Belief in life after death is one of the oldest expressions of spiritual belief1, prevalent across various world civilizations and religious traditions2. Most major religions acknowledge an immaterial aspect of human existence3, highlighting the enduring significance of the idea that life continues beyond physical death. Exploring afterlife beliefs is therefore essential for understanding religious tenets and their broader impact on individuals and societies.

Most existing evidence on afterlife beliefs comes from studies conducted within a Western Christian context. Afterlife beliefs shape our priorities and values and have been associated with several potential benefits4. For instance, in the face of existential threats, belief in an afterlife is linked to lower levels of death anxiety and greater acceptance of mortality5. The idea of reuniting with loved ones in an afterlife has been found to reduce depressive symptoms and anger among bereaved spouses 6 and 18 months after their loss6. Afterlife beliefs align with a belief in a just world rather than a cynical one7. Furthermore, they are also inversely related to anxiety and positively associated with feelings of tranquility8, which may offer coping strategies during financial hardship9. In a national sample of 1,403 adult Americans, afterlife beliefs were inversely associated with psychiatric symptoms like anxiety, depression, and paranoia, even when controlling for demographic factors, stress, and social support10.

In some cases, evidence has been more mixed. For instance, among terminal cancer patients, the association between afterlife belief and lower levels of end-of-life despair disappeared when factors like meaning and peace were accounted for11. Similarly, one cross-national study has shown null associations between afterlife belief and happiness, particularly in the more secular countries of Denmark and Czech Republic12. These findings suggest that the effects of afterlife beliefs may depend on the specific nature of the beliefs and how they are integrated into an individual’s broader worldview.

In addition to potential psychological benefits, afterlife beliefs can have moral implications. They can motivate virtuous or, in some cases, harmful actions. Glock & Stark13 categorized afterlife belief within the consequential dimension of religiosity. Beliefs about moral transgressions are often tied to ideas of supernatural monitoring and punishment, supporting the argument that these beliefs may promote cooperation within human groups14. Moral objections to suicide, including concerns about potential negative consequences in the afterlife, have emerged as significant protective factors against suicidal behaviors15,16. On the other hand, the prospect of posthumous rewards can sometimes distort moral judgment, potentially justifying harmful actions, as seen in cases of suicide terrorism17.

Predictors of afterlife belief

Given the complexity of afterlife beliefs and their effects, it is important to understand the factors that shape them. One of the primary influences is of course religious practice, as religion provides a framework for the transmission of afterlife beliefs. Attending religious services allows individuals to assimilate these beliefs through social interactions. Comprehensive cross-sectional data support the association between religious service attendance and belief in an afterlife. A recent study using the Global Flourishing Study (GFS), which pooled data from 22 countries through a random effects meta-analysis, found that belief in an afterlife was highest among individuals who attended religious services at least once per week and lowest among those who never attended18. Furthermore, modest evidence from longitudinal studies has supported a link between religiousness and afterlife beliefs among adolescents19. However, there is still a lack of evidence showing lagged associations between religious practices and afterlife beliefs across different countries, leaving the important question of how afterlife beliefs develop over the life course largely unaddressed.

Gender also plays a role, with women more likely than men to believe in an afterlife20. However, women consistently report placing greater importance on religious beliefs21, necessitating a formal test to distinguish the effects of gender from those of religious practice on afterlife beliefs. Age is another significant factor; older adults tend to report stronger beliefs in an afterlife compared to younger individuals22. In the GFS18, meta-analytic results showed that belief in the afterlife generally increased with age, peaking among those aged 80 and older. However, this trend is not universal; in some countries, younger adults reported higher belief levels than middle-aged and older individuals (e.g., Japan, Germany). Without accounting for religious practice, it is challenging to disentangle cohort effects—older generations may have lived in times when religion was more widely practiced, potentially explaining their stronger afterlife beliefs.

