https://orcid.org/0000-0002-3295-0698
Figures
Abstract
Religion has contributed to societal divides regarding COVID-19 mRNA vaccines. In this study, we conducted a secondary analysis of a survey of U.S. adults (N = 4939) focused on how religious affiliations, beliefs, and practices impact attitudes toward genetic and genomic activities, one of which was mRNA vaccines. The dataset included large samples of participants from six religious groups in the U.S. (Black Protestant, Catholic, Evangelical Protestant, Jewish, Mainline Protestant, and Muslim), as well as individuals who were atheist, agnostic, or spiritual. ANCOVA results indicated that Evangelical Protestant participants showed significantly less support for mRNA vaccines than other groups, while atheist participants were the most supportive. Muslim participants had the highest concerns, whereas atheist participants had the lowest. Regression analyses indicated the strongest predictors of support for mRNA vaccines were more spiritual community support for community health, followed by higher acceptance of evolution, more liberal political orientation, less distrust toward the healthcare system, higher frequency of attending religious activities, higher income, lower fundamentalist religious beliefs, and more spiritual community support for liberal reproductive and end of life views. The strongest predictors of concerns about mRNA vaccines were more distrust toward the healthcare system and more conservative political orientation, followed by less spiritual community support for community health, stronger beliefs about God in the body, more fundamentalist religious beliefs, and lower knowledge of genetics. The large sample size and examination of a broad array of religious variables alongside distrust and political orientation offer new insights. These findings add to the literature on the culture wars surrounding mRNA vaccines, and can perhaps aid in future efforts to build trust and relationships between public health and religious communities.
Citation: Solomon ED, Chin EG, Baldwin K, Baker LL, DuBois JM (2026) Religious beliefs and practices, political orientation, and distrust in healthcare predict attitudes toward mRNA vaccines in the United States. PLOS Glob Public Health 6(9): e0006330. https://doi.org/10.1371/journal.pgph.0006330
Editor: Stephen Olaide Aremu, Nasarawa State University, NIGERIA
Received: April 7, 2026; Accepted: July 15, 2026; Published: September 9, 2026
Copyright: © 2026 Solomon et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The data that support the findings of this study are openly available in Open ICPSR at https://www.openicpsr.org/openicpsr/project/247241/version/V1/view.
Funding: This research was supported by the National Human Genome Research Institute (5R01HG012830, PI: DuBois) through March 11, 2025 and supported by Washington University institutional funds thereafter. Dr. Baker was supported by the National Center For Advancing Translational Sciences of the National Institutes of Health under Award Number UL1 TR002345 (PI: Powderly). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. The authors declare no conflicts of interest.
Competing interests: The authors have declared that no competing interests exist.
Introduction
In a recent article, Gostin and Wetter lamented that evidence-based practices that save lives—including specifically Covid-19 vaccines—are caught in the cross hairs of culture wars [1]. They characterize these wars as political, reflecting partisan divides. Davison Hunter, the sociologist who coined the term “culture war” believed they arose primary between conservative religious factions and secular society [2]. Most Covid-19 vaccines are mRNA vaccines, and were the first mRNA vaccines to be administered widely [3–6]. In this paper, we draw upon data from a survey of nearly 5,000 adults in the U.S. to explore how religion, politics, distrust, and other demographic variables influence attitudes toward mRNA vaccines [7].
The modest body of literature on religion and vaccines is primarily theoretical rather than empirical. One recent framework suggests that religious factors influencing vaccine attitudes generally fit into a few types [8]. One type involves an individual or faith group’s religious worldview. For example, some religions rely on faith and prayer for healing, rather than medical interventions like vaccines, because they view disease as being caused by sin or evil or they emphasize trust in God for healing [8–10]. Another type centers around the idea of purity. Some religions believe that vaccines are impure because they are unnatural or the ingredients used to manufacture vaccines are impure, and as such are unfit for the human body [8,11,12].
Morality is another religious factor influencing vaccine attitudes. Some religions view vaccines as problematic because of the use of gelatin in the vaccine and vaccine development, which can be derived from sacred animals, or animals that were not slaughtered in accordance with religious practices [8,13,14]. Beliefs about the sanctity of life for embryos and fetuses are rooted in moral concerns about the use of embryonic or fetal tissue in vaccine development [8,13]. However, though religious beliefs such as these are cited as concerns about vaccines, prior research has also suggested that at least in some cases, religious concerns about vaccines actually reflected concerns about vaccine safety or personal beliefs [14].
Religions have been described as “moral communities,” in which communal beliefs can significantly shape individual behavior [15]. When a religion indicates that vaccines are morally problematic or should not be used, it demonstrates how an individual-level decision can be influenced by the broader religious context. Additionally, much of our moral decision making is based on moral intuition or intuitive ethics—a fast, heuristic, outside of conscious awareness type of thinking that involves quick feelings of either approval or disapproval in relation to moral situations (i.e., a “gut reaction”) [16,17]. This type of thinking reinforces how ingrained religious beliefs and communal values can unconsciously shape individual decision making [16,17]. Thus, adopting a theoretical approach to understanding attitudes toward mRNA vaccines may fail to explain real-world attitudes and behaviors.
