- Email: [email protected]
Suicide risk after cancer diagnosis among older adults: A nationwide retrospective cohort study
JGO-00865; No. of pages: 6; 4C: Journal of Geriatric Oncology xxx (2019) xxx
Contents lists available at ScienceDirect
Journal of Geriatric Oncology
Suicide risk after cancer diagnosis among older adults: A nationwide retrospective cohort study Jae Woo Choi a, Eun-Cheol Park b,⁎ a b
College of Pharmacy, Yonsei Institute of Pharmaceutical Sciences, Yonsei University, Incheon, South Korea Department of Preventive Medicine, Yonsei University College of Medicine, Seoul, South Korea
a r t i c l e
i n f o
Article history: Received 13 August 2019 Received in revised form 9 September 2019 Accepted 20 November 2019 Available online xxxx Keywords: Cancer Suicide risk Prediagnosis mental disorders Older adults
a b s t r a c t Objectives: Cancer increases suicide risk, but little is known about associations between cancer and suicide among older adults. We aimed to investigate suicide risk within 1 year after cancer diagnosis in older adults. Methods: Using National Health Insurance Service-Senior Cohort data, we included newly diagnosed older patients with cancer and older adults without cancer, selected by a 1:3 propensity score matching using sex, age, Charlson comorbidity index, and index year from 2004 to 2012 and followed them up throughout 2013. We used Cox proportional hazard models to estimate adjusted hazard ratios (AHR) of suicide risk. Results: In the total sample of 259,688 older adults (aged 62–115 years), the highest proportion of observed suicide deaths occurred within one year after cancer diagnosis (36.1% of total); 64,922 older patients with cancer showed higher suicide risk after cancer diagnosis compared to non-cancer participants (AHR 2.05; 95% CI 1.64–2.56). Patients with mental disorder diagnosis before cancer diagnosis (AHR 2.98; 95% CI 2.12–4.18) had increased suicide risk than those without mental disorder diagnosis (AHR 1.78; 95% CI 1.38–2.29) compared to non-cancer participants. Suicide risk among older patients with bladder, head and neck, liver, lung, and stomach cancers was higher than in non-cancer participants (AHR 4.77, 2.28, 2.99, 1.98, 2.40; 95% CI 2.53–8.99, 1.47–3.54, 1.63–5.49, 1.15–3.40, 1.57–2.24), respectively. Conclusions: Older patients with cancer have a higher suicide risk. Early mental health support needs to be provided with consideration of prediagnosis mental disorders during follow-up after cancer diagnosis. © 2019 Elsevier Ltd. All rights reserved.
1. Introduction Suicide accounts for 1.4% of all deaths worldwide, making it the 18th leading cause of death in 2016 [1]. South Korea is ranked 10th globally for suicide rate in 2016 (20.2 per 100,000 people) and the suicide rate among older adults (60 years old and older) was 53.3 per 100,000 people in 2016, which was the highest among Organization for Economic Cooperation and Development countries [2,3]. Given the higher proportion of the aging population (14.3% of total population in 2018) than that of the worldwide population (8.2%), and its rapidly growing rate (average 4.1% annually) in South Korea [4], suicide prevention among older adults has become a significant public health issue. Previous studies have reported risk factors associated with suicide risk among older adults that include male sex [5], poverty [5], education level [6], physical activity [7], interpersonal trust [8], social support [9], discrimination [10], status loss [11], elder mistreatment [12], mental illness such as depression [13,14], functional disability [15].
⁎ Corresponding author at: Division of Medical Science Research Affairs & President, University-Industry Foundation of Yonsei University Health System, South Korea. E-mail address: [email protected] (E.-C. Park).
The risk of suicide can also be greatly increased by the occurrence of major medical conditions in adulthood [16] and previous research has showed that patients with severe illnesses such as cancer, cardiovascular disease, and cerebrovascular disease are at increased suicide risk after diagnosis [17–19]. Among these, suicide risk after cancer diagnosis has been well established in many countries including the USA [20–22], England [18], Norway [23], Lithuania [24], Australia [25], Korea [26,27], Taiwan [28], Japan [29], Sweden [30], and Italy [31]. However, little is known about suicide risk following cancer diagnosis among older adults. A previous case control study showed that the risk of suicide in older adults was higher among patients with cancer than among patients with other medical illnesses [32]. Another case-control study based on interviews of primary informants for suicides showed that there was a statistically significant association between suicide and cancer among older adults [33]. However, these studies included patients who had been diagnosed with cancer a relatively long time ago and their suicide death may not have been directly caused by cancer diagnosis. It is necessary to study the risk of suicide within short periods after diagnosis of cancer in older adults, considering that the first leading cause of death among older adults in South Korea is cancer [34], as well as the increasing older adult population and high older adults suicide rate. Previous
https://doi.org/10.1016/j.jgo.2019.11.006 1879-4068/© 2019 Elsevier Ltd. All rights reserved.
Please cite this article as: J.W. Choi and E.-C. Park, Suicide risk after cancer diagnosis among older adults: A nationwide retrospective cohort study, J Geriatr Oncol, https://doi.org/10.1016/j.jgo.2019.11.006
2
J.W. Choi, E.-C. Park / Journal of Geriatric Oncology xxx (2019) xxx
research has indicated highest suicide risk during the first year after cancer diagnosis [22,27,28]. Furthermore, mental disorder diagnosis is a strong risk factor for suicide death [35]. However, little is known about the suicide risk after cancer while considering the role of mental disorder diagnosis before cancer diagnosis among older adults. Therefore, we aimed to investigate suicide risk within 1 year after cancer diagnosis among older adults using nationally representative cohort data. We performed separate analyses for risk of suicide according to prediagnosis mental disorders of older patients with cancer. We also analyzed variations in this risk by subtypes of cancer site (head and neck, stomach, colorectal, lung, liver, and other cancers). 2. Methods 2.1. Data Source This study used the National Health Insurance Service–Senior Cohort (NHIS–SC) data, an elderly-population-based cohort provided by the Korean National Health Insurance Service. The NHIS–SC data were extracted using a random sampling method to construct a representative dataset and a representative sample of 558,147 older adults was randomly selected, which was composed of about 10% of the older adult population aged ≥60 years old (5.5 million) in 2002. The NHIS–SC data was followed for 12 years until 2013, unless the participant was disqualified because of immigration or death. Information on each individual's cause of death was extracted from Statistics Korea and was linked with our database. The NHIS–SC data includes general demographics and clinical information for diagnoses, treatments, and prescribed medications for all visits to medical institutions. 2.2. Study Sample Among 528,655 participants from NHIS–SC between 2004 and 2012, we included outpatients in NHIS–SC (n = 526,278), excluded older adults without cancer (n = 447,725), and selected outpatients with a principal diagnosis of cancer (n = 78,553). We defined patients with cancer as patients diagnosed with a malignant tumor (International Classification of Diseases, 10th Revision [ICD-10] codes C00–C97, excluding C44, nonmelanoma skin cancer) from January 1, 2004, to December 31, 2012. We selected older patients with a first principal diagnosis of cancer and the first diagnosis date of cancer was defined as the index date in this study. A previous study reported that distinguishable peaks for cancer recurrence occurred in the first couple of years after diagnosis [36]; thus, we utilized the first two years of the database (2002−2003) as a washout period to exclude all patients with cancer with possible relapse or recurrence within the two years. Close radiographic follow-up after curative resection is also a widely recommended practice, which is executed more frequently within the first two years following diagnosis or treatment [37], and we selected the final study subjects with newly diagnosed cancer (n = 64,922). For comparison of suicide risk among the group without cancer, we selected older adults without cancer (n = 450,102) and performed a 1:3 propensity score matching using sex, age, Charlson Comorbidity Index, and index year as variables. The participants were divided into youngold (60–74 years old) and old-old (N 75 years old) adults, based on previous studies [38,39]. The Charlson comorbidity index corresponds to the sum of the weighted scores assigned to various major health conditions [40], and was calculated by screening the one-year period preceding the participants' index year. The first diagnosis year of cancer in the study period was defined as the index year and we matched each comparison group in the same index year of patients with cancer. After calculating the estimated propensity score, patients with cancer were matched with patients without cancer who had the same or similar propensity scores. Finally, older patients with cancer (n = 64,922) and older patients without cancer (n = 194,766) were selected as the final sample (supplementary Fig. 1).
