Effects of Religious Involvement on HIV Management Outcomes Among HIV-Positive Adults in Central North Carolina

Abstract

Objectives:

The goals of this pilot cross-sectional study were to determine the feasibility of and begin measuring the effect of religious institution affiliation on human immunodeficiency virus (HIV) clinical outcomes in the southern United States, a region marked by later initiation of antiretroviral therapy, higher HIV-related morbidity, and higher mortality rates than people living with HIV (PLWH) elsewhere in the country. It also is a region with a high density of religious institutions, which may facilitate improved health outcomes through leveraged social capital. Because spirituality is a personal construct and PLWH constitute a vulnerable population, we wanted to determine whether it would be feasible to survey patients about the topic. We hypothesized that PLWH would be willing to participate and that PLWH who report involvement in religious institutions would be more likely to have suppressed HIV viral loads (VLs) and better engagement in care than PLWH not involved in a religious institution.

Methods:

Eligible participants were enrolled from the Wake Forest Infectious Diseases Specialty Clinic to complete structured interviews using validated measures of religious institution affiliation, spiritual well-being, social support, and HIV-related stigma. HIV VL and engagement in care (clinic no-show rate) data were abstracted from the electronic medical record. Descriptive statistics calculated the prevalence of religious institution involvement, outcomes of interest, and potential confounders. t Tests compared continuous outcomes assuming normality, χ² tests compared binary outcomes, and the Wilcoxon Mann–Whitney test compared outcomes for non-normal data.

Results:

Fifty participants completed the study (55% participation rate); 72% identified as male and 28% identified as female. A total of 48% of participants identified as black/African American and 44% identified as white. Participants who identified as men who have sex with men made up 34%. More black/African American participants than white participants reported religious institution affiliation (23%; P = 0.15). There was no statistically significant relation between religious institution affiliation and CD4 or VL; however, higher levels of social support and spiritual well-being predicted a lower clinic no-show rate (P = 0.0077 and 0.0195, respectively). There was a trend toward greater perceived HIV-related stigma and CD4 (P = 0.0845) as well as more emergency department visits (P = 0.0976).

Conclusion:

PLWH in a southern US clinic were willing to answer questions about their spirituality. Religious institution affiliation was not significantly related to virologic suppression or CD4 in this sample. Higher levels of self-reported social support (P = 0.0077) and spiritual well-being (P = 0.0195) predict better clinic attendance. These results suggest that religious affiliation alone does not imply positive benefits for all. Social support and spiritual well-being, however obtained, predict engagement in care. The next steps should include a fully powered study to define the relations among social support, spiritual well-being, and relevant clinical outcomes. Our results also support further investigation of perceived HIV-related stigma and healthcare utilization, based on the trend toward significance between emergency department visits and stigma.

Human immunodeficiency virus (HIV) has increasingly become a health crisis in the southern United States, a 16-state region henceforth referred to as the South. People living with HIV (PLWH) in the South initiated antiretroviral therapy (ART) later, experienced substantially more HIV-related morbidity, and had higher mortality rates than individuals living in other areas of the country.¹ This observation was particularly evident in nonwhite southern populations, who had a 2.1- to 4.6-fold higher HIV-related morbidity compared with other groups. Such findings, paired with epidemiological data from the Centers for Disease Control and Prevention describing the disproportionate burden of new HIV infections/acquired immunodeficiency syndrome diagnoses in the South, demonstrate the necessity for better characterization of factors leading to poor outcomes in the South, as well as novel strategies for improving clinical outcomes among PLWH in this region.^1–3

One such noted strategy has been the use of religious institutions, primarily Christian churches, to serve as “health networks” that connect members with resources to improve health outcomes.⁴ Given the dense concentration and profound influence of religious institutions in the South, this approach is aimed at using these grassroots organizations with established community rapport to reach populations at high risk for HIV—chiefly, rural nonwhites of lower socioeconomic status.⁵ Religious institutions in this capacity serve as a key source of social capital for these populations, equipped with the potential to build HIV awareness, reduce stigma, and improve engagement with HIV care services.⁶

Faith-based community health partnership is not a novel idea. Multiple faith-based HIV education and intervention programs have been drafted and studied.^7–13 These programs are based upon numerous studies citing better mental and physical health outcomes for people who attend religious programs. In one review of the literature, we identified 101 studies that analyzed the association between religious involvement and depression, 65% of which found a statistically significant connection between religious involvement and lower rates of depression/ depressive symptoms.¹⁴ Reviews for association between religious participation and decreased anxiety revealed similar positive results, as did studies to detect levels of suicidal ideation among depressed older adults.^14,15 Mechanisms describing these outcomes are proposed to involve both the religious and social support elements that participants experience when attending a faith-based activity.¹⁶ Religious involvement also has been shown to affect physical health, suggesting that higher levels of spiritual wellness were significantly related to lower systolic and diastolic ambulatory blood pressure, lower high-sensitivity C-reactive protein, lower fasting glucose, and marginally lower triglycerides.¹⁷ Similar studies reveal the positive correlation among physical well-being, spirituality, and religious service attendance.^18–20

