- Email: [email protected]
Cigarette Smoking and Completed Suicide among Middle-aged Men: A Population-based Cohort Study in Japan
Cigarette Smoking and Completed Suicide among Middle-aged Men: A Population-based Cohort Study in Japan MOTOKI IWASAKI, MD, PHD, TATSUO AKECHI, MD, PHD, YOSUKE UCHITOMI, MD, PHD, AND SHOICHIRO TSUGANE, MD, DRMS, FOR THE JAPAN PUBLIC HEALTH CENTER-BASED PROSPECTIVE STUDY ON CANCER AND CARDIOVASCULAR DISEASE (JPHC STUDY) GROUP*
PURPOSE: To clarify whether cigarette smoking was associated with completed suicide in a cohort of middle-aged Japanese men. METHODS: A total of 45,209 out of 57,714 men, aged 40 to 69 years, in nine public health centers across Japan responded to a self-administered questionnaire that included questions regarding their smoking history. The questionnaire was distributed in 1990 (Cohort I) and in 1993 and 1994 (Cohort II). By the end of 2000, 173 suicides were identified by death certificates. RESULTS: After adjusting for potential confounders, current smokers had a marginally higher risk of suicide than never smokers. Current smokers with more than 60 pack-years of cigarette smoking had a more than two-fold higher risk of suicide than never smokers (pooled multivariate relative risk = 2.1; 95% confidence interval, 1.1, 4.0), but a significant dose-response relationship between pack-years of smoking and suicide was not found. The risk of suicide increased significantly with increasing numbers of cigarettes smoked per day (p for trend = 0.036). Smoking status, duration of quitting, duration of smoking, and age at the start of smoking were not significantly associated with suicide risk. CONCLUSION: Our results suggest that heavy smoking, particularly a large number of cigarettes per day at baseline was associated with an increased risk of suicide independent of several potential confounders. Ann Epidemiol 2005;15:286–292. Ó 2004 Elsevier Inc. All rights reserved. KEY WORDS:
Cigarette Smoking, Middle-aged Men, Prospective Study, Suicide.
INTRODUCTION Several previous studies and a meta-analysis have shown a significant association between cigarette smoking and suicide (1–14). Most prospective studies, however, investigated only the daily consumption of cigarettes and/or smoking status (e.g., non- vs. ex- vs. current smoker) (1–12, 14) and did not adjust for possible confounders (6, 9, 10, 11). Moreover, these studies were performed in Western countries among specific populations, such as army soldiers (1), health professionals (2, 6, 7, 9), twins (10), and volunteers for an interventional trial (12). From the Epidemiology and Prevention Division, Research Center for Cancer Prevention and Screening, National Cancer Center, Tokyo, Japan (M.I., S.T.); Psycho-Oncology Division, National Cancer Center Research Institute East, Chiba, Japan (T.A., Y.U.); and Psychiatry Division, National Cancer Center Hospital East, Chiba, Japan (T.A., Y.U.). *Study Group members are listed at the end of this article. Address for correspondence: Motoki Iwasaki, M.D., Ph.D., Epidemiology and Prevention Division, Research Center for Cancer Prevention and Screening, National Cancer Center, 5-1-1 Tsukiji, Chuo-ku, Tokyo 1040045, Japan. Tel.: C81-3-3542-2511 (ext. 3389); Fax: C81-3-35478578. E-mail: [email protected] Motoki Iwasaki is an awardee of a Research Resident Fellowship from the Foundation for Promotion of Cancer Research in Japan. Received February 9, 2004; accepted August 13, 2004. Ó 2004 Elsevier Inc. All rights reserved. 360 Park Avenue South, New York, NY 10010
To our knowledge, the association between cigarette smoking and suicide has not been studied prospectively in Asian countries, which have different suicide rates, smoking prevalences, and sociocultural contexts compared with Western countries (15–17). Japan has one of the highest suicide rates in the world and also has a high prevalence of smokers (18, 19). In 1998, the number of suicides in Japan reached 31,755 (25.3 per 100,000). Since then, approximately 30,000 people have committed suicide annually in Japan (23.8 per 100,000 persons in 2002) (18), and more than 40% of Japanese suicide victims are middleaged (20). Determining the association between smoking and suicide could provide a useful indicator of persons at risk for suicide. The purpose of the present study was to clarify the association between cigarette smoking and suicide among middle-aged Japanese men. The following issues were examined: 1) the effect of smoking status and the dose–response relationship, such as pack-years or number of cigarettes smoked per day; 2) the effect of smoking cessation; and 3) the effect of age at the start of smoking. To explore these issues, data from the Japan Public Health Center-based Prospective Study on Cancer and Cardiovascular Disease (JPHC Study) was analyzed. 1047-2797/05/$–see front matter doi:10.1016/j.annepidem.2004.08.011
AEP Vol. 15, No. 4 April 2005: 286–292
Selected Abbreviations and Acronyms JPHC study Z The Japan Public Health Center-based Prospective Study on Cancer and Cardiovascular Disease Study PHC Z public health center RR Z relative risk CI Z confidence interval
SUBJECTS AND METHODS Study Cohort This study was conducted as part of the JPHC study, which began in 1990 for Cohort I and in 1993 for Cohort II. The study population was defined as all residents aged 40 to 59 years in Cohort I and 40 to 69 years in Cohort II at baseline who had registered their address in 27 municipalities supervised by nine public health centers (PHC). A populationbased cohort of 57,714 men (27,063 in Cohort I and 30,651 in Cohort II) was established. Details of the study design have been reported elsewhere (21). The JPHC study was approved by the institutional review board of the National Cancer Center, Tokyo, Japan. Baseline Questionnaire Survey A self-administered questionnaire was distributed by hand or mail to the JPHC study subjects in 1990 (Cohort I) and in 1993 and 1994 (Cohort II). The subjects were asked about their personal and familial medical histories, anthropometric factors, smoking history, habitual intake of foods and beverages (including alcohol), physical activity, and other lifestyle factors. Among the eligible subjects, 45,452 men (79%) (20,658 [77%] in Cohort I and 24,794 [81%] in Cohort II) returned the questionnaire. Questions concerning current or former smoking status, age at the start of smoking, number of cigarettes consumed per day, and age at the cessation of smoking were included in both questionnaires, although the style of the questions differed slightly between the two cohorts. The subjects were then categorized as never, former, or current smokers. Former smokers were grouped according to the duration of smoking cessation: 5 years or less or more than 5 years. Current smokers were grouped according to pack-years (1– 29, 30–59, or 60 or more), the number of cigarettes smoked per day (1–19, 20–29, 30–39, or 40 or more), the duration of smoking (1–24, 25–29, 30–34, or 35 or more years), and age at the start of smoking (less than 19, 19–21, or 22 years and older). Subjects with a past medical history were defined as those with a self-reported history of at least one of the following chronic diseases: diabetes mellitus, ischemic heart disease, cerebrovascular disease, cancer, or chronic liver disease. The use of any medications at baseline was noted. The body mass index of the subjects was calculated from the selfreported height and weight, and was divided into quartiles.
Iwasaki et al. CIGARETTE SMOKING AND SUICIDE IN JAPANESE MEN
287
Regarding alcohol drinking, the subjects were categorized as non-drinkers (! 1 day/month), occasional drinkers (1–3 days/month), and regular drinkers (1–2 days/week or more), based on the frequency of alcohol consumption. Among the regular drinkers, the weekly ethanol consumption was calculated by combining the frequency per week and the usual daily amount of alcoholic beverages. Regular drinkers were classified into four groups according to the weekly ethanol intake quartiles (! 138 g/week, 138–251 g/week, 252–413 g/week, > 414 g/week). The frequency of engaging in sports and/or physical exercise during leisure time was classified into three categories: ! 1 day/month, 1 to 3 days/ month, and > 1 day/week. Self-reported stress was assessed by the item, ‘‘Do you usually suffer from stress (little, average, much)?’’ Follow-up and Identification of Suicide All registered subjects were followed from January 1 of the year of the baseline questionnaire survey until December 31, 1999 (Cohort I) and December 31, 2000 (Cohort II). Migration data were obtained annually from the residential registry. Among the questionnaire respondents at baseline, 1966 (4.3%) moved out of the study area and 36 subjects (0.08%) were lost to follow-up within the follow-up period. When subjects remained in their original area, data on their deaths were based on death certificates obtained from their PHC. For subjects who died after they had moved from their original PHC area, the cause of death was determined from death certificates kept by the Ministry of Health, Labor, and Welfare, Japan. Death from suicide was defined according to the International Classification of Diseases, 10th Revision (ICD–10), as codes X60–X84. Statistical Analysis Subjects with incomplete smoking data were excluded from the present analyses (243 men [0.5%], including one suicide case). Person-years of follow-up were calculated from baseline until the date of suicide, the most recent date when the survival status had been confirmed in censored cases, or the end of the study periods (December 31, 1999, for Cohort I and December 31, 2000, for Cohort II) for survivors, whichever occurred first. Suicide rates were calculated using person-years as the denominator and were standardized using the 5-year age distribution at baseline in each cohort. For each cohort, the Cox proportional hazards model was used to estimate the relative risk (RR) and 95% confidence interval (CI) of suicide according to smoking habits at baseline using the SAS program (PROC PHREG) (SAS Institute Inc., Cary, NC). Linear trends were tested using the Cox proportional hazards model by treating the categories as ordinal variables. All p-values reported are two-sided, and the significance level was set at p ! 0.05.
