Trends in cancer-related suicide in the United States: a population-based cohort study spanning 40 years of data

doi:10.21203/rs.3.rs-3305611/v1

Download PDF

Article

Trends in cancer-related suicide in the United States: a population-based cohort study spanning 40 years of data

https://doi.org/10.21203/rs.3.rs-3305611/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 27 May, 2024

Read the published version in Translational Psychiatry →

Version 1

posted

You are reading this latest preprint version

There is a lack of large cohort studies examining trends in cancer-related suicide. This study addresses this gap by analyzing data from the Surveillance, Epidemiology, and End Results (SEER) database, encompassing a total of 4,870,410 patients diagnosed with cancer from 1975 to 2017. Age-adjusted suicide rates were examined using joinpoint regression to estimate the annual percent change (APC) and average annual percentage change (AAPC). In the past 40 years, studies have revealed a gradual increase in cancer-related suicide rates from 1975 to 1989, followed by a gradual decrease from 1989 to 2013, and a marked decrease from 2013 to 2017. These trends indicate the potential impact of advancements in psychosocial care for patients with cancer in contributing to the observed decrease in suicide rates.

Health sciences/Diseases

Scientific community and society/Scientific community

Suicide

Trend

Cancer patients

Cancer, the second leading cause of death across all age groups in the United States (US) (National Center for Injury Prevention and Control, February 20, 2020), is often accompanied by a higher risk of suicide. Suicide is the tenth leading cause of death among individuals of all ages, with 47,511 suicide cases recorded in 2019 (National Center for Injury Prevention and Control, February 20, 2020). Furthermore, suicide is recognized as a significant contributor to premature mortality (Fazel and Runeson, 2020, Hedegaard et al., 2020, WISQARS, 2021). Patients with cancer face an elevated risk of suicide, primarily attributable to psychological distress, advert effects of multiple treatments, diminished quality of life, and uncontrolled and intense cancer-related pain (Anguiano et al., 2012, Das, 2019, Misono et al., 2008, Zaorsky et al., 2019). The shift toward a bio-psycho-social medical model has underscored the importance of assessing and managing the risk of suicide and implementing suicide prevention strategies and psychological interventions for cancer patients (Fazel and Runeson, 2020, Kumar et al., 2017, Sullivan et al., 2018, Teno et al., 2013).

With advancements in cancer screening and treatments and increased awareness regarding the importance of providing psychological support to patients with cancer, the trends of cancer-related suicide among patients with cancer in the US have changed. However, current evidence regarding these trends is scarce. Only two previous studies have attempted to investigate this issue. One study conducted by Han et al. using the Multiple Cause of Death database reported a decreasing trend in suicides among individuals with cancer from 1999 to 2018. While this study is interesting, it does not specifically reflect the trends among patients with cancer; hence, its findings may not be applicable (Han et al., 2021). Another study by Violette et al. analyzed data from the Surveillance, Epidemiology, and End Results (SEER) database, focusing on 467,368 women diagnosed with uterine cancer, ovarian cancer, or cervical cancer from 1973 to 2013. They reported a downward trend in suicide rates in this specific subset of gynecologic malignancies (Violette et al., 2019). However, the broader profile of suicide trends among patients with cancer in the US over the past decades remains unknown.

Based on the current research status, we propose the hypothesis that trends of cancer-related suicide in the US exhibit corresponding changes in line with shifts in cancer treatment and prevention strategies. Analyzing the suicide trends within the cancer population can provide valuable insights into potential suicide risks, identify changes in trends, and guide clinical practices within the context of the social psychological medical model. Furthermore, such analyses can promote the improvement of relevant policies aimed at improving the well-being and care of patients with cancer.

2.1. Data Source and Study Population

We examined the age-adjusted rates and trends of cancer-related suicide in the US over a 40-year period, specifically from 1975 to 2017. All data were accessed from the SEER Research Plus Data (Scosyrev et al., 2012), which accounts for approximately 10% of the US population (based on the 2010 census). The SEER database is a publicly available, federally funded cancer reporting system that represents a collaboration between the US Centers for Disease Control and Prevention, the National Cancer Institute, and regional and state cancer registries. In contrast to other commonly used datasets, the SEER database is considered population-based, because it collects information on all cancer cases within a specific region and/or defined racial/ethnic population. Therefore, it serves as a nationally representative, population-based cancer reporting system that includes all cancer cases within specific geographic regions of the US (Doll et al., 2018, Wu et al., 2021).

Age-adjusted suicide rates for the total US population were examined through the US mortality data, which is collected and maintained by the National Center for Health Statistics. The inclusion criteria for the study were as follows: only patients diagnosed with cancer, those with cancer originating from any site, and those diagnosed from 1975 to 2017. Finally, a total of 4,870,410 patients were identified, among whom 8,114 died by suicide. Cases were identified as suicide-related deaths only if the “COD to site record” class in the SEER program was coded as “Cause of Death (COD) and Follow-up. COD to site recode” = “Suicide and Self-Inflicted Injury'”. This study was approved by the institutional review board at the National Cancer Center of China, and the requirement of obtaining written informed consent from the participants was waived as the study used a deidentified publicly available registry database.

