Social Support Buffers the Effect of Social Deprivation on Comorbidity Burden in Adults with Cancer

Abstract

Background

Adults with cancer have higher rates of comorbidity compared to those without cancer, with excess burden in people from lower socioeconomic status (SES). Social deprivation, based on geographic indices, broadens the focus of SES to include the importance of “place” and its association with health. Further, social support is a modifiable resource found to have direct and indirect effects on health in adults with cancer, with less known about its impact on comorbidity.

Purpose

We prospectively examined associations between social deprivation and comorbidity burden and the potential buffering role of social support.

Methods

Our longitudinal sample of 420 adults (M_age = 59.6, SD = 11.6; 75% Non-Hispanic White) diagnosed with cancer completed measures at baseline (~6 months post-diagnosis) and four subsequent 3-month intervals for 1 year.

Results

Adjusting for age, cancer type, and race/ethnicity, we found a statistically significant interaction between social support and the effect of social deprivation on comorbidity burden (β = −0.11, p = 0.012), such that greater social support buffered the negative effect of social deprivation on comorbidity burden.

Conclusion

Implementing routine screening for social deprivation in cancer care settings can help identify patients at risk of excess comorbidity burden. Clinician recognition of these findings could trigger a referral to social support resources for individuals high on social deprivation.

Introduction

Cancer-related health disparities persist in the USA, as demonstrated by the unequal burden of cancer incidence, prevalence, and outcomes among socially disadvantaged groups [1, 2]. These disparities are influenced by complex interrelated social determinants of health (SDOH), including economic, social, healthcare access and policy, and deeply embedded structural inequities [3]. Decades of research show stepwise socioeconomic gradients of health by income, education, and occupation, which generally reflect the health benefits of greater economic resources, including adequate housing, healthy nutrition, and greater resources to cope with daily stressors [4–6]. Less understood within the context of cancer is the role of “place” or neighborhood/community level disadvantage on cancer-related health disparities, including comorbidity burden.

Adults with cancer have higher rates of comorbidity compared to those without cancer [7, 8], with excess burden in people of lower socioeconomic status (SES) [1, 9]. The presence of comorbidities is well-documented to result in delay in diagnosis [10], suboptimal cancer treatment [11], poor quality of life [12], and increased mortality risk [13]. People residing in socially disadvantaged communities may be particularly vulnerable to excess comorbidity burden [2, 14]. Aligned with Healthy People 2030, the National Institute on Minority Health and Health Disparities (NIMHHD), and the NIH Social Determinants of Health Research Coordinating Committee which call for more research on SDOH strategies to improve the health and well-being of socially disadvantaged groups [15], our study examines the role of social support as a potential buffer between neighborhood-level deprivation and comorbidity burden.

The stress, social support, and buffering model [16] suggests that higher levels of perceived social support can lessen the negative physical and psychological effects of stress. Studies show higher social support has positive direct effects on cancer patients’ physical [17] and emotional health [18] as well as survival [19]. In both cancer and non-cancer contexts, social support has been found to buffer the health-damaging effects of stress [20, 21], and stress is one proposed mechanism that can help explain the association between social deprivation and comorbidity burden [22]. While social support is a modifiable and beneficial resource found to affect different health outcomes in adults with cancer [23], little is known about its impact on comorbidity burden. It is possible that adequate perceived support can reduce the physiological stress response of social deprivation, facilitate healthy behaviors, or positively impact health care utilization, all of which could contribute to a lower comorbidity burden.

To address these gaps, this study, informed by the stress, social support, and buffering model, aimed to prospectively examine the relationship between social deprivation and comorbidity burden as well as the potential buffering role of social support. We hypothesize that individuals experiencing higher levels of social deprivation will exhibit a greater comorbidity burden, and this relationship will be mitigated by the presence of greater social support.

Methods

Participants and Procedures

We analyzed data from the YUCAN longitudinal, observational study of 569 adults with cancer as they transition from primary treatment to early survivorship [24, 25]. The YUCAN study, which was originally designed to examine mechanisms and outcomes of resilience trajectories, but also affords opportunities for secondary analyses such as these. The original study recruited and enrolled adults (aged 18–80 at diagnosis) between 2019 and 2022 recently diagnosed with breast, prostate, or colorectal cancer, with no prior history of cancer, using the Yale Cancer Center Rapid Case Ascertainment program. Additional inclusion criteria included stages I–III, within 6 months of diagnosis, and able to read and speak English. Participants completed comprehensive physical and psychosocial health questionnaires (online via RedCap), provided hair and nail samples for cortisol analysis, and wore an ActiGraph accelerometer device (for 7-day intervals) to capture physical activity and sleep patterns, at baseline (within 6 months of diagnosis) and at 3-month intervals for 1 year. Participants were compensated with a $50 gift card for each completed questionnaire and an additional $50 bonus if they completed all five-time points (see Bellizzi et al. [24 for additional recruitment and procedural). Our analyses focused on variables collected by questionnaire (2019–2022) as well as patient medical records.

