- Email: [email protected]
Stroke mortality associated with environmental tobacco smoke among never-smoking Japanese women: A prospective cohort study
Preventive Medicine 67 (2014) 41–45
Contents lists available at ScienceDirect
Preventive Medicine journal homepage: www.elsevier.com/locate/ypmed
Stroke mortality associated with environmental tobacco smoke among never-smoking Japanese women: A prospective cohort study Yoshikazu Nishino a,⁎, Ichiro Tsuji b, Hideo Tanaka c, Tomio Nakayama d, Haruo Nakatsuka e, Hidemi Ito c, Takaichiro Suzuki d, Kota Katanoda f, Tomotaka Sobue g, Suketami Tominaga h, for the Three-Prefecture Cohort Study Group a
Division of Cancer Epidemiology and Prevention, Miyagi Cancer Center Research Institute, 47-1 Nodayama, Medeshima-Shiode, Natori, Miyagi 981-1293, Japan Division of Epidemiology, Department of Public Health and Forensic Medicine, Tohoku University Graduate School of Medicine, 2-1 Seiryo-machi, Aoba-ku, Sendai, Miyagi 980-8575, Japan c Division of Epidemiology and Prevention, Aichi Cancer Center Research Institute, 1-1 Kanokoden, Chikusa-ku, Nagoya, Aichi 464-8681, Japan d Center for Cancer Control and Statistics, Osaka Medical Center for Cancer and Cardiovascular Diseases, 1-3-3 Nakamichi, Higashinari-ku, Osaka 537-8511, Japan e School of Nursing, Miyagi University, 1 Gakuen, Taiwa, Kurokawa-gun, Miyagi 981-3298, Japan f Surveillance Division, Center for Cancer Control and Information Services, National Cancer Center, 5-1-1 Tsukiji, Chuo-ku, Tokyo 104-0045, Japan g Environmental Medicine and Population Sciences, Department of Social and Environmental Medicine, Graduate School of Medicine, Osaka University, 2-2 Yamadaoka, Suita, Osaka 565-0871, Japan h Aichi Cancer Center, 1-1 Kanokoden, Chikusa-ku, Nagoya, Aichi 464-8681, Japan b
a r t i c l e
i n f o
Available online 28 June 2014 Keywords: Tobacco smoke pollution Stroke Mortality Cohort studies Japan
a b s t r a c t Objective. This study examined the association of exposure to environmental tobacco smoke (ETS) during adulthood with stroke and its subtypes using data from a large-scale prospective cohort study in Japan. Methods. The study population included 36,021 never-smoking Japanese women who were enrolled between 1983 and 1985 and were followed-up for 15 years. We used Cox proportional hazard regression models to estimate hazard ratios (HRs) for stroke death associated with ETS exposure at home during adulthood. Results. A total of 906 cases of stroke death were observed during 437,715 person-years of follow-up. Compared with never-smoking women without smoking family members, HRs for stroke mortality among never-smoking women living with smoking family members in all subjects, in those aged 40–79 years, and in those aged ≥80 years were 1.14 (95% confidence interval: 0.99–1.31), 1.24 (95% CI: 1.05–1.46), and 0.89 (95% CI: 0.66–1.19), respectively, after adjustment for possible confounders. The risk was most evident for subarachnoid hemorrhage [HR: 1.66 (95% CI: 1.02–2.70) in all subjects]. Conclusion. This study suggests that exposure to ETS at home during adulthood is associated with an increased risk of stroke among never-smoking Japanese women. © 2014 Elsevier Inc. All rights reserved.
The adverse health effects of environmental tobacco smoke (ETS) are conclusive in several smoking-related diseases, such as lung cancer and ischemic heart disease (US Department of Health and Human Services, 2006). Recent evidence suggests that ETS exposure increases the risk of stroke. The 2014 US Surgeon General's report concluded that the evidence is sufficient to infer a causal relationship between exposure to secondhand smoke and increased risk of stroke and that the estimated increase in risk for stroke is about 20%–30% (US Department of Health and Human Services, 2014). However, some issues remain to be addressed. First, few studies have investigated the association between ETS and stroke subtypes; thus the effects of ETS on stroke subtypes are not well defined. Second, the most previous studies were conducted in Western countries and Australasia; studies are relatively few in East Asia, where smoking prevalence among males is still high and control of ETS exposure is generally insufficient ⁎ Corresponding author. E-mail address: [email protected] (Y. Nishino).
http://dx.doi.org/10.1016/j.ypmed.2014.06.029 0091-7435/© 2014 Elsevier Inc. All rights reserved.
(World Health Organization, 2013). In such conditions, ETS may have a greater effect on the risk of stroke than is estimated in the Surgeon General's report. Therefore, elucidating the association of ETS exposure with stroke in the East Asian region has important public health consequences. We investigated the effect of ETS exposure during adulthood on mortality from stroke and its subtypes using data from a largescale population-based Japanese cohort study in which participants were enrolled between 1983 and 1985. Methods Study population and baseline survey The Three-Prefecture Cohort Study was a prospective study to evaluate the long-term effects of air pollution and lifestyle factors on mortality. Details of the study population and baseline survey methods are described elsewhere (Katanoda et al., 2011). In brief, the study areas consisted of selected locations in Miyagi, Aichi, and Osaka prefectures, Japan, and both polluted (urban) and control (rural) areas were selected in each prefecture. The study population
42
Y. Nishino et al. / Preventive Medicine 67 (2014) 41–45
included all residents living in those areas who were aged ≥ 40 years. A self-administered questionnaire was delivered to 118,820 subjects according to residential registries in cooperation with municipal governments between 1983 and 1985. In total, 100,615 individuals (84.7%) responded to the questionnaire; of these, 54,192 were women. We restricted the sample in this study to never-smoking women. Current smokers (n = 5199), former smokers (n = 1747), and individuals whose smoking status was missing (n = 9964) were thus excluded from analysis. In addition, we excluded 1033 women for whom information regarding the presence of smokers among family members living together was missing or inconsistent and 228 women with a past history of stroke. Accordingly, the final study population comprised 36,021 women. Follow-up Vital status and migration out of the study area were ascertained via the residential registry in each area. Causes of death were verified by vital statistics acquired with official permission. The outcome measures used were death from stroke (ICD-9 codes 430–438; ICD-10 codes I60–I69), subarachnoid hemorrhage (ICD-9 code 430; ICD-10 codes I60 and I69.0), intracerebral hemorrhage (ICD-9 code 431; ICD-10 codes I61 and I69.1), and cerebral infarction (ICD-9 codes 433–434; ICD-10 codes I63 and I69.3), classified according to the underlying cause of death. Cohort members were followed-up for 15 years from the baseline survey. The follow-up period was set according to the original protocol of this cohort study. Statistical analysis Person-years of follow-up for all participants were counted from the baseline survey in each study area until the date of stroke death, the date of another cause of death, the date of migration out of the study area, or the end of the 15-year follow-up period, whichever occurred first. The baseline questionnaire in this study included questions regarding the presence of smokers among family members living together and, if smokers were present, their relationship with the participant (husband, wife, father, mother, children, and other family members). In this study, the answer was used as the index of ETS exposure at home. Specifically, subjects with at least 1 smoker among family members living together at baseline were regarded as those exposed to ETS at home. We used Cox proportional hazards regression models to estimate hazard ratios (HRs) and 95% confidence intervals (CIs) for stroke death associated with ETS exposure at home among never-smoking women. We considered the following variables as potential confounders: age (continuous); region of residence (6 locations: urban and rural areas of Miyagi, Aichi, and Osaka); health insurance as an indicator of the socioeconomic status (national health insurance, government or union-managed health insurance, mutual-aid association health insurance, and others); history of hypertension (yes or no); history of diabetes (yes or no); body mass index (BMI) (b 19.0, 19.0–21.9, 22.0–24.9, 25.0–29.9, ≥30.0 kg/m2); alcohol consumption (never, former, current occasional, current almost every day); green and yellow vegetable consumption; non-green and non-yellow vegetable consumption; fruit consumption; miso soup consumption (≤1–2 times/month, 1–2 times/week, 3–4 times/week, almost every day); and pickled vegetable consumption (scarcely any, 1–2 times/month, 1–2 times/week, 3–4 times/week, almost every day). In addition, the smoking status of the father and mother during the participant's childhood (smoker, non-smoker, or unknown) was taken into consideration in order to independently evaluate the effect of ETS during adulthood. Missing values were treated as an additional variable category. Age-adjusted HRs and the HRs after adjustment for all categories listed above were calculated. We also estimated HRs for stroke death stratified by age groups and according to the source of ETS exposure at home (husband and other family members). In addition, we conducted separate analyses after classifying the cause of stroke death according to stroke subtypes: subarachnoid hemorrhage, intracerebral hemorrhage, and cerebral infarction. All analyses were performed using SAS version 8.2 statistical software (SAS Institute, Cary, NC).