2.2. Statistical Analysis

To examine the trend of cancer-related suicide in the general US population, we used a standardized method using the SEER*stat software(Fox, 1987). Age-adjusted rates were calculated and standardized to the US standard population in the year 2000 and expressed per 100,000 person-years, as this approach is widely accepted and statistically sound. Annual percent change (APC) and average annual percentage change (AAPC) were calculated using the standardized method implemented in the Joinpoint Trend Analysis Software (Kim et al., 2000). APC measures the rate of change in cancer rates over time, assuming a constant percent change relative to the rate of the previous year. AAPC, on the other hand, is a summary measure of the average APCs over a pre-specified fixed interval, allowing for a single number to represent the trend over multiple years. In our analyses, we selected the best-fitting log-linear regression model as the final model to identify joinpoints where APCs changed significantly, and the maximum joinpoints were set as 2. Student’s t-tests were performed to determine whether the APCs were statistically significant. All statistical analyses were performed using the Surveillance Research Program, National Cancer Institute SEER*Stat software (seer.cancer.gov/seerstat) version 8.3.9, the Joinpoint Regression Program, Version 4.8.0.1 - April 2020) developed by the Statistical Methodology and Applications Branch, Surveillance Research Program, National Cancer Institute, and R Statistical Software version 4.0.3 (R Foundation for Statistical Computing). All statistical tests were two-sided, and P < .05 was considered as the cutoff value for determining statistical significance.

3.1 Overall Findings and Baseline Characteristics

Among the 4,870,410 patients (51.2% female and 48.8% male) diagnosed with cancer from 1975 to 2017, a total of 8,114 patients died by suicide. The majority of these were male (81.7%), white (92.8%), and aged between 50 and 79 years (72.5%). The age-adjusted rates of cancer-related suicide categorized by the characteristics of the study population are shown in Table 1. Overall, we found that the trends of suicide rates varied considerably between patients with cancer and the general US population (Fig. 1). The results of joinpoint regression analysis are shown in Table 2. We found a significant increasing trend of suicide rates in the US population as a whole since the year 2000 (APC, 1.7%; 95%CI, 1.5–1.9%; P < .001). Notably, there has been a decreasing trend in cancer-related suicide rates since 1989, especially in the most recent five years (AAPC, − 27.3%; 95%CI, − 37.2% to − 15.8%, P < .05). We also found variations in the trends of cancer-related suicide rates based on sex, race, age, and registry center. In the most recent five years (2013–2017), significant declines in cancer-related suicide rates were observed among male patients (AAPC, − 20.4%; 95%CI, − 27.0% to − 13.1%; P < .05), white patients (AAPC, − 25.5%; 95%CI, − 35.7% to − 13.8%, P < .05), and patients aged 15 to 49 years (AAPC, − 16.8%; 95%CI, − 26.3% to − 6.2%; P < .05) and 60 to 79 years (AAPC, − 34.0%; 95%CI,−48.5% to − 15.4%; P < .05 for 60–69 years and 95CI% −29.0–7.2% for 70–79 years). Similar declining trends were observed in specific registry centers, including San Francisco-Oakland SMSA, Connecticut, Utah, and Atlanta (Metropolitan) (Table 2).

