Social Drivers of Cancer Risk and Outcomes among African American Men

INTRODUCTION

According to the National Center for Health Statistics, men from racial minority groups continue to experience poor health outcomes compared to non-minority men. For instance, the life expectancy for African American men (67.6) is lower compared to the life expectancy for white men (74.8). ¹ Over the past two decades, effort have been made to identify factors that contribute to the lower life expectancy among African American men; the national coverage of the acute and chronic stressors that African American men face has raised awareness about the stressful lived experiences and challenges these men can face daily. Consistent with this, stress and stress responses have been recognized as important drivers of mental and physical outcomes among African Americans for decades, ² but these factors are just now being incorporated into conceptual models of minority health and cancer health disparities. For instance, Linnenbringer and colleagues integrated findings from epidemiological studies, health disparities research, and stress biology to develop a conceptual framework to illustrate the pathways through which physiological and behavioral stress responses contribute to disparities in breast cancer sub-types in African American and white women. ³ In addition to being included in conceptual frameworks of cancer health disparities, stress responses are also a focus of national research initiatives such as the Science of Behavior Change Network. ^4,5

Animal and human studies have also been conducted to examine the contribution of social factors, and the physiological underpinnings of responses to adverse social conditions, to cancer health disparities. One hypothesis being tested across studies is that social conditions and physiological responses to social stressors influence biological mechanisms that are involved in the initiation and progress of cancer and responses to treatment. ⁶ This hypothesis is based on data from animal studies which have shown that rats who are exposed to a social stressor (e.g., social isolation) had a greater and more prolonged corticosterone response compared to those who were not exposed to the social stressor. ⁷ Further, rats who demonstrated a longer and elevated corticosterone response also had a greater tumor burden. ⁷ Other reports have described the ways in which exposure to social stressors (e.g., disadvantaged neighborhoods, racial discrimination) could be a link between biological processes and disease risk, and contribute to risk exposure behaviors such as physical inactivity, unhealthy eating, and cigarette smoking. ^8–10 However, exposure to racial discrimination is only one type of social stressor and there are many other types of stressors to which racial minorities can be exposed. Accordingly, research is also examining racial disparities in physiological responses to chronic stress exposure using more comprehensive measures.

Allostatic load is an indicator of biological dysregulation across several physiological systems in response to chronic stress exposure. ¹¹ Research is now being conducted to examine the nature, distribution, and impact of allostatic load on disease risk and outcomes as part epidemiological research, health services studies, and transdisciplinary and translational centers among cancer patients and other groups. ⁵ For instance, African American women in the National Health and Nutrition Examination Survey who had breast cancer had higher allostatic load compared to those who did not have a personal history of disease. ¹² Allostatic load was also positively associated with mortality in a cross-sectional observational cohort study among patients who had metastatic non-small cell lung cancer. ¹³ Further, among prostate cancer patients who were being treated with androgen deprivation therapy (ADT), non-Hispanic Black patients had higher allostatic load before being diagnosed compared to non-Hispanic white patients. ¹⁴ While there were no racial differences in changes in allostatic load following treatment, those who were treated with ADT had significantly greater monthly increases in allostatic load compared to patients who were not treated with ADT. Examining changes in allostatic load during cancer treatment and evaluating the association between allostatic load and cancer mortality is critical to expanding the evidence base about effects of allostatic load on disease risk, response to treatment, and outcomes of cancer care. These findings are needed to determine the utility of using allostatic load as a framework for measuring the physiological consequences of exposure to stressors and to develop precision strategies that address cancer health disparities. However, individuals from racial/ethnic minority groups are likely to experience several different types of stressors before, during, and after being diagnosed with and treated for cancer. Further, exposure to adverse social stressors may be necessary but not sufficient to contribute to disease risk and poor outcomes. That is, not all African Americans who are exposed to social stressors develop advanced stage cancer and individuals who have limited exposure to stressors may develop advanced stage disease. This is because stress reactivity, or how an individual responds physiologically to a stressor, is highly individualized and depends on psychological and social factors.

Basic behavioral science research is now being conducted to characterize stress reactivity in African American breast and prostate cancer patients and other clinical groups and populations as part of the priorities that have been identified for minority health and health disparities. ¹⁵ To increase the precision and quality of cancer care for particularly African American men, it is important to identify specific types of stressors, understand perceptions of stress, and characterize the strategies these men are likely to use to manage stress and stress responses. This report summarizes the results from qualitative and quantitative studies that have examined stress exposure, stress perceptions, and coping strategies among African American. Data from studies that were conducted as part of a transdisciplinary and translational research center to also presented to provide additional insight about the nature of racial differences in specific stressors among African American and white prostate cancer patients.