Measures

Sociodemographic variables, including age, gender, race, ethnicity, education, employment status, and marital status, were collected from participant self-report at baseline. Clinical variables, including cancer type, disease stage at diagnosis, diagnosis date, and treatment information, were extracted from patient medical records.

Using participant zip code from patient medical records, a social deprivation index (SDI) score was calculated based on seven demographic characteristics [26]. The seven demographic characteristics include percent living in poverty, percent with less than 12 years of education, percent single-parent household, percent living in rented housing units, percent living in overcrowded housing units, percent of households without a car, and percentage of non-employed adults under 65 years of age. Scores can range from 0 to 100 with higher scores reflecting greater neighborhood-level social deprivation as opposed to the individual level. The SDI reflects the socioeconomic challenges and environmental hardships faced by a specific geographic area and was developed to examine relationships between levels of social disadvantage and health and healthcare [26]. The SDI has demonstrated good criterion validity, converging as expected with cancer mortality, poor health care access, and poorer health outcomes [27, 28]. Other validity criteria are currently unavailable.

Social support was measured using the 19-item MOS Social Support Scale [29] that provides an overall composite score comprising perceived available emotional/informational support, tangible support, affectionate support, and positive social interaction. Response options range from 1 (none of the time) to 5 (all of the time). Higher scores indicate greater overall social support (Cronbach’s alpha in our sample = 0.969).

The 15-item Self-Administered Comorbidity (SCQ) scale [30] was used to measure comorbidity burden. For each medical condition, a participant is asked if the condition is present, receives treatment for it, and if the condition causes limitations in functioning (for a maximum of 3 points per condition). Higher scores indicate a higher comorbidity burden. The SCQ is modestly correlated with the medical record-based Charlson Index [31] and demonstrates adequate criterion validity with health status and resource utilization over time [30].

Statistical Approach

All analyses were conducted in R (version 4.3.2). Descriptive statistics were used to examine means, standard deviations, and frequencies of study variables. To test our hypothesis, we first mean-centered SDI and social support scales. Next, we fit a multiple linear regression model to examine whether the relationship between social deprivation (at baseline) and comorbidity burden (at time 5) differs depending on the level of social support (at baseline) after adjusting for age, race/ethnicity, and cancer type. Simple slopes at the mean, 1 SD below the mean, and 1 SD above the mean of social support were plotted to examine the interaction effect. All statistical analyses were performed with a two-tailed statistical significance at p < 0.05. Due to individual-level random attrition in the YUCAN longitudinal study, we compared the final analytic sample to those who dropped out or who had incomplete data at time five, by age, sex, SDI, and cancer site.

Results

The analytic sample includes 420 adults (M_age = 59.6, SD = 11.6) diagnosed with pathology-confirmed breast (58.8%), prostate (33.1%), or colorectal cancer (8.1%), with predominantly early stage disease (84.4% stage I or II for breast and colorectal cancer; Gleason score M = 6.9 for prostate cancer) (see Table 1 for additional participant characteristics). Participants in the sample (n = 420) had higher SDI scores (t(243) = −2.43, p = 0.017) and included fewer colorectal cancer cases (X² (2) = 8.35, p = 0.015) compared to those who dropped out of the study or who had incomplete data at time 5 (n = 149). There were no statistically significant differences between the groups by age or sex.

Participants with lower social deprivation index scores were less likely to identify as non-Hispanic White (t(148) = −3.43, p < 0.001), be partnered (t(184) = 4.33, p < 0.001), or have earned a higher degree (F(3, 406) = 4.01, p = 0.008). There were no statistically significant differences in social deprivation scores by gender, age, or cancer type.

The average social deprivation index score in the sample was 26.91 (SD = 28.66) while the mean scores for comorbidity burden and social support were 3.29 (SD = 3.71) and 81.53 (SD = 19.95), respectively. Results of the moderation model are presented in Table 2. The overall multiple regression model testing the interaction between social deprivation and social support on comorbidity burden while controlling for age, race/ethnicity, and cancer type was statistically significant (F(6, 413) = 11.24, p < 0.001, R² = 0.128). There was a significant main effect of both social deprivation (β = 0.21, p < 0.001) and social support (β = −0.16, p < 0.001) at time 1, on comorbidity burden 12 months later. Additionally, there was a statistically significant interaction effect between social deprivation and social support (β = −0.11, p = 0.012), demonstrating that social support buffered the impact of social deprivation on comorbidity burden in our sample of adults with cancer. Specifically, greater social deprivation contributed to less comorbidity burden among individuals with high levels of social support when compared to individuals with low levels of social support (Fig. 1).

