Impact of Positive Childhood Experiences (PCEs): A Systematic Review of Longitudinal Studies

Positive Childhood Experiences (PCEs) have been described as safe, stable, nurturing relationships and environments. They form an important subset of resilience factors that could moderate the impact of ACEs [7]. Influenced by ACEs research, there has been an increasing interest in developing a PCE framework to inform cumulative experiences that could moderate the impact of ACEs [8, 9]. However, currently, there is no consensus on what factors should be included in such a PCE framework. Table 1 lists the four emerging dominant conceptual frameworks and the accompanying measures for PCEs.

The existing frameworks all emphasize that PCEs are not just the absence of ACEs but are positive experiences, assets, or resources that provide the necessary foundation for overall development. They agree that the PCEs occur at multiple levels of a child’s environment (family, school, neighborhood, and community) in line with Bronfenbrenner’s ecological theory and highlight the interrelatedness of those domains [16]. This signifies the importance of cumulative measures to prevent overestimation of the impact of individual factors. While all include safe, stable, nurturing relationships and environments in their framework [8], others additionally include resources such as (i) a clean, safe home with enough food, (ii) opportunities to learn, (iii) hobbies or intellectual pursuit, (iv) organized sport, (v) family routine or consistent rules [10] and vi) positive individual attributes [6, 14]. A recent scoping review of PCEs that focused on measurement and evidence highlighted the heterogeneity in the measurement of PCEs and emphasized the need for consistent operationalization of PCEs [17].

If cumulative PCEs are proven effective in moderating the impact of ACEs using valid study designs, they could have a significant public health impact. Researchers can then explore whether enhancing non-family PCEs, such as the availability of a caring non-parent adult, positive community or neighborhood, school, and peer factors, could change the trajectory of children from disadvantaged backgrounds, such as those in out-of-home care or those exposed to high familial ACEs. This public health approach could potentially improve the resilience and well-being of large groups of young people at a time when mental health problems in youth are increasing.

One of the earlier publications on PCEs used data from the ACEs study and showed that those with greater family strengths had reduced rates of risky sexual behaviors and teenage pregnancy [18]. Research evaluating individual domains of PCEs, such as parenting practices [19], school connectedness [20], non-parent adult [21] or positive neighborhood factors [22] indicated a significant association with better adolescent or adult health outcomes and psychosocial functioning. As PCE types (e.g., parenting, family, school, teacher, peer, neighborhood, and self-capacities and perception) tend to cluster together, studying one PCE type alone might lead to an overestimation of the impact [23].

Studies that used cumulative PCE scores also found a reduced risk of mental illness [8], cardiovascular disease [24], reconviction, and rearrest rates [25] but utilized retrospective reports of childhood experiences. A recent systematic review of 58 studies on the impact of PCEs on adult outcomes found that higher levels of PCEs were associated with lower odds for major depressive disorder, personality disorders, cigarette use, risky alcohol use, and general perceived stress. They also predicted better cardiovascular health, increased engagement in exercise, higher self-esteem, and better psychosocial functioning in adulthood [26].

Despite these important findings, this review had two major limitations [26]. First, all the included studies, except three, were cross-sectional and used retrospective measures to estimate PCEs. While retrospective reporting can allow an adult to recall their whole childhood experience, they are subject to recall bias and have a poor correlation with prospective reports of abuse and neglect [27]. Second, the review did not focus on the cumulative effects of individual PCEs. While the impact of one type of PCE could be small, the question of interest is whether the cumulative effect of PCEs could be substantial to both the individual and groups of individuals and offer opportunities for large-scale public health interventions.

Twenty-three studies in the systematic review by Han et al.[26] used statistical interactions and other analytical methods to evaluate the protective effects of PCEs among individuals exposed to childhood adversities. While only seven out of those twenty-three studies provided evidence for protective effects, such that the association between childhood adversity and adverse outcomes was lower in those exposed to high PCEs, seven others indicated worse outcomes with high PCEs, and the rest provided insignificant or mixed results. Despite the mixed findings, the authors concluded that PCEs are more likely to promote positive outcomes rather than moderate the effects of adversities [26].

In a scoping review on PCEs, 61 of the 66 publications measured PCEs retrospectively. Of the longitudinal studies, only one cohort study reported across three publications evaluated the effects of cumulative PCEs, and two other studies estimated the impact of individual PCEs. They noted that all included studies found evidence for promotive and/or protective effects of PCEs on behavioral and health outcomes [17].