While many prior studies have focused on individual differences, fewer have examined how early life experiences and environmental factors shape afterlife beliefs. As childhood is a formative period for development, exposure to risk factors (e.g., abuse) or protective factors (e.g., parental warmth) can have lasting effects on well-being in adulthood23,24. Parental relationships, in particular, play a key role in shaping children’s religious development. The religious transmission hypothesis suggests that children tend to adopt the religious beliefs and practices of their parents, with parental church attendance being a major influence25. However, few studies have specifically examined how early childhood experiences influence afterlife beliefs later in life.

These factors align with Bronfenbrenner’s26 ecological perspective on human development, which recognizes that an individual’s belief system is shaped by earlier salient experiences within a multilayered network of environments. The microsystem represents the most immediate and influential context, such as a child’s relationship with parents. The mesosystem encompasses interactions between elements of the microsystem, such as parental marital status and the child’s perceptions of treatment in their environment. The exosystem refers to external contexts that indirectly influence the child through impacts on the microsystem, such as family financial status and religious practices. Finally, the macrosystem includes broader sociocultural factors, like social attitudes on immigration, that shape developmental experiences on a societal level.

The present study

This preregistered study (https://osf.io/qczrb) takes an initial step toward addressing key gaps in understanding childhood predictors of afterlife belief later in life. Using nationally representative data from 22 countries on a sample of over 200,000 individuals in the first wave of the GFS, we simultaneously examined 13 potential childhood predictors. This approach allows us to disentangle shared variances and address possible confounding effects due to overlapping influences. The predictors include religious service attendance, individual characteristics (e.g., birth year, gender, childhood health), and early life experiences (e.g., abuse, feeling like an outsider, and parental relationships). We expected these childhood factors to show unique, meaningful associations with afterlife beliefs in adulthood across different countries, even after controlling for each other’s influence. However, given the diverse sociocultural contexts of the 22 countries, we anticipated variation in the patterns of associations between childhood predictors and afterlife beliefs across nations.

Results

Descriptive statistics

The distribution of descriptive statistics for the sample is shown in Table 1. The sample represents the entire adult lifespan (18 to ≥ 80 years), with the largest cohort (20%) consisting of individuals born between 1983 and 1993 (currently aged 30–39 years). Most participants were born in the country where the data were collected (94%). A majority reported that their parents were married when they were about 12 years old (75%), they had a very good relationship with their mother (63%) and father (53%) while growing up, their family was financially stable or lived comfortably around age 12 (76%), they enjoyed very good or excellent health during childhood (64%), and they attended religious services at least once a week at age 12 (41%). A minority indicated experiences of abuse (14%) or feeling like an outsider in their family during childhood (14%).

Table 1 Nationally representative childhood descriptive statistics of the observed sample (N = 202,898).
Full size table

Nationally representative descriptive statistics for each country are presented in Tables S1a to S22a, revealing some variability across countries. For example, the percentage of participants who attended religious services at least once a week at age 12 was substantially higher in Kenya (81%, Table S10a) and Nigeria (87%, Table S12a) than in Japan (1.9%, Table S9a) and Sweden (6.3%, Table S21a). This suggests that early life experiences differ across countries, reflecting unique sociocultural contexts.

Cross-country childhood predictors of afterlife belief

Results of the random-effects meta-analyses are presented in Table 2, while country-specific results can be found in Tables S1b-S22b in the Supplementary Materials. Across the countries, the 95% confidence intervals (CIs) for the effect estimates showed an association with afterlife belief in adulthood for 6 of the candidate predictors (though there was evidence, from the global p-values, that each of the childhood predictors was associated with afterlife belief in at least one country). The strongest predictor was the frequency of religious service attendance around age 12. Compared to individuals who never attended, the likelihood of afterlife belief was higher among those who attended less than once a month (RR = 1.11, 95% CI: 1.05, 1.18), 1–3 times a month (RR = 1.26, 95% CI: 1.15, 1.38), or at least once a week (RR = 1.39, 95% CI: 1.25, 1.55).