2.3. Measurements The dependent variable was death by suicide within 1 year after the index date and we defined death as suicide if the code for cause of death was X60-X84 (intentional self-harm) based on the ICD-10. The length of survival was calculated in days, and all participants enrolled in the study were followed up until death by suicide, withdrawal from the insurance system, death from other causes, or the end of 2013, whichever occurred earlier. Potential confounding factors were residential area, household income, insurance type, disability, and history of mental disorder diagnosis. Information for confounding factors was based on the participants' index date. The residential region was classified as metropolitan (capital), urban (local government where N1 million people live), or rural (otherwise). Household income was classified as (1) high (81st–100th percentile), (2) middle (41st–80th percentile), or (3) low income (covered by Medical Aid and below the 40th percentile). The insurance type was either National Health Insurance or Medical Aid, which are public medical support programs assisting poor individuals by a proportional allocation. The prediagnosis mental disorders were defined by the occurrence of a principal diagnosis of mental disorder preceding the index date (ICD-10: F00–F99). 2.4. Statistical Analysis The baseline characteristics of study subjects were summarized by cancer type. The statistical significance of the characteristic differences between patients with and without cancer was tested by Pearson's χ2 test. We used Cox regression to assess the suicide risk within 1 year after the index date. We reported results as estimates of adjusted hazard ratio (AHR) with a 95% confidence interval (CI). Person years were calculated for each group, and the starting point was the index date. The index date of older patients with cancer was first diagnosis date of cancer in the study period and that of matched older patients without cancer was allocated by the same index date as each matched older patient with cancer. In the survival analysis, events other than suicide within 1 year after the index date were censored. First, we estimated the suicide risk among older patients with cancer compared to the non-cancer group. Next, we performed separate analyses for suicide risk according to prediagnosis mental disorders in older patients with cancer. Finally, we conducted separate analyses for the most common cancer occurrence sites/groups to explore suicide risk by cancer type compared to the non-cancer group. All data-extraction and statistical analyses were performed using the SAS 9.4 program, and the institutional review board of the study site exempted this study from the requirement of ethical approval. 3. Results Fig. 1 shows the number of suicide deaths per period after cancer among older adults. Among 64,570 patients with cancer, 263 male subjects and 89 female subjects died by suicide over the entire follow-up period. The median time from cancer diagnosis to suicide was 1.74 years (range, 0.01–9.46 years), and the suicide rate decreased progressively with time after cancer diagnosis. One hundred twenty three suicides (34.9% of the total) occurred during one year after cancer, 66 (18.8%) during the second year, 48 (13.6%) during the third year, 32 (9.1%) during the fourth year, 26 (7.4%) during the fifth year, 20 (5.7%) during the sixth year, and 37 (10.5%) after the sixth year. The study subjects' baseline characteristics (2004–2012) are presented in Table 1. Before propensity score matching, there were significant differences in all confounding factors between the older patients with and without cancer. After matching, about 56% of the subjects were male, and 60.7% and 39.3% were young–old and old–old adults, respectively, in both the older patients with cancer and matched
Please cite this article as: J.W. Choi and E.-C. Park, Suicide risk after cancer diagnosis among older adults: A nationwide retrospective cohort study, J Geriatr Oncol, https://doi.org/10.1016/j.jgo.2019.11.006
J.W. Choi, E.-C. Park / Journal of Geriatric Oncology xxx (2019) xxx
3
Fig. 1. The number of suicide deaths per period after cancer diagnosis.
comparison groups. Approximately 60% of subjects in both groups had comorbidities. Prediagnosis mental disorders were present in 22.3% of older patients with cancer and 22.8% of older patients without cancer. The proportion of low household income was higher in older patients with cancer than in the matched comparison group (35.5% vs. 35.0%). The proportion of people receiving Medical Aid was higher among older patients with cancer than the non-cancer group (10.1% vs. 8.2%). Older patients with cancer were less likely to live in a metropolitan area compared to their non-cancer counterparts (17.6% vs. 18.1%). There was no statistically significant difference between older patients with and without cancer in having a disability. Table 2 shows the risk of suicide within 1 year after cancer diagnosis compared with the non-cancer group. Model 1 indicates that older patients with cancer were more likely to die by suicide than the noncancer group (AHR 2.05; 95% CI 1.64–2.56). In model 2, older patients with cancer and prediagnosis mental disorders (AHR 2.98; 95% CI 2.12–4.18) had an increased suicide risk than those without a history of mental disorder diagnosis before cancer diagnosis (AHR 1.78; 95% CI 1.38–2.29) compared to the non-cancer group. The AHR of suicide death among older patients with cancer was estimated according to the six most common cancer subtypes (Table 3). A higher risk of suicide was observed in bladder (AHR, 4.77; 95% CI, 2.53–8.99), head and neck (AHR, 2.28; 95% CI, 1.47–3.54), liver (AHR, 2.99; 95% CI, 1.63–5.49), lung (AHR, 1.98; 95% CI, 1.15–3.40), stomach (AHR, 2.40; 95% CI, 1.57–3.66), and other cancers (AHR, 1.50; 95% CI, 1.01–2.24) compared to the non-cancer group. No statistically significant association was observed between suicide risk and colorectal cancer (AHR, 1.56; 95% CI, 0.93–2.64).
4. Discussion There were three main findings in this study. First, among older patients with cancer, the highest proportion of observed suicide deaths occurred within 1 year after cancer diagnosis. Second, suicide risk within 1 year after cancer among older patients was higher than in the noncancer group, especially among those with a history of mental disorder diagnosis before cancer. Third, among the six most common cancer
subtypes, patients with bladder, head and neck, liver, lung, and stomach cancers had a higher suicide risk than those who did not have cancer. Among all suicide deaths associated with cancer, 123 (34.9%) suicide events occurred within 1 year of cancer diagnosis, which seems to be consistent with other reports. Wang et al. (40.8%), Ahn et al. (40.0%), and Saad et al. (31.7%) reported similar suicide rates within 1 year after cancer in adults [22,27,28]. Wang et al. revealed that suicide risk was found to be highest within one month of a new cancer diagnosis and elevated risk persisted thereafter but gradually decreased in intensity over the next 11 months. Ahn et al. showed that the standardized mortality ratio for suicide within 1 year after cancer was significantly higher than in the general population (SMR, 1.65; 95% CI, 1.40–1.94). Saad et al. implied that the risk of suicide increases significantly within the first year after a diagnosis of cancer in comparison with the general population (SMR, 2.52; 95% CI, 2.39–2.64). These findings indicate that clinicians specializing in cancer care should provide special attention to older patients with newly diagnosed cancer to reduce their risk of suicide. Although most previous studies showed an increased risk of suicide among patients with cancer, the analyses did not consider history of mental disorder diagnosis [18,20–31]. Mental disorder diagnosis is a major predictor of suicide and our findings showed that older patients with cancer and prediagnosis mental disorders had a higher suicide risk than those without mental disorder diagnosis before cancer diagnosis compared to the non-cancer group, which is consistent with previous research [41]. The previous study examined suicide risk of adults (aged ≥18 years) with cancer and among study subjects who utilized outpatient psychiatric care before cancer diagnosis, the patients with cancer were significantly more likely to die by suicide compared with non-cancer controls. The results implied that mental disorder diagnosis before cancer diagnosis is important consideration to potentially reduce the risk of suicide death among older patients with cancer. Assessing the association between specific tumor sites and suicidal risk among older patients with cancer, this research is consistent with previous studies suggesting increased suicidal risk among patients with bladder [42–44], head and neck [45], lung [46], liver [26], and stomach cancer [18,27]. Surgical treatment for head and neck cancer (extirpation with or without extensive neck lymphadenectomy) and
Please cite this article as: J.W. Choi and E.-C. Park, Suicide risk after cancer diagnosis among older adults: A nationwide retrospective cohort study, J Geriatr Oncol, https://doi.org/10.1016/j.jgo.2019.11.006
4
J.W. Choi, E.-C. Park / Journal of Geriatric Oncology xxx (2019) xxx
Table 1 Study subjects' general characteristics (2004–2012). Variables
Cancer
Total Sex Male Female Age 60–74 ≥75 Index year 2004 2005 2006 2007 2008 2009 2010 2011 2012 Charlson comorbidity index 0 1 2 ≥3 Prediagnosis mental disorders Yes No Household income Low Middle High Insurance type National health insurance Medical aid Residential area Metropolitan Urban Rural Disability Yes No a b
After propensity score matchingb
Before propensity score matching p-valuea
Non-cancer
N
%
N
%
64,922
12.6
450,102
87.4
p-valuea
Non-cancer N
%
194,766
37.8
109,794 84,972
56.4 43.6
118,131 76,635
60.7 39.3
23,613 24,932 25,005 23,747 22,404 21,624 18,309 18,157 16,975
12.1 12.8 12.8 12.2 11.5 11.1 9.4 9.3 8.7
77,886 59,927 32,314 24,639
40.0 30.8 16.6 12.7
44,449 150,317
22.8 77.2
65,785 60,878 68,103
33.8 31.3 35.0
178,796 15,970
91.8 8.2
35,314 43,989 115,463
18.1 22.6 59.3
2083 192,683
1.1 98.9
b0.001 36,598 28,324
56.4 43.6
172,922 277,180
38.4 61.6
39,377 25,545
60.7 39.3
330,207 119,895
73.4 26.6
7871 8309 8335 7914 7469 7208 6103 6053 5660
12.1 12.8 12.8 12.2 11.5 11.1 9.4 9.3 8.7
386,577 44,775 11,156 3392 1868 1091 653 345 245
85.9 9.9 2.5 0.8 0.4 0.2 0.1 0.1 0.1
25,962 19,975 10,773 8212
40.0 30.8 16.6 12.6
280,314 103,703 41,909 24,176
62.3 23.0 9.3 5.4
14,450 50,472
22.3 77.7
40,244 409,858
8.9 91.1
21,610 20,260 23,052
33.3 31.2 35.5
164,588 149,176 136,338
36.6 33.1 30.3
58,396 6526
89.9 10.1
412,065 38,037
91.5 8.5
11,448 13,921 39,553
17.6 21.4 60.9
78,523 99,864 271,715
17.4 22.2 60.4
713 64,209
1.1 98.9
3674 446,428
0.8 99.2
1.000
b0.001
1.000
b0.001
1.000
b0.001
1.000
b0.001
0.003
b0.001
0.023
b0.001
b0.001
b0.001
b0.001
b0.001
0.539
p-value b0.05 is significant. After propensity score matching for sex, age, Charlson Comorbidity Index, and year of cancer diagnosis (index year).
