Associations between religiosity and medical mistrust: An age-stratified analysis of survey data from Black adults in Chicago

Abstract

Medical mistrust is associated with poor health outcomes, ineffective disease management, lower utilization of preventive care, and lack of engagement in research. Mistrust of healthcare systems, providers, and institutions may be driven by previous negative experiences and discrimination, especially among communities of color, but religiosity may also influence the degree to which individuals develop trust with the healthcare system. The Black community has a particularly deep history of strong religious communities, and has been shown to have a stronger relationship with religion than any other racial or ethnic group. In order to address poor health outcomes in communities of color, it’s important to understand the drivers of medical mistrust, which may include one’s sense of religiosity. The current study used data from a cross-sectional survey of 537 Black individuals living in Chicago to understand the relationship between religiosity and medical mistrust, and how this differs by age group. Descriptive statistics were used to summarize data for our sample. Adjusted stratified linear regressions, including an interaction variable for age group and religiosity, were used to model the association between religiosity and medical mistrust for younger and older people. The results show a statistically significant relationship for younger individuals. Our findings provide evidence for the central role the faith-based community may play in shaping young peoples’ perceptions of medical institutions.

INTRODUCTION

Medical mistrust has been defined as “distrust of health care providers, the healthcare system, medical treatments, and the government as a steward of public health,” and is often a response to historical, interpersonal and structural experiences related to racism and discrimination [1–3]. Mistrust of medicine and healthcare systems has been documented extensively in the literature, and can lead to poor health outcomes [4, 5], ineffective disease management [6, 7], and lower use of preventive care [8–10]. Evidence also suggests that medical mistrust is associated with lower engagement in clinical research [11], particularly among individuals of color who have been historically marginalized and mistreated by healthcare institutions [7, 9]. Given the potential for poor health outcomes associated with medical mistrust, it is critical to understand what may predict or influence medical mistrust to promote better health behavior and ultimately advance health equity.

Religiosity is an important predictor of healthcare decision making and the religious community plays a significant role in promoting public health practices [12–15]. However, there are several dimensions of religiosity, which have unique relationships to varying health outcomes. Hill and Hood identified 10 major dimensions of religiosity and spirituality: denomination, beliefs, attitudes, organizational religious activity, private religious activity, religious/spiritual salience, religious/spiritual orientation, religious/spiritual coping, religious/spiritual history, religious/spiritual experience, and religious/spiritual development [16]. The strongest associations documented in the literature suggest that religious attendance is linked to improved mental [17] and physical health [18], utilization of preventive services [19], and reduced mortality [18, 20, 21]. On the other hand, there is outdated, weak or limited evidence of associations between other religiosity dimensions and health outcomes. A 2000 Cochrane Systematic Review examined 10 studies which included data from more than 7,500 randomized individuals. The authors found no statistically significant relationship between prayer and mortality, general clinical state, or clinical service use (e.g. readmission, emergency room visit, hospitalization) [22]. Similarly, there have been mixed results describing the association between religious salience and patient outcomes. Park et al. found significant associations between religious salience and hospitalization, satisfaction, and mental health, but null associations to health-related quality of life [23].

Religion is at the center of many Black communities, offering an opportunity to explore spirituality and providing a safe space to discuss general concerns [24]. The 2014 U.S. Religious Landscape Study found that Black Americans attend church more frequently than any other racial/ethnic group [25] and research has shown that religious participation among Black individuals is linked to improved health-seeking behavior [26], self-rated health [27], and clinical outcomes [28, 29]. Many people look to religion not only for comfort and solace, but also for explanations during difficult times. Religious leaders are highly respected members of their community who act as a source of information and guidance, and who are often able to connect with hard-to-reach populations [30, 31].

