Life course insights into social relationship quality: a cross-national analysis of 22 countries

Close personal relationships are a significant source of meaning, happiness, and overall well-being in the lives of most people. High-quality friendships, family ties, and marital or romantic relationships are all important forms of social connection that have been shown to contribute to long-term mental health¹, physical health², longevity³, and success at work⁴. Social relationships involve structural features like marital status and the size of one’s social network, as well as functional aspects like perceived and received social support from others⁵. Here, we consider social relationship quality in terms of a subjective, global evaluation of one’s relationships⁶, often involving an individual’s appraisal of their level of satisfaction, happiness, and contentedness with their close social ties. These ties may include one’s spouse or partner, children, parents and other family members, friends, and co-workers. We favor a broader conception due to its simplicity, allowing for easier exploration of the factors that lead to better social relationships overall⁷. In assessments of overall social relationship quality, people are likely to consider daily interaction patterns across their relationships, including the frequency and intensity of conflict and criticism, satisfaction with communication, and role obligations. They also evaluate the level of perceived and received instrumental and emotional support, the extent to which relationships are energizing⁸, and the presence and balance of these features over the longevity of their relationships.

The ability to form and maintain high-quality relationships may have its foundation in the early life course. A life course perspective draws attention to the upstream factors—like early childhood experiences—that can shape adult outcomes⁹. Childhood is a critical life stage when developmental processes and socioeconomic factors have the potential to produce differences in experience that lead to long-term advantage or disadvantage¹⁰. In particular, an individual’s experiences in their family of origin may be important influences on their later interpersonal relationships^11,12. Prior research suggests that negative childhood conditions (e.g., abuse, economic hardship, harsh parenting) are associated with lower adult relationship quality^{13,14,15,16,17,18,19}, while positive childhood experiences (e.g., supportive or high-quality parent-child relationships) are related to higher adult relationship quality¹⁷and relational health²⁰. Prior studies have some methodological limitations, including drawing primarily on small convenience samples mostly within the United States, and they investigate only a limited set of childhood experiences and family and social circumstances and their links to adult social relationship quality. The extent to which specific aspects of a child’s upbringing that predict social relationship quality in adulthood may differ across cultural contexts is unknown.

To address these gaps, our study examines a broad range of social and economic factors, adverse events and experiences, health status, and sociodemographic characteristics from childhood and their associations with social relationship quality in adulthood, and whether differences exist across diverse country contexts. Our approach focuses on breadth over depth is intended to be exploratory and descriptive, as we are reporting findings for the first time in many countries. The results of this study are intended to serve as an empirical foundation for further exploration into the most salient childhood antecedents of adult social relationship quality, as well as the mechanisms and potential heterogeneity in our observed associations, and greater conceptual understanding of these relationships and variation across nations. To summarize, our objectives are to (1) determine which childhood factors are salient around the world for predicting adult social relationship quality, and which are unique to specific countries, (2) provide the first published results on this topic in several countries, and (3) provide an empirical foundation for future in-depth examinations of each significant childhood factor.

Childhood conditions have an enduring impact on adult outcomes across life domains. According to Elder’s life course paradigm⁹, childhood is a pivotal life stage during which family context and other social, economic, and cultural factors shape life trajectories and subsequent outcomes. Early-life advantages such as strong social connections within one’s family and community and the presence of other social, economic, and health resources are expected to provide benefits that accumulate and compound over time, leading to more advantageous outcomes¹⁰. Conversely, childhood disadvantages like poor-quality or absent social ties, experiences of abuse, health challenges, and economic hardship can increase the risk of subsequent disadvantage leading to worse long-term outcomes.

Bowlby’s attachment theory²¹ provides insight into the early-life socioemotional processes that may contribute to the development of social relationship quality in adulthood. Attachment theory suggests that humans have an innate need to create close social bonds with others, and when an individual’s attachment needs are met, they are more capable of being emotionally healthy. Specifically, a child’s experiences with caregivers in their family of origin—including caregivers’ responsiveness, availability, and warmth—are carried forward beyond childhood to form an internal working model that influences the child’s ability to create and maintain strong attachments later in life. Accordingly, individuals who had family relationships that they would characterize as positive, loving, and without abuse while growing up are more likely to have stronger attachments—and relatedly, higher evaluations of the quality of their social relationships—compared to those who had family relationships that could be considered negative, or in which they felt unloved or were abused.