Sources and Impact of Stress Exposure

According to the American Psychological Association, stress is defined as the psychological or physiological response to a stressor; stressors can be internal or external. ¹⁶ Both qualitative and quantitative studies have examined exposure and reactions to stressors among African American men. For instance, Chung and colleagues examined sources of stress in a community-based sample of African American men in Los Angeles, CA (n=295) and identified the resources these men used to manage stressors. ¹⁷ This study found that more than 90% of men reported experiencing some type of stress and the most common types of stressors were related to money and finances (60.8%) and racism (43.2%). Consistent with this, African American men in a community-based sample in the midwestern region of the US reported that navigating stress at home and work were part of their daily, lived experiences and stress was experienced as a result of attempting to fulfill their roles as the economic provider, managing concerns about potential unemployment, and managing multiple roles and obligations simultaneously. ¹⁸ In a related line of work, Bauer and colleagues found that 36% of men who were receiving prostate cancer care in the Veterans Affairs Health System reported that they did not have enough money left over at the end of the month. ¹⁹ Men also reported experiencing stress as a result of being a black man in the US. These studies highlight the importance of clarifying the context and populations within which stress exposure and responses are being measured. Compared to men in general community samples, patients who are receiving health care in an equal access health care system may have different financial constraints and stressors. Consistent with this, geographic factors related to where individuals live are now being examined in cancer health disparities research along with clinical factors and health care characteristics.

Studies are also now examining the association between studies and the association between residential location and cancer risk and outcomes because residency in a particular geographic region has implications for the types of health care services that individuals can access, the policies that govern when and how services are obtained, and the resources that exist for health promotion and disease control. Recent research has shown that African American men who have a personal history of prostate cancer are more likely to live in geographic areas that have high levels of social deprivation compared to white men who have a personal history of this disease. ²⁰ Zeigler-Johnson et al. also found that census level neighborhood characteristics and self-reported socioeconomic factors interacted synergistically with risk of biochemical recurrence among African American and white prostate cancer patients. ²¹ African American and white prostate cancer patients who lived in neighborhoods with greater economic deprivation had more aggressive disease, but this association was most pronounced among African American patients.

African American men may be exposed to some types of stressors, regardless of where they live, however. Edwards et al. demonstrated that the risk of being killed by police is highest among African American men compared to men and women from other racial/ethnic groups (e.g., White, Asian/Pacific Islander, Latinx) and based on their prediction models, African American men are 2.5 times more likely to be killed by the police over their life course compared to white men. ²² Relatedly, in a community-based study that included African American and white men, African American men were more likely to report a personal experience with being treated unfairly by police (35.9%) compared to white men (15.5%).²³ Experiencing personal or vicarious experiences with unfair treatment by the police was also associated with shorter telomeres only among African American men in this study. ²³ Telomeres are located at the end of chromosomes and protect DNA from degrading; ²⁴ shorter telomeres have been associated with acute and chronic conditions. Telomeres shorten naturally as part of the aging process, but premature shortening of telomeres is a biological response to cellular stress. ²⁵ Studies are now being conducted to examine the association between telomere length and exposure to neighborhood stressors. ^26,27 Research has also been conducted to examine the relationship between perceptions of stress and psychological and behavioral outcomes. This work has shown that greater perceived stress is negatively associated with the likelihood of meeting recommended guidelines for physical activity²⁸ and more cancer-related psychological distress. ²⁹

As illustrated in the sections above, research on stress exposure have focused on understanding psychological reactions in terms of perceived stress. However, there is physiological response to stress. In a related line of research, Hoyt et al. found that prostate cancer patients who used avoidant coping strategies had a greater dysregulated cortisol response. ³⁰ Allen et al. found that African American men were more likely than white men to exhibit blunted cortisol responses using data from the MIDUS II study. ³¹