Results During 437,715 person-years of follow-up involving 36,021 subjects, we documented 906 cases of stroke death, including 87 cases of
subarachnoid hemorrhage, 147 cases of intracerebral hemorrhage, and 467 cases of cerebral infarction. Table 1 presents the characteristics of the study subjects at baseline according to their families' smoking status. Subjects with smoking family members constituted a high percentage of those who lived in rural areas in Miyagi and Osaka and were more likely to be obese compared with subjects without smoking family members. In addition, subjects with smoking family members had a lower frequency of consumption of vegetables and fruits and a higher frequency of consumption of salty food such as miso soup and pickled vegetables. Table 1 Characteristics of study subjects according to familial smoking status at baseline among Three-Prefecture Cohort Study participants followed-up for 15 years.
No. at risk Mean age, years (SD) Region of residence (%) Miyagi, urban Miyagi, rural Aichi, urban Aichi, rural Osaka, urban Osaka, rural Health insurance type (%) National health insurance Government/union-managed health insurance Mutual aid associations health insurance Others History of hypertension (%) History of diabetes (%) Body mass index (BMI) (%) b19.0 kg/m2 19.0–21.9 kg/m2 22.0–24.9 kg/m2 25.0–29.9 kg/m2 ≥30.0 kg/m2 Alcohol drinking (%) Never Former Current occasional Current almost everyday Green and yellow vegetable consumption (%) ≤1–2 times/month 1–2 times/week 3–4 times/week Almost every day Non-green and non-yellow vegetable consumption (%) ≤1–2 times/month 1–2 times/week 3–4 times/week Almost every day Fruit consumption (%) ≤1–2 times/month 1–2 times/week 3–4 times/week Almost every day Miso soup consumption (%) ≤1–2 times/month 1–2 times/week 3–4 times/week Almost every day Pickled vegetable consumption (%) Scarcely any 1–2 times/month 1–2 times/week 3–4 times/week Almost every day Smoking status of father during childhood (%) Smoking Non-smoking Unknown Smoking status of mother during childhood (%) Smoking Non-smoking Unknown
Smoker (−)
Smoker (+)
13,834 56.7 (11.9)
22,187 55.9 (11.3)
17.8 7.8 29.3 10.7 18.3 16.1
14.8 14.5 19.4 11.0 19.2 21.1
44.8 42.9 10.5 1.7 21.3 3.3
47.3 42.8 8.6 1.4 22.4 3.7
13.3 37.0 32.7 15.4 1.5
10.3 33.1 35.4 19.2 2.1
63.7 1.5 29.7 5.1
65.0 1.3 29.1 4.6
2.5 14.1 28.7 54.7
3.9 17.2 31.2 47.7
1.3 8.6 24.7 65.5
1.7 9.7 26.1 62.5
3.0 10.0 19.7 67.4
4.3 12.0 22.5 61.1
7.9 18.6 20.9 52.6
6.5 16.1 20.1 57.2
5.4 6.1 13.2 16.3 59.1
3.5 4.1 10.3 13.5 68.8
66.2 27.8 6.0
68.7 24.1 7.3
9.1 87.8 3.1
10.2 85.6 4.2
Y. Nishino et al. / Preventive Medicine 67 (2014) 41–45
Table 2 presents the association between ETS exposure at home and stroke death. After multivariate adjustment, we found a 14% increased risk of stroke death in subjects with ETS exposure at home. The results stratified by age groups showed that the risk of stroke death increased significantly in the group aged 40–79 years (multivariate-adjusted HR: 1.24; 95% CI: 1.05–1.46). No increase in stroke mortality risk was observed in the group aged ≥80 years. Table 3 shows the results of analysis stratified by stroke subtypes. After multivariate adjustment, we found significantly increased mortality on account of subarachnoid hemorrhage in persons exposed to ETS at home (HR: 1.66; 95%: CI 1.02–2.70). We did not find a statistically significant increase in HRs for intracerebral hemorrhage or cerebral infarction. Even in the group aged 40–79 years, HRs for only subarachnoid hemorrhage increased to a statistically significant extent. HRs for stroke mortality according to the smoking status of the husband and other family members are presented in Table 4. The risks of death from total stroke or its subtypes tended to be highest for subjects with both a smoking husband and another smoking family member in the household. Discussion This study shows that the risk of mortality from total stroke associated with ETS exposure at home during adulthood increased significantly in the group of never-smoking Japanese women aged 40–79 years at baseline and did not elevate in the group aged ≥80 years. Analysis by stroke subtypes showed that risk elevation was most apparent for subarachnoid hemorrhage. Previous studies showed that the magnitude of the effect of active smoking on stroke is generally larger in younger individuals (Honjo et al., 2010; Shinton and Beevers, 1989). Similar results were obtained in our study on never-smoking Japanese women. In addition, selection bias caused by premature deaths among participants exposed to ETS may also have contributed to the lack of effect in the older age group. Furthermore, this study evaluated ETS exposure at home on the basis of the smoking status of family members living together at baseline but did not consider earlier exposure before the study began. This misclassification could have occurred more frequently in the older age group, resulting in the difference in risk among age groups. The HR of stroke in subjects aged 40–79 years with ETS exposure at home in this study is comparable with the result of a meta-analysis by Oono et al. that included 20 studies published between 1984 and 2010
Table 2 Hazard ratios of stroke for never-smoking women according to familial smoking status among Three-Prefecture Cohort Study participants followed-up for 15 years.