288
Iwasaki et al. CIGARETTE SMOKING AND SUICIDE IN JAPANESE MEN
AEP Vol. 15, No. 4 April 2005: 286–292
TABLE 1. Baseline characteristics of subjects according to smoking status Cohort I
Cohort II
Smoking status
Smoking status
Characteristics
Never
Former
Current
Never
Former
Current
No. of subjects Age, mean (years) Medication, any medicine (%) Body mass index, mean (kg/m2) Alcohol drinking habits, regular drinkers (at least 1 day/week) (%) Sports and physical exercise during leisure time, >1 day/week (%) Living with spouse, yes (%) Self-reported stress, stressful (%)
4948 49.6 21.9 23.9 59.6 19.1 81.7 23.3
4683 50.4 29.6 23.9 69.5 21.1 85.9 25.9
10943 49.1 20.8 23.2 72.1 15.2 80.1 26.6
6034 54.3 29.8 24.0 62.6 21.8 88.7 17.1
5944 56.8 42.2 23.7 66.7 23.9 92.2 17.4
12657 53.2 27.9 23.1 70.4 16.0 87.8 20.6
The random-effects model developed by DerSimonian and Laird was used to combine the loge RRs; the cohortspecific RRs were weighted by the inverse of their variance (22). Heterogeneity among the cohorts was examined using the asymptotic DerSimonian and Laird Q statistics (22). In the present study, only pooled estimates are given because a similar overall risk trend was observed in the two cohorts (data not shown).
RESULTS Among the 382,390 person-years of observation in 45,209 subjects (average follow-up period: 8.5 years), 173 suicides occurred. Table 1 shows the baseline characteristics according to smoking status. The proportions of current and former smokers were 53.2% and 22.8% for Cohort I and 51.4% and 24.1% for Cohort II, respectively. Current TABLE 2. Relative risk (RR) and 95% confidence interval (95% CI) of suicide according to smoking status Smoking status
Cohort I No. of suicides Person-years Age standardized suicide ratea Cohort II No. of suicides Person-years Age standardized suicide ratea Pooled Multivariate RRb (95% CI) Multivariate RRc (95% CI) a
Never
Former
Current
22 48143 46.0
15 45427 33.1
63 105458 59.7
14 45588 30.4
14 44011 27.8
45 93765 48.3
1.0 1.0
0.8 (0.5, 1.4) 0.7 (0.4, 1.2)
1.4 (0.97, 2.1) 1.3 (0.9, 2.0)
Suicide rate (per 100,000 person-years) standardized by distribution of 5-year age groups at baseline in each cohort. Adjusted for age at baseline (continuous), area (PHCs), past history of any chronic diseases (no, yes), medication (none, yes), body mass index (quartile), physical exercise during leisure time (!1 day/month, 1–3 days/month, >1 day/week), and living with spouse (no, yes). c Further adjusted for alcohol drinking (non-drinkers, occasional drinkers, four groups of regular drinkers according to quartile of ethanol intake [g/week] [0–137, 138–251, 252–413, 414C]) and self-reported stress (little, average, much). b
smokers were less likely to perform sports and physical exercise and more likely to drink alcohol regularly and report stress than never and former smokers in both cohorts. Former smokers were more likely to be receiving medication and living with spouses than never and current smokers. The associations between subsequent suicide and smoking status at baseline, years of cessation among past smokers, and pack-years among current smokers are shown in Tables 2 and 3. Compared to never smokers, current smokers had a marginally higher risk of suicide (Table 2). We failed to observe a statistically significant reduction in suicide risk after smoking cessation, although the estimated pooled RR was lower among former smokers who had quit smoking more than 5 years earlier (Table 3). Current smokers with more than 60 pack-years of cigarette smoking had a more than two-fold higher risk of suicide than never smokers, and the risk of suicide increased with increasing pack-years, although this trend was not statistically significant (Table 3). The associations between suicide and smoking status or pack-years among current smokers were somewhat attenuated by further adjustments for alcohol drinking and selfreported stress. To further assess the suicide risk of current smokers, variables like duration of smoking, the number of cigarettes smoked per day, and the age at the start of smoking were examined (Table 4). No significant association was observed between the duration of smoking and suicide. Although current smokers who smoked more than 40 cigarettes per day had a significantly increased risk of suicide (pooled multivariate RR Z 1.9; 95% CI, 1.04, 3.4), the RR was lowered slightly and the statistical significance disappeared when further adjustments for alcohol drinking and selfreported stress were made. The risk of suicide increased significantly as the number of cigarettes smoked per day increased, even after adjustments for potential confounders, like the duration of smoking, and further adjustments for alcohol drinking and self-reported stress. A significant association was not observed between the age at the start of smoking and suicide.
AEP Vol. 15, No. 4 April 2005: 286–292
Iwasaki et al. CIGARETTE SMOKING AND SUICIDE IN JAPANESE MEN
289
TABLE 3. Pooled multivariate relative risk (RR) and 95% confidence interval (95% CI) of suicide according to years of smoking cessation among former smokers and pack-years among current smokers, compared with never smokers Years of cessation
No. of suicides Multivariate RRa (95% CI) Multivariate RRb (95% CI)
Never smoker
Years !5
5 < Years
p for trend among former smokers
36 1.0 1.0
12 1.2 (0.6, 2.3) 1.0 (0.5, 2.1)
17 0.7 (0.4, 1.3) 0.6 (0.3, 1.2)
0.19 0.19
Pack-years
No. of suicides Multivariate RRa (95% CI) Multivariate RRb (95% CI)
Never smoker
Pack-years!30
30
60
p for trend among current smokers
36 1.0 1.0
45 1.3 (0.8, 2.1) 1.3 (0.8, 2.0)
49 1.4 (0.9, 2.3) 1.3 (0.8, 2.1)
12 2.3 (1.2, 4.5) 2.1 (1.1, 4.0)
0.09 0.16
a Adjusted for age at baseline (continuous), area (PHCs), past history of any chronic diseases (no, yes), medication (none, yes), body mass index (quartile), physical exercise during leisure time (!1 day/month, 1–3 days/month, >1 day/week), and living with spouse (no, yes). b Further adjusted for alcohol drinking (non-drinkers, occasional drinkers, four groups of regular drinkers according to quartile of ethanol intake [g/week] [0–137, 138–251, 252– 413, 414C]) and self-reported stress (little, average, much).
To eliminate the effects of preclinical diseases, further analyses were performed after excluding suicides and censored cases that occurred during the first 2 years of the follow-up period. The associations between subsequent suicide and pack-years among current smokers and the number of cigarettes smoked per day were somewhat attenuated, and the statistical significance disappeared (data not shown). On the other hand, these associations were slightly strengthened when subjects with a past history of chronic diseases were excluded from the analysis (data not shown).
DISCUSSION Four main explanations for the smoking–suicide connection have been proposed (1, 2, 7): 1) depression (a well-known risk factor for suicide) can lead to smoking as a form of selfmedication (23–25); 2) smoking alters brain chemistry (26, 27), leading to depression (28–30); 3) smoking leads to malignant diseases, such as cancer, which increase the risk of suicide (31); and 4) smoking and suicide are not causally related (12, 14). However, the mechanism underlying the smoking–suicide connection remains unclear.
TABLE 4. Pooled multivariate relative risk (RR) and 95% confidence interval (95% CI) of suicide according to duration of smoking, number of cigarettes smoked per day, and age at the start of smoking among current smokers Duration of smoking (years)
No. of suicides Multivariate RRa (95% CI) Multivariate RRb (95% CI)
!25
25–29
30–34
35C
p for trend
32 1.0 1.0
23 1.0 (0.6, 1.9) 1.0 (0.6, 1.9)
22 1.5 (0.7, 3.2) 1.5 (0.7, 3.4)
30 1.4 (0.5, 3.9) 1.4 (0.5, 4.0)
0.52 0.51
!20
20–29
30–39
40C
p for trend
25 1.0 1.0
41 0.9 (0.5, 1.5) 0.9 (0.5, 1.5)
20 1.4 (0.8, 2.6) 1.4 (0.8, 2.6)
21 1.9 (1.04, 3.4) 1.7 (0.9, 3.1)
0.01 0.036
22C
19–21
!19
p for trend
19 1.0 1.0
63 1.2 (0.7, 2.1) 1.4 (0.8, 2.5)
26 1.4 (0.7, 2.6) 1.6 (0.8, 3.0)
0.32 0.18
Number of cigarettes per day
No. of suicides Multivariate RRa (95% CI) Multivariate RRb (95% CI)
Age at start of smoking
No. of suicides Multivariate RRa (95% CI) Multivariate RRb (95% CI)
a Adjusted for age at baseline (continuous), area (PHCs), past history of any chronic diseases (no, yes), medication (none, yes), body mass index (quartile), physical exercise during leisure time (!1 day/month, 1–3 days/month, >1 day/week), living with spouse (no, yes) and duration of smoking for number of cigarettes smoked, or number of cigarettes smoked for duration of smoking and age at the start of smoking. b Further adjusted for alcohol drinking (non-drinkers, occasional drinkers, four groups of regular drinkers according to quartile of ethanol intake [g/week] [0–137, 138–251, 252– 413, 414C]) and self-reported stress (little, average, much).