bladder cancer (radical cystectomy with urinary diversion) is associated with significant morbidity and body dysmorphia [41]. Certainly, this may lead to depression, decreased self-esteem, and suicidal ideation [47]. These 3 malignancies (bladder, head and neck, and lung) are all associated with cigarette smoking, which is an independent risk factor for suicide [48–50]. The cancer type with the lowest survival rate in South Korea was also reported from 2004 to 2013 (study period) to be lung cancer, followed by liver and stomach cancers [51]. The present study has several limitations. First, because of the limited number of suicide deaths in each specific cancer group, we were unable to explore the suicide risk in cancers with low survival rates, such as those of the pancreas and esophagus [18]. Second, we could
not assess the influence of cancer stage at diagnosis on the association between cancer diagnosis and suicide risk. However, previous studies showed an inconsistent correlation between those variables. For example, while some studies reported higher suicide rates among patients with cancer with an advanced disease stage at diagnosis [27], an international study among patients with breast cancer showed that the correlation between increased suicide risk and advanced disease stage was statistically insignificant [52]. Moreover, a Norwegian study showed a higher relative suicide risk among male patients with nonlocalized cancers compared to those with localized cancers [23]. Notwithstanding these limitations, this is the first longitudinal cohort study investigating suicide deaths within 1 year after cancer in
Table 2 Suicide risk after cancer diagnosis among older patients with cancer. Model
Category
No of older patients
No of suicide
Suicide rate per 100,000 person-years
HR
Model 1a
Non-cancer Cancer Non-cancer Cancer without prediagnosis mental disorders Cancer with prediagnosis mental disorders
194,766 64,922 194,766 50,472 14,450
209 123 209 83 40
107.3 189.5 107.3 164.4 276.8
1.00 2.05 1.00 1.78 2.98
Model 2b
Note. HR, hazard ratio; CI, confident interval. a Model 1 was estimated after adjusting for prediagnosis mental disorders, household income, insurance type, residential area, and disability. b Model 2 was estimated after adjusting for household income, insurance type, residential area, and disability.
Please cite this article as: J.W. Choi and E.-C. Park, Suicide risk after cancer diagnosis among older adults: A nationwide retrospective cohort study, J Geriatr Oncol, https://doi.org/10.1016/j.jgo.2019.11.006
95% CI 1.64
2.56
1.38 2.12
2.29 4.18
J.W. Choi, E.-C. Park / Journal of Geriatric Oncology xxx (2019) xxx
5
Table 3 Suicide risk among older patients with cancer by cancer subtypes. Category
No of older patients
No of suicide
Suicide rate per 100,000 person-years
HRa
95% CI
Subtypes of cancer Non-cancer Bladder cancer Colorectal cancer Head and neck cancer Liver cancer Lung cancer Stomach cancer Other cancers
194,766 2,067 9,570 9,000 4,704 9,040 10,796 19,745
209 10 15 22 11 14 24 27
107.3 483.8 156.7 244.4 233.8 154.9 222.3 136.7
1.00 4.77 1.56 2.28 2.99 1.98 2.40 1.50
2.53 0.93 1.47 1.63 1.15 1.57 1.01
8.99 2.64 3.54 5.49 3.40 3.66 2.24
HR, hazard ratio; CI, confidence interval. a The results were estimated after adjusting for prediagnosis mental disorders, household income, insurance type, residential area, and disability.
older adults using nationwide representative data. Older patients with cancer are at increased risk of suicide, especially those with prediagnosis mental disorders. Given the high suicide risk immediately after cancer diagnosis among older adults, early mental health support needs to be provided to older patients with cancer, considering of prediagnosis mental disorders during follow-up after cancer diagnosis. Author Contributions Jae Woo Choi and Eun-Cheol Park designed the study. Jae Woo Choi and Eun-Cheol Park searched the previous literatures. Jae Woo Choi analyses the data. Jae Woo Choi and Eun-Cheol Park wrote the draft. All authors have approved the final manuscript. Declaration of Competing Interest The authors declare that there is no conflict of interest. Acknowledgements Not applicable. Appendix A. Supplementary Data Supplementary data to this article can be found online at https://doi. org/10.1016/j.jgo.2019.11.006. References [1] World Health Organization. Suicide data. [cited 2019 24, June]; Available from https://www.who.int/mental_health/prevention/suicide/suicideprevent/en/; 2019. [2] World Health Organization. Global Health Observatory (GHO) data. [cited 2018 12, December ]; Available from https://www.who.int/gho/mental_health/suicide_ rates_crude/en/; 2018. [3] World Health Organization. 2017 Mortality Database. [cited 2019 05, March]; Available from http://www.who.int/healthinfo/mortality_data/en/; 2018. [4] Statistics Korea. 2018 The Aged Statistics. cited 2019 05, March]; Available from http://www.kostat.go.kr/portal/korea/kor_nw/2/6/1/index.board?bmode=read &aSeq=370779; 2018. [5] Choi JW, Kim TH, Shin J, Han E. Poverty and suicide risk in older adults: a retrospective longitudinal cohort study. Int J Geriatr Psychiatry 2019;34(11):1565–71. [6] Aslan M, Hocaoglu C, Bahceci B. Description of suicide ideation among older adults and a psychological profile: a cross-sectional study in Turkey. Cien Saude Colet 2019;24(5):1865–74. [7] Ghose B, Wang R, Tang S, Yaya S. Engagement in physical activity, suicidal thoughts and suicide attempts among older people in five developing countries. PeerJ 2019;7: e7108. [8] Yu Z, Xu L, Sun L, Zhang J, Qin W, Li J, et al. Association between interpersonal trust and suicidal ideation in older adults: a cross-sectional analysis of 7070 subjects in Shandong. China BMC Psychiatry 2019;19(1):206. [9] Nam EJ, Lee JE. Mediating effects of social support on depression and suicidal ideation in older Korean adults with hypertension who live alone. J Nurs Res 2019;27 (3):e20. [10] Li LW, Gee GC, Dong X. Association of Self-Reported Discrimination and Suicide Ideation in older Chinese Americans. Am J Geriatr Psychiatry 2018;26(1):42–51.