All
40–79 years
80 years+
a
No. at risk Person-years No. of cases HR1 (95% CI)a HR2 (95% CI)b No. at risk Person-years No. of cases HR1 (95% CI)a HR2 (95% CI)b No. at risk Person-years No. of cases HR1 (95% CI)a HR2 (95% CI)b
Smoker (−)
Smoker (+)
13,834 164,354 322 1.00 1.00 13,317 161,088 236 1.00 1.00 517 3266 86 1.00 1.00
22,187 273,361 584 1.27 1.14 21,366 268,294 444 1.38 1.24 821 5067 140 1.04 0.89
(1.11–1.45) (0.99–1.31)
(1.18–1.61) (1.05–1.46)
(0.80–1.36) (0.66–1.19)
HR1: adjusted for age. HR2: adjusted for age, region of residence, health insurance type, history of hypertension, history of diabetes, body mass index (BMI), alcohol drinking, green and yellow vegetable consumption, non-green and non-yellow vegetable consumption, fruit consumption, miso soup consumption, pickled vegetable consumption, smoking status of father during childhood, and smoking status of mother during childhood. b
43
(pooled estimate of risk: 1.25) (Oono et al., 2011). In East Asia, 4 studies in China (1 cohort, 1 case–control, and 2 cross-sectional) showed a positive association of ETS with mortality or prevalence of total stroke in women (He et al., 2008; McGhee et al., 2005; Wen et al., 2006; Zhang et al., 2005); however, a large cohort study in Japan by Hirayama did not show a significant association with cerebrovascular disease (Hirayama, 1987). To the best of our knowledge, this is the second prospective cohort study in Japan and the third in Asia investigating the association of ETS exposure with total stroke. Active smoking is a risk factor for stroke, regardless of its subtypes (US Department of Health and Human Services, 2004). In this study, the effects of ETS exposure by stroke subtype were most evident for subarachnoid hemorrhage, followed by intracerebral hemorrhage; however, the effects for cerebral infarction remain unclear. A study investigating the effect of active smoking on mortality from cardiovascular disease including stroke pooled data from 3 large prospective studies in Japan and showed that the HR among female current smokers was highest for subarachnoid hemorrhage, intracerebral hemorrhage, and cerebral infarction, in that order (Honjo et al., 2010). Our results for ETS were consistent with those of this pooled analysis of active smokers. Few studies have presented results related to stroke subtype (Anderson et al., 2004; Donnan et al., 1989; He et al., 2008; Hirayama, 1987; Iribarren et al., 2004; Qureshi et al., 2005; Yamada et al., 2003; You et al., 1999). For subarachnoid hemorrhage, 2 cohort studies in Japan (Hirayama, 1987; Yamada et al., 2003) and 1 case–control study in Australasia (Anderson et al., 2004) reported no significant effect of ETS. For ischemic stroke, 4 studies showed an increased risk associated with ETS exposure (Donnan et al., 1989; He et al., 2008; Iribarren et al., 2004; You et al., 1999), and 2 studies of them reached statistical significance (He et al., 2008; Iribarren et al., 2004). One cohort study reported no association (Qureshi et al., 2005). No significantly increased risk of hemorrhagic stroke was observed in a Chinese cross-sectional study (He et al., 2008). From the data in Table 4, the effect of husband's smoking on stroke and its subtypes appeared to be lower compared with the effect of smoking by family members other than the husband. Of the study subjects with family members other than the husband who smoked, N 80% had children who smoked. Studies investigating the effect of ETS from family members other than the husband on stroke are scarce, and only 2 studies have demonstrated no association with ETS from parents (Donnan et al., 1989; You et al., 1999). A previous case–control study recruited subjects from 1986 to 1988 in Osaka, Japan, one of the study areas included in this cohort, and showed a significantly increased risk of lung cancer among those whose family members other than the husband smoked and no significant risk elevation for subjects with a smoking husband (Sobue, 1990). Although husband's smoking status has been widely used as a marker of ETS exposure at home, this study indicates that other family members' smoking also may have been an important source of ETS exposure at home, at least in Japan, at that time. In this study, 906 stroke deaths occurred during the follow-up period. This number of events permits a detailed analysis by stroke subtype. In addition, smoking women have a tendency to marry smokers. Thus, misclassification of current smokers as never smokers could have resulted in overestimation of the risk for ETS. However, in this cohort, the prevalence of female smokers was relatively low so that the impact of this misclassification is limited. A study in Japan conducted between 1990 and 1997 investigated the validity of self-reported smoking by the use of the urinary cotinine test as a gold standard (Tsutsumi et al., 2002). In that study, only 1.4% of self-declared never-smoking women were classified as smokers according to urinary cotinine levels in the group among which smoking prevalence based on a self-administrated questionnaire was 8.8%, which is a slight lower than the smoking prevalence in this cohort. Accordingly, this Japanese cohort is considered to have an advantage in evaluating an unbiased health effect of ETS. A considerable number of people with no smoking status or missing data were excluded from this study. Compared with included subjects,
44
Y. Nishino et al. / Preventive Medicine 67 (2014) 41–45
Table 3 Hazard ratios of stroke subtypes for never-smoking women according to familial smoking status among Three-Prefecture Cohort Study participants followed-up for 15 years. Cause of death
ICD-9
ICD-10
Smoker (−)
Smoker (+)
All
Subarachnoid hemorrhage
Intracerebral hemorrhage
Cerebral infarction
430
431
433–434
I60, I69.0
I61, I69.1
I63, I69.3
No. at risk Person-years No. of cases HR1 (95% CI)a HR2 (95% CI)b No. of cases HR1 (95% CI)a HR2 (95% CI)b No. of cases HR1 (95% CI)a HR2 (95% CI)b
13,834 164,354 24 1.00 1.00 45 1.00 1.00 183 1.00 1.00
22,187 273,361 63 1.72 1.66 102 1.55 1.35 284 1.09 0.95
No. at risk Person-years No. of cases HR1 (95% CI)a HR2 (95% CI)b No. of cases HR1 (95% CI)a HR2 (95% CI)b No. of cases HR1 (95% CI)a HR2 (95% CI)b
13,317 161,088 22 1.00 1.00 37 1.00 1.00 129 1.00 1.00
21,366 268,294 59 1.78 1.66 91 1.74 1.46 211 1.24 1.09
(1.08–2.76) (1.02–2.70) (1.09–2.20) (0.94–1.94) (0.91–1.32) (0.78–1.15)
40–79 years
Subarachnoid hemorrhage
Intracerebral hemorrhage
Cerebral infarction
430
431
433–434
I60, I69.0
I61, I69.1
I63, I69.3
(1.09–2.91) (1.00–2.75) (1.19–2.55) (0.99–2.17) (0.99–1.54) (0.87–1.37)
a
HR1: adjusted for age. HR2: adjusted for age, region of residence, health insurance type, history of hypertension, history of diabetes, body mass index (BMI), alcohol drinking, green and yellow vegetable consumption, non-green and non-yellow vegetable consumption, fruit consumption, miso soup consumption, pickled vegetable consumption, smoking status of father during childhood, and smoking status of mother during childhood. b
Table 4 Hazard ratios of stroke for never-smoking women according to smoking status of husband and other family members among Three-Prefecture Cohort Study participants followed-up for 15 years. Smoker (−)
Smoker (+) Husband (−)
Husband (+)
Husband (+)
Others (+)
Others (−)
Others (+)
10,042 125,325 93 0.95 0.93 16 1.31 1.32 18 0.99 0.94 39 0.79 0.78
4728 61,815 82 1.40 1.20 15 2.20 2.04 20 1.91 1.56 33 1.11 0.92
All
Stroke
Subarachnoid hemorrhage
Intracerebral hemorrhage
Cerebral infarction
No. at risk Person-years No. of cases HR1 (95% CI)a HR2 (95% CI)b No. of cases HR1 (95% CI)a HR2 (95% CI)b No. of cases HR1 (95% CI)a HR2 (95% CI)b No. of cases HR1 (95% CI)a HR2 (95% CI)b
13,834 164,354 322 1.00 1.00 24 1.00 1.00 45 1.00 1.00 183 1.00 1.00
7280 84,847 403 1.35 1.19 32 1.85 1.77 60 1.62 1.39 210 1.18 1.01
No. at risk Person-years No. of cases HR1 (95% CI)a HR2 (95% CI)b No. of cases HR1 (95% CI)a HR2 (95% CI)b No. of cases HR1 (95% CI)a HR2 (95% CI)b No. of cases HR1 (95% CI)a HR2 (95% CI)b
13,317 161,088 236 1.00 1.00 22 1.00 1.00 37 1.00 1.00 129 1.00 1.00
6534 80,244 271 1.48 1.30 28 1.93 1.77 52 1.92 1.57 139 1.32 1.14
(1.16–1.56) (1.02–1.39) (1.08–3.16) (1.02–3.08) (1.10–2.40) (0.93–2.08) (0.96–1.44) (0.81–1.24)
(0.75–1.20) (0.73–1.18) (0.68–2.50) (0.69–2.55) (0.57–1.74) (0.54–1.66) (0.55–1.13) (0.55–1.11)