The relationship between religion and medical mistrust vary widely depending on one’s religious denomination, racial/ethnic background, and formal education. People’s perceptions of other individuals and institutions, and how they build trust of others is highly dependent on life experiences, including religious affiliation and other social interactions. Mistrust may also be driven by specific religious sects which are skeptical of biological science grounded in evolutionary theory, and therefore reject medical treatments or procedures such as abortion and vaccines stemming from related cell lines [32]. A study that analyzed data from a nationally representative sample of Americans found that adults who attended religious services at least 2-3 times per month had higher levels of trust in both their physicians and higher confidence in physicians overall compared to those who were infrequent attendees [33]. However, religious denomination can play a significant factor in one’s sense of trust in healthcare. Among Latinos in California, those who attended a Pentecostal church and those identifying as Mexican/Chicano were more less likely to have medical mistrust, found that church denomination made a difference as did length of attendance [34]. Given the significant role of the church in Black communities, this is an opportune setting to leverage to build the trust for medicine and medical personnel and has been done successfully.

Of note, levels of both medical mistrust and religiosity tend to vary with age. While religious attendance is fairly stable throughout adulthood, religious intensity and strength in beliefs may increase as individuals age [35]. The psychological literature has explored potential reasons for changes in religious and spiritual beliefs over the life course such as getting closer to death, facing increased illnesses, and losing loved ones [36, 37]. Although the literature is less robust, studies have also found that younger populations tend to have more mistrust of medical institutions and biomedical interventions [38]. The literature is sparse on the relationship between religiosity and medical mistrust, specifically as it relates to individuals of color in an urban community.

We addressed several of these gaps in the field with data from a convenience sample of Black Chicagoans. Specifically, we: 1) examined the association between religiosity and medical mistrust, and 2) assessed if the relationship differed between younger individuals and older individuals. Based on the existing literature, we hypothesized that higher levels of religiosity would predict higher levels of medical mistrust. We hypothesized that this relationship between religiosity and medical mistrust would be greater among older individuals compared to young individuals. The findings from this study will expand our understanding of medical mistrust, and how it can be influenced by social and environmental factors.

METHODS

Participants and Procedures

Data came from a larger study that surveyed Black adults living in Chicago to understand the relationship between violence, racism, and COVID-19 [39]. The survey was conducted between September 2021 and March 2022. Participants were eligible to complete the survey if they self-identified as Black or African American, were at least 18 years of age or older, lived in Chicago for at least three months, and were able to provide consent. Participants were recruited using a variety of methods including in-person and virtual recruitment at various community events, online advertising through social media and websites, recruitment from members of the project’s community advisory committee, and snowball sampling referrals from other participants. Individuals who indicated they were interested completed a brief screener by phone with a member of the study team. Individuals who were deemed eligible completed the survey using a personalized REDCap link. Upon completion of the survey, participants received a $50 gift card as compensation for their participation. The study was received and approved by the Medical College of Wisconsin Institutional Review Board.

Measures

Medical Mistrust.

Our outcome variable in this analysis was medical mistrust. Medical mistrust was measured using eight items from the Group-Based Medical Mistrust Scale (GBMMS) (Cronbach’s alpha = 0.92) [40]. The GBMMS is a validated scale used to measure health care-related trust within the context of racism and discrimination. For each question, respondents used a 5-point Likert scale to indicate their level of agreement (from 1 – strongly disagree to 5 – strongly agree). Items were averaged, with higher scores indicating greater mistrust. Four of the original 12 items are reverse-coded and were excluded from this analysis due to low correlations between these four items and the scale as a whole.

Religiosity.

Our predictor variable in this analysis was religiosity. Religiosity was measured using the Santa Clara Strength of Religious Faith Questionnaire (SCSRFQ) (Cronbach’s alpha = .97) [41]. The SCSRFQ is a 10-item validated scale used to measure religious engagement and spirituality. Dimensions include religious attendance, sense of community and identity, spirituality, influence on decision making, and connection to God. For each question, respondents used a 4-point Likert scale to indicate their level of agreement (from 1 – strongly disagree to 4 – strongly agree). Items were averaged, with higher scores indicating greater religiosity.

Racism.

Racism was included in our analysis as a covariate based on its proven association with medical mistrust [3]. Connecting with one’s religious community has also been shown to be a solace for people experiencing racism [42]. Racism was measured using the Everyday Discrimination Scale [43], which is a validated 10-item measure of discrimination in employment, housing, and education (Cronbach’s alpha = 0.98). For each question, respondents indicated how frequently they experienced racism in their day-to-day lives (from 0 – never to 5 – almost every day). Items were averaged, with higher scores indicating more frequent experiences of racism.