Religious affiliation and religious beliefs appear to play a major role in shaping attitudes toward vaccines [18–22]. Prior research has shown that individuals and communities higher in religiosity and who frequently attend religious services tend to have lower vaccination rates and lower support for vaccines [23–26]. However, these studies do not explore what specific features of religious engagement might explain vaccine hesitancy. Further, most of the prior literature that examines the influence of religion has focused on specific religious groups and their views of vaccines [8,11,12,27–32]. In rare cases, specific communities such as Orthodox Jews and the Amish, were examined following recent outbreaks of diseases like measles and polio, which occurred due to low vaccination rates [33–35].
mRNA vaccines
Most large empirical studies of Covid-19 mRNA vaccine attitudes have not examined the role of religious affiliation, beliefs, or practices on attitudes or acceptance of vaccines. For example, Iqbal et al report on the Global Listening Project, which conducted an interview-based survey of 70,000 adults from 70 nations on willingness to accept new mRNA Covid vaccines, but never inquired into how religion might influence participant’s willingness [36]. Another global survey of 7,000 adults across 7 nations explored trust and mistrust of new Covid-19 vaccines, but did not inquire into religious variables [37]. In 2022–23, Xu and colleagues reviewed over 541,000 tweets on Twitter related to confidence in mRNA vaccines, but focused solely on sentiments and confidence in the vaccines [38]. In contrast to these studies, Odame et al examined 2,858 Twitter posts using a qualitative interpretative approach, and found that mistrust in vaccine science, politics, and religious attitudes were featured most prominently as reasons for concerns regarding mRNA vaccines [39]. Other research has found that countries with more religious populations tended to have lower COVID-19 vaccination rates [40]. Additionally, some religious groups report higher levels of COVID-19 mRNA vaccination than non-religious individuals (e.g., Church of England, Roman Catholics), while other groups report lower levels (e.g., Muslim, Evangelical Protestant, Pentecostal) [41,42].
The current study
To summarize, much of the literature on religion and vaccines has been theoretical, and while empirical studies have documented an association between religion and vaccine attitudes, few studies have compared religious groups or explored specific aspects of religion that might predict vaccine attitudes. Having a better understanding of the attitudes of major religious groups and which religious beliefs and practices impact attitudes toward mRNA vaccines can inform public health efforts and has the potential to build public trust. Research focused on building trust in public health generally focuses on building trust among racial minorities, [43,44] and very little research has focused specifically on building trust among different religious groups. Thus, it is possible that building trust with religious groups may be an unutilized lever for improving vaccination rates and public health. Additionally, because the U.S. public attends religious services three times more than scientists in the U.S., it is important that scientists study the role of religion in public health because they are at risk of being less informed by personal experience [45].
In the current study, we conducted a secondary analysis of a large sample of U.S. adults [7]. The data was from an online survey of religious affiliation, beliefs, and practices, as well as attitudes toward various genetic and genomic activities, one of which was mRNA vaccines. Analyzing data from a large sample allowed us to 1) compare and contrast various major religions within the U.S. on their attitudes toward mRNA vaccines, as well as 2) examine which specific religious beliefs and practices contribute toward attitudes toward mRNA vaccines.
Materials and methods
Study design
This study was a secondary analysis of cross-sectional survey data, [46] which is appropriate for exploratory studies aimed at testing differences between groups and building predictive models. In the original survey, the data were collected anonymously. The Human Research Protections Office determined the study to be exempt from federal regulations. Participants were provided with an information sheet that disclosed that data would be shared and available to researchers for secondary analyses. The data was accessed for secondary analysis from 20/11/2024-14/03/2025. Further details on the survey and data collection can be found in a prior, open-access publication [7].
The data was originally collected via an online Qualtrics survey battery that consisted of measures assessing participants’ religious and spiritual beliefs and practices, including integration of faith into daily living, religious fundamentalism, views toward creation and evolution, beliefs about God in the body, religious discrimination, beliefs that God controls everything, the healthcare values of their spiritual community, frequency of and time spent in prayer and meditation, and frequency of volunteering in religious group activities [7,47–53]. The survey also included the Attitudes toward Genomics and Precision Medicine (AGPM) measure, measures of genetic knowledge, healthcare system distrust, and participant demographics [7,54–57].
Support for the use of mRNA vaccines was assessed with an AGPM item “I generally support the use of mRNA vaccines.” Concerns with mRNA vaccines were assessed using the mean score of responses to two AGPM items: “I worry that people will be required to get mRNA vaccines,” and “I am concerned that mRNA vaccines might change human bodies in ways that are unknown.” For all three items, response options ranged from 1 (strongly disagree) to 7 (strongly agree).
Because some of our analytical decisions were data-driven, we share some data in our Methods section, while reserving primary findings for the Results section. Further details on the analytical decisions are in the S1 File. The study was approved by the Institutional Review Board (IRB) at Washington University in St. Louis (IRB #202201153).
Sample characteristics
The original sample of participants (N = 4939) were recruited using two panel service companies, Prolific and CloudResearch. From Prolific, participants were recruited who were representative of the U.S. population in terms of age, gender, and race (n = 2999). From CloudResearch, participants were recruited in stratified samples of at least 300 participants from each of six religious groups: Black Protestant, Catholic, Evangelical Protestant, Jewish, Mainline Protestant, and Muslim (n = 1940). To qualify for the original study, participants needed to be 18 years or older and located in the U.S. Demographic details can be found in a prior, open-access publication [7].