290
Iwasaki et al. CIGARETTE SMOKING AND SUICIDE IN JAPANESE MEN
The present study showed that current smokers with more than 60 pack-years of cigarettes had a significantly higher risk of suicide than that of never smokers. Among current smokers, the suicide risk increased as the number of cigarettes smoked per day at baseline increased. This association was statistically significant even after adjusting for potential confounders like the duration of smoking, although most previous studies (1, 2, 4, 6–8, 10, 12) did not control for the duration of smoking. However, the duration of smoking was not significantly associated with the suicide risk in the present study. Our results indicate that heavy smokers, rather than long-term smokers, have a significantly increased risk of suicide. To examine the possibility that the suicide cases had already changed the number of cigarettes smoked per day at baseline as a result of existing factors that could also be related to an increased risk of suicide, such as a specific life event, mental disorder or so on, a preliminary analysis was performed using data from the 5-year follow-up surveys conducted in 1995 for Cohort I and in between 1998 and 1999 for Cohort II. We calculated the change in number of cigarettes smoked per day between the baseline and the 5year follow-up surveys. This change was compared between the suicide cases that occurred after the 5-year follow-up survey and the other non-suicide cases among the current smokers. As the results showed no difference, the subjects who committed suicide during the follow-up period probably did not change the number of cigarettes smoked per day because of factors that could potentially be related to their subsequent suicide. The risk of suicide for current smokers did not reach a significant level in this study, even though a previous study showed a statistically significant increased risk of suicide for current smokers after adjusting for psychological factors, such as symptoms of depressed mood, stress, and anxiety (4). One possible reason for the lack of statistical significance may be misclassification caused by subsequent changes in smoking status as a result of illness or some other events during the long follow-up period. Smoking cessation after the baseline survey had been conducted could lead to an underestimation of the true risk among current smokers who continue to smoke, since the suicide risk decreased as the length of smoking cessation increased. Additionally, since a positive relation was previously observed between indicators of nicotine addiction and the number of cigarettes smoked per day (32) and nicotine dependence has been associated with a risk of major depression (33), the number of cigarettes smoked per day, rather than the duration of smoking, might be more strongly correlated with mental disorders, such as depression, or one or more factors predisposing the subject to mental disorders that increase the risk of suicide (13). Since smokers who are
AEP Vol. 15, No. 4 April 2005: 286–292
less dependent on nicotine are more likely to quit smoking, the proportion of nicotine dependence is probably becoming high among current smokers as the prevalence of current smokers decreases (23). Nevertheless, since smoking used to be socially accepted in Japan and the prevalence of current smokers is still high (19), several non-nicotine dependent subjects who are less likely to suffer from depression than nicotine dependent subjects (34, 35) probably exist among the current smokers in this study. Such a situation may have attenuated the association between smoking status and suicide, possibly explaining the lack of statistical significance. Our findings showed that the associations between subsequent suicide and the pack-years of cigarette smoking and the number of cigarettes smoked per day were somewhat attenuated and the statistical significance disappeared when the suicides and censored cases that occurred during the first 2 years of the follow-up period were excluded. Meanwhile, the statistical significance was preserved when subjects with a past history of chronic diseases were excluded. The lack of statistical significance might simply reflect the reduction in the number of cases, or a lack of statistical power. Alternatively, the exclusion of the suicides and censored cases that occurred during the first 2 years of the follow-up period may have served as a control for the effects of mental illness on the risk of suicide to some extent. The relationships among smoking, mental illness (especially depression), and suicide remain unclear. One study concluded that the association between smoking and suicide accounted for confounders, such as mental well-being (14). On the other hand, adjustments for mental illness or factors related to mental illness would lead to inappropriate results, if these factors were involved in the causal pathway between smoking and suicide (36). Our results were not substantially influenced by adjustments for alcohol drinking and selfreported stress. Further analysis to clarify the possible role of depression on the relationship between smoking and suicide is needed to determine the smoking-suicide connection. However, data on mental conditions, such as depression, were not available in the JPHC study. A previous study suggested a possible protective effect of smoking cessation on the risk of suicide (37). Our results showed that former smokers had the same risk of suicide as never smokers, regardless of the length of smoking cessation. This implies that smoking cessation may be effective for reducing the risk of suicide. However, further research is needed to clarify the possible protective effect of smoking cessation on the risk of suicide because presently available evidence is limited. To our knowledge, this is the first study to examine the effects of age at the start of smoking on the risk of suicide. According to previous studies in adolescents, depression and anxiety were associated with a higher risk of smoking initiation (25), while smoking was also associated with an
AEP Vol. 15, No. 4 April 2005: 286–292
increased risk of the subsequent depressed mood (30). Moreover, smokers who started before the age of 13 years had significantly more depressive episodes than smokers who smoked after the age of 17 years or lifetime nonsmokers (38). These previous studies imply that the age at which smoking is begun is related to mental disorders in adolescents, and vice versa. However, our results showed no association between the start of smoking at an early age and the risk of suicide. One possible reason is that the mean age at which smoking was started was older among Japanese than among Americans (39). In this cohort, the mean ages at which smoking was started among current smokers were 20.3, 21.3, and 21.6 years for men aged 40 to 49, 50 to 59, and 60 to 69 years at baseline, respectively; 90% of the subjects started smoking between the ages of 17 and 28 years. In the Cancer Prevention Study II, the corresponding figures were 17.2, 17.6, and 17.8 years (39). Therefore, the number of subjects who started smoking under the age of 16 years may not have been large enough to establish an association. A second possible reason is that more than 80% of Japanese men smoked cigarettes when the subjects in the JPHC study started smoking (e.g., 82.3% in 1965) (19), and cigarette smoking was a popular and socially-accepted habit in Japan, suggesting that the association between cigarette smoking and mental disorders might be weakened. Our study has several methodological advantages. First, the prospective design enables potential selection or recall biases to be avoided. Second, the study included a large population-based sample with a high response rate (approximately 80%) and a very low rate of loss to follow-up. Third, similar findings were obtained in two different cohorts. On the other hand, the generalizability of this study was a methodological disadvantage. Our findings may not be applicable to urban Japanese populations, to subjects from different generations, or to women, since this study did not include metropolitan areas and the subjects were middleaged Japanese men. Allowing for these limitations, our study showed that heavy cigarette smoking was related to an increased risk of suicide among middle-aged Japanese men. Since progress in suicide prevention has been disappointing (40), useful clinical markers for identifying people with a high risk of suicide are needed. Although psychiatric disorders, especially depression, are a well-known risk factor for suicide, such conditions are often unrecognized, even in medical settings (41). Therefore, smoking, particularly the number of cigarettes smoked per day, may be a useful clinical indicator of a risk for subsequent suicide. In conclusion, our study suggested that cigarette smoking, or at least heavy cigarette smoking, was associated with a risk of suicide that was independent of a past history of chronic disease, heavy alcohol drinking, infrequent exercise, self-reported stress, and an unmarried status, although
Iwasaki et al. CIGARETTE SMOKING AND SUICIDE IN JAPANESE MEN
291
the mechanism underlying the association between cigarette smoking and the risk of suicide was not addressed. Future studies are needed to clarify the associations among smoking, depression, and suicide. We thank all staff members in each study area and in the central offices for their cooperation and technical assistance. We also thank Dr. T. Otani of the Epidemiology and Prevention Division, Research Center for Cancer Prevention and Screening, National Cancer Center, Tokyo, Japan for his helpful comments.