While considerable life-course research has focused on the early origins of adult health^22,23,24, there is a growing body of empirical work that has examined how characteristics of the family environment such as childhood abuse^12,13,14,17, supportive or harsh parenting practices^16,20, overall parent-child relationship quality¹⁵, and composite adverse childhood experiences (including economic hardship)¹⁹are linked with social well-being outcomes like romantic relationship quality in adulthood, with findings generally showing that positive childhood relationships are associated with higher-quality adult relationships. Childhood experiences of parental divorce have been shown to contribute to one’s own likelihood of divorce²⁵, and parental divorce has also been linked with poorer adult intimate relationship quality and satisfaction with social support among women²⁶. Additionally, a study using 43 dyadic longitudinal datasets and machine learning found that family history (e.g., childhood abuse, family structure, parental education) and parents’ relationship quality (e.g., marital status, parental conflict) accounted for 25% and 13%, respectively, of the variance in predicting adult relationship satisfaction and commitment, compared to other individual factors like mental health, personality, and income²⁷.

Research has also investigated links between childhood health and adult relationship quality, demonstrating a negative impact of chronic pain during adolescence on relationship quality in adulthood²⁸, and no differences in adult relationship quality among children who grew up with a chronic illness compared to those who did not²⁹. The bulk of studies have focused exclusively on romantic relationship outcomes, which leaves unclear how social conditions from childhood influence the quality of social relationships considered broadly (e.g., friendships, parent-child and other family relationships, work relationships, etc.), not just the quality of relationships of those romantically involved. However, one study examined adolescent religious service attendance as predictor of psychosocial adjustment in high school and found that higher frequency of attendance predicted lower subsequent friendship quality, and it showed mixed results with parent-child relationship quality³⁰.

Cross-cultural variation in the early origins of adult relationship quality

Variation may also exist in the associations between childhood factors and adult social relationship quality across countries and cultures. To our knowledge, no empirical work has explored the childhood antecedents of adult social relationship quality outside of a Western context, let al.one across nations around the world. The diverse sociocultural, economic, and health contexts characterizing different countries may differentially shape the relationship between childhood experiences and adult social relationship quality. Culture encompasses the values, beliefs, and practices that give meaning to an individual’s experiences^31,32,33, which are often shared within a country and differ across countries, although cultural variation within-countries can and does also occur^34,35. Cultural orientations may be derived from early socialization that occurs in one’s country of birth, and these orientations can have a particularly pervasive influence on individual behavior, although it is increasingly recognized that cultural orientations can be derived from other contextual influences such as living and working outside one’s country of birth, or in a specific region within a country³⁶. For example, culture shapes interaction patterns and provides frames of reference and expectations regarding interpersonal behaviors³⁷and communication preferences³⁸. Those from different cultural contexts often hold different assumptions and preferences regarding relationships, and this applies to how they choose to express themselves and communicate, as well as how they interpret and respond to the expressions and communications of other people⁸.

Thus, the characteristics of an individual’s childhood and the culture and context in which they are embedded are likely to influence later-life outcomes, including the way they form and maintain social bonds, the role of such bonds in their lives, and how they appraise the value and quality of their relationships. Our study is the first to include a diverse, global sample of individuals across 22 countries around the world to examine whether such cross-cultural variation exists.

This study uses nationally representative data from 22 diverse nations around the world to examine links between a broad range of early-life factors and adult social relationship quality. We explore whether array of social and economic factors, adverse events and experiences, health status, and sociodemographic characteristics from childhood are associated with social relationship quality in adulthood across 22 countries included in the sample, and the extent to which variation in these associations exists across these countries. Based on findings from previous research, we expected that several specific childhood factors—specifically, good childhood relationships with mother and father and greater subjective financial security—would be positively associated adult social relationship quality, while other aspects—namely, childhood abuse, parental divorce, and poor health—would be negatively associated with adult social relationship quality. We also expected that the strength of these associations would differ by country, reflecting diverse societal influences.