Statistical analysis
To answer the two research questions, the study was divided into two major analytical sections. The first analytical section was comprised of two Analysis of Covariance (ANCOVA) models: one predicting support for the use of mRNA vaccines, and the other predicting concerns about mRNA vaccines. The second analytical section was comprised of two backward chunkwise elimination model building procedures [58]: one predicting support for the use of mRNA vaccines and the other predicting concerns about mRNA vaccines. Tables 1 and 2 depict the bivariate correlations between predictor variables used in the ANCOVAs and backward chunkwise elimination model building procedures with each outcome measure.
Analysis of Covariances (ANCOVAs).
The ANCOVAs compared each outcome across the nine religious and non-religious groups, accounting for several covariates. The nine religious and non-religious groups were (1) Black Protestant, (2) Catholic, (3) Evangelical Protestant, (4) Mainline Protestant, (5) Jewish, (6) Muslim, (7) atheist, (8) agnostic, and (9) spiritual but not religious (hereafter “spiritual”). Five hundred participants were not able to be included in the ANCOVA models because they did not belong to one of the nine religious/non-religious groups (e.g., Buddhist, Hindu, Mormon) and these subgroup samples were too small to analyze. Based on a review of prior literature, nine demographic covariates were selected: (1) age, (2) education, (3) household income, (4) conservative political orientation, (5) employment, (6) urban/suburban/rural status, (7) gender, (8) race, and (9) ethnicity. See the S1 File for a description of the ANCOVA assumption checking results.
For the ANCOVA predicting support for use of mRNA vaccines, in preliminary checks, we conducted bivariate correlations and one-way ANOVAs between the covariates and support for mRNA vaccines. Only covariates that had a (1) statistically significant relationship with the outcome and (2) evidenced at least a small effect size (e.g., r ≥ .10; partial ≥.0.02) were considered for inclusion in the model [59–61]. Three covariates (i.e., education, p < .001, r = .20; household income, p < .001, r = .14; and political orientation, p < .001, r = -.41) met the these criteria. These three covariates were assessed for collinearity; none of them had large effect sizes (e.g., r ≥ .50; partial
≥.0.26) and thus all were included in the ANCOVA [59,60]. Additionally, 116 participants were not able to be included when their responses could not be used in statistical analyses, such as “prefer not to answer” or “other” income and education respectively, resulting in a total sample size of n = 4,323 for this ANCOVA.
Similarly, for the ANCOVA predicting concerns about mRNA vaccines, we again conducted bivariate correlations and one-way ANOVAs between the covariates and the mean of the two concerns about mRNA vaccines items. Of the nine covariates, only one (i.e., conservative political orientation, p < .001, r = -.46) met the criteria for inclusion in the model. Collinearity was assessed, and conservative political orientation was not strongly related to the religious or non-religious group variable (e.g., r ≥ .50; partial ≥.0.26), thus it was included as a covariate in the ANCOVA model [59,60]. The final sample size for the ANCOVA was n = 4439. See Fig 1 for a depiction of the breakdown of samples for the ANCOVAs.
Because the exclusionary criteria are not mutually exclusive for the ANCOVA subsample predicting support (e.g., a participant may prefer not to state income and listed education as “other”), the total number of number of participants removed (n = 616) does not equal to the total number of participants identified as meeting exclusionary criteria (e.g., 500 + 103 + 29 = 632).
Model building: Backward chunkwise elimination procedure.
The backward chunkwise elimination procedure was conducted to build two regression models: (1) one to predict support for mRNA vaccines, and (2) one to predict concerns about mRNA vaccines. We identified 28 prospective predictor variables to include in the maximum model, which could be broadly divided into four categories: demographic variables, general covariates, religious group, and religious or spiritual variables [7]. Because the model building procedures involved religious or spiritual variables, participants who reported being atheistic or agnostic (n = 1008) were not able to be included in the model building procedures because they did not complete the religious or spiritual survey items. Additionally, participants who could not be categorized into any of the religious or non-religious groups (n = 500) were excluded from the model building procedures, leaving n = 3431 participants. Subsequently, this subset of participants was randomly divided into a model building (training) group (n = 1715) and model reliability testing (holdout) group (n = 1716). See Fig 1 for a depiction of the breakdown of the samples. More details about the backward chunkwise elimination procedure, maximum models, and reliability indices are described in the S1 File [58].
Based on the model building sample, employment was removed from the models because it was deemed to be too collinear with age (p < .001, =.43). Additionally, using the model building sample, we conducted preliminary correlations and ANOVAs between each predictor and each outcome (i.e., support and concerns) to exclude predictors that were either not statistically significant (p < .05) or had a small effect size (r < .1;
<.01; R2 < .02) [60,62].
For the model predicting support, we sequentially removed predictors that produced a minimum test statistic Fp that was smaller than FCRIT, where α = .0023 until the minimum test statistic Fp in the model was larger than FCRIT, leaving us with nine predictors [63]. These predictors were then subjected to a single variable backward elimination procedure using the same inclusion criterion (α = .0023). With the exception of frequency of volunteering, eight of the nine variables met the criterion for inclusion. Thus, these eight predictors were entered into a hierarchical regression, arranged according to previously mentioned conceptual categories [7].
