking status were assessed by general linear model. Person-years were calculated from the date of the baseline assessment to either the date of death, moving out of the study area or loss to follow-up, or scheduled end date for each area, whichever occurred first. The percentages of missing values for all variables were presented as follows: hypertension 8.4%; diabetes 10.2%; BMI 4.6%; drinking 2.9%; mental stress 28.1%; education 24.8%; occupation 14.8%; walking 22.7%; and sports 18.1%. Missing values were categorized as dummy variable and included in the analysis. Cumulative proportion of deaths from aortic diseases according to smoking status was first estimated. Cox proportional hazard regression model was used to compute hazard ratios (HRs) and 95%CIs for mortality from aortic dissection and aneurysm across smoking status, cumulative exposure to cigarettes, and smoking cessation duration categories. All HRs were age- and sex-adjusted in model 1. In model 2, we adjusted further for study areas, BMI (＜18.5, 18.5–22.9, 23–24.9, ≥ 25 kg/m2), history of hypertension and diabetes (yes or no), drinking status (never, former, current drinker), age when completed education (≤ 18 or ≥ 19 years), occupation (employed or self-employed vs. housewife or no job), perceived high mental stress (no, somewhat, or high), walking (＜0.5h or ≥ 0.5 hour per day), sport (＜1 or ≥ 1 hour per week). For cessation analysis, model 2 also included formerly cigarettes smoked per day (≥ 25, 15-24, ＜15 cigs per day) and smoking initiation age (≤ 20, 21-24, ≥ 25 years old). Test for trend was then conducted to detect the existence of a dose-response relationship between current daily cigarette consumption, cumulative exposure to cigarettes, and smoking cessation duration with aortic mortality, respectively. All calculations were performed using SAS version 9.4 (SAS Institute, Cary, NC, USA). All p values for the statistical tests were 2-sided, with p＜0.05 being considered significant.
Table 1 shows the baseline sex- and age-adjusted characteristics of selected cardiovascular risk factors according to smoking status. Never smokers comprise mostly women (85%) while more than 70% of smokers comprise men. Former smokers were the oldest and most physically active and had the highest prevalence of hypertension and diabetes. Current heavy smokers (≥ 25 cigarettes/day) were the youngest and mostly current drinkers. In general, former and current smokers tend to have higher systolic and diastolic blood pressure, higher prevalence of diabetes but low compliance to anti-diabetic medication. They also tend to be current drinkers, employed workers, relatively highly educated, and afflicted with subjective stress. During a mean follow-up of 16.4 years and 1,495,555 person-years, a total of 5,276 participants were lost to follow-up due to moving out of the study areas or missing and a total of 20,939 deaths were recorded. A total of 110 aortic dissection deaths and 112 aortic aneurysm deaths were identified from the 91,141 participants (2.4%). Among aortic aneurysmal deaths, 42 (27 men and 15 women) thoracic, 59 (48 men and 11 women) abdominal, 5 (5 men and 0 women) thoracoabdominal, and 6 (2 men and 4 women) unspecified aneurysms were identified.
| Never- smokers | Cessation duration for former smokers, years | Current smokers, cigs/day | ||||||
|---|---|---|---|---|---|---|---|---|
| ＞15 | 10-15 | 5-9 | 0-4 | ＜15 | 15-24 | ≥ 25 | ||
| Participants, n | 57,699 | 2348 | 2217 | 2317 | 2643 | 5285 | 12,668 | 5,964 |
| Age, y | 57.0±10.0 | 62.8±9.3＊＊＊ | 58.6±10.2＊＊＊ | 58.2±10.1＊＊＊ | 58.2±10.0＊＊＊ | 59.0±10.4＊＊＊ | 56.4±9.8＊＊＊ | 52.4±8.9＊＊＊ |
| Men,％ | 15.0 | 94.7＊＊＊ | 94.2＊＊＊ | 93.1＊＊＊ | 91.3＊＊＊ | 72.8＊＊＊ | 92.1＊＊＊ | 96.8＊＊＊ |
| Systolic blood pressure, mmHg | 131.2±17.4 | 136.9±15.9 | 134.4±15.2 | 135.1±15.6＊＊ | 134.3±16.1 | 133.4±17.3 | 133.5±16.7＊ | 131.3±16.2 |
| Diastolic blood pressure, mmHg | 77.7±10.8 | 80.5±10.7 | 80.7±10.4 | 80.4±10.4 | 80.5±10.2 | 78.8±10.8＊＊＊ | 79.6±10.7＊ | 79.3±10.9＊ |
| BMI, kg/m2 | 22.9±3.1 | 22.7±2.8 | 22.9±2.8 | 23.1±2.8＊＊＊ | 22.9±2.9 | 22.1±3.0＊＊＊ | 22.3±2.8＊＊＊ | 22.9±2.8 |
| History of hypertension,％ | 19.7 | 27.6＊＊＊ | 22.4＊＊＊ | 23.1＊＊＊ | 21.7＊＊＊ | 19.8＊＊＊ | 16.7＊＊＊ | 13.0＊＊＊ |
| Anti-hypertensive medication, ％ | 11.6 | 15.0＊＊＊ | 12.7＊＊＊ | 13.4＊＊＊ | 11.3＊＊＊ | 12.2＊＊＊ | 9.2＊＊＊ | 6.1＊＊＊ |
| History of diabetes,％ | 3.5 | 8.4＊＊＊ | 7.3＊＊＊ | 7.0＊＊＊ | 6.6＊＊＊ | 5.6＊＊＊ | 5.4＊＊＊ | 5.2＊＊＊ |
| Anti-diabetic medication, ％ | 1.9 | 3.2＊＊＊ | 3.2＊＊＊ | 3.5＊＊＊ | 2.5＊＊＊ | 2.6＊＊＊ | 2.3＊＊＊ | 2.0＊＊＊ |
| Current drinker, ％ | 28.5 | 71.9＊＊＊ | 73.3＊＊＊ | 72.4＊＊＊ | 67.5＊ | 65.2＊＊＊ | 73.5＊＊＊ | 75.5＊＊＊ |
| Sports ≥ 5 hr/week,％ | 4.0 | 7.4 | 7.1＊＊＊ | 6.2 | 5.8＊＊ | 5.9 | 5.2 | 4.5 |
| Walking ≥ 1 hr/day,％ | 55.1 | 54.2 | 56.0＊＊＊ | 54.3＊ | 49.5＊＊ | 55.1 | 54.5 | 52.2 |
| Perceived high stress,％ | 14.7 | 12.2 | 17.1＊＊＊ | 17.8 | 16.4 | 14.8 | 15.7 | 21.6＊＊＊ |
| Unemployed,％ | 38.4 | 25.1＊＊＊ | 21.2＊＊＊ | 18.5＊＊＊ | 19.1 | 26.8＊＊＊ | 14.5＊ | 7.3＊＊＊ |
| College or higher education, ％ | 8.6 | 16.4 | 17.6＊＊＊ | 14.0 | 12.2＊＊＊ | 10.5＊ | 11.0 | 12.6＊＊＊ |
Characteristics were presented as mean±SD or proportions. All variables were age- and sex-adjusted except age and sex.