Table 2 Random effects meta-analysis for associations of childhood predictors with belief in life after death in adulthood, and sensitivity to unmeasured confounding.
Full size table

There was also a greater likelihood of afterlife belief in adulthood among those who experienced abuse during childhood compared to those who did not (RR = 1.07, 95% CI: 1.02, 1.11), and among women compared to men (RR = 1.09, 95% CI: 1.03, 1.15). Individuals who reported having a very good or somewhat good relationship with their mother in childhood were more likely to believe in the afterlife as adults, compared to those with a very bad or somewhat bad relationship (RR = 1.03, 95% CI: 1.00, 1.07). Additionally, and in contrast, those who felt like an outsider in their family growing up also had a higher likelihood of afterlife belief compared to those who did not (RR = 1.03, 95% CI: 1.00, 1.06). Year of birth also showed some association, though the pattern was complicated. Compared to individuals born between 1998 and 2005 (currently aged 18–24), those born between 1993 and 1998 (25–29 years old, RR = 1.03, 95% CI: 1.00, 1.06) and between 1973 and 1983 (40–49 years old, RR = 1.05, 95% CI: 1.00, 1.10) were more likely to believe in the afterlife.

The pooled meta-analytic results provided little evidence of an association between other childhood factors and afterlife belief across the countries studied. Specifically, the 95% confidence intervals included 1.00 for the quality of father-child relationship, parental marital status around age 12, subjective financial status of the family around age 12, self-rated health in childhood, and immigration status.

The results of the E-value sensitivity analysis for the random-effects meta-analyses are reported in the last two columns of Table 2 (see Tables S1c–S22c for E-values in each country). For associations that excluded the null, E-values for the effect estimates (ranging from 1.20 to 2.13) and confidence interval limits (ranging from 1.00 to 1.80) suggesting that some of these associations are moderately robust against potential unmeasured confounding. For instance, with regard to the strongest predictor, to explain away the observed association between attending religious service at least once a week (vs. never) and afterlife belief in adulthood, an unmeasured confounder that was associated with both weekly service attendance and afterlife beliefs with risk ratios of 2.13-fold each, beyond what was accounted for by the measured covariates, could suffice, but weaker joint confounder could. For the confidence interval limit, an unmeasured confounder with a risk ratio associations of 1.80 for both weekly service attendance and afterlife beliefs could suffice to shift the interval to include the null, but weaker joint confounder associations could not.

When we repeated the meta-analyses using a population-weighted approach—where each country’s results were weighted by their population size in 2023 (see Table S23)—the patterns of results remained comparable to those from the random-effects meta-analyses. However, some previous weaker effects now included the null (e.g., the 95% CI for the relationship with mother ranged from 0.95 to 1.09).

Similarities and differences across countries

Table 2 highlights cross-national variation in the associations between childhood predictors and afterlife belief in adulthood. Metrics such as the proportion of effects by thresholds, I² values, and global p-values all indicate evidence of heterogeneity. A more detailed description of this heterogeneity is provided in the forest plots, which show country-specific effect estimates and 95% confidence intervals for each category of the childhood predictors (see Figures S1-S27).

Cross-national heterogeneity was observed for all childhood predictors, with some predictors showing more variation than others. For instance, the associations between gender and afterlife belief in adulthood were more heterogeneous across countries (I² = 97.3 to 99.9, estimated proportion of effects by threshold ranged = [0.33 to 0.39]) compared to the associations for the relationship with one’s mother during childhood (I² = 28.1, estimated proportion of effects by threshold ranged = [0.00 to 0.05]). Heterogeneity also varied across the categories of each predictor. For example, regarding the frequency of religious service attendance at age 12 (with “never” as the reference category), the association between attending services at least once per week and afterlife belief in adulthood showed greater heterogeneity across countries compared to the association for attending less than once per month (I² = 96.1, estimated proportion of effects by threshold ranged = [0.00 to 0.77] vs. 78.9, estimated proportion of effects by threshold ranged = [0.05 to 0.45]). These findings suggest that the effects of childhood predictors on afterlife belief in adulthood differ to some extent across sociocultural contexts. Additionally, all global p-values in Table 2 were below the Bonferroni-corrected significance threshold, indicating that each predictor was associated with afterlife belief in adulthood in at least one country. The country-specific results for childhood predictors of afterlife belief are presented in Tables S1b-S22b.