Although there is extensive literature citing the positive health effects of faith-based activities, limited data exist evaluating the correlation between regular religious institution participation and HIV outcomes in the southern United States. Reasons for the lack of data are likely multifactorial; spirituality is a deeply personal construct and powerful HIV-related stigma remains, particularly among religious congregations that display fervent moral objection to injection drug use, homosexuality, and sexual activity extending outside of monogamous heterosexual marital relations.²¹ Van Wagoner et al demonstrated that men who have sex with men (MSM) and who attend church are more likely to present with lower CD4⁺ T-lymphocyte counts than MSM who do not attend church, and hypothesized that this result was the result of fear of rejection by their congregation and subsequent delayed testing.²² This same study found, however, that women who are regular church attendees are more likely to receive early HIV testing and detection than those who do not attend church.²²

Religious coping and spiritual well-being also play important roles among churchgoing PLWH. One multisite longitudinal study showed that the implementation of religious coping strategies (eg, seeking prayer support) were strongly associated with positive HIV management outcomes, whereas spiritual struggle (eg, believing that HIV is a moral punishment from God) was associated with negative outcomes.²³ Such convoluted findings, combined with the sizable volume of faith-based HIV community intervention programs in place, warrant further analysis of the role of faith-based activities and HIV management outcomes.

The overarching goal of the present study was to begin to evaluate the relation between religious institution involvement and key HIV management outcomes, including CD4⁺ count, viral load (VL), and engagement in care. Before initiating a fully powered study, we wanted to assess the feasibility of surveying a population of PLWH in the South about their spirituality. We hypothesized that PLWH would be willing to answer structured interview questions and that individuals who are heavily involved in their religious institution will have more favorable outcomes because of increased social capital through institutional networking, granting them greater access to transportation, financial, and material resources, as well as greater social and spiritual backing. Because spiritual well-being, perceived social support, and HIV-related stigma coexist as variables deserving evaluation, these elements also were analyzed in our study.

Methods

Study Design

We constructed a pilot cross-sectional study to evaluate our hypothesis. Before enrollment and data collection, the study protocol was approved by the Wake Forest School of Medicine institutional review board. Our goal enrollment in this study was N = 50. Inclusion in the study required documented HIV-positive serostatus, ability to provide informed consent, age 18 years or older, ability to understand and speak English, and receipt of an ART regimen for ≤12 months. The language criterion was used because the tools used for the interviews were available only in English. Patients were ineligible to participate if these criteria were not met or if they were currently incarcerated, acutely ill, or in need of urgent medical attention.

We enrolled a convenience sample of PLWH attending routine clinic appointments who met inclusion criteria and gave written informed consent. They were initially surveyed for demographics that included a question about whether the participant had a religious institution that they attended. Based upon the response to that question, participants were split into two groups. The group reporting involvement in a religious institution was administered four additional short surveys: components of the Multidimensional Measure of Religiousness/Spirituality, the Functional Assessment of Chronic Illness Therapy-Spiritual Well-Being (FACIT), the Multidimensional Scale of Perceived Social Support (MSPSS), and the Berger HIV Stigma Scale (Berger). The group who answered “no” to the question of religious institutional attendance, nonattendees, was administered three of the surveys listed above: the FACIT, the MSPSS, and the Berger Scale (Fig.).

Fig.

Algorithm for study procedure allocation. FACIT-Sp-12, Functional Assessment of Chronic Illness Therapy-Spiritual Well-Being; MMRS, Multidimensional Measure of Religiousness/Spirituality; MSPSS, Multidimensional Scale of Perceived Social Support.