[11] Dombrovski AY, Aslinger E, Wright AGC, Szanto K. Losing the battle: perceived status loss and contemplated or attempted suicide in older adults. Int J Geriatr Psychiatry 2018;33(7):907–14. [12] Dong X. Associations between the differential definitions of elder mistreatment and suicidal ideation outcomes in U.S. Chinese older adults: do the definitions matter? J Gerontol A Biol Sci Med Sci 2017;72(suppl_1):S82–9. [13] Rossom RC, Simon GE, Coleman KJ, Beck A, Oliver M, Stewart C, et al. Are wishes for death or suicidal ideation symptoms of depression in older adults? Aging Ment Health 2019;23(7):912–8. [14] Pascal de Raykeer R, Hoertel N, Blanco C, Olfson M, Wall M, Seigneurie AS, et al. Effects of psychiatric disorders on suicide attempt: similarities and differences between older and younger adults in a national cohort study. J Clin Psychiatry 2018: 79(6). [15] Lutz J, Fiske A. Functional disability and suicidal behavior in middle-aged and older adults: a systematic critical review. J Affect Disord 2018;227:260–71. [16] Scott KM, Hwang I, Chiu WT, Kessler RC, Sampson NA, Angermeyer M, et al. Chronic physical conditions and their association with first onset of suicidal behavior in the world mental health surveys. Psychosom Med 2010;72(7):712–9. [17] Wu VC, Chang SH, Kuo CF, Liu JR, Chen SW, Yeh YH, et al. Suicide death rates in patients with cardiovascular diseases - a 15-year nationwide cohort study in Taiwan. J Affect Disord 2018;238:187–93. [18] Henson KE, Brock R, Charnock J, Wickramasinghe B, Will O, Pitman A. Risk of Suicide After Cancer Diagnosis in England. JAMA Psychiatry 2018;76(1):51–60. [19] Chan SS, Lyness JM, Conwell Y. Do cerebrovascular risk factors confer risk for suicide in later life? A case-control study. Am J Geriatr Psychiatry 2007;15(6):541–4. [20] Zaorsky NG, Zhang Y, Tuanquin L, Bluethmann SM, Park HS, Chinchilli VM. Suicide among cancer patients. Nat Commun 2019;10. [21] Misono S, Weiss NS, Fann JR, Redman M, Yueh B. Incidence of suicide in persons with cancer. J Clin Oncol 2008;26(29):4731–8. [22] Saad AM, Gad MM, Al-Husseini MJ, AlKhayat MA, Rachid A, Alfaar AS, et al. Suicidal death within a year of a Cancer diagnosis: a population-Based study. Cancer 2019; 125(6):972–9. [23] Hem E, Loge JH, Haldorsen T, Ekeberg O. Suicide risk in cancer patients from 1960 to 1999. J Clin Oncol 2004;22(20):4209–16. [24] Kaceniene A, Krilaviciute A, Kazlauskiene J, Bulotiene G, Smailyte G. Increasing suicide risk among cancer patients in Lithuania from 1993 to 2012: a cancer registrybased study. Eur J Cancer Prev 2017;26:S197–203. [25] Vyssoki B, Gleiss A, Rockett IRH, Hackl M, Leitner B, Sonneck G, et al. Suicide among 915,303 Austrian cancer patients: who is at risk? J Affect Disord 2015;175:287–91. [26] Ahn E, Shin DW, Cho SI, Park S, Won YJ, Yun YH. Suicide rates and risk factors among Korean Cancer patients, 1993-2005. Cancer Epidemiol Biomarkers Prev 2010;19(8): 2097–105. [27] Ahn MH, Park S, Lee HB, Ramsey CM, Na R, Kim SO, et al. Suicide in cancer patients within the first year of diagnosis. Psycho-Oncology 2015;24(5):601–7. [28] Wang SM, Chang JC, Weng SC, Yeh MK, Lee CS. Risk of suicide within 1 year of cancer diagnosis. Int J Cancer 2018;142(10):1986–93. [29] Yamauchi T, Inagaki M, Yonemoto N, Iwasaki M, Inoue M, Akechi T, et al. Death by suicide and other externally caused injuries following a cancer diagnosis: the Japan public health center-based prospective study. Psycho-Oncology 2014;23(9):1034–41. [30] Fang F, Fall K, Mittleman MA, Sparen P, Ye WM, Adami HO, et al. Suicide and cardiovascular death after a Cancer diagnosis. N Engl J Med 2012;366(14):1310–8. [31] Crocetti E, Buzzoni C, Caldarella A, Intrieri T, Manneschi G, Sacchettini C, et al. Suicide mortality among cancer patients. Epidemiol Prev 2012;36(2):83–7. [32] Miller M, Mogun H, Azrael D, Hempstead K, Solomon DH. Cancer and the risk of suicide in older Americans. J Clin Oncol 2008;26(29):4720–4. [33] Cole TB, Bowling JM, Patetta MJ, Blazer DG. Risk factors for suicide among older adults with cancer. Aging Ment Health 2014;18(7):854–60. [34] Statistics Korea. The causes of death statistics. [cited 2019 04 March]; Available from http://kosis.kr/index/index.do; 2018. [35] Choi JW, Lee KS, Kim TH, Choi J, Han E. Suicide risk after discharge from psychiatric care in South Korea. J Affect Disord 2019;251:287–92. [36] Demicheli R, Fornili M, Ambrogi F, Higgins K, Boyd JA, Biganzoli E, et al. Recurrence dynamics for non-small-cell lung cancer: effect of surgery on the development of metastases. J Thorac Oncol 2012;7(4):723–30. [37] Chang CF, Gould M. Playing the odds: lung cancer surveillance after curative surgery. Curr Opin Pulm Med 2017;23(4):298–304.
Please cite this article as: J.W. Choi and E.-C. Park, Suicide risk after cancer diagnosis among older adults: A nationwide retrospective cohort study, J Geriatr Oncol, https://doi.org/10.1016/j.jgo.2019.11.006
6
J.W. Choi, E.-C. Park / Journal of Geriatric Oncology xxx (2019) xxx
[38] Kim H, Ahn JS, Kim H, Cha YS, Lee J, Kim MH, et al. Sociodemographic and clinical characteristics of old-old suicide attempters compared with young-old and middle-aged attempters. Int J Geriatr Psychiatry 2018;33(12):1717–26. [39] Pothier K, Benguigui N, Kulpa R, Chavoix C. Multiple object tracking while walking: similarities and differences between young, young-old, and old-old adults. J Gerontol B-Psychol 2015;70(6):840–9. [40] Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987;40(5):373–83. [41] Klaassen Z, Wallis CJD, Chandrasekar T, Goldberg H, Sayyid RK, Williams SB, et al. Cancer diagnosis and risk of suicide after accounting for prediagnosis psychiatric care: a matched-cohort study of patients with incident solid-organ malignancies. Cancer 2019;125(16):2886–95. [42] Klaassen Z, Jen RP, DiBianco JM, Reinstatler L, Li Q, Madi R, et al. Factors associated with suicide in patients with genitourinary malignancies. Cancer 2015;121(11): 1864–72. [43] Klaassen Z, Goldberg H, Chandrasekar T, Arora K, Sayyid RK, Hamilton RJ, et al. Changing trends for suicidal death in patients with bladder cancer: a 40+ year population-level analysis. Clin Genitourin Cancer 2018;16(3):206–12 [e1]. [44] Klaassen Z, DiBianco JM, Jen RP, Harper B, Yaguchi G, Reinstatler L, et al. The impact of radical cystectomy and urinary diversion on suicidal death in patients with bladder Cancer. J Wound Ostomy Continence Nurs 2016;43(2):152–7.
[45] Kam D, Salib A, Gorgy G, Patel TD, Carniol ET, Eloy JA, et al. Incidence of suicide in patients with head and neck Cancer. JAMA Otolaryngol Head Neck Surg 2015;141 (12):1075–81. [46] Urban D, Rao A, Bressel M, Neiger D, Solomon B, Mileshkin L. Suicide in lung Cancer who is at risk? Chest 2013;144(4):1245–52. [47] Schneider KL, Shenassa E. Correlates of suicide ideation in a population-based sample of cancer patients. J Psychosoc Oncol 2008;26(2):49–62. [48] Iwasaki M, Akechi T, Uchitomi Y, Tsugane S, Ctr-Based JPH. Cigarette smoking and completed suicide among middle-aged men: a population based cohort study in Japan. Ann Epidemiol 2005;15(4):286–92. [49] Tanskanen A, Tuomilehto J, Viinamaki H, Vartiainen E, Lehtonen J, Puska P. Smoking and the risk of suicide. Acta Psychiatr Scand 2000;101(3):243–5. [50] Sankaranarayanan A, Mancuso S, Castle D. Smoking and suicidality in patients with a psychotic disorder. Psychiatry Res 2014;215(3):634–40. [51] Oh CM, Won YJ, Jung KW, Kong HJ, Cho H, Lee JK, et al. Cancer statistics in Korea: incidence, mortality, survival, and prevalence in 2013. Cancer Res Treat 2016;48(2): 436–50. [52] Schairer C, Brown LM, Chen BE, Howard R, Lynch CF, Hall P, et al. Suicide after breast cancer: an international population-based study of 723,810 women. J Natl Cancer Inst 2006;98(19):1416–9.