Table 1

Trends in cancer-related suicide rates in the United States from 1975 to 2017
	Trend 1				Trend 2			Trend 3			Fixed interval 2013–2017
		Years	APC,% (95% CI)	P	Years	APC,% (95% CI)	P	Years	APC,% (95% CI)	P	AAPC,% (95% CI)
Overall	1975–1989		2.6 (1.4,3.8)	< .001	1989–2013	−3.9 (− 4.4,−3.4)	< .001	2013–2017	−27.3 (− 37.2,−15.8)	< .001	−27.3* (− 37.2,−15.8)
Sex
Male	1975–1989		3.5 (2.3,4.8)	< .001	1989–2012	−4.2 (− 4.7,−3.6)	< .001	2012–2017	−20.4 (− 27.0,−13.1)	< .001	−20.4* (− 27.0,−13.1)
Female	1975–2007		−2.0 (− 2.8, − 1.3)	< .001	2007–2017	−9.8 (− 15.4,−3.8)	.002				−9.8* (− 15.4,−3.8)
Race
White	1975–1989		2.9 (1.6,4.2)	< .001	1989–2013	−3.7 (− 4.3,−3.2)	< .001	2013–2017	−25.2 (− 35.7,−13.8)	< .001	−25.5* (− 35.7,−13.8)
Black	1975–1994		1.6 (− 2.0,5.3)	.373	1994–2017	−5.6 (− 8.2,−2.9)	< .001				−5.6* (− 8.2,−2.9)
Other	1975–2000		0.0 (− 1.9,1.9)	.996	2000–2017	−7.6 (− 10.4,−4.7)	< .001				−7.6* (− 10.4, − 4.7)
Age
15–49	1975–1986		1.4 (− 1.8,4.8)	.385	1986–2010	−3.5 (− 4.5,−2.5)	< .001	2010–2017	−16.8 (− 26.3,−6.2)	.004	−16.8* (− 26.3,−6.2)
50–59	1975–2002		−0.5 (− 1.4,0.4)	.306	2002–2017	−7.3 (− 9.8,−4.8)	< .001				−7.3* (− 9.8,−4.8)
60–69	1975–1991		2.4 (0.9,4.0)	.002	1991–2013	−4.9 (− 5.9,−3.9)	< .001	2013–2017	−34.0 (− 48.5,−15.4)	.002	−34.0* (− 48.5,−15.4)
70–79	1975–1986		6.0 (2.2,9.9)	.002	1986–2011	−3.3 (− 4.3,−2.4)	< .001	2011–2017	−18.9 (− 29,−7.2)	.003	−18.9* (− 29.0,−7.2)
>80	1975–1988		5.6 (2.1,9.1)	.002	1988–2017	−3.5 (− 4.3,−2.7)	< .001				−3.5* (− 4.3,−2.7)
Registry
San Francisco-Oakland SMSA	1975–1989		1.7 (− 0.4,4.1)	.168	1989–2014	−5.2 (− 6.4,−4.0)	< .001	2014–2017	−45.3 (− 73.9,14.7)	.107	−37.2 (− 63.3,7.3)
Connecticut	1975–1979		25.2 (− 0.2,57.8)	.052	1979–2013	−2.5 (− 3.3,−1.6)	< .001	2013–2017	−44.2 (− 65.9,−8.4)	.022	−44.2* (− 65.9,−8.4)
Detroit (Metropolitan)	1975–1983		8.1 (1.6, 15.1)	.016	1983–2007	−2.6 (− 3.7,−1.4)	< .001	2007–2017	−10.5 (− 16.1,−4.5)	.001	−10.5* (− 16.1,−4.5)
Hawaii	1975–1991		2.3 (− 2.2,7.0)	.310	1991–2017	−4.7 (− 6.6,−2.8)	< .001				−4.7* (− 6.6,−2.8)
Iowa	1975–1980		7.8 (− 3.2,20.0)	.165	1980–2006	−1.2 (− 2.0,−0.3)	.007	2006–2017	−8.6 (− 12.3,−4.8)	< .001	−8.6* (− 12.3,−4.8)
New Mexico	1975–2003		1.5 (0.2,2.8)	.030	2003–2017	−11.1 (− 14.4,−7.7)	< .001				−11.1* (− 14.4,−7.7)
Seattle (Puget Sound)	1975–1987		4.7 (1.6,7.9)	.004	1987–2011	−4.3 (− 5.3,−3.3)	< .001	2011–2017	−13.7 (− 22.9,−3.5)	.011	−13.7* (− 22.9,−3.5)
Utah	1975–2013		−1.6 (− 2.4,−0.9)	< .001	2013–2017	−34.1 (− 55.5,−2.3)	.038				−34.1* (− 55.5,−2.3)
Atlanta (Metropolitan)	1975–1991		3.4 (1.0,5.9)	.006	1991–2014	−6.0 (− 7.3,−4.6)	< .001	2014–2017	−38.8 (− 65.0,7.0)	.083	−31.8 (− 54.5,2.1)

Table 2

Demographic characteristics and age-adjusted rates of suicide among cancer survivors per SEER Data, 1975–2017
Characteristics	Cancer Patients	Suicides, No. (%)	Age-adjusted rate, per 100,000 (95% CI)
Overall	4870410 (100)	8114 (100)	0.76 (0.74–0.78)
Sex
Male	2492120 (51.2)	6633 (81.7)	1.4 (1.36–1.43)
Female	2378290 (48.8)	1481 (18.3)	0.26 (0.25–0.28)
Race
White	4065373 (83.5)	7517 (92.8)	0.86 (0.84–0.88)
Black	456176 (9.4)	253 (3.1)	0.26 (0.23–0.30)
Other	324454 (6.7)	327 (4.1)	0.33 (0.29–0.36)
Age
15–49	690099 (14.2)	1438 (17.8)	0.27 (0.26–0.29)
50–59	846062 (17.4)	1479 (18.3)	1.18 (1.13–1.25)
60–69	1273999 (26.2)	2294 (28.4)	2.59 (2.48–2.70)
70–79	1239232 (25.4)	2079 (25.8)	3.69 (3.53–3.85)
>80	786521 (16.1)	782 (9.7)	2.39 (2.23–2.57)
Registry
San Francisco-Oakland SMSA	754866 (15.5)	1528 (18.8)	0.9 (0.86–0.95)
Connecticut	744029 (15.3)	743 (9.2)	0.48 (0.45–0.52)
Detroit (Metropolitan)	840258 (17.3)	1092 (13.5)	0.64 (0.61–0.68)
Hawaii	207104 (4.3)	312 (3.8)	0.62 (0.55–0.69)
Iowa	634916 (13.0)	959 (11.8)	0.7 (0.66–0.75)
New Mexico	273923 (5.6)	765 (9.4)	1.09 (1.01–1.17)
Seattle (Puget Sound)	746086 (15.3)	1511 (18.6)	0.97 (0.92–1.02)
Utah	268980 (5.5)	582 (7.2)	0.84 (0.77–0.91)
Atlanta (Metropolitan)	400248 (8.2)	622 (7.7)	0.69 (0.64–0.75)