The Multidimensional Measure of Religiousness/Spirituality is a validated and widely used resource in health research for measuring the religious/spiritual facets of patients’ lives.²⁴ The 12 survey domains include religious coping, commitment, support, and organizational involvement, and come in long and short forms. For this study, the religious support subtest long form (12 questions) and the organizational religiousness questions on “church/institution attendance” and “church/institution fit” (seven questions) were used. The FACIT (12-item survey) is a tool that was developed for the quick and accurate assessment of spiritual well-being among patients who live with a chronic illness.²⁵ The MSPSS (12 questions) has been used extensively in HIV studies as a concise tool that accurately records perceived levels of social support experienced by participants.^26,27 The Berger (40 questions) is an instrument designed to measure the stigma perceived by PLWH. It was developed based on the literature on stigma and psychosocial aspects of having HIV.²⁸

The study period was defined as the 12 months preceding and including the date of enrollment in the study. After completing the interview, the following HIV-related outcomes data were abstracted from participants’ electronic medical records: last CD4 count in the study period, CD4 nadir in the study period, last VL in the study period, number of hospitalizations within the study period, number of emergency department (ED) visits within the study period, and engagement in care, operationalized as clinic no-show rate (NSR). NSR was calculated as the number of medical visits not attended (nor canceled or rescheduled by the participant) divided by the total number of scheduled visits during the 12-month study period. AVL was considered detectable if it was above zero copies during the study period.

Study Context

Study participants were recruited from the Infectious Diseases Specialty Clinic (IDSC) of Wake Forest Baptist Medical Center, an approximately 880-bed tertiary-care research hospital located in Winston-Salem, North Carolina. The IDSC cares for approximately 2070 PLWH, 33% of whom are female, 64% are male, and 1.4% self-identify as transgender. Clinic patients are predominantly African American (64%), and most of the women receiving HIV primary care in the IDSC are African American (78.4%). Six percent of patients are Latino/a. Forty-one percent of our patients are older than 50 years. Thirty-two percent of the clinic patients are uninsured and 41% of the patients receive Medicaid and/or Medicare benefits. Of patients at least 17 years old with >1 medical visit in the past 12 months, 95% are receiving ART and 82% are virologically suppressed. The most commonly reported transmission risk factors are high-risk heterosexual contact (57%) or being MSM (39%); 8% of patients report intravenous drug use as their HIV acquisition risk, and these data mirror risk profiles in demographically similar regions. As of December 2013, 8% of patients lived in counties with <250,000 people and 10.5% lived in rural areas as defined by the US Office of Management and Budget.

Recruitment and Consent

Subjects were contacted and recruited during routine IDSC clinic visits by institutional review board–approved study staff between June 8, 2016 and July 12, 2016. There was no information collected before obtaining informed consent. All of the patients who met the inclusion criteria were approached for inclusion in the study until the target enrollment of 50 was met. Potential participants were approached for recruitment in the privacy of their examination rooms, ensuring confidentiality. There was no compensation offered to participants. This was explained to participants during the informed consent process. Two trained interviewers (S.Y. and T.Y.) administered the surveys. Subjects received printed versions of the interview tools and the interviewer read the questions and answers aloud, ensuring accuracy of comprehension and results across diverse levels of literacy.

Statistical Methods

Baseline characteristics between those affiliated and not affiliated were compared using t tests for continuous measures and the Fisher exact test for categorical variables. To test for differences in clinical outcomes by religious affiliation status, the Fisher exact test was used. A comparison of average scores on the MSPSS, FACIT, and Berger between those with and without negative clinical outcomes was done using t tests (Table 3). All of the tests were considered significant at P = 0.05.

Table 3.

Clinical outcomes vs perceived social support, spiritual well-being, HIV-related stigma

	MSPSS score, mean (SD)	FACIT score, mean (SD)	Berger Scale score, mean (SD)
VL detected
No (n = 40)	65.6 (12.6)	37.8 (9.6)	96.4 (19.4)
Yes (n = 10)	66.5 (17.2)	35.4 (12.0)	88.8 (18.0)
P	0.8521	0.4953	0.2696
CD4 nadir
0 (n=3)	72.0 (10.8)	41.3 (5.8)	76.3 (27.7)
≥200 (n = 47)	65.4 (13.6)	37.1 (10.2)	96.0 (18.3)
P	0.4143	0.4848	0.0845
Hospitalizations
None (n = 43)	66.0 (13.2)	37.9 (9.2)	93.4 (19.7)
≥1 (n = 7)	64.6 (16.1)	33.9 (14.4)	103.9 (13.5)
P	0.8006	0.3241	0.1822
No. ED visits
0 (n = 32)	67.1 (13.3)	38.4 (7.8)	91.5 (17.5)
≥1 (n = 18)	63.4 (13.8)	35.6 (13.2)	100.8 (21.0)
P	0.3510	0.3453	0.0976
NSR, %
0 (n = 32)	69.5 (10.7)	39.8 (8.2)	91.7 (19.5)
>0 (n = 18)	59.2 (15.4)	33.0 (11.7)	100.5 (17.7)
P	0.0077	0.0195	0.1183

ED, emergency department; FACIT, Functional Assessment of Chronic Illness Therapy-Spiritual Well-Being; MSPSS, Multidimensional Scale of Perceived Social Support; NSR, no-show rate; SD, standard deviation; VL, viral load.