Please cite this article as: J.W. Choi and E.-C. Park, Suicide risk after cancer diagnosis among older adults: A nationwide retrospective cohort study, J Geriatr Oncol, https://doi.org/10.1016/j.jgo.2019.11.006
Contents lists available at ScienceDirect
Journal of Geriatric Oncology
Suicide risk after cancer diagnosis among older adults: A nationwide retrospective cohort study Jae Woo Choi a, Eun-Cheol Park b,⁎ a b
College of Pharmacy, Yonsei Institute of Pharmaceutical Sciences, Yonsei University, Incheon, South Korea Department of Preventive Medicine, Yonsei University College of Medicine, Seoul, South Korea
a r t i c l e
i n f o
Article history: Received 13 August 2019 Received in revised form 9 September 2019 Accepted 20 November 2019 Available online xxxx Keywords: Cancer Suicide risk Prediagnosis mental disorders Older adults
a b s t r a c t Objectives: Cancer increases suicide risk, but little is known about associations between cancer and suicide among older adults. We aimed to investigate suicide risk within 1 year after cancer diagnosis in older adults. Methods: Using National Health Insurance Service-Senior Cohort data, we included newly diagnosed older patients with cancer and older adults without cancer, selected by a 1:3 propensity score matching using sex, age, Charlson comorbidity index, and index year from 2004 to 2012 and followed them up throughout 2013. We used Cox proportional hazard models to estimate adjusted hazard ratios (AHR) of suicide risk. Results: In the total sample of 259,688 older adults (aged 62–115 years), the highest proportion of observed suicide deaths occurred within one year after cancer diagnosis (36.1% of total); 64,922 older patients with cancer showed higher suicide risk after cancer diagnosis compared to non-cancer participants (AHR 2.05; 95% CI 1.64–2.56). Patients with mental disorder diagnosis before cancer diagnosis (AHR 2.98; 95% CI 2.12–4.18) had increased suicide risk than those without mental disorder diagnosis (AHR 1.78; 95% CI 1.38–2.29) compared to non-cancer participants. Suicide risk among older patients with bladder, head and neck, liver, lung, and stomach cancers was higher than in non-cancer participants (AHR 4.77, 2.28, 2.99, 1.98, 2.40; 95% CI 2.53–8.99, 1.47–3.54, 1.63–5.49, 1.15–3.40, 1.57–2.24), respectively. Conclusions: Older patients with cancer have a higher suicide risk. Early mental health support needs to be provided with consideration of prediagnosis mental disorders during follow-up after cancer diagnosis. © 2019 Elsevier Ltd. All rights reserved.
1. Introduction Suicide accounts for 1.4% of all deaths worldwide, making it the 18th leading cause of death in 2016 [1]. South Korea is ranked 10th globally for suicide rate in 2016 (20.2 per 100,000 people) and the suicide rate among older adults (60 years old and older) was 53.3 per 100,000 people in 2016, which was the highest among Organization for Economic Cooperation and Development countries [2,3]. Given the higher proportion of the aging population (14.3% of total population in 2018) than that of the worldwide population (8.2%), and its rapidly growing rate (average 4.1% annually) in South Korea [4], suicide prevention among older adults has become a significant public health issue. Previous studies have reported risk factors associated with suicide risk among older adults that include male sex [5], poverty [5], education level [6], physical activity [7], interpersonal trust [8], social support [9], discrimination [10], status loss [11], elder mistreatment [12], mental illness such as depression [13,14], functional disability [15].
⁎ Corresponding author at: Division of Medical Science Research Affairs & President, University-Industry Foundation of Yonsei University Health System, South Korea. E-mail address: [email protected] (E.-C. Park).
The risk of suicide can also be greatly increased by the occurrence of major medical conditions in adulthood [16] and previous research has showed that patients with severe illnesses such as cancer, cardiovascular disease, and cerebrovascular disease are at increased suicide risk after diagnosis [17–19]. Among these, suicide risk after cancer diagnosis has been well established in many countries including the USA [20–22], England [18], Norway [23], Lithuania [24], Australia [25], Korea [26,27], Taiwan [28], Japan [29], Sweden [30], and Italy [31]. However, little is known about suicide risk following cancer diagnosis among older adults. A previous case control study showed that the risk of suicide in older adults was higher among patients with cancer than among patients with other medical illnesses [32]. Another case-control study based on interviews of primary informants for suicides showed that there was a statistically significant association between suicide and cancer among older adults [33]. However, these studies included patients who had been diagnosed with cancer a relatively long time ago and their suicide death may not have been directly caused by cancer diagnosis. It is necessary to study the risk of suicide within short periods after diagnosis of cancer in older adults, considering that the first leading cause of death among older adults in South Korea is cancer [34], as well as the increasing older adult population and high older adults suicide rate. Previous
https://doi.org/10.1016/j.jgo.2019.11.006 1879-4068/© 2019 Elsevier Ltd. All rights reserved.
Please cite this article as: J.W. Choi and E.-C. Park, Suicide risk after cancer diagnosis among older adults: A nationwide retrospective cohort study, J Geriatr Oncol, https://doi.org/10.1016/j.jgo.2019.11.006
2
J.W. Choi, E.-C. Park / Journal of Geriatric Oncology xxx (2019) xxx
research has indicated highest suicide risk during the first year after cancer diagnosis [22,27,28]. Furthermore, mental disorder diagnosis is a strong risk factor for suicide death [35]. However, little is known about the suicide risk after cancer while considering the role of mental disorder diagnosis before cancer diagnosis among older adults. Therefore, we aimed to investigate suicide risk within 1 year after cancer diagnosis among older adults using nationally representative cohort data. We performed separate analyses for risk of suicide according to prediagnosis mental disorders of older patients with cancer. We also analyzed variations in this risk by subtypes of cancer site (head and neck, stomach, colorectal, lung, liver, and other cancers). 2. Methods 2.1. Data Source This study used the National Health Insurance Service–Senior Cohort (NHIS–SC) data, an elderly-population-based cohort provided by the Korean National Health Insurance Service. The NHIS–SC data were extracted using a random sampling method to construct a representative dataset and a representative sample of 558,147 older adults was randomly selected, which was composed of about 10% of the older adult population aged ≥60 years old (5.5 million) in 2002. The NHIS–SC data was followed for 12 years until 2013, unless the participant was disqualified because of immigration or death. Information on each individual's cause of death was extracted from Statistics Korea and was linked with our database. The NHIS–SC data includes general demographics and clinical information for diagnoses, treatments, and prescribed medications for all visits to medical institutions. 2.2. Study Sample Among 528,655 participants from NHIS–SC between 2004 and 2012, we included outpatients in NHIS–SC (n = 526,278), excluded older adults without cancer (n = 447,725), and selected outpatients with a principal diagnosis of cancer (n = 78,553). We defined patients with cancer as patients diagnosed with a malignant tumor (International Classification of Diseases, 10th Revision [ICD-10] codes C00–C97, excluding C44, nonmelanoma skin cancer) from January 1, 2004, to December 31, 2012. We selected older patients with a first principal diagnosis of cancer and the first diagnosis date of cancer was defined as the index date in this study. A previous study reported that distinguishable peaks for cancer recurrence occurred in the first couple of years after diagnosis [36]; thus, we utilized the first two years of the database (2002−2003) as a washout period to exclude all patients with cancer with possible relapse or recurrence within the two years. Close radiographic follow-up after curative resection is also a widely recommended practice, which is executed more frequently within the first two years following diagnosis or treatment [37], and we selected the final study subjects with newly diagnosed cancer (n = 64,922). For comparison of suicide risk among the group without cancer, we selected older adults without cancer (n = 450,102) and performed a 1:3 propensity score matching using sex, age, Charlson Comorbidity Index, and index year as variables. The participants were divided into youngold (60–74 years old) and old-old (N 75 years old) adults, based on previous studies [38,39]. The Charlson comorbidity index corresponds to the sum of the weighted scores assigned to various major health conditions [40], and was calculated by screening the one-year period preceding the participants' index year. The first diagnosis year of cancer in the study period was defined as the index year and we matched each comparison group in the same index year of patients with cancer. After calculating the estimated propensity score, patients with cancer were matched with patients without cancer who had the same or similar propensity scores. Finally, older patients with cancer (n = 64,922) and older patients without cancer (n = 194,766) were selected as the final sample (supplementary Fig. 1).