The percentages of missing values were as follows : hypertension 8.4%; diabetes 10.2%; BMI 4.6%; drinking 2.9%; mental stress 28.1%; education 24.8%; occupation 14.8%; walking 22.7%; sports 18.1%
P values for difference were compared to never-smokers. ＊P value ＜0.05, ＊＊P value ＜0.01, ＊＊＊P value ＜0.001
Fig.1 illustrates the cumulative proportions and 95% confidence intervals of deaths from aortic diseases according to smoking status along with follow-up years. Table 2 shows aortic mortality risk according to smoking status. Since there was no interaction between daily cigarette consumption and sex concerning mortality from aortic disease (p=0.61), we pooled men and women for further analysis. Compared with never smokers, multivariable hazard ratios (95% CIs) of total aortic diseases for former smokers and current light, intermediate, heavy smokers were 1.84 (1.04-3.15), 3.26 (1.95-5.44), 4.25 (2.66-6.77), 3.91 (2.12-7.20), respectively. The test for trend significantly showed a dose-response relationship between daily cigarette consumption and aortic mortality (p＜0.0001). HRs of mortality from aortic dissection and aneurysm showed a similar pattern despite the association related to former smoking was non-significant (HR:1.34, 95%CI: 0.63-2.87). Table 3 shows mortality risk related to cumulative exposure to cigarettes. No interaction between cumulative exposure and sex was detected (p=0.55). Multivariable HRs (95%CI) for mortality from all aortic diseases were 2.39 (1.40-4.08), 3.57 (2.19-5.83), 3.92 (2.37-7.87) for mild exposure (＜20 pack-years), intermediate exposure (20-39 pack-years), and severe exposure (≥ 40 pack-years), respectively, compared to that for never smokers. Mortality risk from aortic dissection and aneurysm reached the peak for severe exposure category, with HRs being 3.62 (1.67-7.87) and 3.98 (2.04-7.76), respectively. Test for trend significantly showed a dose-response relationship between cumulative exposure and aortic mortality (p＜0.0001) and per 10 pack-year increment has increased the risk to 1.09 (1.01-1.18).
Cumulative proportions and 95% confidence intervals of deaths from aortic dissection and aneurysm according to smoking status
| Never-smokers | Former smokers | Cigarettes smoked per day for current smokers | ||||
|---|---|---|---|---|---|---|
| ＜15 | 15-24 | ≥ 25 | P for trend | |||
| Pearson-years | 967,390 | 148,287 | 80,200 | 200,316 | 99,362 | |
| Participants, n | 57,699 | 9,525 | 5,285 | 12,668 | 5,964 | |
| Total aortic diseases | ||||||
| Cases, n | 89 | 29 | 25 | 60 | 19 | |
| Model 1 | 1.00 (ref ) | 1.81 (1.07-3.08) | 3.04 (1.82-5.06) | 3.81 (2.40-6.06) | 3.57 (1.94-6.55) | ＜.0001 |
| Model 1§ | 1.00 (ref ) | 1.77 (1.05-2.98) | 2.80 (1.68-4.66) | 3.46 (2.25-5.33) | 3.21 (1.75-5.89) | ＜.0001 |
| Model 2 | 1.00 (ref ) | 1.84 (1.07-3.15) | 3.26 (1.95-5.44) | 4.25 (2.66-6.77) | 3.91 (2.12-7.20) | ＜.0001 |
| Model 2§ | 1.00 (ref ) | 1.84 (1.09-3.13) | 3.04 (1.81-5.10) | 3.96 (2.54-6.16) | 3.62 (1.94-6.76) | ＜.0001 |
| Aortic dissection | ||||||
| Cases, n | 55 | 15 | 9 | 23 | 8 | |
| Model 1 | 1.00 (ref ) | 2.59 (1.24-5.45) | 2.61 (1.21-5.66) | 3.70 (1.88-7.30) | 3.52 (1.42-8.73) | 0.0004 |
| Model 1§ | 1.00 (ref ) | 2.55 (1.34-4.85) | 2.43 (1.18-5.02) | 3.42 (1.97-5.94) | 3.23 (1.38-7.60) | ＜.0001 |
| Model 2 | 1.00 (ref ) | 2.75 (1.28-5.90) | 2.90 (1.33-6.33) | 4.28 (1.62-8.55) | 4.07 (1.62-10.21) | 0.0001 |
| Model 2§ | 1.00 (ref ) | 2.71 (1.41-5.23) | 2.70 (1.29-5.66) | 3.99 (2.22-7.17) | 3.80 (1.56-9.25) | ＜.0001 |
| Aortic aneurysm | ||||||
| Cases, n | 34 | 14 | 16 | 37 | 11 | |
| Model 1 | 1.00 (ref ) | 1.37 (0.64-2.92) | 3.32 (1.65-6.69) | 4.02 (2.12-7.60) | 3.92 (1.72-8.90) | ＜.0001 |
| Model 1§ | 1.00 (ref ) | 1.33 (0.59-2.96) | 3.02 (1.41-6.45) | 3.56 (1.84-6.90) | 3.41 (1.43-8.17) | ＜.0001 |
| Model 2 | 1.00 (ref ) | 1.34 (0.63-2.87) | 3.49 (1.74-7.03) | 4.41 (2.34-8.31) | 4.23 (1.86-9.61) | ＜.0001 |
| Model 2§ | 1.00 (ref ) | 1.38 (0.61-3.09) | 3.26 (1.53-6.96) | 4.08 (2.11-88) | 3.86 (1.60-9.31) | 0.0002 |
Model 1 adjusted for age and sex; Model 2 further adjusted for area, body mass index, history of hypertension and diabetes, drinking status, age when completed education, occupation, perceived mental stress, walking, and sports
§Model 1and 2 adjusted for competing risk of death from lung cancer and other major cardiovascular diseases (coronary heart disease, stroke, myocardial infarction, heart failure, and arrhythmia)
| Never-smokers | Cumulative exposure, pack-year | P for trend | Per 10 pack-year increment | |||
|---|---|---|---|---|---|---|
| ＜20 | 20-39 | ≥ 40 | ||||
| Pearson-years | 967,390 | 140,723 | 238,378 | 149,064 | ||
| Participants, n | 57,699 | 8,568 | 14,779 | 10,095 | ||
| Total aortic diseases | ||||||
| Cases, n | 89 | 22 | 56 | 55 | ||