The description of the methods below has been adapted from VanderWeele et al. (2024)³⁹. Further methodological detail is available elsewhere^{40,41,42,43,44,45}.

This study used data from the Global Flourishing Study (GFS), which examines the distribution of determinants of well-being across a sample of 202,898 participants from 22 geographically and culturally diverse countries, with nationally representative sampling within each country. Wave 1 of the data included the following countries and territories: Argentina, Australia, Brazil, Egypt, Germany, Hong Kong (Special Administrative Region of China), India, Indonesia, Israel, Japan, Kenya, Mexico, Nigeria, the Philippines, Poland, South Africa, Spain, Sweden, Tanzania, Turkey, United Kingdom, and the United States. These 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 for Wave 1 was conducted by Gallup Inc. primarily in 2023, although some countries began data collection in 2022; the exact timing of data collection varied by country⁴⁴. The precise sampling design to ensure nationally representative samples also varied by country, and further details are available elsewhere⁴⁴. The data are publicly available through the Center for Open Science (https://www.cos.io/gfs). The translation process adhered to the TRAPD model (translation, review, adjudication, pretesting, and documentation) for cross-cultural survey research (https://www.ccsg.isr.umich.edu/chapters/translation/overview). Further details about the GFS study methodology and survey development were reported earlier^43,44.

The present study used data from all participants in Wave 1 (N= 202,898). Poststratification and nonresponse adjustments were performed to ensure the sample was representative of the adult population in each country^42,44. This study was pre-registered at the Center for Open Science (https://www.osf.io/wke9f). Ethical approval was granted by the institutional review boards at Baylor University and Gallup, and all participants provided informed consent. All research was performed in accordance with relevant guidelines and regulations.

Social relationship quality was assessed by two items: “I am content with my friendships and relationships” and “my relationships are as satisfying as I would want them to be.” Response options were a Likert-type scale ranging from 0 (strongly disagree) to 10 (strongly agree). We analyzed the mean of these two items as a continuous variable, with higher scores indicating higher social relationship quality.

Childhood predictors

Relationships with parents. Participants were asked to assess the quality of their relationships with parents using separate questions that asked about their mother and father: “Please think about your relationship with your [mother/father] 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” to reduce collinearity in regression models. “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.

Parent marital status. Marital status of parents during childhood was assessed by the question: “Were your parents married to each other when you were around 12 years old?” Response options included “married,” “divorced,” “never married,” and “one or both of them had died.”

Subjective financial status growing up. Participants reported their subjective assessment of the financial status of their family while growing up with the question: “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?” Response options included “lived comfortably,” “got by,” “found it difficult,” and “found it very difficult.”

Childhood abuse. Abuse was assessed using the question: “Were you ever physically or sexually abused when you were growing up?” Response options were “yes” and “no.”

Outsider in family growing up. Participants were asked about their feelings of family belonging during childhood using the question: “When you were growing up, did you feel like an outsider in your family?” Response options were “yes” and “no.”

Childhood health. Self-rated health was assessed using responses to the question: “In general, how was your health when you were growing up? Was it excellent, very good, good, fair, or poor?”

Immigration status. Participants reported their immigration status with the following question: “Were you born in this country, or not?” Response options were “yes” and “no.”

Childhood religious service attendance. The frequency of religious attendance during childhood was assessed using the question: “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?” Response options included “at least once per week,” “one-to-three times per month,” “less than once per month,” or “never.”

Birth year/current age. Participants were asked to report their current age (in years), and birth year was calculated based on the year of data collection minus their current age. We created the following age groups: 18–24, 25–29, 30–39, 40–49, 50–59, 60–69, 70–79, and 80 or older (see tables for birth year ranges).

Gender. Participants reported their gender, with “male,” “female,” or “other” as response options.