Similarly, for the model predicting concerns, we sequentially removed predictors that produced a minimum test statistic Fp that was smaller than FCRIT, where α = .0023 until the minimum test statistic Fp in the model was larger than FCRIT, leaving us with seven predictors [63]. These predictors were then subjected to a single variable backward elimination procedure using the same inclusion criterion (α = .0023). With the exception of their perception of healthcare values of their spiritual community relating to reproductive and end of life issues, six of the seven variables met the criterion for inclusion. Thus, these six predictors were entered into a hierarchical regression, arranged according to previously mentioned conceptual categories.
Results
Dispersion of mean support scores
Fig 2 presents the distribution of participants’ Likert ratings for level of support for the use of mRNA vaccines with higher scores indicating higher levels of support. The mean score of 5.48 (SD = 1.70) roughly corresponds to a response between “somewhat agree” (a score of 5) and “agree” (a score of 6). Only 24.35% of respondents had mean ratings at or below neutral.
Fig 3 presents the distribution of participants’ mean level of concerns about the use of mRNA vaccines with higher scores indicating higher levels of concern. The mean score of 3.70 (SD = 1.89) roughly corresponds to a response of “neither agree nor disagree” (a score of 4). The majority of respondents had mean ratings at or below neutral (61.30%).
Analysis 1: ANCOVA comparing support for mRNA vaccines among religious and non-religious groups
The ANCOVA results indicated medium sized differences (partial =.019) across the religious and non-religious groups in terms of support for mRNA vaccines, F (8,4323) = 43.23, p < .001). Pairwise comparison between groups suggested that the atheist group had a significantly higher mean for support for use of mRNA vaccines compared to all other groups (
M ranged from 0.35 to 0.74, with all ps < .05), except for the agnostic group (
M = 0.18, p = .92). These differences were much larger (e.g., 6.34 vs. 4.70 when comparing atheists to Evangelical Protestants) prior to adjusting for the covariates in the model (i.e., education, household income, and conservative political orientation) but remained even after using adjusted means. Table 3 depicts the adjusted means and standard error (i.e., adjusted for covariates), as well as the unadjusted means, standard deviation, and corresponding number of participants in each group.
Analysis 2: ANCOVA comparing concerns about mRNA vaccines among religious and non-religious groups
The ANCOVA results indicated statistically significant small sized differences (partial =.06) across the religious and non-religious groups in terms of concerns about mRNA vaccines, F (8,4439) = 36.83, p < .001). Pairwise comparison between groups suggested that the Muslim group had a significantly higher mean for concerns about the use of mRNA vaccines compared to all other groups (
M ranged from 0.37 to 1.61, with all ps < .05), except for the Black Protestant group (
M = 0.36, p > .05). As in the prior model, group differences were much larger (e.g., 2.28 vs 4.09 when comparing atheists to Muslims) prior to adjusting for the covariate in the model (i.e., conservative political orientation) but remained even after using adjusted means. Table 4 depicts the adjusted means and standard error (i.e., adjusted for covariates), as well as the unadjusted means, standard deviation, and corresponding number of participants in each group.
Analysis 3: What religious beliefs and practices predict support for mRNA vaccines?
The eight predictors selected for inclusion were entered into a hierarchical regression using the model building (training) sample. We then evaluated the reliability of the model in the reliability testing (holdout) sample. The shrinkage on cross-validation was.018, which was smaller than.10 suggested threshold, supporting the conclusion that this was a reliable model [58].
As seen in Table 5, model had a moderate effect size (ΔR2 = .13) [62]. The religious variables that remained statistically significant predictors (ps < .0023) were: more spiritual community support for community health ( = 0.23, p < .001), higher acceptance of evolution (
= 0.19, p < .001), higher frequency of attendance in religious or spiritual activities (
= 0.12, p < .001), lower fundamentalist religious beliefs (
= -0.08, p = .001), and more spiritual community support for liberal reproductive and end of life views (
= 0.07, p < .001). Additionally, two covariates and one demographic variable significantly predicted support: lower conservative political orientation (
= -0.19, p < .001), less distrust towards the healthcare system
= -0.16, p < .001), and higher household income
= 0.11, p < .001).
Analysis 4: What religious beliefs and practices predict concerns about mRNA vaccines?
Similar to the prior regression, we first entered the six selected predictors into a hierarchical regression using the model building (training) sample, then evaluated the reliability of the model in the model reliability testing (holdout) sample. The shrinkage on cross-validation was.050, which was smaller than.10 suggested threshold [58], supporting the conclusion that this was a reliable model.
As seen in Table 6, model had a small effect size (ΔR2 = .06) [62]. The religious variables that remained statistically significant predictors (ps < .0023) were: less spiritual community support for community health ( = -0.15, p < .001), stronger beliefs about God in the body (
= 0.14, p < .001), and more fundamentalist religious beliefs (
= 0.13, p < .001). The strongest predictors of concerns were some of the covariates: more distrust toward the healthcare system (
= 0.32, p < .001) and higher conservative political orientation (
= 0.23, p < .001). Lower genetic knowledge (
= -0.11, p < .001) was also a significant predictor.