Case Example: Exposure to Specific Stressors in Prostate Cancer Patients

As efforts are being made to increase the precision of cancer care through personalized therapies, similar attempts are being developed to increase the quality of cancer care through an improved understanding of the patient’s social risk factors and unmet needs. These efforts are being driven in part by mandate from federal organizations and recommendations from professional groups to screen patients for social risk factors and other non-medical drivers of disparities in disease risk and outcomes. ³² Social isolation, for instance, is an important correlate of morbidity and mortality. ³³ Social isolation refers to the number of social connections and the extent to which individuals perceive that the quality and quantity of their social relationships are sufficient to meet their needs. ³⁴ As part of the Transdisciplinary Collaborative Center in Precision Medicine and Minority Men’s Health, ³⁵ a multilevel strategy was used to understand social isolation among prostate cancer patients. In addition to development machine learning tools to automatically, identify patients who had social isolation documented in their electronic health record, ³⁶ participants (n=124) in a retrospective cohort of African American and white prostate cancer patients also provided self-reported data on social isolation using the Short-Form of the Loneliness Scale. ³⁷ A one-sample t-test showed that patients reported levels of social isolation that were significantly different from zero (M=3.62, SD=1.2 (95% CI=3.40, 3.83; t=33.04, p=0.0001). Scores for social isolation were then re-coded into a dichotomous variable (≥4 versus <4) for subsequent analyses because they were skewed. African American patients (43%) were more likely to report social isolation compared to white patients (20%) (Chi Square=6.18, p<0.01). Social isolation also differed between patients based on their income and education. Only income had a significant independent association with social isolation in the multivariate logistic regression analysis. Patients who had incomes less than $50,000 had an increased likelihood of reporting high levels of social isolation compared to those whose household income was greater than $50,000 (OR=8.27, 95% CI=2.69, 25.41, p=0.0002).

Although race did not have significant independent association with social isolation in the multivariate logistic regression model, it is important to note that race and income are highly correlated; lack of financial resources is a source of stress among African American men. ¹⁷ At the same time, limited financial resources is likely to be stressor among all individuals regardless of their racial or ethnic background. Social determinants of health include economic stability, education, and access to quality health care; patients can be asked about their income, education, and employment status as part of screening for social risk factors. A possible explanation for the positive association between lower income and social isolation is because patients who have limited economic resources have less free or flexible time to engage in social companionship activities because of myriad reasons such as the physical burden of their jobs and family responsibilities. Further, the stress of balancing a household budget with limited economic resources or being overwhelmed with job and family responsibilities can create a sense of facing the world alone. The findings from the case example emphasize the importance of asking prostate cancer patients about their social connections and their perceptions about the quality of these relationships.

CONCLUSIONS AND FUTURE DIRECTIONS

Now more than ever, health care systems and cancer care providers are expected to identify patients who have social risk factors and unmet needs. ^38,39 Screening for social drivers of health is now a quality metric that will be used by CMS tools as part of reimbursement calculations for health care delivery. ⁴⁰ The CMS tool screens for financial strain, employment, education, family and community support, physical activity, substance abuse, mental health, and disability. ⁴¹ Screening for these domains is consistent with the findings in this report. Financial strain, because of employment instability, is a source of stress reported by African American men. While there were no differences in social isolation among prostate cancer patients who were enrolled in a retrospective cohort, lower levels of income were associated with an increased likelihood of social isolation. The CMS screening tool also asks about stress (e.g., feeling tense, restless, or nervous) in addition to screening for depression. As work continues to screen patients for social drivers as part of cancer care, it will be important to examine differences in these variables based on racial background and other social and clinical characteristics. With the ability to generate allostatic load using data in the electronic health records, it will also be important to examine the association between social drivers measured as part of screening and allostatic load at critical points in the trajectory of cancer care (e.g., diagnosis, treatment, and recovery). These analyses may need to prioritize understanding the distribution of social drivers among African American men because their lived experiences with social stressors such as unfair police treatment may result in increased cancer risk and poorer outcomes. Work is now being done to understand the lived experiences of African American men; ⁴² as these methods are validated and demonstrate acceptable reliability and validity in terms of capturing men’s experiences with discrimination, microaggressions, and other socially-based stressors, it will be important to integrate these instruments and measures into multilevel research on stress responses and reactivity and cancer care delivery.

Sample Characteristics (n=124) and Bivariate Analysis^‡

Social Isolation
Variable	Level	n (%)	% Social Isolation	Chi Square
Race	African American White	30 (24%) 93 (76%)	43% 20%	6.18**
Marital Status	Married Not Married	108 (88%) 15 (12%)	25% 40%	1.51
Education Level	≥Some college ≤High school graduate	84 (70%) 36 (30%)	21% 42%	5.18*
Employment Status	Employed Not Employed	47 (41%) 67 (59%)	32% 27%	0.34
Income†	≤$50,000 >$50,000	41 (33%) 83 (67%)	54% 13%	22.94***
Chronic Disease Status	At Least One None	96 (77%) 28 (23%)	25% 32%	0.57
Age§	Mean (SD)	66.7 (6.0)

^‡n may not equal 124 because of missing data.

^†Median household income at county level was used in the analysis for patients with missing data on self-reported income.

^§High Social Isolation Mean (SD)=65.4 (5.8) versus Low Social Isolation Mean (SD)=67.2 (6.1); t=1.43, p=0.16

^***p<0.001;

^**p<0.01;

^*p<0.05