| Model 1 | 1.00 (ref ) | 2.24 (1.32-3.80) | 3.30 (2.03-5.36) | 3.60 (2.19-5.92) | ＜.0001 | 1.09 (1.01-1.17) |
| Model 1§ | 1.00 (ref ) | 2.16 (1.25-3.72) | 3.06 (1.94-4.81) | 3.25 (1.99-5.30) | ＜.0001 | 1.07 (1.00-1.15) |
| Model 2 | 1.00 (ref ) | 2.39 (1.40-4.08) | 3.57 (2.19-5.83) | 3.92 (2.37-6.48) | ＜.0001 | 1.09 (1.01-1.18) |
| Model 2§ | 1.00 (ref ) | 2.33 (1.33-4.07) | 3.38 (2.11-5.41) | 3.66 (2.21-6.05) | ＜.0001 | 1.08 (1.00-1.16) |
| Aortic dissection | ||||||
| Cases, n | 55 | 12 | 25 | 18 | ||
| Model 1 | 1.00 (ref ) | 2.57 (1.28-5.16) | 3.59 (1.80-7.15) | 3.19 (1.49-6.81) | 0.005 | 1.04 (0.91-1.19) |
| Model 1§ | 1.00 (ref ) | 2.47 (1.29-4.75) | 3.34 (1.97-5.64) | 2.89 (1.50-5.64) | 0.002 | 1.03 (0.91-1.17) |
| Model 2 | 1.00 (ref ) | 2.86 (1.41-5.80) | 4.02 (1.98-8.16) | 3.62 (1.67-7.87) | 0.003 | 1.04 (0.91-1.19) |
| Model 2§ | 1.00 (ref ) | 2.74 (1.39-5.42) | 3.79 (2.16-6.64) | 3.33 (1.69-6.56) | 0.001 | 1.03 (0.90-1.17) |
| Aortic aneurysm | ||||||
| Cases, n | 34 | 10 | 31 | 37 | ||
| Model 1 | 1.00 (ref ) | 1.90 (0.85-4.23) | 3.10 (1.57-6.09) | 3.74 (1.91-7.32) | 0.0001 | 1.12 (1.02-1.22) |
| Model 1§ | 1.00 (ref ) | 1.84 (0.75-4.51) | 2.84 (1.36-5.91) | 3.32 (1.58-7.00) | 0.001 | 1.10 (1.01-1.19) |
| Model 2 | 1.00 (ref ) | 1.97 (0.88-4.39) | 3.27 (1.66-6.43) | 3.98 (2.04-7.76) | ＜.0001 | 1.12 (1.02-1.23) |
| Model 2§ | 1.00 (ref ) | 1.96 (0.80-4.81) | 3.11 (1.49-6.50) | 3.74 (1.77-7.90) | 0.0001 | 1.11 (1.02-1.20) |
Model 1 adjusted for age and sex; Model 2 further adjusted for area, body mass index, history of hypertension and diabetes, drinking status, age when completed education, occupation, perceived mental stress, walking, and sports
§Model 1and 2 adjusted for competing risk of death from lung cancer and other major cardiovascular diseases (coronary heart disease, stroke, myocardial infarction, heart failure, and arrhythmia)
Table 4 shows mortality risk across cessation duration of former smokers. No interaction effect was detected between smoking cessation and sex (p=0.48). Decreased risk was found in all former smokers but was only significant for those who quitted for at least 10 years. Multivariable HRs (95%CI) for total aortic mortality were 0.42 (0.18-0.97), 0.27 (0.11-0.66) for those who quitted for 10-15 years, and more than 15 years, respectively, compared to that of 0.24 (0.13-0.44) for never smokers. HRs for aortic aneurysm mortality were 0.20 (0.05-0.82) and 0.21 (0.06-0.67) for those who quitted for 10-15 years and more than 15 years, respectively, with the latter one being close to that of 0.18 (0.08-0.41) for never smokers. However, this pattern was not significant for aortic dissection even with cessation of more than 15 years (HR: 0.35, 95%CI: 0.08-1.47). Although decreased mortality risk with per 10-year cessation increment was non-significant, test for trend significantly supported an inverse linear association between smoking cessation and aortic mortality (p=0.0007). Sex-specific aortic mortality risk associated with smoking status, cumulative pack-year, and smoking cessation were also provided (Supplemental Tables 1, 2, 3).
| Current smokers | Cessation duration for former smokers, year | Never-smokers | P for trend | Per 10-year cessation | ||||
|---|---|---|---|---|---|---|---|---|
| 0-4 | 5-9 | 10-15 | ＞15 | |||||
| Pearson-years | 379,878 | 41,437 | 36,320 | 35,086 | 35,444 | 967,390 | ||
| Participants, n | 23,917 | 2,643 | 2,317 | 2,217 | 2,348 | 57,699 | ||
| Total aortic diseases | ||||||||
| Cases, n | 104 | 9 | 9 | 6 | 5 | 89 | ||
| Model 1 | 1.00 (ref ) | 0.64 (0.32-1.26) | 0.74 (0.37-1.46) | 0.48 (0.21-1.09) | 0.29 (0.19-0.44) | 0.29 (0.19-0.44) | 0.001 | 0.63 (0.39-0.99) |
| Model 1§ | 1.00 (ref ) | 0.66 (0.33-1.30) | 0.78 (0.39-1.54) | 0.52 (0.23-1.19) | 0.31 (0.13-0.76) | 0.32 (0.21-0.47) | 0.004 | 0.64 (0.42-0.98) |
| Model 2 | 1.00 (ref ) | 0.58 (0.29-1.14) | 0.67 (0.34-1.33) | 0.42 (0.18-0.97) | 0.27 (0.11-0.66) | 0.24 (0.13-0.44) | 0.001 | 0.69 (0.43-1.10) |
| Model 2§ | 1.00 (ref ) | 0.60 (0.31-1.17) | 0.70 (0.35-1.39) | 0.47 (0.20-1.09) | 0.29 (0.12-0.73) | 0.26 (0.14-0.50) | 0.003 | 0.69 (0.45-1.07) |
| Aortic dissection | ||||||||
| Cases, n | 40 | 5 | 4 | 4 | 2 | 55 | ||
| Model 1 | 1.00 (ref ) | 0.99 (0.39-2.50) | 0.92 (0.33-2.57) | 0.92 (0.33-2.59) | 0.35 (0.08-1.45) | 0.31 (0.17-0.56) | 0.189 | 0.53 (0.26-1.09) |
| Model 1§ | 1.00 (ref ) | 1.02 (0.40-2.58) | 0.96 (0.35-2.69) | 0.99 (0.35-2.78) | 0.38 (0.09-1.59) | 0.34 (0.21-0.54) | 0.227 | 0.55 (0.30-1.01) |
| Model 2 | 1.00 (ref ) | 0.92 (0.36-2.35) | 0.87 (0.31-2.44) | 0.84 (0.30-2.38) | 0.35 (0.08-1.47) | 0.31 (0.12-0.78) | 0.214 | 0.59 (0.28-1.23) |
| Model 2§ | 1.00 (ref ) | 0.95 (0.39-2.32) | 0.89 (0.32-2.53) | 0.92 (0.32-2.64) | 0.38 (0.09-1.61) | 0.33 (0.14-0.79) | 0.274 | 0.58 (0.33-1.00) |
| Aortic aneurysm | ||||||||
| Cases, n | 64 | 4 | 5 | 2 | 3 | 34 | ||