Discussion
We found that participants across religious groups generally supported mRNA vaccines and had modest levels of concerns, however there were significant differences between the religious groups. The strongest predictor of support for mRNA vaccines was their spiritual community’s support for community health, a more liberal political orientation, and higher acceptance of evolution. Regarding concerns with mRNA vaccines, the covariates of distrust toward the healthcare system and more conservative political orientation were the strongest predictors.
While we observed support for mRNA vaccines across the sample, there were significant differences between the religious groups. In particular, Evangelical Protestant participants had significantly lower support for mRNA vaccines than all other religious groups, while atheist participants had the most support. Similarly, Muslim participants reported the highest level of concerns about mRNA vaccines, while atheist participants reported the lowest concerns. These findings align with past research suggesting that some religious groups, such as Protestant and spiritual individuals, have lower support for vaccines [23,24,64].
For both support and concerns, each religious group mean had a relatively large standard deviation, indicating considerable variability within each group. Given the wide range of attitudes within each group, these findings suggest that religious affiliation alone is not necessarily a strong predictor of one’s level of support or concerns about mRNA vaccines.
Among those who identified with a religion, the strongest predictor of support for mRNA vaccines was their spiritual community’s support for community health. This indicates that religious individuals place significant value on their community’s perspective regarding health-related matters. Religious individuals tend to be higher in group-focused “binding” moral values, which support and enable group cohesion [15,65]. This finding supports the notion that their spiritual community, with which religious individuals feel connected, impacts their attitudes toward mRNA vaccines. This is noteworthy, considering vaccination decisions are often perceived as individual health decisions. Public health efforts to build trust among religious groups should consider engaging spiritual leaders and communities in order to increase mRNA vaccine acceptance and uptake.
The second strongest predictor of support was more acceptance of evolution. Views on evolution may reflect an individual’s acceptance of science and beliefs in creationism, thus, higher levels of acceptance of evolution being associated with more support for mRNA vaccines is understandable [50]. This aligns with prior research indicating these two variables (i.e., their spiritual community’s support for community health and acceptance of evolution) predict support for genetic and genomic healthcare more broadly [7]. Other significant predictors were more liberal political orientation, lower levels of distrust toward the healthcare system, more frequently attending religious activities, and lower fundamentalist religious beliefs.
Regarding concerns about mRNA vaccines, distrust toward the healthcare system was the strongest predictor, followed by more conservative political orientation. Both of these covariates were stronger predictors of concerns than they were of support for mRNA vaccines. They were also stronger predictors of concerns than any of the religious or demographic variables in the study. This is in contrast with previous research findings that religious variables were stronger predictors of attitudes genomic and genetic healthcare than political orientation, distrust, or demographics [7].
Other significant predictors of concerns for mRNA vaccines included lower genetic knowledge, more fundamentalist religious beliefs, beliefs about God in the body, and lower beliefs that your spiritual community supports community health. This makes good sense, as higher fundamentalist religious beliefs may lead an individual to prioritize faith-based healing such as prayer or other spiritual practices over medical interventions [8]. Higher beliefs of God in the body, that one’s body is a gift from God or is created in God’s image, may lead individuals to view bodily interventions like vaccines as unnatural or as disturbing the purity of the body [8]. And similar to earlier, it makes sense that lower levels of their spiritual community’s support for community health would predict concerns with mRNA vaccines, as religious individuals are likely to value on their community’s perspective regarding vaccines.
The finding that political orientation and distrust in the healthcare system were predictors of both support and concerns about mRNA vaccines underscores the culture war around mRNA vaccines within society and distrust of science and healthcare [1,2]. These findings indicate a need for not just for engaging religious and spiritual communities in trust-building efforts, but also the need to target and try to reduce the politicization of mRNA vaccines. These findings are also interesting in the sense that these may be instances of “gut reactions” type of thinking involved in moral decision-making [16,17]. If participants had been asked directly why they have concerns about mRNA vaccines, they may have provided various religious beliefs and practices as reasons for their concerns. In reality, at least according to the current study’s results, their political leanings and distrust of healthcare were stronger factors in their concerns than any of the religious beliefs and practices in the study, lending support to the idea that religious concerns about vaccines may actually reflect personal beliefs rather than theological objections [14]. At the same time, these findings also indicate that public health strategies relying on secular scientific messaging may not be effective if individuals perceive their concerns as being religiously motivated.
Last, the dispersion of scores for support vs. concerns of mRNA vaccines were very different. The distribution for support was skewed such that a majority of participants indicated they supported mRNA vaccines. However, the dispersion for concerns was much different, with the distribution spanning the full range of responses. These findings suggest that support and concerns about mRNA vaccines are separate constructs, and do not likely represent opposite sides of the same spectrum. When asking about overall support, it’s possible that participants provided a socially desirable response indicating general support, or that in theory they support mRNA vaccines. However, when asked about concerns, their responses indicate a wide variety of responses, perhaps reflective of the wide variety of views on mRNA vaccines in society. The concerns items were also more specific than the support item, and participants may support mRNA vaccines, but still have considerable concerns that may stop them from engaging with them.
This secondary analysis had several strengths. The dataset that was used was very large and contained large numbers of many major religious groups in the U.S. This allowed for comparisons of these major religious groups on their attitudes toward mRNA vaccines, whereas most of the literature in this area has been theoretical or conducted with specific religious communities [8,11,12,27–32]. Additionally, the dataset contained a large number of religious beliefs and practices, which allowed for analyses to determine which of these beliefs and practices were most influential in regard to attitudes toward mRNA vaccines. Finally, the dataset was largely representative of the U.S. in terms of age, gender, and race, which aids in the generalizability of these findings.