Study group members Members of the JPHC Study Group (principal investigator: S. Tsugane): M. Inoue, T. Hanaoka, S. Tsugane, S. Yamamoto, T. Sobue, Research Center for Cancer Prevention and Screening, National Cancer Center, Tokyo; J. Ogata, S. Baba, T. Mannami, National Cardiovascular Center, Suita; K. Miyakawa, F. Saito, A. Koizumi, Y. Sano, I. Hashimoto, Iwate Prefectural Ninohe Public Health Center, Ninohe; Y. Miyajima, N. Suzuki, S. Nagasawa, Y. Furusugi, Akita Prefectural Yokote Public Health Center, Yokote; H. Sanada, Y. Hatayama, F. Kobayashi, H. Uchino, Y. Shirai, T. Kondo, R. Sasaki, Y. Watanabe, Nagano Prefectural Saku Public Health Center, Saku; Y. Kishimoto, E. Takara, T. Fukuyama, M. Kinjo, M. Irei, Okinawa Prefectural Ishikawa (Chubu) Public Health Center, Ishikawa; K. Imoto, H. Yazawa, T. Seo, A. Seiko, F. Ito, Katsushika Public Health Center, Tokyo; A. Murata, K. Minato, K. Motegi, T. Fujieda, Ibaraki Prefectural Kasama (Mito) Public Health Center, Mito; K. Matsui, T. Abe, Niigata Prefectural Kashiwazaki Public Health Center, Kashiwazaki; M. Doi, A. Terao, Y. Ishikawa, Kochi Prefectural Tosayamada (Chuo-higashi) Public Health Center, Tosayamada; H. Sueta, H. Doi, M. Urata, N. Okamoto, F. Ide, Nagasaki Prefectural Arikawa (Kamigoto) Public Health Center, Arikawa; H. Sakiyama, N. Onga, H. Takaesu, Okinawa Prefectural Miyako Public Health Center, Hirara; F. Horii, I. Asano, H. Yamaguchi, K. Aoki, S. Maruyama, M. Ichii, Osaka Prefectural Suita Public Health Center, Suita; S. Matsushima, S. Natsukawa, Saku General Hospital, Usuda; S. Watanabe, M. Akabane, Tokyo University of Agriculture, Tokyo; M. Konishi, K. Okada, Ehime University, Matsuyama; H. Iso, Y. Honda, Tsukuba University, Tsukuba; H. Sugimura, Hamamatsu University, Hamamatsu; Y. Tsubono, Tohoku University, Sendai; M. Kabuto, National Institute for Environmental Studies, Tsukuba; S. Tominaga, Aichi Cancer Center Research Institute, Nagoya; M. Iida, W. Ajiki, Osaka Medical Center for Cancer and Cardiovascular Disease, Osaka; S. Sato, Osaka Medical Center for Health Science and Promotion, Osaka; N. Yasuda, Kochi Medical School Nankoku; S. Kono, Kyushu University, Fukuoka; K. Suzuki, Research
292
Iwasaki et al. CIGARETTE SMOKING AND SUICIDE IN JAPANESE MEN
Institute for Brain and Blood Vessels Akita, Akita; Y. Takashima, Kyorin University, Mitaka; E. Maruyama, Kobe University, Kobe; the late M. Yamaguchi, Y. Matsumura, S. Sasaki, National Institute of Health and Nutrition, Tokyo; and T. Kadowaki, Tokyo University, Tokyo. REFERENCES 1. Miller M, Hemenway D, Bell NS, Yore MM, Amoroso PJ. Cigarette smoking and suicide: A prospective study of 300,000 male active-duty Army soldiers. Am J Epidemiol. 2000;151:1060–1063. 2. Miller M, Hemenway D, Rimm E. Cigarettes and suicide: A prospective study of 50,000 men. Am J Public Health. 2000;90:768–773. 3. Leistikow BN, Martin DC, Samuels SJ. Injury death excesses in smokers: A 1990-95 United States national cohort study. Inj Prev. 2000;6:277–280. 4. Tanskanen A, Tuomilehto J, Viinamaki H, Vartiainen E, Lehtonen J, Puska P. Smoking and the risk of suicide. Acta Psychiatr Scand. 2000;101: 243–245. 5. Tanskanen A, Tuomilehto J, Viinamaki H, Vartiainen E, Lehtonen J, Puska P. Joint heavy use of alcohol, cigarettes, and coffee and the risk of suicide. Addiction. 2000;95:1699–1704. 6. Doll R, Peto R, Wheatley K, Gray R, Sutherland I. Mortality in relation to smoking: 40 years’ observations on male British doctors. BMJ. 1994;309: 901–911. 7. Hemenway D, Solnick SJ, Colditz GA. Smoking and suicide among nurses. Am J Public Health. 1993;83:249–251. 8. Tverdal A, Thelle D, Stensvold I, Leren P, Bjartveit K. Mortality in relation to smoking history: 13 years’ follow-up of 68,000 Norwegian men and women 35–49 years. J Clin Epidemiol. 1993;46:475–487. 9. Doll R, Gray R, Hafner B, Peto R. Mortality in relation to smoking: 22 years’ observations on female British doctors. BMJ. 1980;280:967–971. 10. Friberg L, Cederlof R, Lorich U, Lundman T, De Faire U. Mortality in twins in relation to smoking habits and alcohol problems. Arch Environ Health. 1973;27:294–304. 11. Paffenbarger RS Jr, King SH, Wing AL. Chronic disease in former college students. IX. Characteristics in youth that predispose to suicide and accidental death in later life. Am J Public Health. 1969;59:900–908. 12. Smith GD, Phillips AN, Neaton JD. Smoking as ‘‘independent’’ risk factor for suicide: Illustration of an artifact from observational epidemiology? Lancet. 1992;340:709–712. 13. Harris EC, Barraclough B. Suicide as an outcome for mental disorders. A meta-analysis. Br J Psychiatry. 1997;170:205–228. 14. Hemmingsson T, Kriebel D. Smoking at age 18–20 and suicide during 26 years of follow-up: How can the association be explained? Int J Epidemiol. 2003;32:1000–1004. 15. Forey B, Hamling J, Lee P, Wald N. International smoking statistics: A collection of historical data from 30 economically developed countries. 2nd ed. New York: Oxford University Press; 2002. 16. Statistics and Information Department, Minister’s Secretariat, Ministry of Health, Labor, and Welfare. Special Report of Vital Statistics on Suicide Mortality Statistics [in Japanese]. Tokyo: Health and Welfare Statistics Association; 1999. 17. Domino G, Takahashi Y. Attitudes toward suicide in Japanese and American medical students. Suicide Life Threat Behav. 1991;21:345–359. 18. Statistics and Information Department, Minister’s Secretariat, Ministry of Health, Labor, and Welfare. Vital Statistics of Japan [in Japanese]. Tokyo: Health and Welfare Statistics Association; 2002. 19. Ministry of Health, Labor, and Welfare. Smoking and Health. New edition. Report from the Commission on Smoking and Health, Japan [in Japanese]. Tokyo: Hoken Dojin sha; 2000. 20. Lamar J. Suicides in Japan reach a record high. BMJ. 2000;321:528.
AEP Vol. 15, No. 4 April 2005: 286–292
21. Watanabe S, Tsugane S, Sobue T, Konishi M, Baba S. Study design and organization of the JPHC study. J Epidemiol. 2001;11:S3–S7. 22. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials. 1986;7:177–188. 23. Hughes JR. Clonidine, depression, and smoking cessation. JAMA. 1988; 259:2901–2902. 24. Glassman AH. Cigarette smoking: Implications for psychiatric illness. Am J Psychiatry. 1993;150:546–553. 25. Patton GC, Carlin JB, Coffey C, Wolfe R, Hibbert M, Bowes G. Depression, anxiety, and smoking initiation: A prospective study over 3 years. Am J Public Health. 1998;88:1518–1522. 26. Malone KM, Waternaux C, Haas GL, Cooper TB, Li S, Mann JJ. Cigarette smoking, suicidal behavior, and serotonin function in major psychiatric disorders. Am J Psychiatry. 2003;160:773–779. 27. Kirch DG, Gerhardt GA, Shelton RC, Freedman R, Wyatt RJ. Effect of chronic nicotine administration on monoamine and monoamine metabolite concentrations in rat brain. Clin Neuropharmacol. 1987;10:376–383. 28. Newhouse PA, Sunderland T, Tariot PN, Blumhardt CL, Weingartner H, Mellow A, et al. Intravenous nicotine in Alzheimer’s disease: A pilot study. Psychopharmacology (Berl.) 1988;95:171–175. 29. Green BH, Copeland JR, Dewey ME, Sharma V, Saunders PA, Davidson IA, et al. Risk factors for depression in elderly people: A prospective study. Acta Psychiatr Scand. 1992;86:213–217. 30. Wu LT, Anthony JC. Tobacco smoking and depressed mood in late childhood and early adolescence. Am J Public Health. 1999;89:1837–1840. 31. Tanaka H, Tsukuma H, Masaoka T, Ajiki W, Koyama Y, Kinoshita N, et al. Suicide risk among cancer patients: Experience at one medical center in Japan, 1978–1994. Jpn J Cancer Res. 1999;90:812–817. 32. Centers for Disease Control and Prevention. Indicators of nicotine addiction among womendUnited States, 1991–1992. JAMA. 1995;273:842. 33. Breslau N, Kilbey MM, Andreski P. Nicotine dependence and major depression. New evidence from a prospective investigation. Arch Gen Psychiatry. 1993;50:31–35. 34. Breslau N, Kilbey M, Andreski P. Nicotine dependence, major depression, and anxiety in young adults. Arch Gen Psychiatry. 1991;48:1069– 1074. 35. Schmitz N, Kruse J, Kugler J. Disabilities, quality of life, and mental disorders associated with smoking and nicotine dependence. Am J Psychiatry. 2003;160:1670–1676. 36. Leistikow B. Commentary: Questionable premises, overadjustment, and a smoking/suicide association in younger adult men. Int J Epidemiol. 2003;32:1005–1006. 37. Leistikow BN, Shipley MJ. Might stopping smoking reduce injury death risks? A meta-analysis of randomized, controlled trials. Prev Med. 1999;28:255–259. 38. Hanna EZ, Grant BF. Parallels to early onset alcohol use in the relationship of early onset smoking with drug use and DSM-IV drug and depressive disorders: Findings from the National Longitudinal Epidemiologic Survey. Alcohol Clin Exp Res. 1999;23:513–522. 39. Thun MJ, Day-Lally C, Myers DG, Calle EE, Flanders WD, Zhu BP, et al. Trends in tobacco smoking and mortality from cigarette use in Cancer Prevention Studies I (1959 through 1965) and II (1982 through 1988). In: Shopland DR, Burns DM, Garfinkel L, Samet J, eds. National Cancer Institute, Smoking and Tobacco Control, Monograph 8: Changes in Cigarette-related Disease Risks and their Implication for Prevention and Control. NIH Publication No. 97-4213. Washington, DC: National Institutes of Health; 1997; 305–382. 40. Gunnell D, Frankel S. Prevention of suicide: Aspirations and evidence. BMJ. 1994;308:1227–1233. 41. Pignone MP, Gaynes BN, Rushton JL, Burchell CM, Orleans CT, Mulrow CD, et al. Screening for depression in adults: A summary of the evidence for the US Preventive Services Task Force. Ann Intern Med. 2002;136: 765–776.