3.2 Variations in Suicide Trends by Sex

We further examined the profile and variations in suicide trends by sex between patients with cancer and the general US population (Fig. 2). Among the general US population, we identified two joinpoints for males; suicide rates increased from 1975 to 1990, then considerably decreased from 1990 to 2003, but have been markedly increasing since then (Fig. 2A). Among females in the general US population, suicide rates decreased from 1975 to 2000, but markedly increased in the past 17 years (Fig. 2B). In contrast, the cancer-related suicide trends showed a significantly different pattern. Notably, cancer-related suicide rates among female patients have been consistently decreasing over the past 40 years, with a remarkable decrease observed since 2007. Overall, cancer-related suicide rates have been decreasing among both male and female patients in recent years, whereas they have been increasing in the general US population. Moreover, in the most recent five years, male patients exhibited a greater decline in cancer-related suicide rates compared to female patients (Table 2).

3.3 Variations in Suicide Trends by Race

We further analyzed the profile and variations of suicide trends by race between patients with cancer and the general US population (Fig. 3). In the general US population, the trends of suicide rates varied among different races. We identified two joinpoints for the white race; suicide rates gradually decreased from 1975 to 1995 and significantly decreased from 1995 to 2000, followed by a marked increase since the year 2000 (Fig. 3A). In the black race, we observed no significant changing trend in suicide rates from 1975 to 1993, followed by a decrease from 1993 to 2007, and a marked increase in the past decade (Fig. 3B and 3C). In the case of patients with cancer, we found that the trend of cancer-related suicide rates among white patients was similar to the overall trend among all patients (Fig. 3D and 3E). Among black patients, cancer-related suicide rates gradually increased from 1975 to 1994 and then markedly decreased since 1994. In summary, cancer-related suicide rates among all races have been decreasing in recent years, albeit following different patterns, whereas they have been increasing among the general US population. Specifically, from 2013 to 2017, white patients exhibited a larger decline in cancer-related suicide rates compared to other races (Table 2).

3.4 Variations in Suicide Trends by Age

We compared the profile and variations of suicide trends by age between patients with cancer and the general US population (Fig. S1). In the general US population, suicide rates experienced different patterns across different age groups. Among individuals aged 15 to 49 years, suicide rates gradually decreased from 1975 to 1996, markedly decreased from 1996 to 2000, and markedly increased thereafter (Figs. S1A–E). In contrast, we found significantly different patterns of cancer-related suicide trends compared to the general US population. Among patients aged 15 to 49 years, cancer-related suicide rates gradually decreased from 1998 to 2010, followed by a marked decrease thereafter (Fig. S1F). Similarly, among patients aged 50 to 59 years, cancer-related suicide rates gradually decreased from 1975 to 2002, followed by a marked decrease thereafter (Fig. S1G). Interestingly, a similar changing trend was observed in terms of cancer-related suicides among individuals aged 60 to 69 years and 70 to 79 years (Figs. S1H and I). We found that cancer-related markedly increased from 1975 to 1988 among these age groups, which is consistent with the trend observed in the overall US population. However, there has been a marked decrease in suicide rates since then (Fig. S1J). From 2013 to 2017, patients aged 60–79 years and 15–49 years exhibited a larger decline in cancer-related suicide rates compared to those in other age groups.

In our study, we performed in-depth analyses of the trends of cancer-related suicide in the US over a 40-year period. To the best of our knowledge, this is the first study to depict the changing pattern of cancer-related suicide trends and its variations based on the characteristics of patients and compare them with the general US population. By examining the data by sex, race, age, and registry center, we observed variations in cancer-related suicide rates among different subgroups of patients. One significant finding is the decreasing trend of cancer-related suicide rates among patients since 1989, with a remarkable decrease in the most recent five years. While our study does not establish a causal relationship, it suggests that improvements in psychosocial care over the last few decades, along with advancements in cancer treatment and prevention, may have contributed to this downward trend. It is important to note that despite the decreasing cancer-related trend among patients, overall suicide rates in the general US population have increased.