Discussion

These results suggest greater social deprivation, in adults with breast, prostate, or colorectal cancer, is associated with a higher comorbidity burden, but greater social support appears to play a protective role, after adjusting for age, race/ethnicity, and type of cancer.

Our findings complement prior knowledge in many important ways. First, our study expands on what is known about the gradient of increasing comorbidity burden with lower SES [1], by examining “place” or neighborhood/community level deprivation which has not been examined in relation to comorbidity burden in adults with cancer. Second, a recent scoping review of studies examining the association between multi-morbidity and socioeconomic status tend to be cross-sectional [1], which has methodological drawbacks such as limits to causal inference and lack of temporal sequence. Our analyses took advantage of the YUCAN longitudinal design which contributes to a more accurate and reliable understanding of the role of social deprivation on comorbidity burden over time. Third, the comorbidity measure used in this study provides a more nuanced assessment of comorbidity by incorporating the severity or impact of each comorbid condition opposed to a simple numerical tally of the number of conditions. Finally, the identification of a modifiable resource (i.e., social support) that buffers the effect of social deprivation on comorbidity burden is novel with several potential implications discussed below.

The excess comorbidity burden in socially disadvantaged groups can have devastating consequences at every phase of cancer care, from screening and diagnosis to treatment and survivorship [10–13]. Healthy People 2030, the National Institute on Minority Health and Health Disparities (NIMHHD), and the NIH Social Determinants of Health Research Coordinating Committee urgently call for more research addressing SDOH strategies to improve the health and well-being of socially disadvantaged groups [15]. Our findings have important clinical implications that support these national priorities. First, implementing routine screening, at the time of diagnosis and treatment, for social deprivation, an important SDOH, in cancer care settings can help identify patients at risk, early in their care, so that healthcare providers can make appropriate referrals to address their social needs. The time between diagnosis and treatment is often considered a “teachable moment” when many patients express interest in strategies to help minimize recurrence and late health effects [32]. Second, promoting the integration of evidence-based social support interventions into cancer care [33, 34], by collaborating with social workers and community care settings can contribute to improved and more equitable health outcomes for all. Related, the integration of social prescribing, popular in the UK, Australia, and Japan, into cancer care plans to facilitate non-medical interventions can support the social needs and well-being of at-risk individuals affected by cancer [35, 36]. Social prescribing programs systematically link and support people, at risk, to access outside networks and community services to meet unmet social needs. Our findings add further support to the stress, social support, and buffering model [16]. It is possible that adequate perceived support can reduce the physiological stress response of social deprivation, facilitate healthy behaviors, or positively impact health care utilization, all of which could contribute to a lower comorbidity burden. These possible mechanisms might be an important area of future study.

Limitations of the study need to be considered. First, our study focused on adults primarily diagnosed with early stage breast and prostate cancer. Moreover, those who dropped out of the study were more likely to be participants with colorectal cancer. As such, replicating this study in different age groups, cancer sites, and later-stage disease is warranted. Second, SDI measures neighborhood-level disadvantage; hence individual participants may have more or less disadvantage compared to the neighborhood-level social disadvantage. Third, other important interrelated SDOH factors, including access to healthcare, health literacy, and chronic stress, were not collected or analyzed, but are likely important when examining comorbidity burden.

Understanding the association between social deprivation, comorbidity burden, and the moderating role of social support has important implications for both research and clinical practice. It highlights the need for a holistic approach to cancer care that considers the social context in addition to medical factors.

Acknowledgments

Thanks to the YUCAN Study graduate and undergraduate research assistants for their invaluable time and effort, including Dr. Kate Dibble, Dr. Sinead Sinnott, Katherine Gnall, & Zachary Magin. Special thanks to Dr. Tara Sanft, MD, and Rajni Mehta at the Yale School of Medicine. Importantly, thanks to the study participants who shared their experiences to make this research possible. This work was funded by the National Cancer Institute (NCI) grant UH3CA220642.

Compliance with Ethical Standards

Authors’ Statement of Conflict of Interest and Adherence to Ethical Standards Authors Keith M. Bellizzi, Emily Fritzson, Kaleigh Ligus, and Crystal L. Park declare that they have no conflict of interest. All procedures, including the informed consent process, were conducted in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1975, as revised in 2000.

Author Contributions Project administration, supervision, and funding aquisition includes Keith Bellizzi and Crystal Park. All authors contributed to the study’s conceptualization and methodology. Material preparation, data curation, and analysis were performed by all members. The first draft of the manuscript was written by Keith Bellizzi and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

References