PURPOSE: To clarify whether cigarette smoking was associated with completed suicide in a cohort of middle-aged Japanese men. METHODS: A total of 45,209 out of 57,714 men, aged 40 to 69 years, in nine public health centers across Japan responded to a self-administered questionnaire that included questions regarding their smoking history. The questionnaire was distributed in 1990 (Cohort I) and in 1993 and 1994 (Cohort II). By the end of 2000, 173 suicides were identified by death certificates. RESULTS: After adjusting for potential confounders, current smokers had a marginally higher risk of suicide than never smokers. Current smokers with more than 60 pack-years of cigarette smoking had a more than two-fold higher risk of suicide than never smokers (pooled multivariate relative risk = 2.1; 95% confidence interval, 1.1, 4.0), but a significant dose-response relationship between pack-years of smoking and suicide was not found. The risk of suicide increased significantly with increasing numbers of cigarettes smoked per day (p for trend = 0.036). Smoking status, duration of quitting, duration of smoking, and age at the start of smoking were not significantly associated with suicide risk. CONCLUSION: Our results suggest that heavy smoking, particularly a large number of cigarettes per day at baseline was associated with an increased risk of suicide independent of several potential confounders. Ann Epidemiol 2005;15:286–292. Ó 2004 Elsevier Inc. All rights reserved. KEY WORDS:
Cigarette Smoking, Middle-aged Men, Prospective Study, Suicide.
INTRODUCTION Several previous studies and a meta-analysis have shown a significant association between cigarette smoking and suicide (1–14). Most prospective studies, however, investigated only the daily consumption of cigarettes and/or smoking status (e.g., non- vs. ex- vs. current smoker) (1–12, 14) and did not adjust for possible confounders (6, 9, 10, 11). Moreover, these studies were performed in Western countries among specific populations, such as army soldiers (1), health professionals (2, 6, 7, 9), twins (10), and volunteers for an interventional trial (12). From the Epidemiology and Prevention Division, Research Center for Cancer Prevention and Screening, National Cancer Center, Tokyo, Japan (M.I., S.T.); Psycho-Oncology Division, National Cancer Center Research Institute East, Chiba, Japan (T.A., Y.U.); and Psychiatry Division, National Cancer Center Hospital East, Chiba, Japan (T.A., Y.U.). *Study Group members are listed at the end of this article. Address for correspondence: Motoki Iwasaki, M.D., Ph.D., Epidemiology and Prevention Division, Research Center for Cancer Prevention and Screening, National Cancer Center, 5-1-1 Tsukiji, Chuo-ku, Tokyo 1040045, Japan. Tel.: C81-3-3542-2511 (ext. 3389); Fax: C81-3-35478578. E-mail: [email protected] Motoki Iwasaki is an awardee of a Research Resident Fellowship from the Foundation for Promotion of Cancer Research in Japan. Received February 9, 2004; accepted August 13, 2004. Ó 2004 Elsevier Inc. All rights reserved. 360 Park Avenue South, New York, NY 10010
To our knowledge, the association between cigarette smoking and suicide has not been studied prospectively in Asian countries, which have different suicide rates, smoking prevalences, and sociocultural contexts compared with Western countries (15–17). Japan has one of the highest suicide rates in the world and also has a high prevalence of smokers (18, 19). In 1998, the number of suicides in Japan reached 31,755 (25.3 per 100,000). Since then, approximately 30,000 people have committed suicide annually in Japan (23.8 per 100,000 persons in 2002) (18), and more than 40% of Japanese suicide victims are middleaged (20). Determining the association between smoking and suicide could provide a useful indicator of persons at risk for suicide. The purpose of the present study was to clarify the association between cigarette smoking and suicide among middle-aged Japanese men. The following issues were examined: 1) the effect of smoking status and the dose–response relationship, such as pack-years or number of cigarettes smoked per day; 2) the effect of smoking cessation; and 3) the effect of age at the start of smoking. To explore these issues, data from the Japan Public Health Center-based Prospective Study on Cancer and Cardiovascular Disease (JPHC Study) was analyzed. 1047-2797/05/$–see front matter doi:10.1016/j.annepidem.2004.08.011
AEP Vol. 15, No. 4 April 2005: 286–292
Selected Abbreviations and Acronyms JPHC study Z The Japan Public Health Center-based Prospective Study on Cancer and Cardiovascular Disease Study PHC Z public health center RR Z relative risk CI Z confidence interval
SUBJECTS AND METHODS Study Cohort This study was conducted as part of the JPHC study, which began in 1990 for Cohort I and in 1993 for Cohort II. The study population was defined as all residents aged 40 to 59 years in Cohort I and 40 to 69 years in Cohort II at baseline who had registered their address in 27 municipalities supervised by nine public health centers (PHC). A populationbased cohort of 57,714 men (27,063 in Cohort I and 30,651 in Cohort II) was established. Details of the study design have been reported elsewhere (21). The JPHC study was approved by the institutional review board of the National Cancer Center, Tokyo, Japan. Baseline Questionnaire Survey A self-administered questionnaire was distributed by hand or mail to the JPHC study subjects in 1990 (Cohort I) and in 1993 and 1994 (Cohort II). The subjects were asked about their personal and familial medical histories, anthropometric factors, smoking history, habitual intake of foods and beverages (including alcohol), physical activity, and other lifestyle factors. Among the eligible subjects, 45,452 men (79%) (20,658 [77%] in Cohort I and 24,794 [81%] in Cohort II) returned the questionnaire. Questions concerning current or former smoking status, age at the start of smoking, number of cigarettes consumed per day, and age at the cessation of smoking were included in both questionnaires, although the style of the questions differed slightly between the two cohorts. The subjects were then categorized as never, former, or current smokers. Former smokers were grouped according to the duration of smoking cessation: 5 years or less or more than 5 years. Current smokers were grouped according to pack-years (1– 29, 30–59, or 60 or more), the number of cigarettes smoked per day (1–19, 20–29, 30–39, or 40 or more), the duration of smoking (1–24, 25–29, 30–34, or 35 or more years), and age at the start of smoking (less than 19, 19–21, or 22 years and older). Subjects with a past medical history were defined as those with a self-reported history of at least one of the following chronic diseases: diabetes mellitus, ischemic heart disease, cerebrovascular disease, cancer, or chronic liver disease. The use of any medications at baseline was noted. The body mass index of the subjects was calculated from the selfreported height and weight, and was divided into quartiles.
Iwasaki et al. CIGARETTE SMOKING AND SUICIDE IN JAPANESE MEN
287
Regarding alcohol drinking, the subjects were categorized as non-drinkers (! 1 day/month), occasional drinkers (1–3 days/month), and regular drinkers (1–2 days/week or more), based on the frequency of alcohol consumption. Among the regular drinkers, the weekly ethanol consumption was calculated by combining the frequency per week and the usual daily amount of alcoholic beverages. Regular drinkers were classified into four groups according to the weekly ethanol intake quartiles (! 138 g/week, 138–251 g/week, 252–413 g/week, > 414 g/week). The frequency of engaging in sports and/or physical exercise during leisure time was classified into three categories: ! 1 day/month, 1 to 3 days/ month, and > 1 day/week. Self-reported stress was assessed by the item, ‘‘Do you usually suffer from stress (little, average, much)?’’ Follow-up and Identification of Suicide All registered subjects were followed from January 1 of the year of the baseline questionnaire survey until December 31, 1999 (Cohort I) and December 31, 2000 (Cohort II). Migration data were obtained annually from the residential registry. Among the questionnaire respondents at baseline, 1966 (4.3%) moved out of the study area and 36 subjects (0.08%) were lost to follow-up within the follow-up period. When subjects remained in their original area, data on their deaths were based on death certificates obtained from their PHC. For subjects who died after they had moved from their original PHC area, the cause of death was determined from death certificates kept by the Ministry of Health, Labor, and Welfare, Japan. Death from suicide was defined according to the International Classification of Diseases, 10th Revision (ICD–10), as codes X60–X84. Statistical Analysis Subjects with incomplete smoking data were excluded from the present analyses (243 men [0.5%], including one suicide case). Person-years of follow-up were calculated from baseline until the date of suicide, the most recent date when the survival status had been confirmed in censored cases, or the end of the study periods (December 31, 1999, for Cohort I and December 31, 2000, for Cohort II) for survivors, whichever occurred first. Suicide rates were calculated using person-years as the denominator and were standardized using the 5-year age distribution at baseline in each cohort. For each cohort, the Cox proportional hazards model was used to estimate the relative risk (RR) and 95% confidence interval (CI) of suicide according to smoking habits at baseline using the SAS program (PROC PHREG) (SAS Institute Inc., Cary, NC). Linear trends were tested using the Cox proportional hazards model by treating the categories as ordinal variables. All p-values reported are two-sided, and the significance level was set at p ! 0.05.
