Comparing role of religion in perception of the COVID-19 vaccines in Africa and Asia Pacific

Rapid development and deployment of COVID-19 vaccines since late 2020 have reduced disease severity and population harms, yet inequities in vaccine access and variable acceptance hinder universal immunization. By mid-2023, over 80% of the Asia Pacific population had received at least one dose, versus under 40% in Africa, reflecting regional disparities. Beyond access, vaccine hesitancy—now often framed as indecision—remains a critical obstacle shaped by perceptions of risk, confidence in vaccines and authorities, and socio-demographic factors including gender, education, and religion. Religiosity has been linked to reduced support for vaccination; conflicts with religious teachings (e.g., concerns over porcine derivatives in Islam) and misinformation circulating in religious contexts can fuel doubts. Prior work often focuses on single countries, limiting comparative insight across diverse faiths and regions. Given Africa’s predominance of Christians and Muslims and the Asia Pacific’s more heterogeneous religious landscape (including Buddhism and Hinduism), this study investigates how religion relates to perceptions of vaccine–religion compatibility and COVID-19 vaccine acceptance across Africa and the Asia Pacific at two time points post-rollout. The aim is to quantify the interplay among region, religion, time, and socio-demographic factors, and how these dynamics evolved between survey waves.

Existing literature shows a strong association between religiosity and lower vaccination support, with mechanisms including perceived conflicts with doctrine (e.g., Islamic prohibitions related to pork-derived components) and the spread of vaccine misinformation within religious communities. Cross-national studies document variability in vaccine hesitancy determinants, while global surveys highlight heterogeneous vaccine confidence across regions. Prior research frequently examines single-country contexts or specific religious groups, providing limited comparative analysis across multiple religions and regions. Notably, Christianity and Islam dominate in Africa, whereas the Asia Pacific has more diverse affiliations, including substantial Buddhist and Hindu populations. Studies have also linked political ideology and religiosity to COVID-19 vaccine acceptance, and reported mixed associations between education and vaccine confidence, sometimes negative in higher-income settings. This study builds on these insights by directly comparing multiple religious groups across Africa and the Asia Pacific over time.

Design and data sources: Cross-sectional analyses using two waves (mid-2021 to 2022) of Vaccine Confidence Project surveys from 15 countries in Africa and the Asia Pacific. Approximately 1000 adult residents per country per wave were surveyed via telephone, face-to-face, or online modes. Asia Pacific waves mainly occurred in June 2021 and May 2022; Africa waves occurred around February and August 2022. All waves followed vaccine rollout (though vaccine supply remained limited in some African settings). Sample and grouping: Religions with at least 50 respondents in either region were included. Region-specific religious groups were defined, yielding eight groups: Christian (Africa), Muslim (Africa), and in Asia Pacific: Animist, Atheist, Buddhist, Christian, Hindu, Muslim. Respondents with incomplete demographics (age, gender, education), non-binary gender, or missing responses to the religious compatibility statement were excluded (~1%). Final samples: Wave 1 n=14,121 and Wave 2 n=14,107, with ~48% from Africa. Outcomes: (1) Perception of compatibility between vaccines and respondent’s religious beliefs (binary: agree vs other); (2) COVID-19 vaccine acceptance (binary: vaccinated or willing to accept if beneficial for self/family/friends/at-risk groups). Primary models (multivariate logistic regression using survey weights): - Compatibility: compatible ~ time + religion + education + gender + age. - Acceptance: acceptance ~ time + religion + education + gender + age. - Acceptance with compatibility: acceptance ~ time + religion + education + gender + age + compatible. Interaction models to assess heterogeneity across subpopulations and time (non-significant interactions removed): I. Compatibility: compatible ~ religion × time × education + gender × age. II. Acceptance: acceptance ~ religion × time × education + gender × age. III. Acceptance with compatibility: acceptance ~ religion × time × compatibility + religion × time × education + gender × age + age × compatible. Statistical analysis: Conducted in R using survey (weighted analyses) and grid packages. Univariate and bivariate regressions supplemented multivariable models. Point estimates with 95% CIs are reported. Ethics approvals were obtained from the University of Hong Kong (Africa: EA230420) and LSHTM (Asia Pacific: LSHTM 26636).