Religious affiliation. Childhood religious tradition/affiliation was assessed using the question: “What was your religion when you were 12 years old?” Response options included 15 major religious traditions (e.g., Christianity, Islam, Hinduism, Buddhism, Judaism) “some other religion,” and “no religion/atheist/agnostic.” Precise response categories varied by country⁴⁶. To reduce data sparsity, in regression models the response categories of religious affiliation with a prevalence < 3% were collapsed. The “no religion/atheist/agnostic” group was used as the reference group when at least 3% of the observed sample within the country endorsed this category; otherwise, the most prominent religious group was used as the reference category.

Race and ethnicity. Race and ethnicity were assessed in most but not all countries (Germany, Japan, Spain, and Sweden had restrictions on collecting such data), and response categories were unique to each country. Race and ethnicity was dichotomized as racial/ethnic plurality (the category with the largest proportion) and minority (collapsing other categories) in each country.

The descriptive analyses examine the distributions of childhood antecedents in each individual country and in the total sample, weighted to be nationally representative within each country. In the country-specific analyses, a weighted linear regression model with complex survey adjusted standard errors was used to regress social relationship quality on all childhood predictors simultaneously. A Wald-type test was conducted to obtain a global (joint) test of the effect of all categories within a childhood predictor resulting in a single global p-value of the effect of each childhood predictor.

In the primary analyses, random effects meta-analysis was performed to pool regression coefficients from the country-specific analyses^47,48, along with confidence intervals, lower and upper limits of 95% prediction intervals, heterogeneity (τ), and I² for estimating variation within a given childhood antecedent category across countries⁴⁹. Forest plots of estimates are shown in the online supplement (Figures S1 to S27). Religious affiliation and race and ethnicity (when available) were used as control variables within countries, but these coefficients themselves were not included in the meta-analyses because their response options varied by country. A pooled p-value⁴⁸ was reported across countries to evaluate if a given association holds within at least one country. Bonferroni corrected p-value thresholds are provided based on the number of childhood predictor variables^50,51. For each childhood predictor, in both the meta-analyses and country-specific analyses, we calculated E-values to evaluate the sensitivity of the results to potential 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 the observed association⁵⁰. As a supplementary analysis, population weighted meta-analysis was performed to pool the regression coefficients from the country-specific analyses. We also replicated the overall meta-analytic results of Table 1 for each of the two indicators (i.e., relationship contentment and relationship satisfaction) considered separately.

All meta-analyses were conducted in R (R Core Team, 2024) using the metafor package⁵², and country-specific analyses were performed using StataMP 17. All statistical tests were 2-sided. The 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/wke9f); all code to reproduce analyses are openly available in an online repository⁴³.

Because multiple imputation is often a more flexible approach to handling missing data than other methods, missing data on all variables was imputed using multivariate imputation by chained equations, and five imputed datasets were used^53,54,55,56. To account for variation in the assessment of certain variables across countries (e.g., religious affiliation, race and 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 models 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 needs⁴⁴. All analyses accounted for the complex survey design components by including weights, primary sampling units, and strata. Additional methodological details are provided elsewhere⁴³.

Descriptive analyses

Within the full sample (Table 2), most participants reported having “very good” or “somewhat good” relationships with their mother (89%) and father (80%), that their parents were married (75%), and their family “lived comfortably” (35%) or “got by” (41%) financially during their childhood. The majority of participants reported experiencing no abuse (82%) and assessed their health as “excellent” or “very good” (64%) when growing up, and they responded negatively to the question about whether they felt like an outsider in their family (84%). A large proportion of participants attended religious services at least once per week during childhood (41%), and most were born in the country in which they responded to the survey (94%). The distribution across birth year/age cohorts was even apart from fewer participants who were in the oldest categories (i.e., born before 1953, and aged 70 or older), and there was a slightly higher proportion of women (51%). The sample size for each country ranged from 1,473 participants in Turkey to 38,312 participants in the United States. Participant characteristics for each country are reported in Supplementary Tables S1a to S22a.