| Model 1 | 1.00 (ref ) | 0.43 (0.16-1.19) | 0.63 (0.25-1.56) | 0.24 (0.06-0.97) | 0.23 (0.07-0.72) | 0.27 (0.15-0.48) | 0.002 | 0.71 (0.39-1.30) |
| Model 1§ | 1.00 (ref ) | 0.45 (0.16-1.23) | 0.67 (0.27-1.66) | 0.26 (0.06-1.07) | 0.26 (0.08-0.82) | 0.30 (0.16-0.56) | 0.009 | 0.74 (0.42-1.28) |
| Model 2 | 1.00 (ref ) | 0.38 (0.14-1.04) | 0.54 (0.22-1.35) | 0.20 (0.05-0.82) | 0.21 (0.06-0.67) | 0.18 (0.08-0.41) | 0.001 | 0.75 (0.39-1.44) |
| Model 2§ | 1.00 (ref ) | 0.40 (0.15-1.09) | 0.58 (0.23-1.45) | 0.23 (0.06-0.95) | 0.24 (0.07-0.77) | 0.21 (0.08-0.51) | 0.006 | 0.77 (0.39-1.50) |
Model 1 adjusted for age and sex; Model 2 further adjusted for area, body mass index, history of hypertension and diabetes, drinking status, age when completed education, occupation, perceived mental stress, walking, sports, cigarettes smoked per day, and smoking initiation age
§Model 1and 2 adjusted for competing risk of death from lung cancer and other major cardiovascular diseases (coronary heart disease, stroke, myocardial infarction, heart failure, and arrhythmia)
| Never-smokers | Former smokers | Cigarettes smoked per day for current smokers | P for trend | |||
|---|---|---|---|---|---|---|
| ＜15 | 15-24 | ≥ 25 | ||||
| Male | ||||||
| Pearson-years | 143,367 | 138,562 | 57,659 | 184,470 | 96,354 | |
| Participants, n | 8,636 | 8,879 | 3,845 | 11,667 | 5,773 | |
| Total aortic diseases | ||||||
| Cases, n | 12 | 27 | 18 | 59 | 17 | |
| Model 1 | 1.00 (ref ) | 1.97 (1.00-3.89) | 3.12 (1.50-6.48) | 4.32 (2.32-8.05) | 3.56 (1.68-7.52) | ＜.0001 |
| Model 2 | 1.00 (ref ) | 2.08 (1.05-4.13) | 3.71 (1.78-7.73) | 5.03 (2.69-9.42) | 4.02 (1.90-8.52) | ＜.0001 |
| Aortic dissection | ||||||
| Cases, n | 2 | 14 | 6 | 22 | 7 | |
| Model 1 | 1.00 (ref ) | 6.50 (1.48-28.63) | 6.63 (1.34-32.87) | 9.01 (2.12-38.33) | 6.89 (1.42-33.41) | 0.004 |
| Model 2 | 1.00 (ref ) | 6.89 (1.55-30.53) | 7.80 (1.56-38.89) | 9.89 (2.31-42.29) | 7.70 (1.58-37.53) | 0.003 |
| Aortic aneurysm | ||||||
| Cases, n | 10 | 13 | 12 | 37 | 10 | |
| Model 1 | 1.00 (ref ) | 1.12 (0.49-2.55) | 2.46 (1.06-5.68) | 3.49 (1.73-7.04) | 3.02 (1.24-7.37) | 0.001 |
| Model 2 | 1.00 (ref ) | 1.16 (0.51-2.68) | 2.87 (1.23-6.70) | 4.15 (2.05-8.42) | 3.43 (1.40-8.39) | 0.0002 |
| Female | ||||||
| Pearson-years | 824,022 | 9,725 | 22,541 | 15,846 | 3,008 | |
| Participants, n | 49,063 | 646 | 1,440 | 1,001 | 191 | |
| Total aortic diseases | ||||||
| Cases, n | 77 | 2 | 7 | 1 | 2 | |
| Model 1 | 1.00 (ref ) | 1.71 (0.42-6.98) | 3.45 (1.59-7.48) | 0.93 (1.13-6.70) | 10.88 (2.67-44.30) | 0.001 |
| Model 2 | 1.00 (ref ) | 1.81 (0.44-7.50) | 3.49 (1.60-7.65) | 0.97 (0.13-7.06) | 12.85 (3.11-53.06) | 0.002 |
| Aortic dissection | ||||||
| Cases, n | 53 | 1 | 3 | 1 | 1 | |
| Model 1 | 1.00 (ref ) | 1.29 (0.18-9.32) | 2.16 (0.67-6.90) | 1.28 (0.18-9.24) | 7.52 (1.04-54.46) | 0.071 |
| Model 2 | 1.00 (ref ) | 1.47 (0.20-10.82) | 2.26 (0.70-7.30) | 1.46 (0.20-10.73) | 9.45 (1.28-69.49) | 0.044 |
| Aortic aneurysm | ||||||
| Cases, n | 24 | 1 | 4 | 0 | 1 | |
| Model 1 | 1.00 (ref ) | 2.56 (0.35-18.94) | 6.17 (2.14-17.81) | — | 18.34 (2.48-135.76) | 0.003 |
| Model 2 | 1.00 (ref ) | 2.32 (0.30-17.98) | 6.01 (2.08-17.91) | — | 25.35 (3.33-193.25) | 0.010 |
Model 1 adjusted for age and sex; Model 2 further adjusted for area, body mass index, history of hypertension and diabetes, drinking status, age when completed education, occupation, perceived mental stress, walking, and sports
| Never-smoker | Cumulative exposure, pack-year | P for trend | Per 10 pack-year increment | |||
|---|---|---|---|---|---|---|
| ＜20 | 20-39 | ≥ 40 | ||||
| Male | ||||||
| Pearson-years | 143,367 | 104,091 | 227,003 | 145,952 | ||
| Participants, n | 8,636 | 6,287 | 14,006 | 9,871 | ||
| Total aortic diseases | ||||||
| Cases, n | 12 | 13 | 55 | 53 | ||
| Model 1 | 1.00 (ref ) | 1.85 (0.84-4.05) | 3.42 (1.83-6.40) | 3.53 (1.89-6.62) | ＜.0001 | 1.08 (1.00-1.17) |
| Model 2 | 1.00 (ref ) | 2.09 (0.95-4.61) | 3.85 (2.05-7.23) | 3.94 (2.10-7.40) | ＜.0001 | 1.08 (0.99-1.17) |
| Aortic dissection | ||||||
| Cases, n | 2 | 7 | 25 | 17 | ||
| Model 1 | 1.00 (ref ) | 5.55 (1.15-26.76) | 8.66 (2.05-36.60) | 7.17 (1.65-31.07) | 0.008 | 1.04 (0.90-1.19) |
| Model 2 | 1.00 (ref ) | 6.21 (1.28-30.13) | 9.50 (2.23-40.38) | 7.83 (1.80-34.04) | 0.005 | 1.03 (0.89-1.19) |
| Aortic aneurysm | ||||||
| Cases, n | 10 | 6 | 30 | 36 | ||
| Model 1 | 1.00 (ref ) | 1.07 (0.39-2.95) | 2.38 (1.16-4.89) | 2.87 (1.42-5.77) | 0.001 | 1.12 (1.01-1.22) |
| Model 2 | 1.00 (ref ) | 1.21 (1.44-3.35) | 2.68 (1.30-5.54) | 3.20 (1.58-6.48) | 0.000 | 1.11 (1.01-1.23) |
| Female | ||||||
| Pearson-years | 824,022 | 36,633 | 11,376 | 3,112 | ||
| Participants, n | 49,063 | 2,281 | 773 | 224 | ||
| Total aortic diseases | ||||||
| Cases, n | 77 | 9 | 1 | 2 | ||