The finding of an increasing trend of cancer-related suicide in the US followed by a peak in 1989 and subsequent descent is interesting. The significant decline in suicide rates, especially among male patients, white patients, and those aged 60 to 69 years, in the more recent years adds to the complexity of the trend. We speculate that this is attributable to the following reasons. Firstly, there was a bottleneck in antitumor medicine for a long period since chemotherapy was first used in 1948 (Lippman and Hawk, 2009). Secondly, as advancements were made in cancer treatment and surgical techniques, the landscape of cancer care began to change. Traditional open surgery can have several challenges and potential complications including intensive postoperative pain, large body trauma, slow healing, influence on cardiopulmonary function, prolonged hospitalization, and high costs, which could seriously affect the quality of life of patients undergoing such procedures (Wyld et al., 2015). Thirdly, the financial burden associated with cancer care cannot be overlooked. The financial strain experienced by patients with cancer may extend beyond medical expenses to include caregiving, transportation, supplies, and childcare. Moreover, compared to individuals without a cancer history, those with cancer are more likely to face unique psychological and behavioral challenges, including emotional distress and material, medical, and financial adversity (Yabroff et al., 2018, Zheng et al., 2019).

The sharp slump in the rate of cancer-related suicide between 2013 and 2017 may be explained by the following reasons. The expansion of Medicaid has been associated with increased access to healthcare services and improved affordability of medications and treatments for many individuals, including patients with cancer (Han et al., 2020, Hendryx and Luo, 2018). In addition to the promising advances in medical treatments for malignancies, this period witnessed an evolving role of psycho-oncology care and palliative and hospice care, leading to the promotion and increased utilization of these services by patients with cancer, enhancing their overall quality of life (Council, 2001, Kumar et al., 2017, Page AE 2008, Sullivan et al., 2018, Teno et al., 2013). Moreover, the development of integrated care models, including collaborative care models, has provided a more comprehensive and coordinated approach to cancer care (Aaronson et al., 2014a, Aaronson et al., 2014b, Alfano et al., 2012, Stanton et al., 2015). The implementation of all these measures may have contributed to the decline in suicide rates among patients with cancer between 2013 and 2017.

Zaorsky et al. reported an overall suicide rate of 28.58 per 100,000 person-years among patients with cancer, which is significantly higher than the suicide rates reported in the general population (Zaorsky et al., 2019). They also found that the risk of suicide among patients with cancer was 4.4 times higher than that of the general population, which is broadly consistent with our previous work (Liu et al., 2022). These findings highlight the urgent need for addressing the mental health issues of patients with cancer and providing appropriate support and intervention strategies. In contrast, the primary objective of our current study was to examine the trends in cancer-related suicides over the past few decades using joinpoint models. We aimed to analyze the trends in cancer-related suicides over time rather than directly comparing the rates to those observed in the general population, or discussing the absolute numbers reported by Zaorsky et al. Nevertheless, our study aligns with the broader understanding that patients with cancer face a significantly higher risk of suicide and highlights the need for continued efforts to improve mental health support and care for these patients.

Several studies suggest a link between suicide risk and the type or severity of cancer. A Swedish cohort study conducted from 1965 to 1999 revealed that female patients with cancer had a higher suicide risk compared to male patients and that there was a strong inverse correlation between survival and suicide rate. This suggests that cancer types with a worse prognosis may be associated with higher suicide rates (Björkenstam et al., 2005). Choi et al. 's study conducted in South Korea also indicated that suicide risk among patients with cancer varied according to the anatomical site of the cancer (Choi and Park, 2021). However, it did not find an association between suicide risk in cancer patients and the prognosis of cancer. Anderson et al. reviewed data from the SEER database regarding patients diagnosed with cancer of the digestive system from 2000 to 2014, and the results revealed that patients with pancreatic and esophageal cancer had more than five times the risk of suicide compared to the general population, while those with other digestive system cancers had about twice the risk of suicide compared to the general population (Anderson et al., 2018). This suggests that the specific type of cancer can affect suicide risk, with worse prognosis potentially contributing to the higher risk, possibly due to feelings of depression and hopelessness accompanying the challenging prognosis (Heinrich et al., 2022, Li and Rodin, 2022). A study by Kahn et al. analyzed data from the Mental Health Research Network also reported that patients diagnosed with cancer with a poor prognosis in the past year had a nearly five-fold increased risk of suicide compared to the general population, while cancers with an average five-year survival rate of > 70% did not significantly increase the suicide risk (Kahn et al., 2023). These findings highlight the importance of tailoring support and prevention strategies based on the prognosis of the cancer. Patients with cancers that have a very poor prognosis, such as pancreatic cancer, should receive better care, support, and psychological counseling.