The variable for religious attendance was the strongest predictor that evidenced consistent associations with a higher likelihood of afterlife belief in adulthood across a wide range of countries (with 95% CIs excluding the null). Specifically, attending religious services at least once per week was associated with afterlife belief in 17 out of 22 countries (see Figure S16), attending 1 to 3 times a month in 12 countries (see Figure S17), and attending less than once per month in 7 countries (see Figure S18). Hong Kong, Israel, Japan, Spain, and Sweden were the most consistent in showing these associations for all three categories. There was no association in the opposite direction.

Several predictors showed mixed patterns across countries, with associations in opposite directions. While the pooled random effects meta-analysis indicated a general positive association between being female and a higher likelihood of afterlife belief, cross-country results varied (see Figure S26). Women reported higher levels of belief in 10 out of 22 countries (e.g., Sweden and Japan) but lower levels in 3 countries (Egypt, Tanzania, and Nigeria). Similarly, experiencing abuse was generally associated with a higher likelihood of afterlife belief (in 7 countries, including India, Japan, and Sweden), but in Kenya, it was linked to a lower likelihood (see Figure S9). Feeling like an outsider growing up was associated with higher belief in 4 countries (e.g., Japan and Spain) but lower in one country (United States, see Figure S10).

The patterns of association for birth year categories were also mixed. For example, being born between 1963 and 1973 (ages 50–59) was linked to higher afterlife belief in 7 countries and lower belief in 3 (see Figure S22). Being born between 1953 and 1963 (ages 60–69) was associated with higher belief in 5 countries and lower in 4 (see Figure S23). Being born between 1943 and 1953 (ages 70–79) was associated with higher belief in 4 countries and lower in 4 (see Figure S24). Notably, in Poland and Kenya, older age was linked to a higher belief in the afterlife, whereas in Japan and Sweden, older age was associated with lower belief.

There were some country-specific associations that did not appear in the pooled meta-analyses. For instance, there was notable evidence that self-rated health as “very good” during childhood (compared to “good”) was linked to higher belief in 3 countries and lower belief in 2 (see Figure S12). Conversely, some associations observed in the pooled meta-analysis were absent in specific countries. A notable example is the relationship with one’s mother during childhood, which only showed a significant association with afterlife belief in Japan, with no other country’s results excluding the null (see Figure S1), but the directions were sufficiently consistent across countries that evidence emerges from the pooled meta-analytic results.

Country-specific models also estimated the associations between religious affiliation at age 12 and racial/ethnic identity (where available) with afterlife belief in adulthood. These variables were not pool in meta-analysis due to varying measurement across countries (see Tables S1b-S22b). In 14 countries, religious affiliation—whether comparing those with a religion to those without, or between two religions—was associated with varying levels of afterlife belief. For example, in India, being Muslim at age 12 was linked to higher levels of adult afterlife belief compared to being Hindu. In contrast, in 7 countries, no significant association was found between religious affiliation at age 12 and adult afterlife belief. For instance, in Tanzania and South Africa, having a religion did not differ from having no religion at age 12, while in Kenya and the Philippines, Christianity and Islam at age 12 showed no significant difference regarding afterlife belief.

Additionally, in 8 countries, the 95% CIs for effect estimates indicated associations between racial/ethnic minority status and afterlife belief in adulthood. In 3 countries—Argentina, Poland, and Israel—minority status was linked to a higher likelihood of belief in afterlife. However, in 5 countries—Australia, India, Israel, Nigeria, and the Philippines—minority status was associated with a lower likelihood of afterlife belief.

Discussion

Leveraging data from a geographically and culturally diverse set of countries, this study examined 13 candidate childhood predictors of afterlife belief in adulthood. These predictors were simultaneously entered into the regression model, allowing for the control of their shared effects. Meta-analytic effect estimates across countries indicated evidence for associations with afterlife belief in adulthood, when pooled across all 22 countries, for 6 o f the 11 predictors. These predictors include a combination of religious service attendance, early life experiences (e.g., abuse, feeling like an outsider in the family, and positive relationship with mother), and individual characteristics (e.g., gender and age). The magnitude of effect estimates varied across predictors, and there was considerable cross-national variation in the patterns of association observed.