| Model 1 | 1.00 (ref ) | 3.14 (1.57-6.26) | 0.89 (0.12-6.39) | 4.33 (1.06-17.71) | 0.010 | 1.14 (0.94-1.38) |
| Model 2 | 1.00 (ref ) | 3.09 (1.52-6.27) | 0.93 (0.13-6.76) | 4.74 (1.13-19.79) | 0.009 | 1.24 (0.94-1.63) |
| Aortic dissection | ||||||
| Cases, n | 53 | 5 | 0 | 1 | ||
| Model 1 | 1.00 (ref ) | 2.49 (0.99-6.22) | — | 3.37 (0.46-24.51) | 0.281 | 1.10 (0.77-1.56) |
| Model 2 | 1.00 (ref ) | 2.66 (1.04-6.81) | — | 4.43 (0.60-32.88) | 0.183 | 1.13 (0.68-1.86) |
| Aortic aneurysm | ||||||
| Cases, n | 24 | 4 | 1 | 1 | ||
| Model 1 | 1.00 (ref ) | 4.58 (1.59-13.19) | 2.85 (0.39-21.11) | 6.00 (0.80-44.77) | 0.006 | 1.16 (0.92-1.46) |
| Model 2 | 1.00 (ref ) | 3.91 (1.32-11.63) | 2.53 (0.33-19.24) | 4.70 (0.59-37.31) | 0.024 | 1.47 (0.95-2.27) |
Model 1 adjusted for age and sex; Model 2 further adjusted for area, body mass index, history of hypertension and diabetes, drinking status, age when completed education, occupation, perceived mental stress, walking, and sports
| Current smokers | Cessation duration for ex- smokers, year | Never-smokers | P for trend | Per 10-year cessation | ||||
|---|---|---|---|---|---|---|---|---|
| 0-4 | 5-9 | 10-15 | ＞15 | |||||
| Male | ||||||||
| Pearson-years | 338,483 | 37,807 | 33,961 | 33,243 | 33,552 | 143,367 | ||
| Participants, n | 21,285 | 2,312 | 2,156 | 2,088 | 2,223 | 8,636 | ||
| Total aortic diseases | ||||||||
| Cases, n | 94 | 9 | 9 | 5 | 4 | 12 | ||
| Model 1 | 1.00 (ref ) | 0.69 (0.35-1.37) | 0.79 (0.40-1.57) | 0.42 (0.17-1.04) | 0.23 (0.09-0.63) | 0.25 (0.14-0.47) | 0.001 | 0.57 (0.35-0.94) |
| Model 2 | 1.00 (ref ) | 0.62 (0.31-1.24) | 0.71 (0.36-1.42) | 0.37 (0.15-0.93) | 0.22 (0.08-0.60) | 0.23 (0.11-0.48) | 0.001 | 0.62 (0.37-1.04) |
| Aortic dissection | ||||||||
| Cases, n | 35 | 5 | 4 | 3 | 2 | 2 | ||
| Model 1 | 1.00 (ref ) | 1.13 (0.44-2.89) | 1.02 (0.36-2.88) | 0.78 (0.23-2.44) | 0.40 (0.10-1.67 | 0.12 (0.03-0.52) | 0.200 | 0.51 (0.24-1.09) |
| Model 2 | 1.00 (ref ) | 1.11 (0.43-2.86) | 0.98 (0.34-2.80) | 0.75 (0.23-2.47) | 0.42 (0.10-1.78) | 0.16 (0.03-0.74) | 0.212 | 0.54 (0.24-1.22) |
| Aortic aneurysm | ||||||||
| Cases, n | 59 | 4 | 5 | 2 | 2 | 10 | ||
| Model 1 | 1.00 (ref ) | 0.47 (0.17-1.28) | 0.67 (0.27-1.67) | 0.25 (0.06-1.04) | 0.16 (0.04-0.66) | 0.32 (0.16-0.62) | 0.002 | 0.63 (0.32-1.23) |
| Model 2 | 1.00 (ref ) | 0.40 (0.14-1.12) | 0.58 (0.23-1.46) | 0.20 (0.05-0.85) | 0.14 (0.04-0.59) | 0.23 (0.10-0.54) | 0.001 | 0.65 (0.32-1.33) |
| Female | ||||||||
| Pearson-years | 41,395 | 3,631 | 2,359 | 1,843 | 1,892 | 824,022 | ||
| Participants, n | 2,632 | 231 | 161 | 129 | 125 | 49,063 | ||
| Total aortic diseases | ||||||||
| Cases, n | 10 | 0 | 0 | 1 | 1 | 77 | ||
| Model 1 | 1.00 (ref ) | — | — | 1.52 (0.19-11.89) | 1.09 (0.14-8.55) | 0.33 (0.17-0.64) | 0.965 | 1.46 (0.45-4.68) |
| Model 2 | 1.00 (ref ) | — | — | 1.41 (0.18-11.33) | 1.19 (0.15-9.83) | 0.10 (0.02-0.59) | 0.512 | — |
| Aortic dissection | ||||||||
| Cases, n | 5 | 0 | 0 | 1 | 0 | 53 | ||
| Model 1 | 1.00 (ref ) | — | — | 3.18 (0.37-27.27) | — | 0.46 (0.18-1.16) | 0.762 | 0.94 (0.11-8.22) |
| Model 2 | 1.00 (ref ) | — | — | 2.79 (0.32-24.92) | — | — | 0.818 | — |
| Aortic aneurysm | ||||||||
| Cases, n | 5 | 0 | 0 | 0 | 1 | 24 | ||
| Model 1 | 1.00 (ref ) | — | — | — | 1.84 (0.21-15.82) | 0.20 (0.08-0.52) | 0.800 | 1.98 (0.47-8.40) |
| Model 2 | 1.00 (ref ) | — | — | — | 1.85 (0.20-17.53) | 0.02 (0.002-0.20) | 0.066 | — |
Model 1 adjusted for age and sex; Model 2 further adjusted for area, body mass index, history of hypertension and diabetes, drinking status, age when completed education, occupation, perceived mental stress, walking, and sports, cigarettes smoked per day, smoking initiation age
We observed significantly higher mortality from aortic dissection and aneurysm associated with active cigarette smoking. A positive dose-response relationship was found for cumulative exposure to cigarettes and aortic dissection and aneurysm death. On the other hand, an inverse dose-response relationship was found between smoking cessation and total aortic death, through which at least 10 years of cessation led to the reduction of mortality risk.
Although many studies thoroughly investigated and revealed that active cigarette smoking is a major risk factor for aortic aneurysms, few have focused on aortic dissection. We found that active smoking greatly increased the mortality risk of aortic dissection, which strengthens the finding from a case-control study among Japanese that reported a positive association between smoking and aortic dissection (OR:3.48, 95%CI:1.58-7.66)21). However, we didn’t detect a significant association between smoking cessation and aortic dissection mortality. Our study enriched the present limited knowledge of the association between smoking and aortic dissection among Asian populations.