By providing insights into the temporal trends of cancer-related suicides, our study contributes to the existing knowledge on the subject and may inform future interventions and support strategies for patients with cancer at risk of suicide. Our study has several strengths. Firstly, our study is the largest and most comprehensive in characterizing the profile of cancer-related suicide trends over a 40-year period. Studying a diverse population of patients with cancer, rather than focusing on specific complications or single-system malignancies, adds to the generalizability and applicability of the findings. Secondly, this study is the first to describe the 40-year changing pattern of cancer-related suicide rates through joinpoint regression analysis. These joinpoints enable us to accurately capture and describe the changing patterns over time. This approach provides a comprehensive map of the cancer-related suicide rate patterns among patients, and comparing these patterns with the suicide rates in the general US population further enhances the understanding of the unique challenges faced by these patients.

Our study has some limitations that should be acknowledged. We have extensively reviewed a large amount of data and literature related to the SEER program (D Park et al., 2012). However, the findings of our study indicate a potential bias. On the one hand, SEER pathology terminology has undergone changes over time due to advances in tumor classification, which has led to long-term challenges regarding the comparability of SEER data captured over more than 40 years. Because of advances in diagnostic testing, recent pathologic diagnoses captured by the SEER database are expected to be more precise than those made 40 years ago. This situation poses a challenge in terms of aligning with new diagnostic criteria and tumor classification systems. Moreover, the National Cancer Institute (NCI) adopted the new behavioral code ICD-O-3 in 2021, which may have an impact on our study. On the other hand, source data inaccuracies can arise from data coding errors transmitted to the NCI by regional registration authorities or incorrect data provided to regional registration services for coding. In addition, although SEER takes into account the proportion of each ethnic group and even oversamples certain racial and ethnic minorities when sampling populations to improve the representation of diverse populations, the proportion of the US population has undergone significant changes over the past 40 years due to various social factors such as economic development and increased migration. This demographic shift poses another limitation to our study. Besides, any underestimation of suicide rates in our study is likely to be similar in patients with cancer and the general population, which reduces the likelihood of major deviations when calculating relevant data. The observed association between cancer and suicide may be confounded by psychiatric disorders and medication use. Unfortunately, we were unable to control for these factors in the current work.

Acknowledgments

The authors thank the National Cancer Institute, Information Management Services, and SEER program registries for creating and maintaining the database.

Ethics approval and consent to participate

The study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines. This study was approved by the institutional review board at the National Cancer Center of China, and the requirement of obtaining written informed consent from the participants was waived as the study used a deidentified publicly available registry database.

Consent for publication

All authors read and approved the final manuscript for publication.

Availability of data and materials

All supporting data are included in the manuscript and supplemental files. Additional data are available upon reasonable request to the corresponding author.

Competing interests

The authors declare no competing interests.

Funding

This work was supported by the China National Key Research and Development Program (No. 2020AAA0105000 and 2020AAA0105004), Natural Science Foundation of China (No. 82173328, No. 81872160, No. 82072940, No. 82103047, No. 82102887), Beijing Municipal Natural Science Foundation (Key Project) (No. 7191009, No. 7204293), CAMS Innovation Fund for Medical Sciences (CIFMS; No. 2021-I2M-CT-A-012, No. 2021-I2M-CT-B-044, No. 2022-I2M-CT-B-087), Non-profit Central Research Institute Fund of Chinese Academy of Medical Sciences (No. 2022-JKCS-04), Beijing Association for Science and Technology's "Golden-Bridge Seed Funding Program" (No. ZZ22027), Beijing Hope Run Special Fund of Cancer Foundation of China (No. LC2019B03, No. LC2019L07, No. LC2020L01), 2021 Chaoyang District Social Development Science and Technology Plan Project (Medical and Health Field) (No. CYSF2115), Chinese Young Breast Experts Research project (No. CYBER-2021-005), XianSheng Clinical Research Special Fund of China International Medical Foundation (No. Z-2014-06-2103), and Beijing Xisike Clinical Oncology Research Foundation (No. Y-Young2021-0017).

Author contributions

Conception and design: Qiang Liu, Jing Wang

Collection and assembly of data: Qiang Liu, Yihang Qi, Wenxiang Zhang

Data analysis and interpretation: Qiang Liu, Yihang Qi, Xiangyu Wang

Manuscript writing: Qiang Liu, Yihang Qi, Xiangyu Wang, Wenxiang Zhang, Xiangyi Kong