There were some limitations to the study. One limitation was that the study was a secondary analysis of an existing dataset. Thus, research questions were limited to exploring the variables present in the dataset. Because of this, our measures of support and concerns for mRNA vaccines were relatively brief. Support consisted of one item, and concerns consisted of two items from the AGPM measure. While brief, our prior work validating the AGPM suggests that these measures were reliable and valid, but they may not be capturing the full multi-dimensional nature of mRNA vaccine attitudes. Future research may want to use more robust assessments. Additionally, the online sampling methods used in the original data collection may underrepresent religious individuals who are more socially isolated or who hold more fundamentalist views, given the technological literacy required to participate. Furthermore, these data were collected in 2023, and evolving public discourse, pandemic fatigue, and shifting political narratives may impact the generalizability of these findings. As such, future research should continue to examine the potentially evolving relationship between religion and vaccine attitudes.
While the dataset used for this analysis contained large enough numbers to compare several major religious groups in the United States, it did not allow for comparison or insights into smaller religious groups. For example, from this dataset, there were too few Buddhist, Hindu, Mormon, or other common religious groups present in the U.S. to allow for inclusion in the analysis. Furthermore, the use of ANCOVA in non-experimental research is an important consideration when interpreting these findings. Because religious groups are not randomly assigned, covariates such as political orientation or income may be intrinsically tied to some religious groups more than others. Because ANCOVA assumes that the relationship between covariates and outcomes is consistent across groups (i.e., homogeneity of regression slopes), the adjusted means (Tables 3 and 4) should be interpreted with these considerations in mind.
In conclusion, the study identified differences in support and concerns with mRNA vaccines among several religious groups in the U.S., finding that within group differences were often larger than between group differences. Furthermore, we aimed to determine which religious beliefs and practices predicted support and concerns with mRNA vaccines and found that their spiritual community’s support for community health was the strongest religious factor impacting their attitudes, while political orientation and distrust in the healthcare system were also important factors. These findings add to the literature on the culture wars surrounding mRNA vaccines, and can perhaps aid in future efforts to build trust between public health and religious communities.
Supporting information
S1 File. Supplemental information on statistical analyses and samples.
https://doi.org/10.1371/journal.pgph.0006330.s001
(PDF)
References
- 1. Gostin LO, Wetter S. Public health in the crosshairs of culture wars. JAMA Health Forum. 2024;5(2):e240465. pmid:38386336
- 2.
Hunter JD. Culture wars: The struggle to control the family, art, education, law, and politics in America. New York: Basic Books. 1991.
- 3. Verbeke R, Lentacker I, De Smedt SC, Dewitte H. The dawn of mRNA vaccines: The COVID-19 case. Journal of Controlled Release. 2021;333:511–20.
- 4. Abbasí J. First mRNA HIV vaccine clinical trial launches. JAMA. 2022;327(10):909.
- 5. Schlake T, Thess A, Fotin-Mleczek M, Kallen K-J. Developing mRNA-vaccine technologies. RNA Biol. 2012;9(11):1319–30. pmid:23064118
- 6. Wang Y, Zhang Z, Luo J, Han X, Wei Y, Wei X. mRNA vaccine: a potential therapeutic strategy. Mol Cancer. 2021;20(1):33. pmid:33593376
- 7. DuBois JM, Chin EG, Solomon ED, Harris JK, Hill P, Baldwin K, et al. Religious factors predict support for genomic medicine more strongly than politics, education, or trust: A survey of 4,939 adults in the United States. Front Genet. 2025;16:1587774. pmid:40534838
- 8. Trangerud HA. “What is the problem with vaccines?” A typology of religious vaccine skepticism. Vaccine X. 2023;14:100349. pmid:37484867
- 9. Daniels CL, Ellison CG, Shattuck EC, Sunil TS, Xu X. God knows best: exploring the god locus of health control, race-ethnicity, and risk-taking and protective behaviors interrelationship in the USA. J Relig Health. 2025;64(6):4847–66. pmid:39753995
- 10. Upenieks L, Ford-Robertson J, Robertson JE. Trust in god and/or science? sociodemographic differences in the effects of beliefs in an engaged god and mistrust of the COVID-19 vaccine. J Relig Health. 2022;61(1):657–86. pmid:34843011
- 11. Meier BP, Dillard AJ, Fetterman AK, Ji L-J, Lappas CM. Religiosity and the naturalness bias in drug and vaccine choices. J Relig Health. 2023;62(1):702–19. pmid:36394691
- 12. Boone BC, Johnson RT, Rolando LA, Talbot TR. Use of a comprehensive program to review religious and personal seasonal influenza vaccination exemption requests by healthcare personnel. Infect Control Hosp Epidemiol. 2021;42(5):507–12. pmid:33118915
- 13. Wombwell E, Fangman MT, Yoder AK, Spero DL. Religious barriers to measles vaccination. J Community Health. 2015;40(3):597–604. pmid:25315714
- 14. Grabenstein JD. What the world’s religions teach, applied to vaccines and immune globulins. Vaccine. 2013;31(16):2011–23. pmid:23499565
- 15. Graham J, Haidt J. Beyond beliefs: religions bind individuals into moral communities. Personality and Social Psychology Review. 2010;14(1):140–50.