(1.09–1.80) (0.94–1.55) (1.14–4.24) (1.04–4.02) (1.12–3.28) (0.90–2.71) (0.76–1.62) (0.62–1.34)
40–79 years
Stroke
Subarachnoid hemorrhage
Intracerebral hemorrhage
Cerebral infarction
a
(1.24–1.76) (1.08–1.56) (1.10–3.38) (0.99–3.16) (1.26–2.94) (1.01–2.43) (1.04–1.68) (0.89–1.46)
10,010 125,044 89 1.03 1.02 16 1.35 1.33 17 1.06 1.00 38 0.94 0.95
(0.80–1.32) (0.80–1.32) (0.70–2.60) (0.69–2.59) (0.59–1.90) (0.55–1.80) (0.65–1.36) (0.66–1.38)
4699 61,611 80 1.60 1.38 15 2.28 2.03 19 2.07 1.64 33 1.40 1.17
(1.23–2.07) (1.06–1.80) (1.17–4.43) (1.02–4.04) (1.18–3.64) (0.92–2.92) (0.95–2.06) (0.78–1.74)
HR1: adjusted for age. HR2: adjusted for age, region of residence, health insurance type, history of hypertension, history of diabetes, body mass index (BMI), alcohol drinking, green and yellow vegetable consumption, non-green and non-yellow vegetable consumption, fruit consumption, miso soup consumption, pickled vegetable consumption, smoking status of father during childhood, and smoking status of mother during childhood. b
Y. Nishino et al. / Preventive Medicine 67 (2014) 41–45
people excluded because of no smoking status included a higher proportion of those who lived in Miyagi and rural areas of Aichi and a higher proportion of those with smoking family members, possibly because of the inclusion of some smokers. Further, compared with included subjects, excluded people for whom information regarding ETS at home was missing included a higher percentage of elderly people. The effects of these exclusions on the results remain unclear. In this study, ETS exposure at home was evaluated on the basis of the presence of smokers among family members living together at the time of the baseline survey. A study by Ozasa et al. in 1997 presented that approximately two-thirds of non-smoking females reported being exposed to ETS at home in Japan at that time (Ozasa et al., 1997). Although the effect of ETS at home, which is the predominant source of exposure for the subjects, was examined in our study, exposure in the workplace and public settings, which are other important exposure settings, was not considered. The baseline survey in this study was conducted between 1983 and 1985. Male smoking prevalence in Japan at that time exceeded 60% (Japan Health Promotion and Fitness Foundation, 2013), and strategies for preventing ETS exposure in the workplace and public settings were quite incomplete, because the adverse health effects of ETS were poorly understood then. Accordingly, subjects in this cohort were considered to be more extensively exposed to ETS in the workplace and public settings than at present; thus, this misclassification of ETS exposure may dilute the association between ETS exposure and stroke in this study. Our study did not make adjustments for several established risk factors for stroke, such as serum cholesterol, which may have affected the results. In addition, socioeconomic factors other than insurance, such as marital status, education level, income, and differences in the quality of health care among study areas, were not considered in this study. These possible confounding factors could have an effect on the relationship between ETS exposure and stroke mortality reported here. Conclusions In our large-scale cohort study, we found an increase in stroke mortality among never-smoking Japanese women who lived with smoking family members during adulthood. These results did not apply to very elderly people, and the risk was most evident for subarachnoid hemorrhage following subgroup analysis by stroke subtype. Our results provide a new perspective concerning the adverse health effect of ETS exposure and encourage further efforts to control ETS exposure in Japan. Conflict of interest statement The authors declare that there are no conflicts of interest.
Acknowledgments This study was funded by the Japanese Ministry of the Environment (former Environment Agency). We sincerely thank the staff within each study area for collecting and processing data. We also express our gratitude to all participants in the study.
45
References Anderson, C.S., Feigin, V., Bennett, D., et al., 2004. Active and passive smoking and the risk of subarachnoid hemorrhage: an international population-based case–control study. Stroke 35, 633–637. Donnan, G.A., McNeil, J.J., Adena, M.A., Doyle, A.E., O'Malley, H.M., Neill, G.C., 1989. Smoking as a risk factor for cerebral ischaemia. Lancet 2 (8664), 643–647. He, Y., Lam, T.H., Jiang, B., et al., 2008. Passive smoking and risk of peripheral arterial disease and ischemic stroke in Chinese women who never smoked. Circulation 118, 1535–1540. Hirayama, T., 1987. Passive smoking and cancer: an epidemiological review. Gann Monogr. Cancer Res. 33, 127–135. Honjo, K., Iso, H., Tsugane, S., et al., 2010. The effects of smoking and smoking cessation on mortality from cardiovascular disease among Japanese: pooled analysis of three large-scale cohort studies in Japan. Tob. Control. 19, 50–57. Iribarren, C., Darbinian, J., Klatsky, A.L., Friedman, G.D., 2004. Cohort study of exposure to environmental tobacco smoke and risk of first ischemic stroke and transient ischemic attack. Neuroepidemiology 23, 38–44. Japan Health Promotion and Fitness Foundation, 2013. The Latest Information on TobaccoAvailable at http://www.health-net.or.jp/tobacco/product/pd090000.html (Accessed 22 August, 2013). Katanoda, K., Sobue, T., Satoh, H., et al., 2011. An association between long-term exposure to ambient air pollution and mortality from lung cancer and respiratory diseases in Japan. J. Epidemiol. 21, 132–143. McGhee, S.M., Ho, S.Y., Schooling, M., et al., 2005. Mortality associated with passive smoking in Hong Kong. BMJ 330, 287–288. Oono, I.P., Mackay, D.F., Pell, J.P., 2011. Meta-analysis of the association between secondhand smoke exposure and stroke. J. Public Health (Oxf.) 33, 496–502. Ozasa, K., Higashi, A., Yamasaki, M., Hayashi, K., Watanabe, Y., 1997. Validity of selfreported passive smoking evaluated by comparison with smokers in the same household. J. Epidemiol. 7, 205–209. Qureshi, A.I., Suri, M.F., Kirmani, J.F., Divani, A.A., 2005. Cigarette smoking among spouses: another risk factor for stroke in women. Stroke 36, e74–e76. Shinton, R., Beevers, G., 1989. Meta-analysis of relation between cigarette smoking and stroke. BMJ 298, 789–794. Sobue, T., 1990. Association of indoor air pollution and lifestyle with lung cancer in Osaka, Japan. Int. J. Epidemiol. 19, S62–S66. Tsutsumi, A., Kagawa, J., Yamano, Y., Nakadate, T., Shimizu, S., 2002. Relation between cotinine in the urine and indices based on self-declared smoking habits. Environ. Health Prev. Med. 6, 240–247. US Department of Health and Human Services, 2004. The Health Consequences of Smoking: A Report of the Surgeon General. US Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health, Atlanta. US Department of Health and Human Services, 2006. The Health Consequences of Involuntary Exposure to Tobacco Smoke: A Report of the Surgeon General. US Department of Health and Human Services, Centers for Disease Control and Prevention, Coordinating Center for Health Promotion, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health, Atlanta. US Department of Health and Human Services, 2014. The Health Consequences of Smoking—50 Years of Progress: A Report of the Surgeon General. US Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health, Atlanta. Wen, W., Shu, X.O., Gao, Y.T., et al., 2006. Environmental tobacco smoke and mortality in Chinese women who have never smoked: prospective cohort study. BMJ 333, 376. World Health Organization, 2013. WHO Report on the Global Tobacco Epidemic, 2013: Enforcing Bans on Tobacco Advertising, Promotion and Sponsorship. World Health Organization, Geneva. Yamada, S., Koizumi, A., Iso, H., et al., 2003. Risk factors for fatal subarachnoid hemorrhage: the Japan Collaborative Cohort Study. Stroke 34, 2781–2787. You, R.X., Thrift, A.G., McNeil, J.J., Davis, S.M., Donnan, G.A., Melbourne Stroke Risk Factor Study (MERFS) Group, 1999. Ischemic stroke risk and passive exposure to spouses' cigarette smoking. Am. J. Public Health 89, 572–575. Zhang, X., Shu, X.O., Yang, G., et al., 2005. Association of passive smoking by husbands with prevalence of stroke among Chinese women nonsmokers. Am. J. Epidemiol. 161, 213–218.