Final approval of manuscript: Qiang Liu, Yi Fang, and Jing Wang

Declaration of Interest Statement

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Aaronson, N. K., Mattioli, V., Minton, O., Weis, J., Johansen, C., Dalton, S. O., Verdonck-de Leeuw, I. M., Stein, K. D., Alfano, C. M., Mehnert, A., de Boer, A. & van de Poll-Franse, L. V. (2014a). Beyond treatment - Psychosocial and behavioural issues in cancer survivorship research and practice. European Journal of Cancer, Supplement 12, 54-64.
Aaronson, N. K., Mattioli, V., Minton, O., Weis, J., Johansen, C., Dalton, S. O., Verdonck-de Leeuw, I. M., Stein, K. D., Alfano, C. M., Mehnert, A., de Boer, A. & van de Poll-Franse, L. V. (2014b). Beyond treatment – Psychosocial and behavioural issues in cancer survivorship research and practice. European Journal of Cancer Supplements 12, 54-64.
Alfano, C. M., Ganz, P. A., Rowland, J. H. & Hahn, E. E. (2012). Cancer survivorship and cancer rehabilitation: Revitalizing the link. Journal of Clinical Oncology 30, 904-906.
Anderson, C., Park, E. M., Rosenstein, D. L. & Nichols, H. B. (2018). Suicide rates among patients with cancers of the digestive system. Psycho-oncology 27, 2274-2280.
Anguiano, L., Mayer, D. K., Piven, M. L. & Rosenstein, D. (2012). A literature review of suicide in cancer patients. Cancer Nurs 35, E14-26.
Björkenstam, C., Edberg, A., Ayoubi, S. & Rosén, M. (2005). Are cancer patients at higher suicide risk than the general population? Scandinavian Journal of Public Health 33, 208-214.
Choi, Y. & Park, E.-C. (2021). Suicide after cancer diagnosis in South Korea: a population-based cohort study. BMJ Open 11, e049358.
Council, N. C. P. B. I. o. M. N. R. (2001). Improving Palliative Care for Cancer: Summary and Recommendations. National Academy Press: Washington, DC.
Das, M. (2019). Patients diagnosed with cancer at increased risk of suicide. The Lancet. Oncology 20, e75.
Doll, K. M., Rademaker, A. & Sosa, J. A. (2018). Practical Guide to Surgical Data Sets: Surveillance, Epidemiology, and End Results (SEER) Database. JAMA Surgery 153, 588-589.
Duggan, M. A., Anderson, W. F., Altekruse, S., Penberthy, L. & Sherman, M. E. (2016). The Surveillance, Epidemiology, and End Results (SEER) Program and Pathology: Toward Strengthening the Critical Relationship. The American Journal of Surgical Pathology 40.
Fazel, S. & Runeson, B. (2020). Suicide. The New England journal of medicine 382, 266-274.
Fox, A. J. (1987). Statistical methods in cancer research. Volume II--The design and analysis of cohort studies. Iarc Scientific Publications 82, 1.
Han, X., Hu, X., Zhao, J., Ma, J., Jemal, A. & Yabroff, K. (2021). Trends of Cancer-Related Suicide in the United States: 1999-2018. Journal of the National Cancer Institute.
Han, X., Jemal, A., Zheng, Z., Sauer, A., Fedewa, S. & Yabroff, K. (2020). Changes in Noninsurance and Care Unaffordability Among Cancer Survivors Following the Affordable Care Act. Journal of the National Cancer Institute 112, 688-697.
Hedegaard, H., Curtin, S. & Warner, M. (2020). Increase in Suicide Mortality in the United States, 1999-2018. NCHS data brief, 1-8.
Heinrich, M., Hofmann, L., Baurecht, H., Kreuzer, P. M., Knüttel, H., Leitzmann, M. F. & Seliger, C. (2022). Suicide risk and mortality among patients with cancer. Nature Medicine 28, 852-859.
Hendryx, M. & Luo, J. (2018). Increased Cancer Screening for Low-income Adults Under the Affordable Care Act Medicaid Expansion. Medical care 56, 944-949.
Kahn, G. D., Tam, S. H., Felton, J. W., Westphal, J., Simon, G. E., Owen-Smith, A. A., Rossom, R. C., Beck, A. L., Lynch, F. L., Daida, Y. G., Lu, C. Y., Waring, S., Frank, C. B., Akinyemi, E. O. & Ahmedani, B. K. (2023). Cancer and psychiatric diagnoses in the year preceding suicide. Cancer Medicine 12, 3601-3609.
Kim, H. J., Fay, M. P., Feuer, E. J. & Midthune, D. N. (2000). Permutation tests for joinpoint regression with applications to cancer rates. Statistics in medicine 19, 335-351.
Kumar, P., Wright, A. A., Hatfield, L. A., Temel, J. S. & Keating, N. L. (2017). Family Perspectives on Hospice Care Experiences of Patients with Cancer. J Clin Oncol 35, 432-439.
Li, M. & Rodin, G. (2022). Suicide prevention in cancer: earlier is better. Trends In Molecular Medicine 28, 437-438.
Lippman, S. M. & Hawk, E. T. (2009). Cancer prevention: from 1727 to milestones of the past 100 years. Cancer Research 69, 5269-5284.