The present study found a relatively stronger effect of current smoking on aortic mortality (HR: 3.79, 95%CI: 2.50-5.74, not listed), compared to CHD or total cardiovascular risk previously published. For instance, a meta-analysis7) pooling 14 cohorts from 25 countries showed a less elevated mortality risk of total CVD for current smokers (HR: 2.07, 95% CI: 1.82-2.36). In addition, another study22) showed the mortality risk of aortic dissection and aneurysm (HR: 8.47, 95% CI: 4.46-16.12) is more than 2 times higher than that of CHD (HR: 3.22, 95% CI: 2.39-4.33) or stroke (HR: 1.59, 95% CI: 0.90-2.81) associated with current smoking. As for smoking cessation, a prospective study23) revealed that the risk decreased significantly by 5 years of cessation for CHD (HR: 0.78, 95% CI: 0.64-0.95) and stroke (HR: 0.67, 95% CI: 0.51-0.88), respectively. Another study24) reported that quitting smoking after a myocardial infarction led to a decline of coronary events risk over time and eventually a risk equal to that for never smokers by 3 years of cessation. Similarly, we found total and specific aortic mortality risk reduced sooner after initiation of smoking cessation but this reduced risk was only found significant by 10 years of cessation for aortic aneurysm rather than for aortic dissection. This insignificance of smoking cessation’s benefit for aortic dissection is probably due to the impact of the paucity of aortic dissection cases among former smokers.
Interestingly, these findings of the seemingly stronger effect of smoking and the delayed benefit of smoking cessation on aortic mortality also suggest that smoking may impact aortic diseases and coronary heart disease through considerately different mechanisms. Atherosclerosis, induced by smoking through a series of effects including aggregation of low-density cholesterol25), endothelial injury, fibrinogen concentration, and formation of thrombosis and plaques26), is a key component triggering CHD as well as aortic diseases, especially aortic aneurysm. Moreover, unlike CHD, degeneration of aorta media (the crucial layer that integrates and stabilizes the aortic wall), is also suggested to play an important role in the pathogenesis of aortic diseases. The media consists of a lamellar unit structured by smooth muscle cells, elastin and collagen and thus loss of any of these components would result in a damaged media in which a tear or dilation happens, hence disrupt the aortic wall and eventually cause aortic dissection or aneurysm27). An extensive body of literature reported that excess cigarette smoke stimulates macrophages’ production of matrix metalloproteinase, thereby degenerates the elastin and collagen and lead to incidence of aortic diseases28-30). Furthermore, some basic research reported aortic dissection preceded the formation of atherosclerosis31), which indicates certain components of the pathogenesis of aortic dissection could be independent from atherosclerosis. Thus we could assume that aortic dissection is under regulation by atherosclerosis or atherosclerosis-independent degeneration of aortic media.
On the other hand, smoking cessation was reported with a certain improved atherosclerotic state. For example, fibrinogen concentration and fibrinogen synthesis rate32) and platelet aggregability33) had reduced within 2 weeks of smoking cessation and platelet volume34) decreased after 3 months of intervention. Also, lipid profile has improved with an increase in high-density cholesterol (HDL) and the HDL/LDL ratio, and a decrease in LDL35) and so did hemodynamics with significant lowered mean arterial pressure and heart rate36). However, the existing loss of normal medial arterial structure is naturally irreversible, given the improvement of atherosclerosis.
Taken together, due to the additional pathogenic component of atherosclerosis-independent degeneration of aortic media, it is reasonable that the magnitude of risk associated with smoking appears to be substantially greater and the cessation duration needed to counteract the detrimental effect of smoking is longer for aortic diseases, compared to CHD. Also, this difference of pathophysiological mechanisms probably explains the inconsistency of the beneficial effect of smoking cessation between aortic aneurysm and aortic dissection observed in the present study. Nevertheless, further research is warranted to elucidate the specific mechanism through which smoking impacts on aortic diseases.
Our study utilized a prospective cohort design that strengthened the estimation of causality. Also, the large number of residents aged 40-79 years from diverse communities across Japan that are considered a favorable representative of general middle-aged to older Japanese population. Nevertheless, several limitations should be noted. First, despite long-term follow-up of a large population, the number of aortic death cases was relatively small, especially among former smokers, which probably impact the accuracy of detecting a significant inverse association between smoking cessation and death from aortic dissection. Secondly, to reduce the reverse causality due to preclinical conditions which may affected smoking status at baseline, we repeated the analysis excluding 3,047 deaths and 1,725 censoring happened within 5 years from the baseline. The results showed no material difference from the major analysis (Supplemental Tables 4, 5, 6), which indicated that the impact of selection bias is acceptable. Thirdly, there are substantial missing for some covariates, especially for mental stress (28.1%) and education (24.8%), that would detriment the accuracy of our estimation. However, we repeated the analysis using multiple imputation37) to replenish missing values and the results showed no material difference from the main results treating missing data as dummy variables (Supplemental Tables 7, 8, 9). Fourthly, smoking is an established risk factor for lung cancer and CVD diseases and thus the competing risk38) of these diseases may not be neglected. Accordingly, we repeated the analysis accounting for competing risk of deaths from lung cancer and other major CVD (CHD, stroke, cardiac arrest, arrhythmia, and heart failure), the associations of cumulative exposure to cigarettes and smoking cessation duration with the risk of aortic mortality were slightly attenuated but remained statistically significant (Tables 2-4 Model 1-2§). Furthermore, information on smoking behavior were obtained only once at baseline; the possible alteration of smoking behavior during the follow-up would impact the accurate estimation. In accordance with the declining trend in smoking prevalence among Japanese for recent decades, some current smokers at baseline in our study were likely to quit smoking or reduce the number of cigarettes smoker per day while never or former smokers at baseline were unlikely to change their smoking status (e.g. 16% of current smokers at baseline reported to quit smoking while only 0.08% of former smokers and 0.008% of never smokers at baseline started smoking after 5 years of follow-up)39). Therefore, the beneficial effect of smoking cessation on risk of aortic disease mortality was probably underestimated. Last but not least, the application of our results based on Japanese population to other population should be cautious with its generalizability.