To summarise, previous reviews reported that cumulative PCEs appear to be associated with positive adolescent or adult outcomes. However, the existing literature is limited by largely cross-sectional studies or retrospective reports from longitudinal samples, and there are inherent challenges in interpreting the true impact of PCEs on long-term outcomes using such studies. There is a need to examine prospective reports of PCEs using longitudinal designs and ascertain which PCEs are likely to be the best predictors of long-term outcomes. Further, previous research on PCEs has predominantly focused on adults. However, adolescence is a critical period of development, with the peak age and median age of onset of mental disorders being 14.5 years and 18 years, respectively [28]. It is also a time when those affected by ACEs are likely to begin engaging in high-risk behaviors [29]. Studying longitudinal adolescent and adult cohorts is thus needed to better understand the protective impact of PCEs on long-term behaviors. Additionally, it is essential to examine studies that evaluate cumulative PCE scores, as an individual exposed to multiple ACEs may require multiple PCEs to cope effectively [30]. To address some of these gaps in the literature, this systematic review was conducted to test the following hypotheses using longitudinal studies that explored the cumulative impact of PCEs prospectively: (1) Cumulative PCEs would positively impact longitudinal mental, physical, and psychosocial outcomes. (2) Cumulative PCEs would moderate the impact of ACEs (3) Certain combinations of PCEs would predict better outcomes in longitudinal samples.

Studies that measured cumulative PCEs prospectively before age 18 and evaluated the influence of those scores on mental health, physical health, or psychosocial outcomes using a longitudinal design were included. Studies conducted in clinical samples, measured PCEs retrospectively, included only one type of PCE or conducted analysis based on individual PCE types rather than cumulative scores were excluded.

A systematic search was conducted in the following databases: Pubmed, EMBASE, CINAHL, Cochrane, PsychINFO, and Scopus, and it was updated on 24 May 2024. Reference lists of eligible papers were screened for additional articles.

The search strategy was designed to find longitudinal studies that examined PCEs and their impact on mental and physical health or psychosocial outcomes. A search strategy was created for PubMed, using various terms for PCEs, child or teen, longitudinal or prospective study, and then modified for other databases. Some search terms were based on those used in previous publications known to the authors. Reference lists of eligible papers were screened for additional papers. The search strategies for each database are presented in Appendix 2.

One author (KK) performed the initial screen (title & abstract) to shortlist articles for in-depth full-text review. Two authors (KK) and (SS) reviewed these articles independently, and decisions were based on consensus using the criteria outlined above. Disagreements were resolved through discussions with a senior author (JS).

One author (KK) extracted the data using a standardized data extraction sheet, as shown in Appendix 3, which was verified by another author (SS).

The quality of each publication was rated independently using criteria based on the Joanna Briggs Institute (JBI) checklist for cohort studies (which is listed in Appendix 4) [32] by two authors (KK and SS) based on the following questions (i) Were the two groups recruited from the same population? (ii) Was exposure measured similarly in people assigned to exposed and unexposed groups? (iii) Were the exposures measured in a valid and reliable way? (iv) Were confounding factors identified? (v) Were strategies to deal with confounders stated? (vi) Were the groups free of the outcome at the start of the study or the moment of exposure? (vii) Were the outcome measures valid and reliable? (viii) Was the follow-up time reported and was it sufficient enough for outcomes to occur? (ix) Was the follow-up complete, and if not, were the reasons for the loss of follow-up described and explored? (x) Were strategies to address incomplete follow-up utilized? (xi) Was appropriate statistical analysis used? Decisions were based on consensus, and differences were resolved by further discussion with a senior author (JS).

As the included papers measured different frameworks of PCEs and evaluated different outcomes using different measures across various age groups, it was not possible to synthesize the results in a meta-analysis.

The protocol for this systematic review was prospectively registered and can be accessed on PROSPERO (ID: CRD42022384775).

The publications included in this review examined the relationship between PCEs and various outcomes across different contexts.

The quality of the included articles was rated against JBI criteria for cohort studies and is shown in Table 4 [32]. All included articles had both exposed and non-exposed (to PCEs) participants within the same group, accounted for confounders, and were free of outcomes when the exposure (to PCEs) started. Three publications adapted items from the validated BCEs scale to develop the PCE measure (Counter-ACEs) but did not evaluate the reliability and validity [33‐35]. Three papers used the PCE measure developed from the HOPE framework and used Confirmatory Factor Analysis (CFA) and found a good fit for the underlying factor structure of the model [15, 38, 39]. One paper estimated intra-class correlation for each domain of PCEs but did not test for construct validity [36]. Wu [41] employed seven PCE items developed by Bethell et al., who verified intra-class correlation and completed Principal Component Analysis (PCA) of those items [8]. All PCE measures had good predictive validity.

While six publications utilized valid and reliable outcome measures [15, 33‐35, 38, 39], two used self-reporting for outcomes without testing for reliability or validity [36, 41]. The duration of follow-up varied widely from four [36] to fourteen years [15, 39] and attrition ranged from 7.4% [34] to 70% [38]. For the biological measures of health, only about 24% of participants in LSAC and ALSPAC had blood samples taken [39].

In this systematic review, cumulative PCEs in childhood were associated longitudinally with modest benefits to mental health and psychosocial outcomes in adolescence and young adulthood (ages 15 to 23 years). We found two publications on physical health outcomes that suggested that PCEs were correlated with a lower risk of persistent insomnia and lower levels of inflammation. A previous systematic review that predominantly included cross-sectional and retrospective studies also showed the beneficial effects of PCEs on physical health, mental health, and psychosocial functioning in adults [26].