Liu, Q., Wang, X., Kong, X., Wang, Z., Zhu, M., Ren, Y., Dong, H., Fang, Y. & Wang, J. (2022). Subsequent risk of suicide among 9,300,812 cancer survivors in US: A population-based cohort study covering 40 years of data. EClinicalMedicine 44, 101295.
Misono, S., Weiss, N. S., Fann, J. R., Redman, M. & Yueh, B. (2008). Incidence of suicide in persons with cancer. J Clin Oncol 26, 4731-8.
National Center for Injury Prevention and Control, C. f. D. C. a. P. (February 20, 2020). National Center for Health Statistics (NCHS), National Vital Statistics System.
Page AE , A. N. (2008). Cancer Care for the Whole Patient: Meeting Psychosocial Health Needs. National Academies Press: Washington, DC.
Park, H. S., Lloyd, S., Decker, R. H., Wilson, L. D. & Yu, J. B. (2012). Limitations and biases of the Surveillance, Epidemiology, and End Results database. Current Problems In Cancer 36, 216-224.
Scosyrev, E., Messing, J., Noyes, K., Veazie, P. & Messing, E. (2012). Surveillance Epidemiology and End Results (SEER) program and population-based research in urologic oncology: an overview. In Urologic Oncology: Seminars and Original Investigations, pp. 126-132. Elsevier.
Stanton, A. L., Rowland, J. H. & Ganz, P. A. (2015). Life after diagnosis and treatment of cancer in adulthood: Contributions from psychosocial oncology research. American Psychologist 70, 159-174.
Sullivan, D., Forsberg, C., Golden, S., Ganzini, L., Dobscha, S. & Slatore, C. (2018). Incidence of Suicide and Association with Palliative Care among Patients with Advanced Lung Cancer. Annals of the American Thoracic Society 15, 1357-1359.
Teno, J. M., Gozalo, P. L., Bynum, J. P., Leland, N. E., Miller, S. C., Morden, N. E., Scupp, T., Goodman, D. C. & Mor, V. (2013). Change in end-of-life care for Medicare beneficiaries: site of death, place of care, and health care transitions in 2000, 2005, and 2009. Jama 309, 470-7.
Violette, C., Mandelbaum, R., Nusbaum, D., Duval, C., Matsuzaki, S., Machida, H., Roman, L. & Matsuo, K. (2019). Temporal trends and characteristics of suicide among women with gynecologic malignancy in the United States. Gynecologic oncology reports 30, 100510.
WISQARS, C. (2021). Ten leading causes of death by age group.
Wu, W.-T., Li, Y.-J., Feng, A.-Z., Li, L., Huang, T., Xu, A.-D. & Lyu, J. (2021). Data mining in clinical big data: the frequently used databases, steps, and methodological models. Military Medical Research 8, 44.
Wyld, L., Audisio, R. A. & Poston, G. J. (2015). The evolution of cancer surgery and future perspectives. Nature Reviews. Clinical Oncology 12, 115-124.
Yabroff, K. R., Zhao, J., Zheng, Z., Rai, A. & Han, X. (2018). Medical Financial Hardship among Cancer Survivors in the United States: What Do We Know? What Do We Need to Know? Cancer Epidemiology Biomarkers & Prevention 27, 1389-1397.
Zaorsky, N. G., Zhang, Y., Tuanquin, L., Bluethmann, S. M., Park, H. S. & Chinchilli, V. M. (2019). Suicide among cancer patients. Nat Commun 10, 207.
Zheng, Z., Jemal, A., Han, X., Guy, G. P., Jr., Li, C., Davidoff, A. J., Banegas, M. P., Ekwueme, D. U. & Yabroff, K. R. (2019). Medical financial hardship among cancer survivors in the United States. Cancer 125, 1737-1747.

The authors have declared there is NO conflict of interest to disclose

FigureS1.tif
Figure S1. Age-adjusted suicide rate per 100,000 person-years and suicide trend varied by age in the past 40 years. Trends of suicide among the total US population by 15-49 years (A), 50-59 years (B), 60-69 years (C), 70-79 years (D), 80+ years (E); Trends of cancer-related suicide in the US by 15-49 years (F), 50-59 years (G), 60-69 years (H), 70-79 years (I), 80+ years (J). APC: annual percentage change. * Indicates that the Annual Percent Change (APC) is significantly different from zero at the alpha=0.05 level.

Download PDF

Journal Publication

published 27 May, 2024

Read the published version in Translational Psychiatry →

Version 1

posted

You are reading this latest preprint version

Trends in cancer-related suicide in the United States: a population-based cohort study spanning 40 years of data

Status:

Journal Publication

Version 1

Abstract

Figures

1. Introduction

2. Methods

2.1. Data Source and Study Population

2.2. Statistical Analysis

3. Results

3.1 Overall Findings and Baseline Characteristics

3.2 Variations in Suicide Trends by Sex

3.3 Variations in Suicide Trends by Race

3.4 Variations in Suicide Trends by Age

4. Discussion

5. Limitations

Declarations

References

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1