2.3. Measurements The dependent variable was death by suicide within 1 year after the index date and we defined death as suicide if the code for cause of death was X60-X84 (intentional self-harm) based on the ICD-10. The length of survival was calculated in days, and all participants enrolled in the study were followed up until death by suicide, withdrawal from the insurance system, death from other causes, or the end of 2013, whichever occurred earlier. Potential confounding factors were residential area, household income, insurance type, disability, and history of mental disorder diagnosis. Information for confounding factors was based on the participants' index date. The residential region was classified as metropolitan (capital), urban (local government where N1 million people live), or rural (otherwise). Household income was classified as (1) high (81st–100th percentile), (2) middle (41st–80th percentile), or (3) low income (covered by Medical Aid and below the 40th percentile). The insurance type was either National Health Insurance or Medical Aid, which are public medical support programs assisting poor individuals by a proportional allocation. The prediagnosis mental disorders were defined by the occurrence of a principal diagnosis of mental disorder preceding the index date (ICD-10: F00–F99). 2.4. Statistical Analysis The baseline characteristics of study subjects were summarized by cancer type. The statistical significance of the characteristic differences between patients with and without cancer was tested by Pearson's χ2 test. We used Cox regression to assess the suicide risk within 1 year after the index date. We reported results as estimates of adjusted hazard ratio (AHR) with a 95% confidence interval (CI). Person years were calculated for each group, and the starting point was the index date. The index date of older patients with cancer was first diagnosis date of cancer in the study period and that of matched older patients without cancer was allocated by the same index date as each matched older patient with cancer. In the survival analysis, events other than suicide within 1 year after the index date were censored. First, we estimated the suicide risk among older patients with cancer compared to the non-cancer group. Next, we performed separate analyses for suicide risk according to prediagnosis mental disorders in older patients with cancer. Finally, we conducted separate analyses for the most common cancer occurrence sites/groups to explore suicide risk by cancer type compared to the non-cancer group. All data-extraction and statistical analyses were performed using the SAS 9.4 program, and the institutional review board of the study site exempted this study from the requirement of ethical approval. 3. Results Fig. 1 shows the number of suicide deaths per period after cancer among older adults. Among 64,570 patients with cancer, 263 male subjects and 89 female subjects died by suicide over the entire follow-up period. The median time from cancer diagnosis to suicide was 1.74 years (range, 0.01–9.46 years), and the suicide rate decreased progressively with time after cancer diagnosis. One hundred twenty three suicides (34.9% of the total) occurred during one year after cancer, 66 (18.8%) during the second year, 48 (13.6%) during the third year, 32 (9.1%) during the fourth year, 26 (7.4%) during the fifth year, 20 (5.7%) during the sixth year, and 37 (10.5%) after the sixth year. The study subjects' baseline characteristics (2004–2012) are presented in Table 1. Before propensity score matching, there were significant differences in all confounding factors between the older patients with and without cancer. After matching, about 56% of the subjects were male, and 60.7% and 39.3% were young–old and old–old adults, respectively, in both the older patients with cancer and matched
Please cite this article as: J.W. Choi and E.-C. Park, Suicide risk after cancer diagnosis among older adults: A nationwide retrospective cohort study, J Geriatr Oncol, https://doi.org/10.1016/j.jgo.2019.11.006
J.W. Choi, E.-C. Park / Journal of Geriatric Oncology xxx (2019) xxx
3
Fig. 1. The number of suicide deaths per period after cancer diagnosis.
comparison groups. Approximately 60% of subjects in both groups had comorbidities. Prediagnosis mental disorders were present in 22.3% of older patients with cancer and 22.8% of older patients without cancer. The proportion of low household income was higher in older patients with cancer than in the matched comparison group (35.5% vs. 35.0%). The proportion of people receiving Medical Aid was higher among older patients with cancer than the non-cancer group (10.1% vs. 8.2%). Older patients with cancer were less likely to live in a metropolitan area compared to their non-cancer counterparts (17.6% vs. 18.1%). There was no statistically significant difference between older patients with and without cancer in having a disability. Table 2 shows the risk of suicide within 1 year after cancer diagnosis compared with the non-cancer group. Model 1 indicates that older patients with cancer were more likely to die by suicide than the noncancer group (AHR 2.05; 95% CI 1.64–2.56). In model 2, older patients with cancer and prediagnosis mental disorders (AHR 2.98; 95% CI 2.12–4.18) had an increased suicide risk than those without a history of mental disorder diagnosis before cancer diagnosis (AHR 1.78; 95% CI 1.38–2.29) compared to the non-cancer group. The AHR of suicide death among older patients with cancer was estimated according to the six most common cancer subtypes (Table 3). A higher risk of suicide was observed in bladder (AHR, 4.77; 95% CI, 2.53–8.99), head and neck (AHR, 2.28; 95% CI, 1.47–3.54), liver (AHR, 2.99; 95% CI, 1.63–5.49), lung (AHR, 1.98; 95% CI, 1.15–3.40), stomach (AHR, 2.40; 95% CI, 1.57–3.66), and other cancers (AHR, 1.50; 95% CI, 1.01–2.24) compared to the non-cancer group. No statistically significant association was observed between suicide risk and colorectal cancer (AHR, 1.56; 95% CI, 0.93–2.64).
4. Discussion There were three main findings in this study. First, among older patients with cancer, the highest proportion of observed suicide deaths occurred within 1 year after cancer diagnosis. Second, suicide risk within 1 year after cancer among older patients was higher than in the noncancer group, especially among those with a history of mental disorder diagnosis before cancer. Third, among the six most common cancer
subtypes, patients with bladder, head and neck, liver, lung, and stomach cancers had a higher suicide risk than those who did not have cancer. Among all suicide deaths associated with cancer, 123 (34.9%) suicide events occurred within 1 year of cancer diagnosis, which seems to be consistent with other reports. Wang et al. (40.8%), Ahn et al. (40.0%), and Saad et al. (31.7%) reported similar suicide rates within 1 year after cancer in adults [22,27,28]. Wang et al. revealed that suicide risk was found to be highest within one month of a new cancer diagnosis and elevated risk persisted thereafter but gradually decreased in intensity over the next 11 months. Ahn et al. showed that the standardized mortality ratio for suicide within 1 year after cancer was significantly higher than in the general population (SMR, 1.65; 95% CI, 1.40–1.94). Saad et al. implied that the risk of suicide increases significantly within the first year after a diagnosis of cancer in comparison with the general population (SMR, 2.52; 95% CI, 2.39–2.64). These findings indicate that clinicians specializing in cancer care should provide special attention to older patients with newly diagnosed cancer to reduce their risk of suicide. Although most previous studies showed an increased risk of suicide among patients with cancer, the analyses did not consider history of mental disorder diagnosis [18,20–31]. Mental disorder diagnosis is a major predictor of suicide and our findings showed that older patients with cancer and prediagnosis mental disorders had a higher suicide risk than those without mental disorder diagnosis before cancer diagnosis compared to the non-cancer group, which is consistent with previous research [41]. The previous study examined suicide risk of adults (aged ≥18 years) with cancer and among study subjects who utilized outpatient psychiatric care before cancer diagnosis, the patients with cancer were significantly more likely to die by suicide compared with non-cancer controls. The results implied that mental disorder diagnosis before cancer diagnosis is important consideration to potentially reduce the risk of suicide death among older patients with cancer. Assessing the association between specific tumor sites and suicidal risk among older patients with cancer, this research is consistent with previous studies suggesting increased suicidal risk among patients with bladder [42–44], head and neck [45], lung [46], liver [26], and stomach cancer [18,27]. Surgical treatment for head and neck cancer (extirpation with or without extensive neck lymphadenectomy) and
Please cite this article as: J.W. Choi and E.-C. Park, Suicide risk after cancer diagnosis among older adults: A nationwide retrospective cohort study, J Geriatr Oncol, https://doi.org/10.1016/j.jgo.2019.11.006
4
J.W. Choi, E.-C. Park / Journal of Geriatric Oncology xxx (2019) xxx
Table 1 Study subjects' general characteristics (2004–2012). Variables
Cancer
Total Sex Male Female Age 60–74 ≥75 Index year 2004 2005 2006 2007 2008 2009 2010 2011 2012 Charlson comorbidity index 0 1 2 ≥3 Prediagnosis mental disorders Yes No Household income Low Middle High Insurance type National health insurance Medical aid Residential area Metropolitan Urban Rural Disability Yes No a b
After propensity score matchingb
Before propensity score matching p-valuea
Non-cancer
N
%
N
%
64,922
12.6
450,102
87.4
p-valuea
Non-cancer N
%
194,766
37.8
109,794 84,972
56.4 43.6
118,131 76,635
60.7 39.3
23,613 24,932 25,005 23,747 22,404 21,624 18,309 18,157 16,975
12.1 12.8 12.8 12.2 11.5 11.1 9.4 9.3 8.7
77,886 59,927 32,314 24,639
40.0 30.8 16.6 12.7
44,449 150,317
22.8 77.2
65,785 60,878 68,103
33.8 31.3 35.0
178,796 15,970
91.8 8.2
35,314 43,989 115,463
18.1 22.6 59.3
2083 192,683
1.1 98.9
b0.001 36,598 28,324
56.4 43.6
172,922 277,180
38.4 61.6
39,377 25,545
60.7 39.3
330,207 119,895
73.4 26.6
7871 8309 8335 7914 7469 7208 6103 6053 5660
12.1 12.8 12.8 12.2 11.5 11.1 9.4 9.3 8.7
386,577 44,775 11,156 3392 1868 1091 653 345 245
85.9 9.9 2.5 0.8 0.4 0.2 0.1 0.1 0.1
25,962 19,975 10,773 8212
40.0 30.8 16.6 12.6
280,314 103,703 41,909 24,176
62.3 23.0 9.3 5.4
14,450 50,472
22.3 77.7
40,244 409,858
8.9 91.1
21,610 20,260 23,052
33.3 31.2 35.5
164,588 149,176 136,338
36.6 33.1 30.3
58,396 6526
89.9 10.1
412,065 38,037
91.5 8.5
11,448 13,921 39,553
17.6 21.4 60.9
78,523 99,864 271,715
17.4 22.2 60.4
713 64,209
1.1 98.9
3674 446,428
0.8 99.2
1.000
b0.001
1.000
b0.001
1.000
b0.001
1.000
b0.001
0.003
b0.001
0.023
b0.001
b0.001
b0.001
b0.001
b0.001
0.539
p-value b0.05 is significant. After propensity score matching for sex, age, Charlson Comorbidity Index, and year of cancer diagnosis (index year).