Contents lists available at ScienceDirect
Preventive Medicine journal homepage: www.elsevier.com/locate/ypmed
Stroke mortality associated with environmental tobacco smoke among never-smoking Japanese women: A prospective cohort study Yoshikazu Nishino a,⁎, Ichiro Tsuji b, Hideo Tanaka c, Tomio Nakayama d, Haruo Nakatsuka e, Hidemi Ito c, Takaichiro Suzuki d, Kota Katanoda f, Tomotaka Sobue g, Suketami Tominaga h, for the Three-Prefecture Cohort Study Group a
Division of Cancer Epidemiology and Prevention, Miyagi Cancer Center Research Institute, 47-1 Nodayama, Medeshima-Shiode, Natori, Miyagi 981-1293, Japan Division of Epidemiology, Department of Public Health and Forensic Medicine, Tohoku University Graduate School of Medicine, 2-1 Seiryo-machi, Aoba-ku, Sendai, Miyagi 980-8575, Japan c Division of Epidemiology and Prevention, Aichi Cancer Center Research Institute, 1-1 Kanokoden, Chikusa-ku, Nagoya, Aichi 464-8681, Japan d Center for Cancer Control and Statistics, Osaka Medical Center for Cancer and Cardiovascular Diseases, 1-3-3 Nakamichi, Higashinari-ku, Osaka 537-8511, Japan e School of Nursing, Miyagi University, 1 Gakuen, Taiwa, Kurokawa-gun, Miyagi 981-3298, Japan f Surveillance Division, Center for Cancer Control and Information Services, National Cancer Center, 5-1-1 Tsukiji, Chuo-ku, Tokyo 104-0045, Japan g Environmental Medicine and Population Sciences, Department of Social and Environmental Medicine, Graduate School of Medicine, Osaka University, 2-2 Yamadaoka, Suita, Osaka 565-0871, Japan h Aichi Cancer Center, 1-1 Kanokoden, Chikusa-ku, Nagoya, Aichi 464-8681, Japan b
a r t i c l e
i n f o
Available online 28 June 2014 Keywords: Tobacco smoke pollution Stroke Mortality Cohort studies Japan
a b s t r a c t Objective. This study examined the association of exposure to environmental tobacco smoke (ETS) during adulthood with stroke and its subtypes using data from a large-scale prospective cohort study in Japan. Methods. The study population included 36,021 never-smoking Japanese women who were enrolled between 1983 and 1985 and were followed-up for 15 years. We used Cox proportional hazard regression models to estimate hazard ratios (HRs) for stroke death associated with ETS exposure at home during adulthood. Results. A total of 906 cases of stroke death were observed during 437,715 person-years of follow-up. Compared with never-smoking women without smoking family members, HRs for stroke mortality among never-smoking women living with smoking family members in all subjects, in those aged 40–79 years, and in those aged ≥80 years were 1.14 (95% confidence interval: 0.99–1.31), 1.24 (95% CI: 1.05–1.46), and 0.89 (95% CI: 0.66–1.19), respectively, after adjustment for possible confounders. The risk was most evident for subarachnoid hemorrhage [HR: 1.66 (95% CI: 1.02–2.70) in all subjects]. Conclusion. This study suggests that exposure to ETS at home during adulthood is associated with an increased risk of stroke among never-smoking Japanese women. © 2014 Elsevier Inc. All rights reserved.
The adverse health effects of environmental tobacco smoke (ETS) are conclusive in several smoking-related diseases, such as lung cancer and ischemic heart disease (US Department of Health and Human Services, 2006). Recent evidence suggests that ETS exposure increases the risk of stroke. The 2014 US Surgeon General's report concluded that the evidence is sufficient to infer a causal relationship between exposure to secondhand smoke and increased risk of stroke and that the estimated increase in risk for stroke is about 20%–30% (US Department of Health and Human Services, 2014). However, some issues remain to be addressed. First, few studies have investigated the association between ETS and stroke subtypes; thus the effects of ETS on stroke subtypes are not well defined. Second, the most previous studies were conducted in Western countries and Australasia; studies are relatively few in East Asia, where smoking prevalence among males is still high and control of ETS exposure is generally insufficient ⁎ Corresponding author. E-mail address: [email protected] (Y. Nishino).
http://dx.doi.org/10.1016/j.ypmed.2014.06.029 0091-7435/© 2014 Elsevier Inc. All rights reserved.
(World Health Organization, 2013). In such conditions, ETS may have a greater effect on the risk of stroke than is estimated in the Surgeon General's report. Therefore, elucidating the association of ETS exposure with stroke in the East Asian region has important public health consequences. We investigated the effect of ETS exposure during adulthood on mortality from stroke and its subtypes using data from a largescale population-based Japanese cohort study in which participants were enrolled between 1983 and 1985. Methods Study population and baseline survey The Three-Prefecture Cohort Study was a prospective study to evaluate the long-term effects of air pollution and lifestyle factors on mortality. Details of the study population and baseline survey methods are described elsewhere (Katanoda et al., 2011). In brief, the study areas consisted of selected locations in Miyagi, Aichi, and Osaka prefectures, Japan, and both polluted (urban) and control (rural) areas were selected in each prefecture. The study population
42
Y. Nishino et al. / Preventive Medicine 67 (2014) 41–45
included all residents living in those areas who were aged ≥ 40 years. A self-administered questionnaire was delivered to 118,820 subjects according to residential registries in cooperation with municipal governments between 1983 and 1985. In total, 100,615 individuals (84.7%) responded to the questionnaire; of these, 54,192 were women. We restricted the sample in this study to never-smoking women. Current smokers (n = 5199), former smokers (n = 1747), and individuals whose smoking status was missing (n = 9964) were thus excluded from analysis. In addition, we excluded 1033 women for whom information regarding the presence of smokers among family members living together was missing or inconsistent and 228 women with a past history of stroke. Accordingly, the final study population comprised 36,021 women. Follow-up Vital status and migration out of the study area were ascertained via the residential registry in each area. Causes of death were verified by vital statistics acquired with official permission. The outcome measures used were death from stroke (ICD-9 codes 430–438; ICD-10 codes I60–I69), subarachnoid hemorrhage (ICD-9 code 430; ICD-10 codes I60 and I69.0), intracerebral hemorrhage (ICD-9 code 431; ICD-10 codes I61 and I69.1), and cerebral infarction (ICD-9 codes 433–434; ICD-10 codes I63 and I69.3), classified according to the underlying cause of death. Cohort members were followed-up for 15 years from the baseline survey. The follow-up period was set according to the original protocol of this cohort study. Statistical analysis Person-years of follow-up for all participants were counted from the baseline survey in each study area until the date of stroke death, the date of another cause of death, the date of migration out of the study area, or the end of the 15-year follow-up period, whichever occurred first. The baseline questionnaire in this study included questions regarding the presence of smokers among family members living together and, if smokers were present, their relationship with the participant (husband, wife, father, mother, children, and other family members). In this study, the answer was used as the index of ETS exposure at home. Specifically, subjects with at least 1 smoker among family members living together at baseline were regarded as those exposed to ETS at home. We used Cox proportional hazards regression models to estimate hazard ratios (HRs) and 95% confidence intervals (CIs) for stroke death associated with ETS exposure at home among never-smoking women. We considered the following variables as potential confounders: age (continuous); region of residence (6 locations: urban and rural areas of Miyagi, Aichi, and Osaka); health insurance as an indicator of the socioeconomic status (national health insurance, government or union-managed health insurance, mutual-aid association health insurance, and others); history of hypertension (yes or no); history of diabetes (yes or no); body mass index (BMI) (b 19.0, 19.0–21.9, 22.0–24.9, 25.0–29.9, ≥30.0 kg/m2); alcohol consumption (never, former, current occasional, current almost every day); green and yellow vegetable consumption; non-green and non-yellow vegetable consumption; fruit consumption; miso soup consumption (≤1–2 times/month, 1–2 times/week, 3–4 times/week, almost every day); and pickled vegetable consumption (scarcely any, 1–2 times/month, 1–2 times/week, 3–4 times/week, almost every day). In addition, the smoking status of the father and mother during the participant's childhood (smoker, non-smoker, or unknown) was taken into consideration in order to independently evaluate the effect of ETS during adulthood. Missing values were treated as an additional variable category. Age-adjusted HRs and the HRs after adjustment for all categories listed above were calculated. We also estimated HRs for stroke death stratified by age groups and according to the source of ETS exposure at home (husband and other family members). In addition, we conducted separate analyses after classifying the cause of stroke death according to stroke subtypes: subarachnoid hemorrhage, intracerebral hemorrhage, and cerebral infarction. All analyses were performed using SAS version 8.2 statistical software (SAS Institute, Cary, NC).