Results

The sample consisted of 50 subjects, 28% women, mean age 51.3 (standard deviation 12.2), 48% black/African American, 44% white, 64% with education beyond high school, 58% unemployed, 36% with a history of substance abuse, and 60% identifying as same gender sexual orientation (including bisexual, 26%; Table 1). Of the 50 participants, 29 (58%) reported being affiliated with a religious institution, whereas the remaining 21 (42%) reported no religious institution affiliation. Subject demographics did not differ significantly between those who did and those who did not report a religious institution affiliation. No differences were found between religious affiliation categories on the MSPSS (social support), or the FACIT (spiritual well-being).

Table 1.

Subject characteristics by religious affiliation

	Involved in religious institution?
	No (n = 21)		Yes (n = 29)
	N	%	N	%	All	P
Age, y, mean (SD)	21	51.9 (12.3)	29	50.4 (12.3)	51.3 (12.2)	0.6605
Sex (%)						0.3408
Male	17	81	19	66	36 (72)
Female	4	19	10	34	14 (28)
Race (%)						0.1520
African American	5	24	19	66	24 (48)
White	12	57	10	34	22 (44)
Multiple	4	19	0	0	4 (8)
Education level (%)						0.3883
7–11 y	3	14	3	10	6 (12)
High school graduate	3	14	9	31	12 (24)
13–15 y	6	29	10	34	16 (32)
College graduate	8	38	5	17	13 (26)
17–18 y	1	5	2	7	3 (6)
Employment status (%)						0.5683
Employed full-time	8	38	8	28	16 (32)
Employed part-time	2	10	3	10	5 (10)
Unemployed	11	52	18	62	29 (58)
Insurance status (%)						0.7327
Uninsured	4	19	9	31	13 (26)
Private insurance	7	33	6	21	13 (26)
Medicaid	4	19	5	17	9 (18)
Medicare	6	29	9	31	15 (30)
Substance abuse history (%)						0.7742
No	14	67	18	62	32 (64)
Yes	7	33	11	38	18 (36)
Sexual orientation (%)						0.3827
MSW	6	29	6	21	12 (24)
MSM	9	43	8	28	17 (34)
MSB	3	14	10	34	13 (26)
WSM	3	14	5	17	8 (16)
MSPSS score, mean (SD)	65.3 (12.6)	66.1 (14.2)		0.8276
FACIT score, mean (SD)	35.6 (7.2)	38.7 (11.6)		0.2540
Berger score, mean (SD)	94.2 (16.8)	95.3 (21.0)		0.8409

FACIT, Functional Assessment of Chronic Illness Therapy-Spiritual Well-Being; MSB, men who have sex with both; MSM, men who have sex with men; MSPSS, Multidimensional Scale of Perceived Social Support; MSW, men who have sex with women; WSM, women who have sex with men; SD, standard deviation.

HIV VL was detectable in 20% of subjects and a CD4 count of ≥200 was detected in 94% of subjects. Because detectable VL is commonly defined as ≥200 copies per milliliter in the literature, we did look at those data, but none of the participants had a peak VL >200 during the study period. Only 14% of participants experienced any hospitalizations during the 12-month period preinterview, but 36% reported at least one ED visit during this period. A majority of subjects missed no clinic visits (64%), 16% missed 1 of 4 visits, and 20% missed 1 of 3 visits. These rates did not differ by religious institution affiliation (Table 2).

Table 2.

Religious affiliation vs clinical outcomes

	Involved in religious institution?
	No (n = 21)		Yes (n = 29)
	N	%	N	%	All (%)	P
VL detected						0.2859
No	15	71	25	86	40 (80)
Yes	6	29	4	14	10 (20)
CD4 nadir						0.5650
0	2	10	1	3	3 (6)
≥200	19	90	28	97	47 (94)
Hospitalizations						1.0
None	18	86	25	86	43 (86)
≥1	3	14	4	14	7 (14)
No. ED visits						0.5946
0	15	71	17	59	32 (64)
1	4	19	6	21	10 (20)
≥2	2	10	6	21	8 (16)
NSR, %						0.4169
0	13	62	19	66	32 (64)
25	5	24	3	10	8 (16)
33	3	14	7	24	10 (20)

ED, emergency department; NSR, no-show rate; VL, viral load.