bladder cancer (radical cystectomy with urinary diversion) is associated with significant morbidity and body dysmorphia [41]. Certainly, this may lead to depression, decreased self-esteem, and suicidal ideation [47]. These 3 malignancies (bladder, head and neck, and lung) are all associated with cigarette smoking, which is an independent risk factor for suicide [48–50]. The cancer type with the lowest survival rate in South Korea was also reported from 2004 to 2013 (study period) to be lung cancer, followed by liver and stomach cancers [51]. The present study has several limitations. First, because of the limited number of suicide deaths in each specific cancer group, we were unable to explore the suicide risk in cancers with low survival rates, such as those of the pancreas and esophagus [18]. Second, we could
not assess the influence of cancer stage at diagnosis on the association between cancer diagnosis and suicide risk. However, previous studies showed an inconsistent correlation between those variables. For example, while some studies reported higher suicide rates among patients with cancer with an advanced disease stage at diagnosis [27], an international study among patients with breast cancer showed that the correlation between increased suicide risk and advanced disease stage was statistically insignificant [52]. Moreover, a Norwegian study showed a higher relative suicide risk among male patients with nonlocalized cancers compared to those with localized cancers [23]. Notwithstanding these limitations, this is the first longitudinal cohort study investigating suicide deaths within 1 year after cancer in
Table 2 Suicide risk after cancer diagnosis among older patients with cancer. Model
Category
No of older patients
No of suicide
Suicide rate per 100,000 person-years
HR
Model 1a
Non-cancer Cancer Non-cancer Cancer without prediagnosis mental disorders Cancer with prediagnosis mental disorders
194,766 64,922 194,766 50,472 14,450
209 123 209 83 40
107.3 189.5 107.3 164.4 276.8
1.00 2.05 1.00 1.78 2.98
Model 2b
Note. HR, hazard ratio; CI, confident interval. a Model 1 was estimated after adjusting for prediagnosis mental disorders, household income, insurance type, residential area, and disability. b Model 2 was estimated after adjusting for household income, insurance type, residential area, and disability.
Please cite this article as: J.W. Choi and E.-C. Park, Suicide risk after cancer diagnosis among older adults: A nationwide retrospective cohort study, J Geriatr Oncol, https://doi.org/10.1016/j.jgo.2019.11.006
95% CI 1.64
2.56
1.38 2.12
2.29 4.18
J.W. Choi, E.-C. Park / Journal of Geriatric Oncology xxx (2019) xxx
5
Table 3 Suicide risk among older patients with cancer by cancer subtypes. Category
No of older patients
No of suicide
Suicide rate per 100,000 person-years
HRa
95% CI
Subtypes of cancer Non-cancer Bladder cancer Colorectal cancer Head and neck cancer Liver cancer Lung cancer Stomach cancer Other cancers
194,766 2,067 9,570 9,000 4,704 9,040 10,796 19,745
209 10 15 22 11 14 24 27
107.3 483.8 156.7 244.4 233.8 154.9 222.3 136.7
1.00 4.77 1.56 2.28 2.99 1.98 2.40 1.50
2.53 0.93 1.47 1.63 1.15 1.57 1.01
8.99 2.64 3.54 5.49 3.40 3.66 2.24
HR, hazard ratio; CI, confidence interval. a The results were estimated after adjusting for prediagnosis mental disorders, household income, insurance type, residential area, and disability.
older adults using nationwide representative data. Older patients with cancer are at increased risk of suicide, especially those with prediagnosis mental disorders. Given the high suicide risk immediately after cancer diagnosis among older adults, early mental health support needs to be provided to older patients with cancer, considering of prediagnosis mental disorders during follow-up after cancer diagnosis. Author Contributions Jae Woo Choi and Eun-Cheol Park designed the study. Jae Woo Choi and Eun-Cheol Park searched the previous literatures. Jae Woo Choi analyses the data. Jae Woo Choi and Eun-Cheol Park wrote the draft. All authors have approved the final manuscript. Declaration of Competing Interest The authors declare that there is no conflict of interest. Acknowledgements Not applicable. Appendix A. Supplementary Data Supplementary data to this article can be found online at https://doi. org/10.1016/j.jgo.2019.11.006. References [1] World Health Organization. Suicide data. [cited 2019 24, June]; Available from https://www.who.int/mental_health/prevention/suicide/suicideprevent/en/; 2019. [2] World Health Organization. Global Health Observatory (GHO) data. [cited 2018 12, December ]; Available from https://www.who.int/gho/mental_health/suicide_ rates_crude/en/; 2018. [3] World Health Organization. 2017 Mortality Database. [cited 2019 05, March]; Available from http://www.who.int/healthinfo/mortality_data/en/; 2018. [4] Statistics Korea. 2018 The Aged Statistics. cited 2019 05, March]; Available from http://www.kostat.go.kr/portal/korea/kor_nw/2/6/1/index.board?bmode=read &aSeq=370779; 2018. [5] Choi JW, Kim TH, Shin J, Han E. Poverty and suicide risk in older adults: a retrospective longitudinal cohort study. Int J Geriatr Psychiatry 2019;34(11):1565–71. [6] Aslan M, Hocaoglu C, Bahceci B. Description of suicide ideation among older adults and a psychological profile: a cross-sectional study in Turkey. Cien Saude Colet 2019;24(5):1865–74. [7] Ghose B, Wang R, Tang S, Yaya S. Engagement in physical activity, suicidal thoughts and suicide attempts among older people in five developing countries. PeerJ 2019;7: e7108. [8] Yu Z, Xu L, Sun L, Zhang J, Qin W, Li J, et al. Association between interpersonal trust and suicidal ideation in older adults: a cross-sectional analysis of 7070 subjects in Shandong. China BMC Psychiatry 2019;19(1):206. [9] Nam EJ, Lee JE. Mediating effects of social support on depression and suicidal ideation in older Korean adults with hypertension who live alone. J Nurs Res 2019;27 (3):e20. [10] Li LW, Gee GC, Dong X. Association of Self-Reported Discrimination and Suicide Ideation in older Chinese Americans. Am J Geriatr Psychiatry 2018;26(1):42–51.