Religious service attendance

Religious service attendance at age 12 emerged as the most robust predictor of afterlife belief in adulthood. These findings are also notable as they control for individual characteristics such as age and gender, which often covary with religiosity. This result aligns with previous cross-sectional studies18 and provides preliminary longitudinal evidence (from retrospective reporting) across a broad range of countries.

The effect of weekly religious service attendance at age 12 (compared to “never”) is notably stronger in countries with lower overall religious service attendance (e.g., Sweden, Israel, and Hong Kong) than in those with higher attendance rates (e.g., South Africa, Indonesia, and India). This suggests a potential interaction effect between societal religious influence and personal religious practice in shaping afterlife beliefs. The lower prevalence of religious service attendance in some countries may heighten the impact of individual participation, whereas in more religiously active societies, the personal practice might be less distinguishing due to the broader cultural context.

Another possible explanation is that attending religious services at age 12 could serve as a marker of being raised in a religious family, which may actively transmit religious values and beliefs to their children. Moreover, consistent exposure to afterlife concepts during religious services, when a person’s worldview is still developing in adolescence, may leave a lasting impact. This notion aligns with our findings that the effect of youth service attendance was much stronger in less religious countries, where adolescents may be less likely to encounter afterlife beliefs through the general culture.

Early life experiences

Experiencing abuse and feeling like an outsider in the family during childhood were found to be positively associated with afterlife belief, although the pooled effect sizes were small and not all countries showed the same direction of effect.

People often engage in a meaning-making process following trauma27, and belief in life after death could be one such ideological commitment that arises from this process. Negative early life experiences, such as abuse and feeling like an outsider, might drive individuals toward adopting afterlife beliefs as a way to cope with their experiences. Abuse, being a profound violation of one’s dignity, can lead victims to turn to religion for hope, particularly the hope of seeing justice done, since full justice may seem unattainable in this life. In this context, belief in an afterlife tied to religious convictions may be a product of seeking meaning and resolution in the face of injustice.

While the positive effect of parental relationships was weak for maternal relationships and nonsignificant for father relationships, these findings align with the general religious transmission hypothesis, which suggests that parenting styles contribute to the adoption of religious beliefs among children25. The mechanisms of religious transmission extend beyond mere exposure to parental beliefs. Credibility-enhancing displays28, where parents demonstrate costly and sincere commitments to their religious beliefs, play a crucial role in transmitting those beliefs29. This implies that children are more likely to adopt their parents’ religion when they witness committed religious behaviors, particularly in societies where religious practices are less prevalent, as such behaviors stand out more and may exert a stronger influence.

Gender and age

While the association between age and afterlife belief varied across countries, some comment on the pooled results also merit attention. At some risk of over-interpretation, it seemed that there may have been some evidence for a general inverted U-shaped trend emerged, suggesting that individuals tend to hold stronger afterlife beliefs as they age, with the trend perhaps beginning to reverse after the age of 60. Since religious service attendance was controlled for in the regression model, this pattern indicates that age influences afterlife beliefs independently of religious activity and may perhaps be shaped by birth cohort effects. This trend may be understood through the concept of gerotranscendence, which suggests that spiritual concerns often deepen with age and that afterlife beliefs can act as a psychological buffer against the fear of death30,31.

Additionally, this study demonstrates that women exhibit higher levels of afterlife belief, aligning with existing literature showing that women tend to be more religious than men21. The effect of being female is most pronounced in less religious countries (e.g., Sweden, Australia, and Japan), suggesting that gender differences in afterlife belief may be more marked in secular environments. This study contributes to the understanding of how gender differences manifest in religious beliefs, particularly in relation to afterlife convictions, across varying sociocultural contexts.

Limitations and conclusion

This study has several limitations. First, the childhood predictors were assessed retrospectively in the same wave as the outcome measure of afterlife belief. Although this allowed for the construction of a synthetic longitudinal design using Wave 1 of the Global Flourishing Study (GFS), the findings may be influenced by recall bias and common method bias. However, for recall bias to completely explain away the observed associations would require that the effect of adult meaning/purpose on biasing retrospective assessments of the childhood predictors would essentially have to be at least as strong as the observed associations themselves, and some of these were quite substantial32. To reduce these biases, future research could adopt prospective longitudinal designs that follow individuals from childhood to adulthood.