Random effects meta-analysis

The results from the random effect meta-analyses pooling regression coefficients from the country-specific analyses (Table 1) suggested that all but one of the childhood factors we examined were associated with social relationship quality in adulthood in at least one country. All global p-values except for one (i.e., parent marital status; p = .021) were below the Bonferroni-corrected significance threshold, which provides evidence that in at least one country, the childhood factors are likely predictive of social relationship quality scores. Specifically, on average across countries in the sample, having a very good/somewhat good (versus very bad/somewhat bad) relationship with one’s mother (β = 0.26, 95% confidence interval [CI] 0.15, 0.37) and father (β = 0.25, 95% CI: 0.18, 0.32) during childhood was associated with greater social relationship quality in adulthood. In contrast, experiencing abuse (β = −0.41, 95% CI: −0.49, −0.33) and feeling like an outsider in one’s family growing up (β = −0.37, 95% CI: −0.46, −0.27) were associated with lower adult social relationship quality.

The results also indicated that self-rated health during childhood was positively associated with social relationship quality in adulthood: compared with participants who had good health during childhood, those with excellent health (β = 0.45, 95% CI: = 0.29, 0.61) or very good health (β = 0.22, 95% CI: 0.13, 0.31) had higher social relationship quality, while those with fair health (β = −0.16, 95% CI: −0.28, −0.04) had lower social relationship quality. Participants who reported higher subjective financial status growing up also had higher social relationship quality as adults: reports that one’s family lived comfortably (versus got by) were associated with higher adult social relationship quality (β = 0.16, 95% CI: 0.09, 0.24), while reporting that one’s family found it difficult financially (β = −0.08, 95% CI: −0.14, −0.03) or found it very difficult financially (β = −0.20, 95% CI: −0.32, −0.09) were negatively associated with adult social relationship quality. All levels of religious service attendance during childhood (versus never attending services) were positively associated with social relationship quality; in particular, religious service attendance at least weekly (versus never) was associated with higher social relationship quality (β = 0.30, 95% CI: 0.19, 0.40).

Older cohorts (i.e., participants born before 1973; age 50 or older) had higher social relationship quality compared to the youngest cohort (i.e., participants born between 1998 and 2005; ages 18–24), with those in the oldest age group (born in 1943 or earlier; age 80 or older) having the highest social relationship quality (β = 0.52, 95% CI: 0.16, 0.89) relative to the youngest age group. Women also reported higher social relationship quality (β = 0.10, 95% CI: 0.04, 0.17), compared to men. The confidence interval of the association for immigration status included the null but the global p-value was < 0.001, which suggests that immigration status was associated with social relationship quality in at least one country but not necessarily on average across countries. This finding was also evidenced for some levels of other childhood factors (e.g., poor versus good childhood health; parents were never married or one or both parents had died, versus parents were married). Heterogeneity in the associations, assessed by τ (i.e., how much the childhood antecedent associations varied across countries), was greater for childhood self-rated health (e.g., excellent versus good health, τ = 0.36), birth year/current age (e.g., oldest age group versus youngest, τ = 0.79), and gender (e.g., other gender versus men, τ = 1.16) compared to other childhood factors.

E-values indicated that the observed associations between the childhood factors we assessed and social relationship quality in the meta-analysis were moderately robust to potential unmeasured confounding (Table 3). For example, to explain away the association between excellent (versus good) childhood health and social relationship quality (β = 0.45), an unmeasured confounder associated with both increased likelihood of excellent childhood health and of higher social relationship quality by risk ratios of 1.69 each, above and beyond all measured covariates (including other childhood predictors), could suffice, but weaker joint confounder associations could not. Additionally, to shift the confidence interval to include the null value, an unmeasured confounder associated with both excellent childhood health and increased social relationship quality by risk ratios of 1.50-fold each could suffice, but weaker joint confounder associations could. In other words, in order to make the observed effect disappear or become insignificant, it would need to exert a moderately large effect on both excellent childhood health and social relationship quality, on top of the other childhood factors we assessed. E-values for several other childhood factors (e.g., abuse, feeling like an outsider growing up) were of similar magnitude to that for excellent childhood health. Overall, E-values results indicated that other factors not assessed in our study would need to exert a moderately large effect to change the results significantly.