Results During 437,715 person-years of follow-up involving 36,021 subjects, we documented 906 cases of stroke death, including 87 cases of
subarachnoid hemorrhage, 147 cases of intracerebral hemorrhage, and 467 cases of cerebral infarction. Table 1 presents the characteristics of the study subjects at baseline according to their families' smoking status. Subjects with smoking family members constituted a high percentage of those who lived in rural areas in Miyagi and Osaka and were more likely to be obese compared with subjects without smoking family members. In addition, subjects with smoking family members had a lower frequency of consumption of vegetables and fruits and a higher frequency of consumption of salty food such as miso soup and pickled vegetables. Table 1 Characteristics of study subjects according to familial smoking status at baseline among Three-Prefecture Cohort Study participants followed-up for 15 years.
No. at risk Mean age, years (SD) Region of residence (%) Miyagi, urban Miyagi, rural Aichi, urban Aichi, rural Osaka, urban Osaka, rural Health insurance type (%) National health insurance Government/union-managed health insurance Mutual aid associations health insurance Others History of hypertension (%) History of diabetes (%) Body mass index (BMI) (%) b19.0 kg/m2 19.0–21.9 kg/m2 22.0–24.9 kg/m2 25.0–29.9 kg/m2 ≥30.0 kg/m2 Alcohol drinking (%) Never Former Current occasional Current almost everyday Green and yellow vegetable consumption (%) ≤1–2 times/month 1–2 times/week 3–4 times/week Almost every day Non-green and non-yellow vegetable consumption (%) ≤1–2 times/month 1–2 times/week 3–4 times/week Almost every day Fruit consumption (%) ≤1–2 times/month 1–2 times/week 3–4 times/week Almost every day Miso soup consumption (%) ≤1–2 times/month 1–2 times/week 3–4 times/week Almost every day Pickled vegetable consumption (%) Scarcely any 1–2 times/month 1–2 times/week 3–4 times/week Almost every day Smoking status of father during childhood (%) Smoking Non-smoking Unknown Smoking status of mother during childhood (%) Smoking Non-smoking Unknown
Smoker (−)
Smoker (+)
13,834 56.7 (11.9)
22,187 55.9 (11.3)
17.8 7.8 29.3 10.7 18.3 16.1
14.8 14.5 19.4 11.0 19.2 21.1
44.8 42.9 10.5 1.7 21.3 3.3
47.3 42.8 8.6 1.4 22.4 3.7
13.3 37.0 32.7 15.4 1.5
10.3 33.1 35.4 19.2 2.1
63.7 1.5 29.7 5.1
65.0 1.3 29.1 4.6
2.5 14.1 28.7 54.7
3.9 17.2 31.2 47.7
1.3 8.6 24.7 65.5
1.7 9.7 26.1 62.5
3.0 10.0 19.7 67.4
4.3 12.0 22.5 61.1
7.9 18.6 20.9 52.6
6.5 16.1 20.1 57.2
5.4 6.1 13.2 16.3 59.1
3.5 4.1 10.3 13.5 68.8
66.2 27.8 6.0
68.7 24.1 7.3
9.1 87.8 3.1
10.2 85.6 4.2
Y. Nishino et al. / Preventive Medicine 67 (2014) 41–45
Table 2 presents the association between ETS exposure at home and stroke death. After multivariate adjustment, we found a 14% increased risk of stroke death in subjects with ETS exposure at home. The results stratified by age groups showed that the risk of stroke death increased significantly in the group aged 40–79 years (multivariate-adjusted HR: 1.24; 95% CI: 1.05–1.46). No increase in stroke mortality risk was observed in the group aged ≥80 years. Table 3 shows the results of analysis stratified by stroke subtypes. After multivariate adjustment, we found significantly increased mortality on account of subarachnoid hemorrhage in persons exposed to ETS at home (HR: 1.66; 95%: CI 1.02–2.70). We did not find a statistically significant increase in HRs for intracerebral hemorrhage or cerebral infarction. Even in the group aged 40–79 years, HRs for only subarachnoid hemorrhage increased to a statistically significant extent. HRs for stroke mortality according to the smoking status of the husband and other family members are presented in Table 4. The risks of death from total stroke or its subtypes tended to be highest for subjects with both a smoking husband and another smoking family member in the household. Discussion This study shows that the risk of mortality from total stroke associated with ETS exposure at home during adulthood increased significantly in the group of never-smoking Japanese women aged 40–79 years at baseline and did not elevate in the group aged ≥80 years. Analysis by stroke subtypes showed that risk elevation was most apparent for subarachnoid hemorrhage. Previous studies showed that the magnitude of the effect of active smoking on stroke is generally larger in younger individuals (Honjo et al., 2010; Shinton and Beevers, 1989). Similar results were obtained in our study on never-smoking Japanese women. In addition, selection bias caused by premature deaths among participants exposed to ETS may also have contributed to the lack of effect in the older age group. Furthermore, this study evaluated ETS exposure at home on the basis of the smoking status of family members living together at baseline but did not consider earlier exposure before the study began. This misclassification could have occurred more frequently in the older age group, resulting in the difference in risk among age groups. The HR of stroke in subjects aged 40–79 years with ETS exposure at home in this study is comparable with the result of a meta-analysis by Oono et al. that included 20 studies published between 1984 and 2010
Table 2 Hazard ratios of stroke for never-smoking women according to familial smoking status among Three-Prefecture Cohort Study participants followed-up for 15 years.