In the current review, four studies that evaluated the moderating effects of cumulative PCEs and one that studied interaction effects yielded positive results. These findings support our second hypothesis that cumulative PCEs may reduce poor mental health and psychosocial outcomes associated with adversities in childhood. This is congruous with previous research suggesting that ACEs are not necessarily deterministic of poor health outcomes [42]. Similarly, two studies that measured PCEs prospectively found that PCEs moderated the impact of childhood adversities [25, 43] but one study that used retrospective measures of PCEs indicated no protective effects [44].

When Han and colleagues [26] evaluated interactions between childhood adversities and PCEs., only seven out of twenty-three studies showed evidence of significant interaction. This could be due to the inclusion of studies that did not evaluate cumulative PCE scores, as an individual exposed to multiple ACEs may require multiple PCEs to cope effectively [30]. Using retrospective measures that were subject to recall bias could be another reason.

Further, the results from three included studies [34, 37, 39] indicated that PCEs may moderate the effects better in those exposed to less severe adversities. These findings are consistent with two retrospective studies that evaluated the impact of cumulative PCEs on various health diagnoses and obesity [45, 46] and a study where the greatest caregiver warmth protected cardiovascular health in those with the lowest child abuse exposure and not in those with the most severe child abuse exposure [47].

The papers in our review used four different PCE frameworks: a modified HOPE framework, a BCEs framework, and a PCE list from Bethell et al. [8], and the PCEs list that was curated for the LONGSCAN data set. Five articles only used items that were part of safe, stable nurturing relationships and environment frameworks, including various social engagements but not positive self-qualities or stable family routines. Interestingly, Guo et al. [15] deliberately excluded the “social and emotional competencies” while validating the HOPE framework, arguing that those could be outcomes secondary to PCEs. Only the study by Crandall et al. [34] the BCE framework also included measures on positive self-qualities and stable family routines, but their results were in line with others. These results indicate that safe, stable, nurturing relationships may form the core component of the PCE construct. This partly verifies our third hypothesis that certain combinations of PCEs would predict better outcomes in longitudinal samples.

Overall, the included studies were of good quality as rated by the criteria of JBI quality rating for cohort studies, but there were some deficits [32]. Two studies did not use valid measures for PCEs [34, 36] and two studies did not use valid or reliable outcome measures for outcomes [36, 37]. As expected, some of these longitudinal studies had large attrition, although research suggests this does not necessarily lead to bias [48].

The research into cumulative PCEs is in its early stages, and some gaps require further study. With the assumption that each type of PCE is equally important for a specific outcome, the potential for differential influence of each PCE type is ignored. It is unclear if non-family PCEs (school, neighborhood, peers, non-parent adult, and positive self-perception) are effective in those without PCEs within their families. There are also uncertainties on how to measure PCEs. One study counted individual PCE domains as present if the participants scored more than the average of the sum for all the items in that domain, even in one wave of data collection [34]. Others had cut-offs dichotomized for each domain at the 75th percentile of summed scores [15, 36] and estimated associations separately for each wave of data collection and in combination [15]. Both approaches may have captured those exposed to higher-intensity PCEs. This highlights the possibility that the mere presence of PCEs may not influence outcomes but depend on the greater quality and intensity of those experiences.

Given the high worldwide prevalence of mental disorders, there is strong interest in translating learnings from research on PCEs into meaningful public health interventions [49]. Healthy Steps, The Strengthening Families Approach and Protective Factors Framework ™, the HOPE framework, the Families and Schools Together program (FAST), Restorative Circles, and Circle of Courage are some of the initiatives and interventions that have been recommended to promote PCEs, to mitigate the impact of ACEs and thereby disrupt the link between adversities and negative stress-related outcomes [50‐55]. This has led to predictions that the promotion of PCEs might improve mental and physical health at a population level [56].

While we were able to investigate cumulative PCEs from longitudinal studies in the current review, there are some limitations. Only eight publications from five longitudinal datasets were available for inclusion, and further research using prospective longitudinal data is required. Inherent to long-term prospective studies, some had high attrition rates. Most studies evaluated outcomes in adolescence but were of high quality. Finally, they were all from developed countries limiting generalizability to other populations.

This systematic review of longitudinal studies that evaluated prospectively measured PCEs strengthens the previous findings from cross-sectional and retrospective studies that PCEs may be associated with long-term positive mental health, physical health, and psychosocial outcomes. In contrast to previous research, this review also highlights the potential protective effects of PCEs against various adverse experiences. Most included studies had high-quality ratings based on the JBI rating for cohort studies, offering us confidence in the results. There is a lack of consensus on what components constitute PCEs. Interestingly, all studies except one included only the safe, stable, nurturing relational component of PCEs, indicating its central role in PCEs. Most studies in this review captured higher exposure to PCEs, suggesting that the mere presence of PCEs may not be adequate to offer long-term benefits, and the results would depend on the quality of those experiences. PCEs could be integrated into public health interventions to protect against the negative effects of adverse childhood experiences and promote overall well-being. However, further research is required to replicate the existing findings, clarify the key components of PCEs, resolve the issues with the measurement of PCEs, and improve the generalizability and validity of PCEs by conducting studies globally, including in developing countries.