- 16. Haidt J, Joseph C. Intuitive ethics: how innately prepared intuitions generate culturally variable virtues. Daedalus. 2004;133(4):55–66.
- 17.
Graham J, Haidt J, Koleva S, Motyl M, Iyer R, Wojcik SP, et al. Moral foundations theory: The pragmatic validity of moral pluralism. Advances in experimental social psychology. Elsevier. 2013. 55–130.
- 18. Garcia LL, Yap JFC. The role of religiosity in COVID-19 vaccine hesitancy. J Public Health (Oxf). 2021;43(3):e529–30. pmid:34080617
- 19. Milligan MA, Hoyt DL, Gold AK, Hiserodt M, Otto MW. COVID-19 vaccine acceptance: influential roles of political party and religiosity. Psychology, Health & Medicine. 2021;:1–11.
- 20. Tiwana MH, Smith J. Faith and vaccination: a scoping review of the relationships between religious beliefs and vaccine hesitancy. BMC Public Health. 2024;24(1):1806. pmid:38971784
- 21. Dubé E, Gagnon D, Nickels E, Jeram S, Schuster M. Mapping vaccine hesitancy--country-specific characteristics of a global phenomenon. Vaccine. 2014;32(49):6649–54. pmid:25280436
- 22. Corcoran KE, Scheitle CP, DiGregorio BD. Christian nationalism and COVID-19 vaccine hesitancy and uptake. Vaccine. 2021;39(45):6614–21. pmid:34629205
- 23. Martens JP, Rutjens BT. Spirituality and religiosity contribute to ongoing COVID-19 vaccination rates: Comparing 195 regions around the world. Vaccine X. 2022;12:100241. pmid:36407820
- 24. Nagar S, Ashaye T. A shot of faith-analyzing vaccine hesitancy in certain religious communities in the United States. Am J Health Promot. 2022;36(5):765–7. pmid:34978209
- 25. Barnack JL, Reddy DM, Swain C. Predictors of parents’ willingness to vaccinate for human papillomavirus and physicians’ intentions to recommend the vaccine. Womens Health Issues. 2010;20(1):28–34. pmid:20123174
- 26. Shelton RC, Snavely AC, Jesús M d, Othus MKD, Allen JD. HPV vaccine decision-making and acceptance: does religion play a role?. Journal of Religion and Health. 2013;52:1120–30.
- 27. Carson PJ, Flood AT. Catholic social teaching and the duty to vaccinate. Am J Bioeth. 2017;17(4):36–43. pmid:28328377
- 28. Muravsky NL, Betesh GM, McCoy RG. Religious doctrine and attitudes toward vaccination in jewish law. J Relig Health. 2023;62(1):373–88. pmid:34708328
- 29. Whitehead AL, Perry SL. How culture wars delay herd immunity: christian nationalism and anti-vaccine attitudes. Socius: Sociological Research for a Dynamic World. 2020;6.
- 30. Jones DG. Religious concerns about COVID-19 vaccines: from abortion to religious freedom. J Relig Health. 2022;61(3):2233–52.
- 31. Alsuwaidi AR, Hammad HAA-K, Elbarazi I, Sheek-Hussein M. Vaccine hesitancy within the Muslim community: islamic faith and public health perspectives. Hum Vaccin Immunother. 2023;19(1):2190716. pmid:36914409
- 32. Kasstan B. “If a rabbi did say ‘you have to vaccinate,’ we wouldn’t”: Unveiling the secular logics of religious exemption and opposition to vaccination. Soc Sci Med. 2021;280:114052. pmid:34051560
- 33. Baugh V, Figueroa J, Bosanquet J, Kemsley P, Addiman S, Turbitt D. Ongoing measles outbreak in Orthodox Jewish community, London, UK. Emerging Infectious Diseases. 2013;19:1707–9.
- 34. Fournet N, Mollema L, Ruijs WL, Harmsen IA, Keck F, Durand JY, et al. Under-vaccinated groups in Europe and their beliefs, attitudes and reasons for non-vaccination; two systematic reviews. BMC Public Health. 2018;18(1):196. pmid:29378545
- 35. Kisjes KH, Duintjer Tebbens RJ, Wallace GS, Pallansch MA, Cochi SL, Wassilak SGF, et al. Individual-based modeling of potential poliovirus transmission in connected religious communities in North America with low uptake of vaccination. J Infect Dis. 2014;210 Suppl 1(Suppl 1):S424-33. pmid:25316864
- 36. Iqbal SM, Rosen AM, Edwards D, Bolio A, Larson HJ, Servin M, et al. Opportunities and challenges to implementing mRNA-based vaccines and medicines: lessons from COVID-19. Front Public Health. 2024;12:1429265. pmid:39175908
- 37. D’Silva C, Fullerton MM, Hu J, Rabin K, Ratzan SC. A global survey to understand general vaccine trust, COVID-19 and influenza vaccine confidence. Front Public Health. 2024;12:1406861. pmid:39635221
- 38. Xu J, Wu Z, Wass L, Larson HJ, Lin L. Mapping global public perspectives on mRNA vaccines and therapeutics. NPJ Vaccines. 2024;9(1):218. pmid:39543153
- 39. Odame EA, Dada O, Nelson J, Ogunyiola A, Haley J. Preparing for future pandemics: A qualitative exploration of social media in light of the COVID-19 pandemic and vaccine hesitancy. PLOS Glob Public Health. 2025;5(7):e0004317. pmid:40622951
- 40. Mao Z-F, Li Q-W, Wang Y-M, Zhou J. Pro-religion attitude predicts lower vaccination coverage at country level. Humanit Soc Sci Commun. 2024;11(1).