Second, the observed associations may be affected by unmeasured confounding factors. For instance, certain personality traits, such as openness to experience33, and cognitive challenges in imagining one’s own non-existence may promote beliefs in continued existence after death34. These factors were not measured in this study and could confound the current findings. It is also important to acknowledge that while age plays a role in afterlife beliefs, both children and adults can entertain ideas about psychological capacities continuing beyond death. Such beliefs are more likely to be fostered in cultures that encourage them, but they can also exist in cultures where such thinking is less prevalent35. The E-value sensitivity analysis partially addresses this, but results may indeed still be subject to unmeasured confounding.

Third, all psychosocial constructs, including afterlife belief, were measured using a single item. While single-item measures are often used in large-scale epidemiological studies for their simplicity and efficiency, they tend to provide narrower conceptual coverage than multi-item measures. Additionally, people may interpret the question “belief in life after death” in various ways, such as bodily survival, psychological continuation, being rewarded in heaven, or reincarnation into a different form. To improve the validity of future findings, more comprehensive measures that capture the full complexity of these constructs should be considered.

In conclusion, this multinational study investigated childhood predictors of afterlife belief in adulthood. The findings provide preliminary evidence that religious service attendance in adolescence, early life experiences, and individual characteristics such as age and gender are associated with afterlife beliefs in adulthood, both within and across countries.

Methods

The description of the methods below has been adapted from VanderWeele et al.36. Further methodological detail is available elsewhere37,38,39,40,41,42,43.

Data

The Global Flourishing Study (GFS) is a study of 202,898 participants from 22 geographically and culturally diverse countries, with nationally representative sampling within each country, concerning the distribution of determinants of well-being. Wave 1 of the data included the following countries and territories: Argentina, Australia, Brazil, Egypt, Germany, Hong Kong (Special Administrative Region of China, with mainland China also included from 2024 onwards), India, Indonesia, Israel, Japan, Kenya, Mexico, Nigeria, the Philippines, Poland, South Africa, Spain, Sweden, Tanzania, Turkey, United Kingdom, and the United States. The countries were selected to (a) maximize coverage of the world’s population, (b) ensure geographic, cultural, and religious diversity, and (c) prioritize feasibility and existing data collection infrastructure. Data collection was carried out by Gallup Inc. Data for Wave 1 were collected principally during 2023, with some countries beginning data collection in 2022 and exact dates varying by country42. Four additional waves of panel data on the participants will be collected annually from 2024 to 2027. The precise sampling design to ensure nationally representative samples varied by country and further details are available42. Survey items included aspects of well-being such as happiness, health, meaning, character, relationships, and financial stability44, along with other demographic, social, economic, political, religious, personality, childhood, community, health, and well-being variables. The data are publicly available through the Center for Open Science (https://www.cos.io/gfs). During the translation process, Gallup adhered to TRAPD model (translation, review, adjudication, pretesting, and documentation) for cross-cultural survey research (ccsg.isr.umich.edu/chapters/translation/overview).