Sociodemographic and Health Background.

Self-reported age was included in our analysis as a stratifying variable. Age was collected as a continuous variable. We considered those under 35 to be younger participants and those 35 years of age and older to be older participants based on the age distribution of our sample. Self-reported gender identity and income were included as covariates. Participants reported their gender identity as woman, man, transgender man, or transgender woman. Transgender men and transgender women were included as men and women, respectively. Participants reported their income on a scale from no income to $75,000 or more. Based on the distribution of the sample, we collapsed all income categories above $35,000. Participants reported their education based on an eight category scale including having never gone to school, 5^th grade or less, 6^th-8^th grade, 9^th-12^th grade (no diploma), high school graduate or GED, some college/technical/vocational degree, Bachelor’s degree, and other advanced degree (Master’s, Doctoral). Participants could also indicate they did not know or preferred not to answer. Based on the distribution of the sample, we collapsed the four lowest categories into one (less than a high school degree) and the two highest categories into one (Bachelor’s and advanced degree). Finally, participants’ self-reported underlying health conditions were used as a covariate. We created a dichotomous variable to indicate if a participant reported at least one of the following health conditions: immunocompromised condition, autoimmune disease, hypertension, diabetes, chronic kidney disease, cancer diagnosis and/or treatment in the past 12 months, cardiovascular disease, asthma, chronic obstructive pulmonary disease, other chronic lung disease, sickle cell anemia, depression, alcohol or substance use disorder, intravenous drug use, other mental health disorder, other chronic condition.

Data Analysis

Descriptive statistics summarized study variables for younger (aged 18-34) and older (aged 35+) participants. Younger and older participants were compared using T-tests and chi-square tests. Linear regression models were used to examine the relationship between religiosity and medical mistrust. We first looked at this relationship in the sample as a whole, including medical mistrust as the outcome, religiosity as the primary predictor, and age group (aged 18-34 vs. aged 35+), gender, income, underlying health conditions, and racism as covariates. We also tested the interaction between age group and religiosity to test whether the association between religiosity and medical mistrust differed across age group. Next, given some evidence of differing relationships between religiosity and medical mistrust across age groups, we stratified the linear regression analysis by age group, again including medical mistrust as the outcome, religiosity as the primary predictor, and gender, income, underlying health conditions, and racism as covariates. We report unstandardized regression coefficients.

A small amount of data (<2%) was missing; this missing data resulted from participants selecting “don’t know” or “prefer not to answer” responses. Nearly 10% of all participants (n = 51) had some missing data, and individual variables had between 0 and 5% missing data. Participants were more likely to be missing some data when they were younger or men, and those with some missing data reported less education, lower household incomes, and fewer experiences of racism (results not shown). We addressed missing data using multiple imputation, which avoids biases associated with using only complete cases or with single imputations [45]. Missing data were imputed using chained equations with the mice package in R 4.2.0 [44]. mice can handle both continuous and categorical variables when imputing. We created 100 imputations, using all analytic variables (including the interaction between religiosity and age) when imputing. We took the “just another variable” (JAV) approach to imputing values for the interaction between age and religiosity, as recommended in the literature when the analysis is a linear regression [46]. The JAV approach may be somewhat biased when data are missing at random (MAR) rather than missing completely at random (MCAR), but this bias is found to generally be less than with other approaches. Multiple imputation results were combined in R using Rubin’s standard rules [44]. Sensitivity analyses showed that the pattern of results did not differ when utilizing other methods for imputing the interaction term (stratification or passive imputation) or when taking a complete case approach (results not shown), likely due to the modest amount of missing data.