288
Iwasaki et al. CIGARETTE SMOKING AND SUICIDE IN JAPANESE MEN
AEP Vol. 15, No. 4 April 2005: 286–292
TABLE 1. Baseline characteristics of subjects according to smoking status Cohort I
Cohort II
Smoking status
Smoking status
Characteristics
Never
Former
Current
Never
Former
Current
No. of subjects Age, mean (years) Medication, any medicine (%) Body mass index, mean (kg/m2) Alcohol drinking habits, regular drinkers (at least 1 day/week) (%) Sports and physical exercise during leisure time, >1 day/week (%) Living with spouse, yes (%) Self-reported stress, stressful (%)
4948 49.6 21.9 23.9 59.6 19.1 81.7 23.3
4683 50.4 29.6 23.9 69.5 21.1 85.9 25.9
10943 49.1 20.8 23.2 72.1 15.2 80.1 26.6
6034 54.3 29.8 24.0 62.6 21.8 88.7 17.1
5944 56.8 42.2 23.7 66.7 23.9 92.2 17.4
12657 53.2 27.9 23.1 70.4 16.0 87.8 20.6
The random-effects model developed by DerSimonian and Laird was used to combine the loge RRs; the cohortspecific RRs were weighted by the inverse of their variance (22). Heterogeneity among the cohorts was examined using the asymptotic DerSimonian and Laird Q statistics (22). In the present study, only pooled estimates are given because a similar overall risk trend was observed in the two cohorts (data not shown).
RESULTS Among the 382,390 person-years of observation in 45,209 subjects (average follow-up period: 8.5 years), 173 suicides occurred. Table 1 shows the baseline characteristics according to smoking status. The proportions of current and former smokers were 53.2% and 22.8% for Cohort I and 51.4% and 24.1% for Cohort II, respectively. Current TABLE 2. Relative risk (RR) and 95% confidence interval (95% CI) of suicide according to smoking status Smoking status
Cohort I No. of suicides Person-years Age standardized suicide ratea Cohort II No. of suicides Person-years Age standardized suicide ratea Pooled Multivariate RRb (95% CI) Multivariate RRc (95% CI) a
Never
Former
Current
22 48143 46.0
15 45427 33.1
63 105458 59.7
14 45588 30.4
14 44011 27.8
45 93765 48.3
1.0 1.0
0.8 (0.5, 1.4) 0.7 (0.4, 1.2)
1.4 (0.97, 2.1) 1.3 (0.9, 2.0)
Suicide rate (per 100,000 person-years) standardized by distribution of 5-year age groups at baseline in each cohort. Adjusted for age at baseline (continuous), area (PHCs), past history of any chronic diseases (no, yes), medication (none, yes), body mass index (quartile), physical exercise during leisure time (!1 day/month, 1–3 days/month, >1 day/week), and living with spouse (no, yes). c Further adjusted for alcohol drinking (non-drinkers, occasional drinkers, four groups of regular drinkers according to quartile of ethanol intake [g/week] [0–137, 138–251, 252–413, 414C]) and self-reported stress (little, average, much). b
smokers were less likely to perform sports and physical exercise and more likely to drink alcohol regularly and report stress than never and former smokers in both cohorts. Former smokers were more likely to be receiving medication and living with spouses than never and current smokers. The associations between subsequent suicide and smoking status at baseline, years of cessation among past smokers, and pack-years among current smokers are shown in Tables 2 and 3. Compared to never smokers, current smokers had a marginally higher risk of suicide (Table 2). We failed to observe a statistically significant reduction in suicide risk after smoking cessation, although the estimated pooled RR was lower among former smokers who had quit smoking more than 5 years earlier (Table 3). Current smokers with more than 60 pack-years of cigarette smoking had a more than two-fold higher risk of suicide than never smokers, and the risk of suicide increased with increasing pack-years, although this trend was not statistically significant (Table 3). The associations between suicide and smoking status or pack-years among current smokers were somewhat attenuated by further adjustments for alcohol drinking and selfreported stress. To further assess the suicide risk of current smokers, variables like duration of smoking, the number of cigarettes smoked per day, and the age at the start of smoking were examined (Table 4). No significant association was observed between the duration of smoking and suicide. Although current smokers who smoked more than 40 cigarettes per day had a significantly increased risk of suicide (pooled multivariate RR Z 1.9; 95% CI, 1.04, 3.4), the RR was lowered slightly and the statistical significance disappeared when further adjustments for alcohol drinking and selfreported stress were made. The risk of suicide increased significantly as the number of cigarettes smoked per day increased, even after adjustments for potential confounders, like the duration of smoking, and further adjustments for alcohol drinking and self-reported stress. A significant association was not observed between the age at the start of smoking and suicide.
AEP Vol. 15, No. 4 April 2005: 286–292
Iwasaki et al. CIGARETTE SMOKING AND SUICIDE IN JAPANESE MEN
289
TABLE 3. Pooled multivariate relative risk (RR) and 95% confidence interval (95% CI) of suicide according to years of smoking cessation among former smokers and pack-years among current smokers, compared with never smokers Years of cessation
No. of suicides Multivariate RRa (95% CI) Multivariate RRb (95% CI)
Never smoker
Years !5
5 < Years
p for trend among former smokers
36 1.0 1.0
12 1.2 (0.6, 2.3) 1.0 (0.5, 2.1)
17 0.7 (0.4, 1.3) 0.6 (0.3, 1.2)
0.19 0.19
Pack-years
No. of suicides Multivariate RRa (95% CI) Multivariate RRb (95% CI)
Never smoker
Pack-years!30
30
60
p for trend among current smokers
36 1.0 1.0
45 1.3 (0.8, 2.1) 1.3 (0.8, 2.0)
49 1.4 (0.9, 2.3) 1.3 (0.8, 2.1)
12 2.3 (1.2, 4.5) 2.1 (1.1, 4.0)
0.09 0.16
a Adjusted for age at baseline (continuous), area (PHCs), past history of any chronic diseases (no, yes), medication (none, yes), body mass index (quartile), physical exercise during leisure time (!1 day/month, 1–3 days/month, >1 day/week), and living with spouse (no, yes). b Further adjusted for alcohol drinking (non-drinkers, occasional drinkers, four groups of regular drinkers according to quartile of ethanol intake [g/week] [0–137, 138–251, 252– 413, 414C]) and self-reported stress (little, average, much).
To eliminate the effects of preclinical diseases, further analyses were performed after excluding suicides and censored cases that occurred during the first 2 years of the follow-up period. The associations between subsequent suicide and pack-years among current smokers and the number of cigarettes smoked per day were somewhat attenuated, and the statistical significance disappeared (data not shown). On the other hand, these associations were slightly strengthened when subjects with a past history of chronic diseases were excluded from the analysis (data not shown).
DISCUSSION Four main explanations for the smoking–suicide connection have been proposed (1, 2, 7): 1) depression (a well-known risk factor for suicide) can lead to smoking as a form of selfmedication (23–25); 2) smoking alters brain chemistry (26, 27), leading to depression (28–30); 3) smoking leads to malignant diseases, such as cancer, which increase the risk of suicide (31); and 4) smoking and suicide are not causally related (12, 14). However, the mechanism underlying the smoking–suicide connection remains unclear.
TABLE 4. Pooled multivariate relative risk (RR) and 95% confidence interval (95% CI) of suicide according to duration of smoking, number of cigarettes smoked per day, and age at the start of smoking among current smokers Duration of smoking (years)
No. of suicides Multivariate RRa (95% CI) Multivariate RRb (95% CI)
!25
25–29
30–34
35C
p for trend
32 1.0 1.0
23 1.0 (0.6, 1.9) 1.0 (0.6, 1.9)
22 1.5 (0.7, 3.2) 1.5 (0.7, 3.4)
30 1.4 (0.5, 3.9) 1.4 (0.5, 4.0)
0.52 0.51
!20
20–29
30–39
40C
p for trend
25 1.0 1.0
41 0.9 (0.5, 1.5) 0.9 (0.5, 1.5)
20 1.4 (0.8, 2.6) 1.4 (0.8, 2.6)
21 1.9 (1.04, 3.4) 1.7 (0.9, 3.1)
0.01 0.036
22C
19–21
!19
p for trend
19 1.0 1.0
63 1.2 (0.7, 2.1) 1.4 (0.8, 2.5)
26 1.4 (0.7, 2.6) 1.6 (0.8, 3.0)
0.32 0.18
Number of cigarettes per day
No. of suicides Multivariate RRa (95% CI) Multivariate RRb (95% CI)
Age at start of smoking
No. of suicides Multivariate RRa (95% CI) Multivariate RRb (95% CI)
a Adjusted for age at baseline (continuous), area (PHCs), past history of any chronic diseases (no, yes), medication (none, yes), body mass index (quartile), physical exercise during leisure time (!1 day/month, 1–3 days/month, >1 day/week), living with spouse (no, yes) and duration of smoking for number of cigarettes smoked, or number of cigarettes smoked for duration of smoking and age at the start of smoking. b Further adjusted for alcohol drinking (non-drinkers, occasional drinkers, four groups of regular drinkers according to quartile of ethanol intake [g/week] [0–137, 138–251, 252– 413, 414C]) and self-reported stress (little, average, much).