[11] Dombrovski AY, Aslinger E, Wright AGC, Szanto K. Losing the battle: perceived status loss and contemplated or attempted suicide in older adults. Int J Geriatr Psychiatry 2018;33(7):907–14. [12] Dong X. Associations between the differential definitions of elder mistreatment and suicidal ideation outcomes in U.S. Chinese older adults: do the definitions matter? J Gerontol A Biol Sci Med Sci 2017;72(suppl_1):S82–9. [13] Rossom RC, Simon GE, Coleman KJ, Beck A, Oliver M, Stewart C, et al. Are wishes for death or suicidal ideation symptoms of depression in older adults? Aging Ment Health 2019;23(7):912–8. [14] Pascal de Raykeer R, Hoertel N, Blanco C, Olfson M, Wall M, Seigneurie AS, et al. Effects of psychiatric disorders on suicide attempt: similarities and differences between older and younger adults in a national cohort study. J Clin Psychiatry 2018: 79(6). [15] Lutz J, Fiske A. Functional disability and suicidal behavior in middle-aged and older adults: a systematic critical review. J Affect Disord 2018;227:260–71. [16] Scott KM, Hwang I, Chiu WT, Kessler RC, Sampson NA, Angermeyer M, et al. Chronic physical conditions and their association with first onset of suicidal behavior in the world mental health surveys. Psychosom Med 2010;72(7):712–9. [17] Wu VC, Chang SH, Kuo CF, Liu JR, Chen SW, Yeh YH, et al. Suicide death rates in patients with cardiovascular diseases - a 15-year nationwide cohort study in Taiwan. J Affect Disord 2018;238:187–93. [18] Henson KE, Brock R, Charnock J, Wickramasinghe B, Will O, Pitman A. Risk of Suicide After Cancer Diagnosis in England. JAMA Psychiatry 2018;76(1):51–60. [19] Chan SS, Lyness JM, Conwell Y. Do cerebrovascular risk factors confer risk for suicide in later life? A case-control study. Am J Geriatr Psychiatry 2007;15(6):541–4. [20] Zaorsky NG, Zhang Y, Tuanquin L, Bluethmann SM, Park HS, Chinchilli VM. Suicide among cancer patients. Nat Commun 2019;10. [21] Misono S, Weiss NS, Fann JR, Redman M, Yueh B. Incidence of suicide in persons with cancer. J Clin Oncol 2008;26(29):4731–8. [22] Saad AM, Gad MM, Al-Husseini MJ, AlKhayat MA, Rachid A, Alfaar AS, et al. Suicidal death within a year of a Cancer diagnosis: a population-Based study. Cancer 2019; 125(6):972–9. [23] Hem E, Loge JH, Haldorsen T, Ekeberg O. Suicide risk in cancer patients from 1960 to 1999. J Clin Oncol 2004;22(20):4209–16. [24] Kaceniene A, Krilaviciute A, Kazlauskiene J, Bulotiene G, Smailyte G. Increasing suicide risk among cancer patients in Lithuania from 1993 to 2012: a cancer registrybased study. Eur J Cancer Prev 2017;26:S197–203. [25] Vyssoki B, Gleiss A, Rockett IRH, Hackl M, Leitner B, Sonneck G, et al. Suicide among 915,303 Austrian cancer patients: who is at risk? J Affect Disord 2015;175:287–91. [26] Ahn E, Shin DW, Cho SI, Park S, Won YJ, Yun YH. Suicide rates and risk factors among Korean Cancer patients, 1993-2005. Cancer Epidemiol Biomarkers Prev 2010;19(8): 2097–105. [27] Ahn MH, Park S, Lee HB, Ramsey CM, Na R, Kim SO, et al. Suicide in cancer patients within the first year of diagnosis. Psycho-Oncology 2015;24(5):601–7. [28] Wang SM, Chang JC, Weng SC, Yeh MK, Lee CS. Risk of suicide within 1 year of cancer diagnosis. Int J Cancer 2018;142(10):1986–93. [29] Yamauchi T, Inagaki M, Yonemoto N, Iwasaki M, Inoue M, Akechi T, et al. Death by suicide and other externally caused injuries following a cancer diagnosis: the Japan public health center-based prospective study. Psycho-Oncology 2014;23(9):1034–41. [30] Fang F, Fall K, Mittleman MA, Sparen P, Ye WM, Adami HO, et al. Suicide and cardiovascular death after a Cancer diagnosis. N Engl J Med 2012;366(14):1310–8. [31] Crocetti E, Buzzoni C, Caldarella A, Intrieri T, Manneschi G, Sacchettini C, et al. Suicide mortality among cancer patients. Epidemiol Prev 2012;36(2):83–7. [32] Miller M, Mogun H, Azrael D, Hempstead K, Solomon DH. Cancer and the risk of suicide in older Americans. J Clin Oncol 2008;26(29):4720–4. [33] Cole TB, Bowling JM, Patetta MJ, Blazer DG. Risk factors for suicide among older adults with cancer. Aging Ment Health 2014;18(7):854–60. [34] Statistics Korea. The causes of death statistics. [cited 2019 04 March]; Available from http://kosis.kr/index/index.do; 2018. [35] Choi JW, Lee KS, Kim TH, Choi J, Han E. Suicide risk after discharge from psychiatric care in South Korea. J Affect Disord 2019;251:287–92. [36] Demicheli R, Fornili M, Ambrogi F, Higgins K, Boyd JA, Biganzoli E, et al. Recurrence dynamics for non-small-cell lung cancer: effect of surgery on the development of metastases. J Thorac Oncol 2012;7(4):723–30. [37] Chang CF, Gould M. Playing the odds: lung cancer surveillance after curative surgery. Curr Opin Pulm Med 2017;23(4):298–304.
Please cite this article as: J.W. Choi and E.-C. Park, Suicide risk after cancer diagnosis among older adults: A nationwide retrospective cohort study, J Geriatr Oncol, https://doi.org/10.1016/j.jgo.2019.11.006
6
J.W. Choi, E.-C. Park / Journal of Geriatric Oncology xxx (2019) xxx
[38] Kim H, Ahn JS, Kim H, Cha YS, Lee J, Kim MH, et al. Sociodemographic and clinical characteristics of old-old suicide attempters compared with young-old and middle-aged attempters. Int J Geriatr Psychiatry 2018;33(12):1717–26. [39] Pothier K, Benguigui N, Kulpa R, Chavoix C. Multiple object tracking while walking: similarities and differences between young, young-old, and old-old adults. J Gerontol B-Psychol 2015;70(6):840–9. [40] Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987;40(5):373–83. [41] Klaassen Z, Wallis CJD, Chandrasekar T, Goldberg H, Sayyid RK, Williams SB, et al. Cancer diagnosis and risk of suicide after accounting for prediagnosis psychiatric care: a matched-cohort study of patients with incident solid-organ malignancies. Cancer 2019;125(16):2886–95. [42] Klaassen Z, Jen RP, DiBianco JM, Reinstatler L, Li Q, Madi R, et al. Factors associated with suicide in patients with genitourinary malignancies. Cancer 2015;121(11): 1864–72. [43] Klaassen Z, Goldberg H, Chandrasekar T, Arora K, Sayyid RK, Hamilton RJ, et al. Changing trends for suicidal death in patients with bladder cancer: a 40+ year population-level analysis. Clin Genitourin Cancer 2018;16(3):206–12 [e1]. [44] Klaassen Z, DiBianco JM, Jen RP, Harper B, Yaguchi G, Reinstatler L, et al. The impact of radical cystectomy and urinary diversion on suicidal death in patients with bladder Cancer. J Wound Ostomy Continence Nurs 2016;43(2):152–7.
[45] Kam D, Salib A, Gorgy G, Patel TD, Carniol ET, Eloy JA, et al. Incidence of suicide in patients with head and neck Cancer. JAMA Otolaryngol Head Neck Surg 2015;141 (12):1075–81. [46] Urban D, Rao A, Bressel M, Neiger D, Solomon B, Mileshkin L. Suicide in lung Cancer who is at risk? Chest 2013;144(4):1245–52. [47] Schneider KL, Shenassa E. Correlates of suicide ideation in a population-based sample of cancer patients. J Psychosoc Oncol 2008;26(2):49–62. [48] Iwasaki M, Akechi T, Uchitomi Y, Tsugane S, Ctr-Based JPH. Cigarette smoking and completed suicide among middle-aged men: a population based cohort study in Japan. Ann Epidemiol 2005;15(4):286–92. [49] Tanskanen A, Tuomilehto J, Viinamaki H, Vartiainen E, Lehtonen J, Puska P. Smoking and the risk of suicide. Acta Psychiatr Scand 2000;101(3):243–5. [50] Sankaranarayanan A, Mancuso S, Castle D. Smoking and suicidality in patients with a psychotic disorder. Psychiatry Res 2014;215(3):634–40. [51] Oh CM, Won YJ, Jung KW, Kong HJ, Cho H, Lee JK, et al. Cancer statistics in Korea: incidence, mortality, survival, and prevalence in 2013. Cancer Res Treat 2016;48(2): 436–50. [52] Schairer C, Brown LM, Chen BE, Howard R, Lynch CF, Hall P, et al. Suicide after breast cancer: an international population-based study of 723,810 women. J Natl Cancer Inst 2006;98(19):1416–9.
Please cite this article as: J.W. Choi and E.-C. Park, Suicide risk after cancer diagnosis among older adults: A nationwide retrospective cohort study, J Geriatr Oncol, https://doi.org/10.1016/j.jgo.2019.11.006