RESULTS

Descriptive Information

Participants included 537 Black/African American individuals living in Chicago. Most were women (70.8%) with an average age of 37 years (range = 18-79). Nearly 40% of participants reported income less than $4,999 per year, including no income. Almost half of the sample (44.5%) reported a high school degree or GED. Forty-four percent of participants reported at least one underlying health condition. The average racism score was 1.8 (SD = 1.45) out of 5, the average medical mistrust score was 2.7 (SD = 0.82) out of 5, and the average religiosity score was 3.1 (SD = 0.67) out of 4. Analyses comparing younger and older participants (Table 1) showed that compared to older participants, young participants were more likely to be women (76.4% vs. 64.4%, p = 0.003) and have reported an income below $4,999 per year (46.9% vs. 27.7%, p < .001). Older individuals were more likely to report at least one underlying health condition compared to younger individuals (63.6% and 26.8%, respectively; p < .001). While perceived racism was approximately the same between older and younger participants, younger participants reported higher medical mistrust (score of 2.9 vs 2.5, p < .001) and lower religiosity (score of 2.9 vs. 3.2, p < .001).

Table 1.

Sample Descriptive Statistics (N = 537)

	Total			Younger Participants (Age < 35)			Older Participants (Age ≥ 35)
	n	%	M (SD)	n	%	M (SD)	n	%	M (SD)	χ2-/t-Statistica	p-value
Total	537			284			253
Gender
Women	380	70.8		217	76.4		163	64.4		9.001	0.003
Men	150	27.9		64	22.5		86	34.0
Missing/Other	7	1.3		3	1.1		4	1.6
Annual Household Income
No income	72	13.4		49	17.3		23	9.1		33.725	<.001
$1-$4,999	131	24.4		84	29.6		47	18.6
$5,000-$14,999	82	15.3		31	10.9		51	20.2
$15,000-$19,999	41	7.6		18	6.3		23	9.1
$20,000-$24,999	65	12.1		38	13.4		27	10.7
$25,000-$34,999	57	10.6		30	10.6		27	10.7
More than $35,000	74	13.3		25	8.8		49	13.8
Missing (Unsure/No Answer)	15	2.8		9	3.2		6	2.8
Education
Less than high school degree	82	15.3		41	14.4		41	16.2		20.935	<.001
High school degree/GED	239	44.5		150	52.8		89	35.2
Some college/technical/vocational degree	161	30.0		69	24.3		92	36.4
Bachelor’s degree or more	50	9.3		20	7.0		30	11.9
Missing (Unsure/No Answer)	5	0.9		4	1.4		1	0.4
Health Condition	237	44.1		76	26.8		161	63.6		74.092	<.001
Racism (range: 0 - 5)	525		1.8 (1.45)	274		1.7 (1.40)	251		1.8 (1.50)	−0.864	0.388
Medical Mistrust (range: 1 - 5)	520		2.7 (0.82)	272		2.9 (0.78)	248		2.5 (0.83)	3.938	<.001
Religiosity (range: 1 - 4)	502		3.1 (0.67)	267		2.9 (0.63)	245		3.2 (0.65)	−5.651	<.001

^aDifferences between younger and older participants were tested

Regression Analyses

In the linear regression model for the entire sample (Table 2, Model A), participants who were more religious reported higher levels of medical mistrust, B = 0.11 (0.05), p = .04. Higher levels of medical mistrust were also reported by those with higher incomes, B = 0.04 (0.02), p = .03, and those who had more experiences of racism, B = 0.10 (0.02), p < .001. Older participants reported lower levels of medical mistrust than younger participants, B = −0.42 (0.08), p < .001. The interaction between age group and religiosity was significant (Model B; p = .04), indicating the relationship between religiosity and medical mistrust differed across age groups.

Table 2.