- 41. Orlandi LB, Febo V, Perdichizzi S. The role of religiosity in product and technology acceptance: evidence from COVID-19 vaccines. Technol Forecast Soc Change. 2022;:122032. pmid:36117496
- 42. Hansen EM, Pickering SD. The role of religion and COVID-19 vaccine uptake in England. Vaccine. 2024;42(13):3215–9. pmid:38677793
- 43. Christopher S, Watts V, McCormick AKHG, Young S. Building and maintaining trust in a community-based participatory research partnership. Am J Public Health. 2008;98(8):1398–406. pmid:18556605
- 44. Meredith LS, Eisenman DP, Rhodes H, Ryan G, Long A. Trust influences response to public health messages during a bioterrorist event. J Health Commun. 2007;12(3):217–32. pmid:17497377
- 45. Ecklund EH, Johnson DR, Scheitle CP, Matthews KR, Lewis SW. Religion among scientists in international context: A new study of scientists in eight regions. Socius. 2016;2:1–9.
- 46.
DuBois JM. Which religious and personal characteristics predict attitudes toward mRNA vaccines in the United States?. Ann Arbor, MI: Inter-university Consortium for Political and Social Research. 2026.
- 47. Hoge R. A Validated Intrinsic Religious Motivation Scale. Journal for the Scientific Study of Religion. 1972;11(4):369.
- 48. Altemeyer B, Hunsberger B. A revised religious fundamentalism scale: The short and sweet of it. The International Journal for the Psychology of Religion. 2004;14(1):47–54.
- 49. Rutledge ML, Warden MA. The development and validation of the measure of acceptance of the theory of evolution instrument. School Sci & Mathematics. 1999;99(1):13–8.
- 50. Rutledge ML, Sadler KC. Reliability of the measure of acceptance of the theory of evolution (MATE) instrument with university students. The American Biology Teacher. 2007;69(6):332–5.
- 51. Mahoney A, Carels RA, Pargament KI, Wachholtz A, Edwards Leeper L, Kaplar M, et al. RESEARCH: “The sanctification of the body and behavioral health patterns of college students”. International Journal for the Psychology of Religion. 2005;15(3):221–38.
- 52. Kawika Allen GE, Wang KT, Richards PS, Ming M, Suh HN. Religious Discrimination scale: development and initial psychometric evaluation. J Relig Health. 2020;59(2):700–13. pmid:29705827
- 53. Wallston KA, Malcarne VL, Flores L, Hansdottir I, Smith CA, Stein MJ. Does god determine your health? the god locus of health control scale. Cognitive Therapy and Research. 1999;23(2):131–42.
- 54. DuBois JM, Mozersky J, Antes A, English T, Parsons MV, Baldwin K. Attitudes toward genomics and precision medicine. J Clin Transl Sci. 2021;5(1):e120. pmid:34267947
- 55. Harris JK, Solomon ED, Baldwin K, Baker LL, Chin EG, DuBois JM. Integrating mRNA vaccines into the attitudes toward genomics and precision medicine scale: A validation study with a sample of 4939 adults in the USA. J Clin Transl Sci. 2025;9(1):e60. pmid:40201642
- 56. Furr LA, Kelly SE. The Genetic Knowledge Index: developing a standard measure of genetic knowledge. Genet Test. 1999;3(2):193–9. pmid:10464667
- 57. Shea JA, Micco E, Dean LT, McMurphy S, Schwartz JS, Armstrong K. Development of a revised Health Care System Distrust scale. J Gen Intern Med. 2008;23(6):727–32. pmid:18369678
- 58.
Kleinbaum DG, Kupper LL, Nizam A, Muller KE. Applied Regression Analysis and Other Multivariable Methods. 4th ed. Duxbury Press. 2008.
- 59.
Cohen J. Statistical Power Analysis for the Behavioral Sciences. 2 ed. Hillsdale, NJ: Lawrence Earlbaum Associates; 1988.
- 60. Cohen J. A power primer. Psychol Bull. 1992;112(1):155–9. pmid:19565683
- 61.
Field A. Discovering Statistics Using IBM SPSS Statistics. 4th ed. Sage. 2013.
- 62. Cohen J. Statistical power analysis for the behavioral sciences. 2013.
- 63. Šidák Z. Rectangular confidence regions for the means of multivariate normal distributions. Journal of the American Statistical Association. 1967;62(318):626–33.
- 64. Zarzeczna N, Većkalov B, Rutjens BT. Spirituality and intentions to engage in Covid‐19 protective behaviours. Social & Personality Psych. 2023;17(8).
- 65. Graham J, Nosek BA, Haidt J, Iyer R, Koleva S, Ditto PH. Mapping the moral domain. J Pers Soc Psychol. 2011;101(2):366–85. pmid:21244182