| Never smokers | Former smokers | Cigarettes per day for current smokers | P for trend | |||
|---|---|---|---|---|---|---|
| ＜15 | 15-24 | ≥ 25 | ||||
| Pearson-years | 960,857 | 146,392 | 78,926 | 197,897 | 98,507 | |
| Participants, n | 55,256 | 8,838 | 4,826 | 11,810 | 5,639 | |
| Total aortic diseases | ||||||
| Cases, n | 82 | 27 | 21 | 47 | 19 | |
| Model 1 | ref | 2.09 (1.20-3.66) | 3.10 (1.79-3.66) | 3.72 (2.25-6.16) | 4.43 (2.36-8.29) | ＜.0001 |
| Model 2§ | ref | 2.10 (1.19-3.70) | 3.33 (1.92-5.78) | 4.19 (2.53-6.96) | 4.89 (2.53-9.19) | ＜.0001 |
| Aortic dissection | ||||||
| Cases, n | 50 | 13 | 8 | 15 | 8 | |
| Model 1 | ref | 2.80 (1.26-6.22) | 2.82 (1.25-6.37) | 2.99 (1.38-6.47) | 4.30 (1.68-11.00) | 0.001 |
| Model 2 | ref | 2.97 (1.31-6.76) | 3.17 (1.39-7.24) | 3.57 (1.63-7.83) | 5.11 (1.97-13.26) | 0.0002 |
| Aortic aneurysm | ||||||
| Cases, n | 33 | 14 | 13 | 32 | 11 | |
| Model 1 | ref | 1.72 (0.79-3.77) | 3.34 (1.57-7.10) | 4.41 (2.23-8.74) | 4.99 (2.14-11.67) | ＜.0001 |
| Model 2 | ref | 1.65 (0.75-3.63) | 3.47 (1.63-7.38) | 4.80 (2.43-9.46) | 5.36 (2.29-12.52) | ＜.0001 |
Model 1 adjusted for age and sex; Model 2 further adjusted for area, body mass index, history of hypertension and diabetes, drinking status, age when completed education, occupation, perceived mental stress, walking, and sports
§P for interaction was 0.424 for Model 1 and 0.475 for Model 2, calculated by the likelihood ratio between sex and smoking status, respectively
| Never smokers | Cumulative exposure for current smokers, pack-year | P for trend | Per 10 pack-year increment | |||
|---|---|---|---|---|---|---|
| ＜20 | 20-39 | ≥ 40 | ||||
| Pearson-years | 960,857 | 139,380 | 235,770 | 146,572 | ||
| Participants, n | 55,256 | 8,048 | 13,852 | 9,213 | ||
| Total aortic diseases | ||||||
| Cases, n | 82 | 20 | 47 | 47 | ||
| Model 1 | ref | 2.41 (1.39-4.20) | 3.42 (2.03-5.75) | 3.86 (2.26-6.59) | ＜.0001 | 1.10 (1.02-1.19) |
| Model 2§ | ref | 2.58 (1.47-4.51) | 3.71 (2.20-6.28) | 4.19 (2.44-7.18) | ＜.0001 | 1.10 (1.01-1.19) |
| Aortic dissection | ||||||
| Cases, n | 50 | 11 | 20 | 13 | ||
| Model 1 | ref | 2.74 (1.32-5.67) | 3.46 (1.63-7.32) | 2.86 (1.23-.6.65) | 0.021 | 1.03 (0.89-1.20) |
| Model 2 | ref | 3.09 (1.47-6.49) | 3.95 (1.83-8.52) | 3.27 (1.38-7.76) | 0.012 | 1.02 (0.88-1.20) |
| Aortic aneurysm | ||||||
| Cases, n | 32 | 9 | 27 | 34 | ||
| Model 1 | ref | 2.09 (0.90-4.84) | 3.43 (1.66-7.09) | 4.43 (2.16-9.07) | ＜.0001 | 1.14 (1.04-1.25) |
| Model 2 | ref | 2.13 (0.92-4.94) | 3.57 (1.73-7.37) | 4.63 (2.27-9.45) | ＜.0001 | 1.14 (1.04-1.25) |
Model 1 adjusted for age and sex; Model 2 further adjusted for area, body mass index, history of hypertension and diabetes, drinking status, age when completed education, occupation, perceived mental stress, walking, and sports
§P for interaction was 0.502 for Model 1 and 0.464 for Model 2, calculated by the likelihood ratio between sex and pack-years, respectively
| Current smokers | Cessation duration for former smokers, year | Never smokers | P for trend | Per 10-year cessation | ||||
|---|---|---|---|---|---|---|---|---|
| 0-4 | 5-9 | 10-15 | ＞15 | |||||
| Pearson-years | 375,330 | 40,835 | 35,887 | 34,728 | 34,941 | 960,857 | ||
| Participants, n | 22,275 | 2,427 | 2,157 | 2,078 | 2,176 | 55,256 | ||
| Total aortic diseases | ||||||||
| Cases, n | 87 | 8 | 8 | 6 | 5 | 82 | ||
| Model 1†a | ref | 0.68 (0.33-1.40) | 0.79 (0.38-1.63) | 0.58 (0.25-1.32) | 0.33 (0.14-0.82) | 0.28 (0.18-0.44) | 0.004 | 0.95 (0.92-0.99) |
| Model 2‡b | ref | 0.70 (0.29-1.24) | 0.70 (0.34-1.45) | 0.50 (0.22-1.14) | 0.32 (0.13-0.78) | 0.22 (0.11-0.43) | 0.003 | 0.95 (0.91-0.98) |
| Aortic dissection | ||||||||
| Cases, n | 31 | 4 | 3 | 4 | 2 | 50 | ||
| Model 1† | ref | 1.03 (0.36-2.91) | 0.90 (0.28-2.95) | 1.21 (0.43-3.45) | 0.47 (0.11-1.96) | 0.33 (0.17-0.61) | 0.222 | 0.98 (0.93-1.03) |
| Model 2‡ | ref | 0.93 (0.32-2.65) | 0.85 (0.26-2.82) | 108 (0.37-3.10) | 0.46 (0.11-1.93) | 0.34 (0.12-0.94) | 0.313 | 0.98 (0.93-1.11) |
| Aortic aneurysm | ||||||||
| Cases, n | 56 | 4 | 5 | 2 | 3 | 32 | ||
| Model 1† | ref | 0.49 (0.18-1.36) | 0.72 (0.29-1.36) | 0.28 (0.07-1.12) | 0.26 (0.08-0.83) | 0.25 (0.13-0.46) | 0.008 | 0.93 (0.89-0.98) |
| Model 2‡ | ref | 0.42 (0.15-1.16) | 0.60 (0.24-1.50) | 0.22 (0.05-0.92) | 0.23 (0.07-0.75) | 0.15 (0.07-0.36) | 0.003 | 0.92 (0.88-0.97) |
Model 1 adjusted for age and sex; Model 2 further adjusted for area, body mass index, history of hypertension and diabetes, drinking status, age when completed education, occupation, perceived mental stress, walking, and sports
§P for interaction was 0.640 for Model 1 and 0.566 for Model 2, calculated by the likelihood ratio between sex and pack-years, respectively
| Never-smokers | Former smokers | Cigarettes per day for current smokers | P for trend | |||
|---|---|---|---|---|---|---|
| ＜15 | 15-24 | ≥ 25 | ||||
| Pearson-years | 967,390 | 148,287 | 80,200 | 200,316 | 99,362 | |
| Participants, n | 57,699 | 9,525 | 5,285 | 12,668 | 5,964 | |
| Total aortic diseases | ||||||
| Cases, n | 89 | 29 | 25 | 60 | 19 | |
| Model 1 | 1.00 (ref ) | 1.81 (1.07-3.08) | 3.04 (1.82-5.06) | 3.81 (2.40-6.06) | 3.57 (1.94-6.55) | ＜.0001 |
| Model 2§ | 1.00 (ref ) | 1.78 (1.04-3.06) | 3.21 (1.92-5.38) | 4.26 (2.67-6.80) | 3.94 (2.14-7.27) | ＜.0001 |
| Aortic dissection | ||||||
| Cases, n | 55 | 15 | 9 | 23 | 8 | |
| Model 1 | 1.00 (ref ) | 2.59 (1.24-5.45) | 2.61 (1.21-5.66) | 3.70 (1.88-7.30) | 3.52 (1.42-8.73) | 0.0004 |
| Model 2 | 1.00 (ref ) | 2.71 (1.26-5.82) | 2.92 (1.34-6.36) | 4.35 (2.17-8.69) | 4.12 (1.64-10.33) | ＜.0001 |
| Aortic aneurysm | ||||||