290
Iwasaki et al. CIGARETTE SMOKING AND SUICIDE IN JAPANESE MEN
The present study showed that current smokers with more than 60 pack-years of cigarettes had a significantly higher risk of suicide than that of never smokers. Among current smokers, the suicide risk increased as the number of cigarettes smoked per day at baseline increased. This association was statistically significant even after adjusting for potential confounders like the duration of smoking, although most previous studies (1, 2, 4, 6–8, 10, 12) did not control for the duration of smoking. However, the duration of smoking was not significantly associated with the suicide risk in the present study. Our results indicate that heavy smokers, rather than long-term smokers, have a significantly increased risk of suicide. To examine the possibility that the suicide cases had already changed the number of cigarettes smoked per day at baseline as a result of existing factors that could also be related to an increased risk of suicide, such as a specific life event, mental disorder or so on, a preliminary analysis was performed using data from the 5-year follow-up surveys conducted in 1995 for Cohort I and in between 1998 and 1999 for Cohort II. We calculated the change in number of cigarettes smoked per day between the baseline and the 5year follow-up surveys. This change was compared between the suicide cases that occurred after the 5-year follow-up survey and the other non-suicide cases among the current smokers. As the results showed no difference, the subjects who committed suicide during the follow-up period probably did not change the number of cigarettes smoked per day because of factors that could potentially be related to their subsequent suicide. The risk of suicide for current smokers did not reach a significant level in this study, even though a previous study showed a statistically significant increased risk of suicide for current smokers after adjusting for psychological factors, such as symptoms of depressed mood, stress, and anxiety (4). One possible reason for the lack of statistical significance may be misclassification caused by subsequent changes in smoking status as a result of illness or some other events during the long follow-up period. Smoking cessation after the baseline survey had been conducted could lead to an underestimation of the true risk among current smokers who continue to smoke, since the suicide risk decreased as the length of smoking cessation increased. Additionally, since a positive relation was previously observed between indicators of nicotine addiction and the number of cigarettes smoked per day (32) and nicotine dependence has been associated with a risk of major depression (33), the number of cigarettes smoked per day, rather than the duration of smoking, might be more strongly correlated with mental disorders, such as depression, or one or more factors predisposing the subject to mental disorders that increase the risk of suicide (13). Since smokers who are
AEP Vol. 15, No. 4 April 2005: 286–292
less dependent on nicotine are more likely to quit smoking, the proportion of nicotine dependence is probably becoming high among current smokers as the prevalence of current smokers decreases (23). Nevertheless, since smoking used to be socially accepted in Japan and the prevalence of current smokers is still high (19), several non-nicotine dependent subjects who are less likely to suffer from depression than nicotine dependent subjects (34, 35) probably exist among the current smokers in this study. Such a situation may have attenuated the association between smoking status and suicide, possibly explaining the lack of statistical significance. Our findings showed that the associations between subsequent suicide and the pack-years of cigarette smoking and the number of cigarettes smoked per day were somewhat attenuated and the statistical significance disappeared when the suicides and censored cases that occurred during the first 2 years of the follow-up period were excluded. Meanwhile, the statistical significance was preserved when subjects with a past history of chronic diseases were excluded. The lack of statistical significance might simply reflect the reduction in the number of cases, or a lack of statistical power. Alternatively, the exclusion of the suicides and censored cases that occurred during the first 2 years of the follow-up period may have served as a control for the effects of mental illness on the risk of suicide to some extent. The relationships among smoking, mental illness (especially depression), and suicide remain unclear. One study concluded that the association between smoking and suicide accounted for confounders, such as mental well-being (14). On the other hand, adjustments for mental illness or factors related to mental illness would lead to inappropriate results, if these factors were involved in the causal pathway between smoking and suicide (36). Our results were not substantially influenced by adjustments for alcohol drinking and selfreported stress. Further analysis to clarify the possible role of depression on the relationship between smoking and suicide is needed to determine the smoking-suicide connection. However, data on mental conditions, such as depression, were not available in the JPHC study. A previous study suggested a possible protective effect of smoking cessation on the risk of suicide (37). Our results showed that former smokers had the same risk of suicide as never smokers, regardless of the length of smoking cessation. This implies that smoking cessation may be effective for reducing the risk of suicide. However, further research is needed to clarify the possible protective effect of smoking cessation on the risk of suicide because presently available evidence is limited. To our knowledge, this is the first study to examine the effects of age at the start of smoking on the risk of suicide. According to previous studies in adolescents, depression and anxiety were associated with a higher risk of smoking initiation (25), while smoking was also associated with an
AEP Vol. 15, No. 4 April 2005: 286–292
increased risk of the subsequent depressed mood (30). Moreover, smokers who started before the age of 13 years had significantly more depressive episodes than smokers who smoked after the age of 17 years or lifetime nonsmokers (38). These previous studies imply that the age at which smoking is begun is related to mental disorders in adolescents, and vice versa. However, our results showed no association between the start of smoking at an early age and the risk of suicide. One possible reason is that the mean age at which smoking was started was older among Japanese than among Americans (39). In this cohort, the mean ages at which smoking was started among current smokers were 20.3, 21.3, and 21.6 years for men aged 40 to 49, 50 to 59, and 60 to 69 years at baseline, respectively; 90% of the subjects started smoking between the ages of 17 and 28 years. In the Cancer Prevention Study II, the corresponding figures were 17.2, 17.6, and 17.8 years (39). Therefore, the number of subjects who started smoking under the age of 16 years may not have been large enough to establish an association. A second possible reason is that more than 80% of Japanese men smoked cigarettes when the subjects in the JPHC study started smoking (e.g., 82.3% in 1965) (19), and cigarette smoking was a popular and socially-accepted habit in Japan, suggesting that the association between cigarette smoking and mental disorders might be weakened. Our study has several methodological advantages. First, the prospective design enables potential selection or recall biases to be avoided. Second, the study included a large population-based sample with a high response rate (approximately 80%) and a very low rate of loss to follow-up. Third, similar findings were obtained in two different cohorts. On the other hand, the generalizability of this study was a methodological disadvantage. Our findings may not be applicable to urban Japanese populations, to subjects from different generations, or to women, since this study did not include metropolitan areas and the subjects were middleaged Japanese men. Allowing for these limitations, our study showed that heavy cigarette smoking was related to an increased risk of suicide among middle-aged Japanese men. Since progress in suicide prevention has been disappointing (40), useful clinical markers for identifying people with a high risk of suicide are needed. Although psychiatric disorders, especially depression, are a well-known risk factor for suicide, such conditions are often unrecognized, even in medical settings (41). Therefore, smoking, particularly the number of cigarettes smoked per day, may be a useful clinical indicator of a risk for subsequent suicide. In conclusion, our study suggested that cigarette smoking, or at least heavy cigarette smoking, was associated with a risk of suicide that was independent of a past history of chronic disease, heavy alcohol drinking, infrequent exercise, self-reported stress, and an unmarried status, although
Iwasaki et al. CIGARETTE SMOKING AND SUICIDE IN JAPANESE MEN
291
the mechanism underlying the association between cigarette smoking and the risk of suicide was not addressed. Future studies are needed to clarify the associations among smoking, depression, and suicide. We thank all staff members in each study area and in the central offices for their cooperation and technical assistance. We also thank Dr. T. Otani of the Epidemiology and Prevention Division, Research Center for Cancer Prevention and Screening, National Cancer Center, Tokyo, Japan for his helpful comments.