Measures

Childhood demographic variables

Relationship with mother during childhood was assessed with the question: “Please think about your relationship with your mother when you were growing up. In general, would you say that relationship was very good, somewhat good, somewhat bad, or very bad?” Responses were dichotomized to very/somewhat good versus very/somewhat bad. An analogous variable was used for relationship with father. “Does not apply” was treated as a dichotomous control variable for respondents who did not have a mother or father due to death or absence. Parental marital status during childhood was assessed with responses of married, divorced, never married, and one or both had died. Financial status was measured with: “Which one of these phrases comes closest to your own feelings about your family’s household income when you were growing up, such as when YOU were around 12 years old?” Responses were lived comfortably, got by, found it difficult, and found it very difficult. Abuse was assessed with yes/no responses to “Were you ever physically or sexually abused when you were growing up?” Participants were separately asked: “When you were growing up, did you feel like an outsider in your family?” Childhood health was assessed by: “In general, how was your health when you were growing up? Was it excellent, very good, good, fair, or poor?” Immigration status was assessed with: “Were you born in this country, or not?” Religious attendance during childhood was assessed with: “How often did YOU attend religious services or worship at a temple, mosque, shrine, church, or other religious building when YOU were around 12 years old?” with responses of at least once/week, one-to-three times/month, less than once/month, or never. Gender was assessed as male, female, or other. Continuous age (year of birth) was classified as 18–24, 25–29, 30–39, 40–49, 50–59, 60–69, 70–79, and 80 or older. Childhood religious tradition/affiliation was had response categories of Christianity, Islam, Hinduism, Buddhism, Judaism, Sikhism, Baha’i, Jainism, Shinto, Taoism, Confucianism, Primal/Animist/Folk religion, Spiritism, African-Derived, some other religion, or no religion/atheist/agnostic; precise response categories varied by country45. When more than 5% of a within-country sample endorsed the no religion/atheist/agnostic category, this was used as the reference category in the country-specific analyses; otherwise, the most prominent religious group was used. Additionally, all religious affiliation categories endorsed by less than 3% of a within-country sample were collapsed into a single religious affiliation category. Racial/ethnic identity was assessed in most (18/22) countries, and response categories varied across countries. Country-specific analyses used a binary racial/ethnic identity variable based on whether an individual was in the most prominent group in the sample versus a minority group. Racial/ethnic identity were assessed in some, but not all, countries, and response categories were unique to each country. Racial/ethnic identity was assessed in some, but not all, countries, with response categories varying by country. For additional details on the assessments see the GFS codebook (COS, https://osf.io/cg76b).

Outcome variable(s)

The afterlife belief measure comes from the following question: “I believe in life after death.” Respondents could answer Yes, No, or Unsure. This variable was dichotomized as Yes (1) vs. No/Unsure (0).

Analysis

Descriptive statistics for the observed sample, weighted to be nationally representative within country, were estimated for each childhood demographic category. A weighted modified Poisson regression model with complex survey adjusted standard errors was fit within each country of afterlife belief on all of the aforementioned childhood predictor variables simultaneously. In the primary analyses, random effects meta-analyses of the regression coefficients46,47 along with confidence intervals, estimate proportions of effects across countries with effect sizes larger than 1.1 and smaller than 0.9, and I2 for evidence concerning variation within a given demographic category across countries48. Forest plots of estimates are available in the online supplement. Religious affiliation/tradition and race/ethnicity were used within country as control variables, when available, but these coefficients themselves were not included in the meta-analyses since categories/responses varied by country. All meta-analyses were conducted in R using the metafor package49. Within each country, a global test of association of each childhood predictor variable group with outcome was conducted, and a pooled p-value50 across countries reported concerning evidence for association within any country. Bonferroni corrected p-value thresholds are provided based on the number of childhood demographic variables51. For each childhood predictor, we calculated E-values to evaluate the sensitivity of results to unmeasured confounding. An E-value is the minimum strength of the association an unmeasured confounder must have with both the outcome and the predictor, above and beyond all measured covariates, for an unmeasured confounder to explain away an association52. As a supplementary analysis, population weighted meta-analyses of the regression coefficients were estimated. All analyses were pre-registered with COS prior to data access, with only slight subsequent modification in the regression analyses due to multicollinearity (https://osf.io/qczrb); all code to reproduce analyses are openly available in an online repository40,42.

Missing data

Missing data on all variables was imputed using multivariate imputation by chained equations, and five imputed datasets were used53,54. To account for variation in the assessment of certain variables across countries (e.g., religious affiliation/tradition and race/ethnicity), the imputation process was conducted separately in each country. This within-country imputation approach ensured that the imputation models accurately reflected country-specific contexts and assessment methods. Sampling weights were included in the imputation model to account for specific-variable missingness that may have been related to probability of inclusion in the study.

Accounting for complex sampling design

The GFS used different sampling schemes across countries based on availability of existing panels and recruitment needs42. All analyses accounted for the complex survey design components by including weights, primary sampling units, and strata. Additional methodological detail, including accounting for the complex sampling design, is provided elsewhere39.