Linear Regression Model Assessing Factors Associated with Levels of Medical Mistrust Among Black Chicagoans (N = 537)

	Simple Regressions		Model A (No Interaction)		Model B (Interaction)
	B (SE)	p	B (SE)	p	B (SE)	p
Age 35+ (ref: Age 18-34)	−0.34 (0.07)	<.001	−0.42 (0.08)	<.001	0.23 (0.34)	.496
Male	−0.12 (0.08)	.148	−0.09 (0.08)	.271	−0.09 (0.08)	.263
Annual Household Income	0.02 (0.02)	.194	0.04 (0.02)	.028	0.04 (0.02)	.015
Education	0.01 (0.04)	.776	−0.02 (0.04)	.645	−0.02 (0.04)	.658
Health Conditions (ref: No Conditions)	−0.04 (0.07)	.620	0.05 (0.08)	.498	0.05 (0.08)	.513
Racism	0.10 (0.02)	<.001	0.10 (0.02)	<.001	0.10 (0.02)	<.001
Religiosity	0.05 (0.05)	.389	0.11 (0.05)	.034	0.23 (0.08)	.004
Age 35+ * Religiosity					−0.21 (0.11)	.047
R 2			0.10		0.11

In models stratified by age (Table 3), younger participants who were more religious reported higher levels of medical mistrust than younger participants who were less religious, B = 0.23 (0.07), p = 0.003. Conversely, there was no significant relationship between religiosity and medical mistrust for older participants, B = 0.01 (0.07), p = 0.848. Although not the focus of this study, it’s worth noting that younger participants with higher incomes, B = 0.06 (0.02), p = 0.011, and underlying health conditions, B = 0.23 (0.11), p = 0.030, had greater medical mistrust. Similarly, older participants who reported more frequent experiences of racism had greater medical mistrust, B = 0.15 (0.03), p < 0.001.

Table 3.

Stratified Linear Regression Models Predicting Levels of Medical Mistrust Among Younger and Older Black Chicagoans

	Younger Participants (Age < 35; n = 284)		Older Participants (Age ≥ 35; n = 253)
	B (SE)	p	B (SE)	p
Male (ref: Female)	0.00 (0.11)	.993	−0.19 (0.11)	.089
Annual Household Income	0.05 (0.02)	.022	0.03 (0.03)	.234
Education	0.02 (0.06)	.769	−0.05 (0.06)	.409
Health Conditions (ref: No Conditions)	0.23 (0.11)	.035	−0.12 (0.11)	.251
Racism	0.06 (0.03)	.092	0.15 (0.03)	<.001
Religiosity	0.22 (0.08)	.003	0.02 (0.08)	.753
R 2	0.08		0.09

DISCUSSION

This study examined the relationship between religiosity and medical mistrust among Black adults during the height of the COVID-19 pandemic. Notably, our data come from a sample of nearly 540 Black individuals living in Chicago. Religiosity had a statistically significant positive association with medical mistrust for younger individuals (under 35 years of age), but not for individuals 35 years of age and older.

Existing research on the relationship between religiosity and medical mistrust has primarily focused on religious attendance, but has not fully explored other aspects of religiosity such as prayer and one’s relationship with God. Cacciatore et al. found a strong correlation between religiosity and general trust in informal sources of information, including religious organization websites, spiritual leaders, and family or friends [47]. Yet, the information shared by these sources may not always be factual and can contradict scientific evidence. Olagoke et. al. found that external health locus of control, the belief that one’s health is influenced by external factors such as a higher power, can serve as a pathway through which religiosity may be associated with health decision-making [48]. Our findings suggest that religiosity, as measured in this study by one’s sense of religion, connection to God, and participation in religious activities like churchgoing, may influence medical mistrust, and that influence may differ based on age. These findings are an important addition to the body of work on religiosity and medical mistrust, however it is important to recognize that because the analysis used a composite score of religiosity, it is not possible to identify associations with specific religiosity dimensions.

As described previously, churches are an important source of community and social cohesion, emotional and spiritual support, and can provide a sense of purpose and meaning within the Black community. Churches often have available meeting space, kitchens, and regular attendees, making it an opportune setting to provide health promotion programming and prevention resources. Partnerships between religious institutions and medicine has been shown to increase healthy behaviors (e.g., physical activity and fruit and vegetable consumption), reduce clinical risk factors for chronic disease (e.g., body weight), and increase preventive activities (e.g., cancer screenings) [49–52]. However, research has also shown that church-based health promotion activities require investment in resources and trustworthy partnerships [49, 53]. Potential activities to bridge these sectors include co-hosting health fairs or health promotion programs, using church leaders and church members to train healthcare providers in culturally competent practices, and even partnering on advocacy efforts that address systemic health inequities affecting minority populations.