Study group members Members of the JPHC Study Group (principal investigator: S. Tsugane): M. Inoue, T. Hanaoka, S. Tsugane, S. Yamamoto, T. Sobue, Research Center for Cancer Prevention and Screening, National Cancer Center, Tokyo; J. Ogata, S. Baba, T. Mannami, National Cardiovascular Center, Suita; K. Miyakawa, F. Saito, A. Koizumi, Y. Sano, I. Hashimoto, Iwate Prefectural Ninohe Public Health Center, Ninohe; Y. Miyajima, N. Suzuki, S. Nagasawa, Y. Furusugi, Akita Prefectural Yokote Public Health Center, Yokote; H. Sanada, Y. Hatayama, F. Kobayashi, H. Uchino, Y. Shirai, T. Kondo, R. Sasaki, Y. Watanabe, Nagano Prefectural Saku Public Health Center, Saku; Y. Kishimoto, E. Takara, T. Fukuyama, M. Kinjo, M. Irei, Okinawa Prefectural Ishikawa (Chubu) Public Health Center, Ishikawa; K. Imoto, H. Yazawa, T. Seo, A. Seiko, F. Ito, Katsushika Public Health Center, Tokyo; A. Murata, K. Minato, K. Motegi, T. Fujieda, Ibaraki Prefectural Kasama (Mito) Public Health Center, Mito; K. Matsui, T. Abe, Niigata Prefectural Kashiwazaki Public Health Center, Kashiwazaki; M. Doi, A. Terao, Y. Ishikawa, Kochi Prefectural Tosayamada (Chuo-higashi) Public Health Center, Tosayamada; H. Sueta, H. Doi, M. Urata, N. Okamoto, F. Ide, Nagasaki Prefectural Arikawa (Kamigoto) Public Health Center, Arikawa; H. Sakiyama, N. Onga, H. Takaesu, Okinawa Prefectural Miyako Public Health Center, Hirara; F. Horii, I. Asano, H. Yamaguchi, K. Aoki, S. Maruyama, M. Ichii, Osaka Prefectural Suita Public Health Center, Suita; S. Matsushima, S. Natsukawa, Saku General Hospital, Usuda; S. Watanabe, M. Akabane, Tokyo University of Agriculture, Tokyo; M. Konishi, K. Okada, Ehime University, Matsuyama; H. Iso, Y. Honda, Tsukuba University, Tsukuba; H. Sugimura, Hamamatsu University, Hamamatsu; Y. Tsubono, Tohoku University, Sendai; M. Kabuto, National Institute for Environmental Studies, Tsukuba; S. Tominaga, Aichi Cancer Center Research Institute, Nagoya; M. Iida, W. Ajiki, Osaka Medical Center for Cancer and Cardiovascular Disease, Osaka; S. Sato, Osaka Medical Center for Health Science and Promotion, Osaka; N. Yasuda, Kochi Medical School Nankoku; S. Kono, Kyushu University, Fukuoka; K. Suzuki, Research
292
Iwasaki et al. CIGARETTE SMOKING AND SUICIDE IN JAPANESE MEN
Institute for Brain and Blood Vessels Akita, Akita; Y. Takashima, Kyorin University, Mitaka; E. Maruyama, Kobe University, Kobe; the late M. Yamaguchi, Y. Matsumura, S. Sasaki, National Institute of Health and Nutrition, Tokyo; and T. Kadowaki, Tokyo University, Tokyo. REFERENCES 1. Miller M, Hemenway D, Bell NS, Yore MM, Amoroso PJ. Cigarette smoking and suicide: A prospective study of 300,000 male active-duty Army soldiers. Am J Epidemiol. 2000;151:1060–1063. 2. Miller M, Hemenway D, Rimm E. Cigarettes and suicide: A prospective study of 50,000 men. Am J Public Health. 2000;90:768–773. 3. Leistikow BN, Martin DC, Samuels SJ. Injury death excesses in smokers: A 1990-95 United States national cohort study. Inj Prev. 2000;6:277–280. 4. Tanskanen A, Tuomilehto J, Viinamaki H, Vartiainen E, Lehtonen J, Puska P. Smoking and the risk of suicide. Acta Psychiatr Scand. 2000;101: 243–245. 5. Tanskanen A, Tuomilehto J, Viinamaki H, Vartiainen E, Lehtonen J, Puska P. Joint heavy use of alcohol, cigarettes, and coffee and the risk of suicide. Addiction. 2000;95:1699–1704. 6. Doll R, Peto R, Wheatley K, Gray R, Sutherland I. Mortality in relation to smoking: 40 years’ observations on male British doctors. BMJ. 1994;309: 901–911. 7. Hemenway D, Solnick SJ, Colditz GA. Smoking and suicide among nurses. Am J Public Health. 1993;83:249–251. 8. Tverdal A, Thelle D, Stensvold I, Leren P, Bjartveit K. Mortality in relation to smoking history: 13 years’ follow-up of 68,000 Norwegian men and women 35–49 years. J Clin Epidemiol. 1993;46:475–487. 9. Doll R, Gray R, Hafner B, Peto R. Mortality in relation to smoking: 22 years’ observations on female British doctors. BMJ. 1980;280:967–971. 10. Friberg L, Cederlof R, Lorich U, Lundman T, De Faire U. Mortality in twins in relation to smoking habits and alcohol problems. Arch Environ Health. 1973;27:294–304. 11. Paffenbarger RS Jr, King SH, Wing AL. Chronic disease in former college students. IX. Characteristics in youth that predispose to suicide and accidental death in later life. Am J Public Health. 1969;59:900–908. 12. Smith GD, Phillips AN, Neaton JD. Smoking as ‘‘independent’’ risk factor for suicide: Illustration of an artifact from observational epidemiology? Lancet. 1992;340:709–712. 13. Harris EC, Barraclough B. Suicide as an outcome for mental disorders. A meta-analysis. Br J Psychiatry. 1997;170:205–228. 14. Hemmingsson T, Kriebel D. Smoking at age 18–20 and suicide during 26 years of follow-up: How can the association be explained? Int J Epidemiol. 2003;32:1000–1004. 15. Forey B, Hamling J, Lee P, Wald N. International smoking statistics: A collection of historical data from 30 economically developed countries. 2nd ed. New York: Oxford University Press; 2002. 16. Statistics and Information Department, Minister’s Secretariat, Ministry of Health, Labor, and Welfare. Special Report of Vital Statistics on Suicide Mortality Statistics [in Japanese]. Tokyo: Health and Welfare Statistics Association; 1999. 17. Domino G, Takahashi Y. Attitudes toward suicide in Japanese and American medical students. Suicide Life Threat Behav. 1991;21:345–359. 18. Statistics and Information Department, Minister’s Secretariat, Ministry of Health, Labor, and Welfare. Vital Statistics of Japan [in Japanese]. Tokyo: Health and Welfare Statistics Association; 2002. 19. Ministry of Health, Labor, and Welfare. Smoking and Health. New edition. Report from the Commission on Smoking and Health, Japan [in Japanese]. Tokyo: Hoken Dojin sha; 2000. 20. Lamar J. Suicides in Japan reach a record high. BMJ. 2000;321:528.
AEP Vol. 15, No. 4 April 2005: 286–292
21. Watanabe S, Tsugane S, Sobue T, Konishi M, Baba S. Study design and organization of the JPHC study. J Epidemiol. 2001;11:S3–S7. 22. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials. 1986;7:177–188. 23. Hughes JR. Clonidine, depression, and smoking cessation. JAMA. 1988; 259:2901–2902. 24. Glassman AH. Cigarette smoking: Implications for psychiatric illness. Am J Psychiatry. 1993;150:546–553. 25. Patton GC, Carlin JB, Coffey C, Wolfe R, Hibbert M, Bowes G. Depression, anxiety, and smoking initiation: A prospective study over 3 years. Am J Public Health. 1998;88:1518–1522. 26. Malone KM, Waternaux C, Haas GL, Cooper TB, Li S, Mann JJ. Cigarette smoking, suicidal behavior, and serotonin function in major psychiatric disorders. Am J Psychiatry. 2003;160:773–779. 27. Kirch DG, Gerhardt GA, Shelton RC, Freedman R, Wyatt RJ. Effect of chronic nicotine administration on monoamine and monoamine metabolite concentrations in rat brain. Clin Neuropharmacol. 1987;10:376–383. 28. Newhouse PA, Sunderland T, Tariot PN, Blumhardt CL, Weingartner H, Mellow A, et al. Intravenous nicotine in Alzheimer’s disease: A pilot study. Psychopharmacology (Berl.) 1988;95:171–175. 29. Green BH, Copeland JR, Dewey ME, Sharma V, Saunders PA, Davidson IA, et al. Risk factors for depression in elderly people: A prospective study. Acta Psychiatr Scand. 1992;86:213–217. 30. Wu LT, Anthony JC. Tobacco smoking and depressed mood in late childhood and early adolescence. Am J Public Health. 1999;89:1837–1840. 31. Tanaka H, Tsukuma H, Masaoka T, Ajiki W, Koyama Y, Kinoshita N, et al. Suicide risk among cancer patients: Experience at one medical center in Japan, 1978–1994. Jpn J Cancer Res. 1999;90:812–817. 32. Centers for Disease Control and Prevention. Indicators of nicotine addiction among womendUnited States, 1991–1992. JAMA. 1995;273:842. 33. Breslau N, Kilbey MM, Andreski P. Nicotine dependence and major depression. New evidence from a prospective investigation. Arch Gen Psychiatry. 1993;50:31–35. 34. Breslau N, Kilbey M, Andreski P. Nicotine dependence, major depression, and anxiety in young adults. Arch Gen Psychiatry. 1991;48:1069– 1074. 35. Schmitz N, Kruse J, Kugler J. Disabilities, quality of life, and mental disorders associated with smoking and nicotine dependence. Am J Psychiatry. 2003;160:1670–1676. 36. Leistikow B. Commentary: Questionable premises, overadjustment, and a smoking/suicide association in younger adult men. Int J Epidemiol. 2003;32:1005–1006. 37. Leistikow BN, Shipley MJ. Might stopping smoking reduce injury death risks? A meta-analysis of randomized, controlled trials. Prev Med. 1999;28:255–259. 38. Hanna EZ, Grant BF. Parallels to early onset alcohol use in the relationship of early onset smoking with drug use and DSM-IV drug and depressive disorders: Findings from the National Longitudinal Epidemiologic Survey. Alcohol Clin Exp Res. 1999;23:513–522. 39. Thun MJ, Day-Lally C, Myers DG, Calle EE, Flanders WD, Zhu BP, et al. Trends in tobacco smoking and mortality from cigarette use in Cancer Prevention Studies I (1959 through 1965) and II (1982 through 1988). In: Shopland DR, Burns DM, Garfinkel L, Samet J, eds. National Cancer Institute, Smoking and Tobacco Control, Monograph 8: Changes in Cigarette-related Disease Risks and their Implication for Prevention and Control. NIH Publication No. 97-4213. Washington, DC: National Institutes of Health; 1997; 305–382. 40. Gunnell D, Frankel S. Prevention of suicide: Aspirations and evidence. BMJ. 1994;308:1227–1233. 41. Pignone MP, Gaynes BN, Rushton JL, Burchell CM, Orleans CT, Mulrow CD, et al. Screening for depression in adults: A summary of the evidence for the US Preventive Services Task Force. Ann Intern Med. 2002;136: 765–776.