All
40–79 years
80 years+
a
No. at risk Person-years No. of cases HR1 (95% CI)a HR2 (95% CI)b No. at risk Person-years No. of cases HR1 (95% CI)a HR2 (95% CI)b No. at risk Person-years No. of cases HR1 (95% CI)a HR2 (95% CI)b
Smoker (−)
Smoker (+)
13,834 164,354 322 1.00 1.00 13,317 161,088 236 1.00 1.00 517 3266 86 1.00 1.00
22,187 273,361 584 1.27 1.14 21,366 268,294 444 1.38 1.24 821 5067 140 1.04 0.89
(1.11–1.45) (0.99–1.31)
(1.18–1.61) (1.05–1.46)
(0.80–1.36) (0.66–1.19)
HR1: adjusted for age. HR2: adjusted for age, region of residence, health insurance type, history of hypertension, history of diabetes, body mass index (BMI), alcohol drinking, green and yellow vegetable consumption, non-green and non-yellow vegetable consumption, fruit consumption, miso soup consumption, pickled vegetable consumption, smoking status of father during childhood, and smoking status of mother during childhood. b
43
(pooled estimate of risk: 1.25) (Oono et al., 2011). In East Asia, 4 studies in China (1 cohort, 1 case–control, and 2 cross-sectional) showed a positive association of ETS with mortality or prevalence of total stroke in women (He et al., 2008; McGhee et al., 2005; Wen et al., 2006; Zhang et al., 2005); however, a large cohort study in Japan by Hirayama did not show a significant association with cerebrovascular disease (Hirayama, 1987). To the best of our knowledge, this is the second prospective cohort study in Japan and the third in Asia investigating the association of ETS exposure with total stroke. Active smoking is a risk factor for stroke, regardless of its subtypes (US Department of Health and Human Services, 2004). In this study, the effects of ETS exposure by stroke subtype were most evident for subarachnoid hemorrhage, followed by intracerebral hemorrhage; however, the effects for cerebral infarction remain unclear. A study investigating the effect of active smoking on mortality from cardiovascular disease including stroke pooled data from 3 large prospective studies in Japan and showed that the HR among female current smokers was highest for subarachnoid hemorrhage, intracerebral hemorrhage, and cerebral infarction, in that order (Honjo et al., 2010). Our results for ETS were consistent with those of this pooled analysis of active smokers. Few studies have presented results related to stroke subtype (Anderson et al., 2004; Donnan et al., 1989; He et al., 2008; Hirayama, 1987; Iribarren et al., 2004; Qureshi et al., 2005; Yamada et al., 2003; You et al., 1999). For subarachnoid hemorrhage, 2 cohort studies in Japan (Hirayama, 1987; Yamada et al., 2003) and 1 case–control study in Australasia (Anderson et al., 2004) reported no significant effect of ETS. For ischemic stroke, 4 studies showed an increased risk associated with ETS exposure (Donnan et al., 1989; He et al., 2008; Iribarren et al., 2004; You et al., 1999), and 2 studies of them reached statistical significance (He et al., 2008; Iribarren et al., 2004). One cohort study reported no association (Qureshi et al., 2005). No significantly increased risk of hemorrhagic stroke was observed in a Chinese cross-sectional study (He et al., 2008). From the data in Table 4, the effect of husband's smoking on stroke and its subtypes appeared to be lower compared with the effect of smoking by family members other than the husband. Of the study subjects with family members other than the husband who smoked, N 80% had children who smoked. Studies investigating the effect of ETS from family members other than the husband on stroke are scarce, and only 2 studies have demonstrated no association with ETS from parents (Donnan et al., 1989; You et al., 1999). A previous case–control study recruited subjects from 1986 to 1988 in Osaka, Japan, one of the study areas included in this cohort, and showed a significantly increased risk of lung cancer among those whose family members other than the husband smoked and no significant risk elevation for subjects with a smoking husband (Sobue, 1990). Although husband's smoking status has been widely used as a marker of ETS exposure at home, this study indicates that other family members' smoking also may have been an important source of ETS exposure at home, at least in Japan, at that time. In this study, 906 stroke deaths occurred during the follow-up period. This number of events permits a detailed analysis by stroke subtype. In addition, smoking women have a tendency to marry smokers. Thus, misclassification of current smokers as never smokers could have resulted in overestimation of the risk for ETS. However, in this cohort, the prevalence of female smokers was relatively low so that the impact of this misclassification is limited. A study in Japan conducted between 1990 and 1997 investigated the validity of self-reported smoking by the use of the urinary cotinine test as a gold standard (Tsutsumi et al., 2002). In that study, only 1.4% of self-declared never-smoking women were classified as smokers according to urinary cotinine levels in the group among which smoking prevalence based on a self-administrated questionnaire was 8.8%, which is a slight lower than the smoking prevalence in this cohort. Accordingly, this Japanese cohort is considered to have an advantage in evaluating an unbiased health effect of ETS. A considerable number of people with no smoking status or missing data were excluded from this study. Compared with included subjects,
44
Y. Nishino et al. / Preventive Medicine 67 (2014) 41–45
Table 3 Hazard ratios of stroke subtypes for never-smoking women according to familial smoking status among Three-Prefecture Cohort Study participants followed-up for 15 years. Cause of death
ICD-9
ICD-10
Smoker (−)
Smoker (+)
All
Subarachnoid hemorrhage
Intracerebral hemorrhage
Cerebral infarction
430
431
433–434
I60, I69.0
I61, I69.1
I63, I69.3
No. at risk Person-years No. of cases HR1 (95% CI)a HR2 (95% CI)b No. of cases HR1 (95% CI)a HR2 (95% CI)b No. of cases HR1 (95% CI)a HR2 (95% CI)b
13,834 164,354 24 1.00 1.00 45 1.00 1.00 183 1.00 1.00
22,187 273,361 63 1.72 1.66 102 1.55 1.35 284 1.09 0.95
No. at risk Person-years No. of cases HR1 (95% CI)a HR2 (95% CI)b No. of cases HR1 (95% CI)a HR2 (95% CI)b No. of cases HR1 (95% CI)a HR2 (95% CI)b
13,317 161,088 22 1.00 1.00 37 1.00 1.00 129 1.00 1.00
21,366 268,294 59 1.78 1.66 91 1.74 1.46 211 1.24 1.09
(1.08–2.76) (1.02–2.70) (1.09–2.20) (0.94–1.94) (0.91–1.32) (0.78–1.15)
40–79 years
Subarachnoid hemorrhage
Intracerebral hemorrhage
Cerebral infarction
430
431
433–434
I60, I69.0
I61, I69.1
I63, I69.3
(1.09–2.91) (1.00–2.75) (1.19–2.55) (0.99–2.17) (0.99–1.54) (0.87–1.37)
a
HR1: adjusted for age. HR2: adjusted for age, region of residence, health insurance type, history of hypertension, history of diabetes, body mass index (BMI), alcohol drinking, green and yellow vegetable consumption, non-green and non-yellow vegetable consumption, fruit consumption, miso soup consumption, pickled vegetable consumption, smoking status of father during childhood, and smoking status of mother during childhood. b
Table 4 Hazard ratios of stroke for never-smoking women according to smoking status of husband and other family members among Three-Prefecture Cohort Study participants followed-up for 15 years. Smoker (−)
Smoker (+) Husband (−)
Husband (+)
Husband (+)
Others (+)
Others (−)
Others (+)
10,042 125,325 93 0.95 0.93 16 1.31 1.32 18 0.99 0.94 39 0.79 0.78
4728 61,815 82 1.40 1.20 15 2.20 2.04 20 1.91 1.56 33 1.11 0.92
All
Stroke
Subarachnoid hemorrhage
Intracerebral hemorrhage
Cerebral infarction
No. at risk Person-years No. of cases HR1 (95% CI)a HR2 (95% CI)b No. of cases HR1 (95% CI)a HR2 (95% CI)b No. of cases HR1 (95% CI)a HR2 (95% CI)b No. of cases HR1 (95% CI)a HR2 (95% CI)b
13,834 164,354 322 1.00 1.00 24 1.00 1.00 45 1.00 1.00 183 1.00 1.00
7280 84,847 403 1.35 1.19 32 1.85 1.77 60 1.62 1.39 210 1.18 1.01
No. at risk Person-years No. of cases HR1 (95% CI)a HR2 (95% CI)b No. of cases HR1 (95% CI)a HR2 (95% CI)b No. of cases HR1 (95% CI)a HR2 (95% CI)b No. of cases HR1 (95% CI)a HR2 (95% CI)b
13,317 161,088 236 1.00 1.00 22 1.00 1.00 37 1.00 1.00 129 1.00 1.00
6534 80,244 271 1.48 1.30 28 1.93 1.77 52 1.92 1.57 139 1.32 1.14
(1.16–1.56) (1.02–1.39) (1.08–3.16) (1.02–3.08) (1.10–2.40) (0.93–2.08) (0.96–1.44) (0.81–1.24)
(0.75–1.20) (0.73–1.18) (0.68–2.50) (0.69–2.55) (0.57–1.74) (0.54–1.66) (0.55–1.13) (0.55–1.11)