Country-specific analyses

In the country-specific analyses (Supplementary Tables S1b to S22c), associations between the childhood factors and social relationship quality varied, though many of the childhood factors were each associated with social relationship quality in at least half of the countries in the sample.

The most consistent associations were observed for experiencing childhood abuse and birth year/current age; these two factors were associated with adult social relationship quality in almost all countries. Subjective financial status growing up, childhood health, gender, and feeling like an outsider in one’s family growing up were associated with social relationship quality in more than half of countries. In terms of the predictors with notable variation across countries, experiencing childhood abuse was associated with the largest decrease in social relationship quality in Poland (β = −0.77) and the smallest decrease in Sweden (β = −0.22), and there was limited evidence for associations in Germany, Hong Kong, and India (childhood abuse was not assessed in Israel). Patterns for birth year/current age indicated that participants in earlier birth year cohorts/older age groups (versus the latest cohort/youngest group) had higher social relationship quality in most countries (e.g., Brazil, Germany, Sweden, U.S., U.K.), although occasionally this pattern included the null for the earliest birth year cohort/oldest age group (e.g., Argentina, Hong Kong, Mexico), which could be related to the comparatively smaller size of the oldest age group. Interestingly, the most recent birth year cohort/youngest age group had higher social relationship quality compared to other birth year cohorts/age groups in Poland, India, Israel, and Tanzania.

Some of the other childhood factors only demonstrated strong evidence for associations with social relationship quality at the most extreme response category. For subjective financial status growing up, only the highest subjective assessment (i.e., lived comfortably versus got by) showed strong evidence for association with social relationship quality in many countries (e.g., Argentina, Hong Kong, Kenya, Mexico, Philippines, Sweden, U.S.), while in other countries, only the lowest subjective assessment (i.e., found it very difficult versus got by) did (e.g., Tanzania, Turkey, U.K.). Similarly, associations for self-rated health with social relationship quality were generally evidenced in the expected direction across most countries (primarily for excellent versus good health), with exceptions in Germany and Israel where poor (versus good) health was associated with higher social relationship quality.

Women compared to men had higher social relationship quality in many counties, except for Kenya in which the pattern was reversed. Identifying with another gender (versus men) was associated with lower social relationship quality in several countries (i.e., Hong Kong, Indonesia, Philippines, Poland, South Africa) and higher social relationship quality in Japan and Nigeria. Among countries for which an association was evidenced, feeling like an outsider in one’s family growing up was associated with the largest decrease in social relationship quality in Australia (β = −0.75) and the smallest decrease in Egypt (β = −0.32).

Additional analyses

The population weighted meta-analysis results (Table S23), which gives considerably more weight to India due to its large population, generally resembled those from the main analysis, although a few associations that included the null in the main analysis did not in the population weighted results (e.g., birth year/current age: 1983–1993, β = 0.16, 95% CI: 0.02, 0.30; other gender: β = −0.65, 95% CI: −1.20, −0.09). The results for the additional analyses examining the individual items (i.e., relationship contentedness [Table S24] and relationship satisfaction [Table 25]) that comprise the combined social relationship quality measure, and the robustness of these results using E-values (Tables S26 and S27), were highly similar to those from the main analysis. A minor difference was that select birth year cohorts/age groups evidenced associations with the individual items while they did not with the composite measure (i.e., 1963–1973; age 50–59 for relationship contentedness, and 1973–1983; age 40–49 for relationship satisfaction).

While having close social relationships is a desired end in itself⁵⁷, the quality of one’s social relationships has also been shown to be an important factor that influences well-being across life domains^1,2,3,4. However, less is known about the factors present early in life that may shape later social functioning, and specifically, social relationship quality. This study contributes to the literature on the childhood antecedents of adult social relationship quality by examining an array of social and economic factors, adverse events and experiences, health status, and sociodemographic characteristics from childhood as potential predictors of adult social relationship quality across 22 culturally diverse countries, with nationally representative samples from within each country.