Mistrust of medical institutions including providers, clinical researchers, pharmaceutical companies, government agencies, and health care institutions can be driven by a host of factors. Our findings contribute to a broader understanding of medical mistrust within the context of Black communities in Chicago, a marginalized population often underrepresented in research. Building upon existing literature highlighting the influence of previous experiences of discrimination and sociodemographic factors like race/ethnicity, income, education, and insurance status, our analysis identifies an additional factor: religiosity [54–56]. This contributes to our understanding of the multifaceted nature of medical mistrust and the extent to which it may be experienced across Black populations. Findings from this analysis highlight the importance of cultural factors, such as religion, in addition to the more traditional demographic factors.

To our knowledge, this is the first study to use age as a stratifying variable in the relationship between religiosity and medical mistrust. Religious affiliation among young adults ages 18-29 has increased since the mid 1980’s (10% in 1986 compared to 38% in 2022) [57]. And although young Americans are more likely to be religiously unaffiliated compared to older Americans, those who are religious may have a greater connection to their religious community or a greater sense of spirituality. We found a measurable and significant association between young participants’ sense of religiosity and medical mistrust, however the drivers of this relationship have yet to be explored.

There is limited research that has sufficiently explored the relationship between age and medical mistrust. The existing literature is primarily qualitative, with young adults and youth reporting a lack of general health care knowledge, a fear of being treated differently based on their health status, and poor understanding of biomedical interventions may be related to their mistrust of the health care system [38, 58]. Mistrust of institutions, more broadly, could be explained by young Black Americans’ experiences or awareness of widespread violence or discrimination toward Black individuals. High-profile instances of violence against young Black individuals in recent years such as Breonna Taylor, Philando Castile, and Freddie Gray have permeated the mainstream media and evidence suggests coverage of these encounters have caused fear and distrust among young Black men [59]. While this study only included racism as a covariate to control for potential confounding between religiosity and medical mistrust, it would be important for future research to better understand how racism fits in the causal pathway from religiosity to medical mistrust.

Although we found a statistically significant relationship between religiosity and medical mistrust for young adults, we did not find the same strength of relationship among older adults. Older adults, on average, have more lived experience than younger adults, and this lived experience may be more influential on their perception of medical institutions and medical mistrust than outside sources such as their community and religious practices. It is also possible that the questions asked within each of these scales were interpreted differently between older and younger individuals.

Strengths and Limitations

The current study has several strengths and limitations. First, although our data comes from a convenience sample of Black Chicagoans, which may not be generalizable to other communities or populations, this is a unique population which is medically underserved and experiences enormous health inequities. Studies like ours provide a more nuanced understanding of the factors related to medical mistrust, and can therefore be used to develop programs and policies to improve access to and utilization of health-related services. Second, the data comes from a cross-sectional survey which prevents us from making causal claims, however our sample size is sufficient to draw the previously stated conclusions. Third, the study uses a single composite measure of religiosity which does not isolate the nuances that may exist between different domains of religiosity included religious attendance, spirituality, and connection to God. Fourth, due to a modest amount of missing data, we needed to utilize multiple imputation to maintain our full sample, which has the potential to create bias when considering interaction terms [46]. However, given differences between participants with and without missing data, we considered multiple imputation a more reasonable approach than complete case analysis. Finally, these findings are solely based on quantitative data captured from a cross-sectional survey. Future research should explore racism’s role in this relationship and employ mixed methods to gain a deeper understanding of why and how the relationship between religiosity and medical mistrust may be stronger in young individuals compared to older individuals.

Conclusion

These findings have practical implications for faith-based organizations and healthcare institutions. Given the influence the religious community may have on young people’s perception of healthcare, faith-based organizations have an important role in providing accurate, relevant, and time-sensitive information to congregants. This may be related to urgent health crises such as COVID-19 or other infectious diseases, but is also important when it comes to preventive care and medicine such as cancer screenings, healthy eating, physical exercise, and regular dental care.