(1.09–1.80) (0.94–1.55) (1.14–4.24) (1.04–4.02) (1.12–3.28) (0.90–2.71) (0.76–1.62) (0.62–1.34)
40–79 years
Stroke
Subarachnoid hemorrhage
Intracerebral hemorrhage
Cerebral infarction
a
(1.24–1.76) (1.08–1.56) (1.10–3.38) (0.99–3.16) (1.26–2.94) (1.01–2.43) (1.04–1.68) (0.89–1.46)
10,010 125,044 89 1.03 1.02 16 1.35 1.33 17 1.06 1.00 38 0.94 0.95
(0.80–1.32) (0.80–1.32) (0.70–2.60) (0.69–2.59) (0.59–1.90) (0.55–1.80) (0.65–1.36) (0.66–1.38)
4699 61,611 80 1.60 1.38 15 2.28 2.03 19 2.07 1.64 33 1.40 1.17
(1.23–2.07) (1.06–1.80) (1.17–4.43) (1.02–4.04) (1.18–3.64) (0.92–2.92) (0.95–2.06) (0.78–1.74)
HR1: adjusted for age. HR2: adjusted for age, region of residence, health insurance type, history of hypertension, history of diabetes, body mass index (BMI), alcohol drinking, green and yellow vegetable consumption, non-green and non-yellow vegetable consumption, fruit consumption, miso soup consumption, pickled vegetable consumption, smoking status of father during childhood, and smoking status of mother during childhood. b
Y. Nishino et al. / Preventive Medicine 67 (2014) 41–45
people excluded because of no smoking status included a higher proportion of those who lived in Miyagi and rural areas of Aichi and a higher proportion of those with smoking family members, possibly because of the inclusion of some smokers. Further, compared with included subjects, excluded people for whom information regarding ETS at home was missing included a higher percentage of elderly people. The effects of these exclusions on the results remain unclear. In this study, ETS exposure at home was evaluated on the basis of the presence of smokers among family members living together at the time of the baseline survey. A study by Ozasa et al. in 1997 presented that approximately two-thirds of non-smoking females reported being exposed to ETS at home in Japan at that time (Ozasa et al., 1997). Although the effect of ETS at home, which is the predominant source of exposure for the subjects, was examined in our study, exposure in the workplace and public settings, which are other important exposure settings, was not considered. The baseline survey in this study was conducted between 1983 and 1985. Male smoking prevalence in Japan at that time exceeded 60% (Japan Health Promotion and Fitness Foundation, 2013), and strategies for preventing ETS exposure in the workplace and public settings were quite incomplete, because the adverse health effects of ETS were poorly understood then. Accordingly, subjects in this cohort were considered to be more extensively exposed to ETS in the workplace and public settings than at present; thus, this misclassification of ETS exposure may dilute the association between ETS exposure and stroke in this study. Our study did not make adjustments for several established risk factors for stroke, such as serum cholesterol, which may have affected the results. In addition, socioeconomic factors other than insurance, such as marital status, education level, income, and differences in the quality of health care among study areas, were not considered in this study. These possible confounding factors could have an effect on the relationship between ETS exposure and stroke mortality reported here. Conclusions In our large-scale cohort study, we found an increase in stroke mortality among never-smoking Japanese women who lived with smoking family members during adulthood. These results did not apply to very elderly people, and the risk was most evident for subarachnoid hemorrhage following subgroup analysis by stroke subtype. Our results provide a new perspective concerning the adverse health effect of ETS exposure and encourage further efforts to control ETS exposure in Japan. Conflict of interest statement The authors declare that there are no conflicts of interest.
Acknowledgments This study was funded by the Japanese Ministry of the Environment (former Environment Agency). We sincerely thank the staff within each study area for collecting and processing data. We also express our gratitude to all participants in the study.
45
References Anderson, C.S., Feigin, V., Bennett, D., et al., 2004. Active and passive smoking and the risk of subarachnoid hemorrhage: an international population-based case–control study. Stroke 35, 633–637. Donnan, G.A., McNeil, J.J., Adena, M.A., Doyle, A.E., O'Malley, H.M., Neill, G.C., 1989. Smoking as a risk factor for cerebral ischaemia. Lancet 2 (8664), 643–647. He, Y., Lam, T.H., Jiang, B., et al., 2008. Passive smoking and risk of peripheral arterial disease and ischemic stroke in Chinese women who never smoked. Circulation 118, 1535–1540. Hirayama, T., 1987. Passive smoking and cancer: an epidemiological review. Gann Monogr. Cancer Res. 33, 127–135. Honjo, K., Iso, H., Tsugane, S., et al., 2010. The effects of smoking and smoking cessation on mortality from cardiovascular disease among Japanese: pooled analysis of three large-scale cohort studies in Japan. Tob. Control. 19, 50–57. Iribarren, C., Darbinian, J., Klatsky, A.L., Friedman, G.D., 2004. Cohort study of exposure to environmental tobacco smoke and risk of first ischemic stroke and transient ischemic attack. Neuroepidemiology 23, 38–44. Japan Health Promotion and Fitness Foundation, 2013. The Latest Information on TobaccoAvailable at http://www.health-net.or.jp/tobacco/product/pd090000.html (Accessed 22 August, 2013). Katanoda, K., Sobue, T., Satoh, H., et al., 2011. An association between long-term exposure to ambient air pollution and mortality from lung cancer and respiratory diseases in Japan. J. Epidemiol. 21, 132–143. McGhee, S.M., Ho, S.Y., Schooling, M., et al., 2005. Mortality associated with passive smoking in Hong Kong. BMJ 330, 287–288. Oono, I.P., Mackay, D.F., Pell, J.P., 2011. Meta-analysis of the association between secondhand smoke exposure and stroke. J. Public Health (Oxf.) 33, 496–502. Ozasa, K., Higashi, A., Yamasaki, M., Hayashi, K., Watanabe, Y., 1997. Validity of selfreported passive smoking evaluated by comparison with smokers in the same household. J. Epidemiol. 7, 205–209. Qureshi, A.I., Suri, M.F., Kirmani, J.F., Divani, A.A., 2005. Cigarette smoking among spouses: another risk factor for stroke in women. Stroke 36, e74–e76. Shinton, R., Beevers, G., 1989. Meta-analysis of relation between cigarette smoking and stroke. BMJ 298, 789–794. Sobue, T., 1990. Association of indoor air pollution and lifestyle with lung cancer in Osaka, Japan. Int. J. Epidemiol. 19, S62–S66. Tsutsumi, A., Kagawa, J., Yamano, Y., Nakadate, T., Shimizu, S., 2002. Relation between cotinine in the urine and indices based on self-declared smoking habits. Environ. Health Prev. Med. 6, 240–247. US Department of Health and Human Services, 2004. The Health Consequences of Smoking: A Report of the Surgeon General. US Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health, Atlanta. US Department of Health and Human Services, 2006. The Health Consequences of Involuntary Exposure to Tobacco Smoke: A Report of the Surgeon General. US Department of Health and Human Services, Centers for Disease Control and Prevention, Coordinating Center for Health Promotion, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health, Atlanta. US Department of Health and Human Services, 2014. The Health Consequences of Smoking—50 Years of Progress: A Report of the Surgeon General. US Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health, Atlanta. Wen, W., Shu, X.O., Gao, Y.T., et al., 2006. Environmental tobacco smoke and mortality in Chinese women who have never smoked: prospective cohort study. BMJ 333, 376. World Health Organization, 2013. WHO Report on the Global Tobacco Epidemic, 2013: Enforcing Bans on Tobacco Advertising, Promotion and Sponsorship. World Health Organization, Geneva. Yamada, S., Koizumi, A., Iso, H., et al., 2003. Risk factors for fatal subarachnoid hemorrhage: the Japan Collaborative Cohort Study. Stroke 34, 2781–2787. You, R.X., Thrift, A.G., McNeil, J.J., Davis, S.M., Donnan, G.A., Melbourne Stroke Risk Factor Study (MERFS) Group, 1999. Ischemic stroke risk and passive exposure to spouses' cigarette smoking. Am. J. Public Health 89, 572–575. Zhang, X., Shu, X.O., Yang, G., et al., 2005. Association of passive smoking by husbands with prevalence of stroke among Chinese women nonsmokers. Am. J. Epidemiol. 161, 213–218.