The results from the random effects meta-analysis support the conclusions of previous work that found that having positive relationships with parents during childhood was positively associated with relationship quality in adulthood^15,16, and experiencing childhood abuse was negatively associated with relationship quality^13,14,17; although to our knowledge, this study is the first to examine links between these childhood factors and a global assessment of the quality of one’s social relationships across relationship types (i.e., not exclusively romantic relationships). The findings from this study add new evidence to the literature by revealing positive associations with several social, economic, and health factors in childhood and social relationship quality in adulthood, including religious service attendance, subjective financial status, and self-rated health. The results also illuminated a strong link between feeling like an outsider in one’s family growing up and lower adult social relationship quality. In terms of demographic characteristics, women and participants born in earlier birth year cohorts also reported higher social relationship quality as adults.

The findings linking childhood factors like relationships with parents, experiencing abuse, and feeling like an outsider in one’s family with adult social relationship quality align with the tenets of attachment theory. Attachment theory posits that having secure attachments with caregivers during childhood shapes an individual’s ability to form and maintain strong, healthy relationships with others. Our study finds evidence that good relationships with mothers and fathers during childhood indeed are associated with better social relationship quality in adulthood, while bonds damaged by abuse or feelings of isolation in one’s family predict lower reports of the quality of one’s social relationships as an adult. The relationships children form with people in their religious communities also may be a mechanism through which religious service attendance in childhood leads to better social relationship quality in adulthood.

Older cohorts (relative to the youngest cohort) and women (compared to men) had higher reports of social relationship quality in adulthood. These findings suggest that older people (i.e., those from earlier birth cohorts) may have a stronger emphasis toward and value for their interpersonal relationships, and subsequently, may invest more time and energy in maintaining or improving their quality. This resonates with socioemotional selectivity theory⁵⁸, suggesting that people tend to prioritize meaningful and emotionally rewarding social interactions as they age. It also could be the case that members of older cohorts may be more likely than those from younger cohorts to have friendships or relationships within their communities (e.g., with neighbors) that involve meeting in person (i.e., physical co-presence), rather than maintaining connection by other modes (e.g., text, telephone, social media), and the “embodied” nature of these social ties contributes more strongly to their sense of the quality of their relationships. However, it is worth noting that in our analysis, we are unable to distinguish between age and cohort effects, and therefore, we cannot evaluate whether the association with social relationship quality is shaped by birth cohort or age differences. For gender, women may be socialized more than men to inherently orient themselves toward interpersonal relationships and mastery of these relationships⁵⁹, and therefore, their deeper investment and success in nurturing relationships could lead to better quality assessments among women compared to men.

Other early-life advantages in terms of financial security and health status appeared to impact subsequent social relationship quality. Higher evaluations of financial security and better self-assessed health were both associated with higher social relationship quality in the meta-analytic results and in most of the countries assessed individually. Better health and higher financial status during childhood may serve as resources that enable young people to create and sustain friendships and relationships with other social figures by instilling positive psychological assets (e.g., sense of control, self-esteem) that help them form bonds with others that are satisfying and mutually beneficial⁶⁰. Those who are in better health and have more economic resources may also be more capable of nurturing close bonds with others through time spent together and enjoying shared activities; conversely, being in worse health and having fewer economic resources may serve as barriers to physical co-presence⁶⁰.

The country-specific results provide insight into the near-universal negative associations of certain experiences like childhood abuse with adult social relationship quality, but the less consistent relationships between other childhood factors and the outcome are interesting and warrant further exploration. For example, considering relationships with parents, associations were evidenced for both mother and father relationship (net of the other) with adult social relationship quality in some countries (i.e., Brazil, Indonesia, Japan, Spain, U.S.), with more evidence for only for mother relationship (i.e., Argentina, Mexico, U.K.) or father relationship (i.e., Australia, Sweden), or neither (e.g., Egypt, Germany, South Africa). This could be due to cultural orientations such as the extent to which communal (typical female norm trait described as helpful, kind, sympathetic, and gentle) as opposed agentic (typical male norm trait described as more assertive, controlling, and confident) orientations dominate in a particular country, versus those countries where both orientations are salient^61,62. Another potential explanation is the concept of gender role orientation which has been characterized as traditional (strong demarcation of the role of men versus women) versus egalitarian (less demarcation such that men and women share roles); these orientations tend to be salient in certain cultures, and not in others⁶³.