- Religious composition: In Africa, Christians were the majority with ~30% Muslims; in the Asia Pacific, Buddhists formed the largest group (~25%), with ~10% each Christian and Atheist. Compositions were stable across waves. - Compatibility perceptions: Atheists and Buddhists in the Asia Pacific consistently showed the lowest agreement that vaccines are compatible with religion (fewer than 60%). Atheists had elevated odds of disagreeing with compatibility (ORs 1.11 [0.99–1.25] Wave 1; 1.66 [1.47–1.89] Wave 2) and were more likely to respond “don’t know/undecided” (ORs 3.65 [3.19–4.17] Wave 1; 2.35 [1.96–2.81] Wave 2). In Wave 1, Muslims (Africa and Asia Pacific) were more likely than African Christians to agree vaccines are compatible with religion; in Wave 2, African Muslims remained more supportive than African Christians, whereas in the Asia Pacific the more supportive groups shifted to Animists and Christians. - Time trends in compatibility: Agreement with compatibility increased from Wave 1 to Wave 2 for all groups except African Muslims and Asia Pacific Atheists, Hindus, and Muslims. Increases among Asia Pacific Christians and Buddhists exceeded those among African Christians; for AP Christians, gains were mainly among those with secondary education. - Education heterogeneity in compatibility: Among AP Animists in Wave 1, those with secondary-or-below education had higher odds of endorsing compatibility than both similarly educated African Christians (OR 1.45, 95% CI 1.05–2.02) and higher-educated AP Animists (OR 1.96, 95% CI 1.51–2.56). AP Buddhists with post-secondary education were less supportive than those with lower education (OR 0.61 [0.45–0.83] Wave 1; 0.37 [0.21–0.64] Wave 2). In contrast, higher education increased compatibility support among AP Muslims, especially in Wave 1 (OR 2.46, 95% CI 1.34–4.52). Older age (45+) and male gender were associated with higher compatibility support. - Vaccine acceptance levels: Acceptance was consistently higher in the Asia Pacific than Africa by at least four percentage points in both waves, with the gap widening in Wave 2. Acceptance increased over time for all groups except African Muslims (decrease; OR 0.81, 95% CI 0.69–0.95) and AP Atheists; AP Muslims showed only mild increases. - Education heterogeneity in acceptance: Gains from Wave 1 to Wave 2 were generally larger among those with secondary-or-below education; among African Christians and AP Muslims, between-wave differences were not evident for post-secondary educated individuals (OR 1.16 [0.97–1.38] and 2.11 [0.94–4.74], respectively). In Wave 1, post-secondary education increased acceptance among AP Christians and Muslims in both regions, whereas AP Buddhists with lower education were more pro-vaccine in both waves. - Role of compatibility in acceptance: Acknowledging vaccine–religion compatibility was positively associated with acceptance overall, but the association weakened in Wave 2, particularly in Africa compared with Asia Pacific groups. Among those agreeing with compatibility, acceptance decreased notably for African Muslims in Wave 2. The positive impact of compatibility on acceptance was stronger among older adults (45+), and age effects on acceptance were mainly evident among those who endorsed compatibility (OR 1.66, 95% CI 1.44–1.89).

The study demonstrates that religion is a salient, context-dependent correlate of COVID-19 vaccine perceptions, with clear regional heterogeneity. Asia Pacific respondents, despite lower perceptions of religious compatibility among Atheists and Buddhists, consistently exhibited higher vaccine acceptance than African respondents, suggesting that factors beyond perceived compatibility—such as vaccine availability, policy environments, and broader social dynamics—shape acceptance. Temporal increases in both compatibility and acceptance were uneven across religious groups, regions, and education strata. Education showed mixed effects: in some groups (e.g., AP Muslims) higher education enhanced perceived compatibility and acceptance, whereas in others (e.g., AP Buddhists) higher education correlated with lower compatibility support and relatively lower acceptance, potentially reflecting differential trust in institutions or media. The generally positive association between perceiving vaccines as compatible with religion and acceptance aligns with prior research, but this relationship weakened over time, likely due to increased vaccine access, rising overall acceptance, and evolving risk perceptions. Age and gender further modulated effects, with older adults and males more likely to endorse compatibility and older age amplifying the compatibility–acceptance link. These findings address the research question by quantifying how religion, region, time, and socio-demographics interact to influence vaccine perceptions, underscoring the need for targeted, faith- and subgroup-sensitive strategies to bolster vaccine confidence and uptake.

This cross-sectional, two-wave, multi-country analysis across Africa and the Asia Pacific shows that vaccine perceptions are strongly shaped by religious affiliation, region, and socio-demographic factors, and that these relationships evolve over time. Atheists and Buddhists in the Asia Pacific were least likely to view vaccines as religiously compatible, while vaccine acceptance remained higher overall in the Asia Pacific than Africa. Perceiving vaccine–religion compatibility fostered acceptance, particularly among older adults, though this effect attenuated in the second wave. Education exerted heterogeneous influences across faiths, with lower-educated groups often driving gains over time. These insights support religion-tailored communication and engagement, with attention to education level, age, and regional context. Future research should leverage longitudinal designs over longer horizons, incorporate country-level policy and epidemiologic contexts, and examine evolving drivers of primary and booster vaccination intent to inform adaptive, faith-sensitive strategies that enhance coverage and reduce disease burden.

- Timing differences between Africa and Asia Pacific waves (shorter intervals in Africa) may have dampened observed changes in Africa and reduced comparability. - Pooled multi-country data within regions may introduce bias from unmeasured country-level factors (e.g., transmission, policies, misinformation) not modeled explicitly. - Limited covariates: employment and socioeconomic status were unavailable; education may only partially capture these influences. - Class imbalance in Wave 2 Asia Pacific (near-universal acceptance among Animists and Hindus) hindered precise estimation of covariate effects. - Potential confounding from increasing public awareness, evolving variants, and changing access over time; cross-sectional design limits causal inference. - Sensitivity analyses suggest robustness to alternative time specifications, but residual bias may remain.