| Cases, n | 34 | 14 | 16 | 37 | 11 | |
| Model 1 | 1.00 (ref ) | 1.37 (0.64-2.92) | 3.32 (1.65-6.69) | 4.02 (2.12-7.60) | 3.92 (1.72-8.90) | ＜.0001 |
| Model 2 | 1.00 (ref ) | 1.30 (0.60-2.78) | 3.39 (1.68-6.84) | 4.42 (2.34-8.36) | 4.26 (1.87-9.72) | ＜.0001 |
Analysis based on 20 replicated datasets using FCS method
Model 1 adjusted for age and sex; Model 2 further adjusted for area, body mass index, history of hypertension and diabetes, drinking status, age when completed education, occupation, perceived mental stress, walking, and sports
§P for interaction was 0.583 for Model 1 and 0.628 for Model 2, calculated by the likelihood ratio between sex and smoking status, respectively
| Never-smoker | Cumulative exposure, pack-year | P for trend | Per 10 pack-year increment | |||
|---|---|---|---|---|---|---|
| ＜20 | 20-39 | ≥ 40 | ||||
| Pearson-years | 967,390 | 140,723 | 238,378 | 149,064 | ||
| Participants, n | 57,699 | 8,568 | 14,779 | 10,095 | ||
| Total aortic diseases | ||||||
| Cases, n | 89 | 22 | 56 | 55 | ||
| Model 1 | 1.00 (ref ) | 2.24 (1.32-3.80) | 3.30 (2.03-5.36) | 3.60 (2.19-5.92) | ＜.0001 | 1.09 (1.01-1.17) |
| Model 2§ | 1.00 (ref ) | 2.35 (1.38-4.02) | 3.55 (2.17-5.81) | 3.88 (2.34-6.43) | ＜.0001 | 1.08 (1.00-1.17) |
| Aortic dissection | ||||||
| Cases, n | 55 | 12 | 25 | 18 | ||
| Model 1 | 1.00 (ref ) | 2.57 (1.28-5.16) | 3.59 (1.80-7.15) | 3.19 (1.49-6.81) | 0.005 | 1.04 (0.91-1.19) |
| Model 2 | 1.00 (ref ) | 2.86 (1.41-5.80) | 4.07 (2.01-8.26) | 3.62 (1.67-7.87) | 0.003 | 1.03 (0.91-1.18) |
| Aortic aneurysm | ||||||
| Cases, n | 34 | 10 | 31 | 37 | ||
| Model 1 | 1.00 (ref ) | 1.90 (0.85-4.23) | 3.10 (1.57-6.09) | 3.74 (1.91-7.32) | 0.0001 | 1.12 (1.02-1.22) |
| Model 2 | 1.00 (ref ) | 1.93 (0.87-4.31) | 3.21 (1.63-6.34) | 3.90 (1.99-7.65) | ＜.0001 | 1.11 (1.02-1.22) |
Analysis based on 20 replicated datasets using FCS method
Model 1 adjusted for age and sex; Model 2 further adjusted for area, body mass index, history of hypertension and diabetes, drinking status, age when completed education, occupation, perceived mental stress, walking, and sports
§P for interaction was 0.599 for Model 1 and 0.537 for Model 2, calculated by the likelihood ratio between sex and smoking status, respectively
| Current smokers | Cessation duration for former smokers, year | Never-smokers | P for trend | Per 10-year cessation | ||||
|---|---|---|---|---|---|---|---|---|
| 0-4 | 5-9 | 10-15 | ＞15 | |||||
| Pearson-years | 379,878 | 41,437 | 36,320 | 35,086 | 35,444 | 967,390 | ||
| Participants, n | 23,917 | 2,643 | 2,317 | 2,217 | 2,348 | 57,699 | ||
| Total aortic diseases | ||||||||
| Cases, n | 104 | 9 | 9 | 6 | 5 | 89 | ||
| Model 1†a | 1.00 (ref ) | 0.64 (0.32-1.26) | 0.74 (0.37-1.46) | 0.48 (0.21-1.09) | 0.29 (0.19-0.44) | 0.29 (0.19-0.44) | 0.001 | 0.63 (0.39-0.99) |
| Model 2‡b | 1.00 (ref ) | 0.54 (0.27-1.08) | 0.67 (0.34-1.33) | 0.41 (0.18-0.94) | 0.26 (0.11-0.64) | 0.24 (0.13-0.44) | 0.0005 | 0.68 (0.42-1.09) |
| Aortic dissection | ||||||||
| Cases, n | 40 | 5 | 4 | 4 | 2 | 55 | ||
| Model 1† | 1.00 (ref ) | 0.99 (0.39-2.50) | 0.92 (0.33-2.57) | 0.92 (0.33-2.59) | 0.35 (0.08-1.45) | 0.31 (0.17-0.56) | 0.189 | 0.53 (0.26-1.09) |
| Model 2‡ | 1.00 (ref ) | 0.85 (0.33-2.19) | 0.87 (0.31-2.45) | 0.83 (0.29-2.36) | 0.34 (0.08-1.43) | 0.30 (0.12-0.78) | 0.149 | 0.59 (0.28-1.23) |
| Aortic aneurysm | ||||||||
| Cases, n | 64 | 4 | 5 | 2 | 3 | 34 | ||
| Model 1† | 1.00 (ref ) | 0.43 (0.16-1.19) | 0.63 (0.25-1.56) | 0.24 (0.06-0.97) | 0.23 (0.07-0.72) | 0.27 (0.15-0.48) | 0.002 | 0.71 (0.39-1.30) |
| Model 2‡ | 1.00 (ref ) | 0.36 (0.13-0.99) | 0.53 (0.21-1.34) | 0.19 (0.05-0.78) | 0.20 (0.06-0.65) | 0.18 (0.08-0.42) | 0.001 | 0.73 (0.38-1.38) |
Analysis based on 20 replicated datasets using FCS method
Model 1 adjusted for age and sex; Model 2 further adjusted for area, body mass index, history of hypertension and diabetes, drinking status, age when completed education, occupation, perceived mental stress, walking, sports, cigarettes smoked per day, and smoking initiation age
§P for interaction was 0.599 for Model 1 and 0.486 for Model 2, calculated by the likelihood ratio between sex and smoking status, respectively
In conclusion, we found strong positive associations of daily cigarette consumption and cumulative exposure to cigarettes associated with aortic disease mortality as well as an inverse association between smoking cessation with aortic aneurysm mortality. Our finding reinforces the significance of tobacco-free and smoking-cessation campaigns on public health.
The authors would like to thank all the staff involved in this research for their valuable help participating baseline surveys and follow-up. The entire list of JACC Study collaborators was presented previously.
This study was supported by Grants-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology of Japan (MEXT, Monbu Kagaku-sho), Tokyo (grant numbers 61010076, 62010074, 63010074, 1010068, 2151065, 3151064, 4151063, 5151069, 6279102, 11181101, 17015022, 18014011, 20014026, 20390156, 26293138), JSPS KAKENHI Grant Number JP 16H06277 (CoBiA) as well as Grant–in–Aid from the Ministry of Health, Labor and Welfare, Health and LaborSciences research grants, Japan (Research on Health Services: H17–Kenkou–007; Comprehensive Research on Cardiovascular Disease and Life–Related Disease: H18–Junkankitou [Seishuu]–Ippan–012; H19–Junkankitou [Seishuu]–Ippan–012; H20–Junkankitou [Seishuu]–Ippan–013; H23–Junkankitou [Seishuu]–Ippan–005; H26-Junkankitou [Seisaku]-Ippan-001; H29–Junkankitou–Ippan–003 and 20FA1002)
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