For the countries for which an association between aspects of parent marital status and social relationship quality were shown, growing up with one or more parents having died (versus having parents who were married) was associated with higher social relationship quality in some countries (i.e., Australia, Japan) and lower social relationship quality in others (i.e., Germany, Nigeria), and many of these associations were moderately robust to potential confounding. Similarly, having divorced (versus married) parents was associated with lower scores on the outcome in some countries (i.e., Israel, Kenya, Philippines, Poland) but higher scores in a few (i.e., Japan, Sweden, Turkey), though these associations are conditional on relationship quality with mother and father. Future research should examine these findings in additional detail.

Limitations and future research directions

The present study has several limitations to be considered. First, our approach examining on an array of potential childhood predictors of adult social relationship quality focuses on breadth over depth is intended to be exploratory and descriptive. We encourage future research to use the findings from this study as a foundation for deeper examination of each of the specific childhood factors and their influence on adult social relationship quality (and different aspects of social connectedness), as well as further investigation into variability in these associations and the pathways connecting childhood factors with adult social relationship quality. We are also eager to see conceptual work that explores how national context shapes the links between specific childhood factors and subsequent relationship quality. Second, the childhood factors we examined were assessed retrospectively during adulthood and therefore may be subject to recall bias, particularly among older adults; however, research has shown that recall of some childhood events and experiences like abuse is stable and accurate over time^64,65. Moreover, for recall bias to completely explain away the observed associations would require that the effect of adult social relationship quality on biasing retrospective assessments of the childhood predictors would essentially have to be at least as strong as the observed associations themselves⁶⁶, and many of these were quite substantial. Third, although we took steps to reduce concerns of collinearity between the childhood factors in the models as they were included simultaneously, such as collapsing response categories or omitting variables from models, collinearity remains a potential issue and may especially affect, for example, parent marital status and maternal and paternal relationship quality.

Fourth, we focus on a subjective, global evaluation of social relationship quality using a combined measure of two items assessing different aspects of relationship quality—contentedness and satisfaction. While this approach has the benefit of capturing a wholistic view of the quality of one’s relationships across relationship types, it does not allow us to examine whether the childhood factors we examined in this study show different associations by relationship type (e.g., romantic relationships, siblings or parent-child relationships, friendships, colleagues, etc.), which is worthy of investigation in subsequent studies. Fifth, although the two items we use to assess social relationship quality—relationship contentedness and satisfaction—have been previously validated in multiple countries⁶⁷, GFS cognitive interviews found that respondents sometimes struggled with understanding what types of relationships were referred to in the statement “My relationships are as satisfying as I would want them to be”⁴⁰. Due to the challenges of survey interpretation and other methodological issues related to performing cross-cultural survey research (e.g., survey item translation, response style differences across countries and cultures, seasonal effects, and different modes of assessment), we caution against overinterpreting cross-cultural differences observed in our results. This study also has many strengths including its use of large representative samples in 22 culturally and geographically diverse countries, measurement of two aspects of social relationship quality, meta-analyses and country-specific analyses investigating childhood predictors of adult social relationship quality, sensitivity analyses to examine the robustness of the results to unmeasured confounding, and the longitudinal design of the Global Flourishing Study which will facilitate continued exploration into this topic in the future.

Given the established positive links between social relationships and health, well-being, and longevity, better understanding about the childhood antecedents of adult social relationship quality–specifically, how protective factors can be promoted, and harmful factors minimized–may lead to improvements in overall well-being. This study is the first to examine associations of a range of childhood social and economic factors, adverse events and experiences, health, and sociodemographic characteristics with adult social relationship quality simultaneously and in multiple countries around the world. The findings from this research highlight several key childhood antecedents of adult social relationship quality and increase our knowledge of the early-life factors that contribute to aspects of social connectedness across the life course, with potential consequences for population health and well-being.