Higher MSPSS scores, indicating higher levels of perceived social support, were associated with lower NSR (P = 0.0077). Similarly, higher FACIT scores, signifying higher levels of spiritual well-being, were associated with lower NSR compared with subjects with lower FACIT scores. A borderline significant association was seen between the perceived HIV-related stigma and CD4 (P = 0.0845), with those having CD4 counts >200 having higher average stigma scores, although only 3 participants had CD4 nadirs <200, so conclusions are limited. There was a trend toward significance between higher perceived stigma and more ED visits (P = 0.0976; Table 3).

Discussion

There was a 55% participation rate in this study, suggesting that patients were willing to answer questions about their spirituality within the context of an HIV clinic visit but that it is not universal. We did not have an expected participation rate because this was an exploratory pilot study. Being affiliated with a religious institution does not appear to be significantly related to virologic suppression or immune health in this sample. The original hypothesis that affiliation with a religious institution may lead to better clinical outcomes was motivated by the belief that this affiliation provided two positive benefits: social support and spiritual well-being. These results raise the possibilities that people may obtain social support and/or spiritual well-being from other venues than religious institutions or that some people who are affiliated with a religious institution are not getting the social support or spiritual well-being expected, so their affiliation with the institution, while stated, was not providing those resources. It is certainly possible that a strong faith could discourage participation in clinical care, or that strong faith could be correlated with high levels of stigma and less social support. As noted in prior research in this area, individuals with strong religious involvement have different outcomes based upon their individual religious convictions (eg, views of God as forgiving vs punishing).²³ Our study design did not separate individuals based upon their specific religion, and thus we are unable to directly address this question.

To explore the role of these specific benefits, we looked at measures of spiritual well-being and social support as direct predictors of engagement in care, a clinical outcome that has been associated with HIV-associated mortality.²⁹ The results did show a significant association between social support, as measured by the MSPSS, and engagement in care (P = 0.0077) as well as between spiritual well-being, as measured by the FACIT, and engagement in care (P = 0.0195). Based on these results, we concluded that religious affiliation alone does not imply positive benefits for all. Social support and spiritual well-being, however obtained, may be more powerful predictors of engagement in care than biologic measures of HIV.

A primary limitation of our study is the pilot nature of our sample size (N = 50). The study was not powered to detect significant relations between key variables, but the presence of the significant associations we did find are notable. This study also may have had selection bias because we used a convenience sample; however, recruitment included patients from the clinic’s urgent care sessions, in which people who have been out of care or lost to follow-up are reengaged in care. In addition, the participant demographics, viral suppression rates, and engagement in care numbers mirror our larger clinic population’s profile, suggesting that this sample is representative of our population. Another limitation was the exclusion of non–English-language speakers, necessitated by the lack of available approved translations for the interview tools. Future studies should include other language groups and recruitment methods outside the confines of the clinic to ensure a more diverse sample. Generalizability of our findings may be limited because data were collected at a single clinic site, and thus, a single geographic location. Future studies should be aimed at including clinic sites in different geographic regions to assess whether there are local or cultural trends.

Practical application of these findings also may be limited because of the sensitive nature of the topic. Some study participants were not pleased with the idea that discussions of spirituality and/or religion could be incorporated into their care as a means of improving their health. Based on our findings that spiritual well-being predicted better engagement in care, clinics could consider providing resource information for ways to create or improve a sense of spiritual well-being for patients. These resources would need to be vetted, particularly for any religious institutions, to ensure that they would be welcoming and non-judgmental given the diversity of populations for PLWH. Any interventions or procurement of information would need to be done with significant engagement of the local community of PLWH to be appropriately sensitive and respectful.

This pilot study suggests that PLWH in an academic HIV care setting are willing to answer questions about their spirituality, although this is not universally true. Our findings suggest that spiritual well-being and social support are important predictors of clinical outcomes, specifically engagement in care. The next steps should include a more definitively powered study to examine the relations among social support, spiritual well-being, and other relevant clinical outcomes, including virologic suppression and inclusion of other HIV clinics in the region for better generalizability. Our preliminary results also would support further investigation of the role of perceived stigma and healthcare utilization, based on the trend toward significance between ED visits and stigma.

Key Points.

• The South is the current epicenter of human immunodeficiency virus (HIV) infection trends in the United States and has worse outcomes for people living with HIV (PLWH) compared with other regions of the country.

• Novel, culturally relevant approaches are needed to improve health outcomes for PLWH in the South.

• Although self-reported affiliation with a religious institution was not related to HIV viral suppression or CD4 count, social support and spiritual well-being significantly predict engagement in care. Fully powered studies are needed to better identify how social support and spiritual well-being can be leveraged to improve outcomes for PLWH in the South.