본문 바로가기
읽기 · 학술 · 공개 라이선스 원문

Parenting interventions to promote early child development in the first three years of life: A global systematic review and meta-analysis

Jeong, Franchett, Ramos de Oliveira, Rehmani, Yousafzai · 2021

공개 학술지 PLoS Medicine 에 실린 총설 논문입니다(2021년). CC BY 4.0 으로 공개된 글이고, 그림 · 표 · 참고 문헌 표시를 뺀 본문을 실었습니다.

출처 Jeong et al. (2021), PLoS Medicine · 라이선스 CC BY 4.0 · 원문

장 4개 올림 이 장 2081낱말
← 앞 장

이 장은 원문만 있습니다. 문장별 해석은 앞 장부터 차례로 붙입니다.

4장 · Parenting interventions to promote early child development in the first three years of life: A global systematic review and meta-analysis — Discussion

To the best of our knowledge, this is the largest and most comprehensive systematic review and meta-analysis that synthesized 102 parenting interventions delivered during early childhood from across 33 countries and estimated the effectiveness on a wide range of ECD and caregiving outcomes. We found that parenting interventions delivered during the first 3 years of life improved early child cognitive, language, motor, and socioemotional development, and attachment, and reduced behavior problems. Parenting interventions additionally improved parenting knowledge, parenting practices, and parent–child interactions; however, they did not significantly reduce parental depressive symptoms. Our findings extend prior evidence in 4 key ways.

First, previous reviews of parenting interventions delivered during early childhood have focused on the effectiveness of select types of parenting interventions on 1 or 2 developmental outcomes. We conducted the most comprehensive synthesis that evaluates intervention effects on 6 distinct child development outcomes that are most relevant to children during the earliest years of life. Overall, pooled effect sizes were relatively greater for cognitive, language, and motor development and attachment, but smaller for socioemotional development and behavior problems. Greater estimated pooled effects on cognitive, language, and motor development are consistent with the strategies used in the majority of reviewed parenting interventions that focused largely on increasing early play and learning opportunities—such as through caregiver engagement in stimulation activities or enhanced verbal responsiveness—which are directly related to these particular outcomes. In contrast, the smaller observed effects on socioemotional development and behavior problems are likely because fewer of the reviewed parenting interventions incorporated program components directly related to the commonly assessed outcome domains such as emotional competence or externalizing problem behaviors.

In order to effectively improve child socioemotional development and reduce behavior problems, interventions require evidence-based curriculum content and strategies that are grounded in social learning theory, such as supporting caregivers’ behavioral management skills, addressing parental mental health, and encouraging nonviolent discipline. Yet, most parenting programs for socioemotional and behavioral development to date have been commonly delivered for preschool-aged children rather than young children during the earliest years of life. New intervention models are needed in order to more effectively improve early child socioemotional and behavioral development particularly for young children during infancy and early childhood. Part of this effort will require establishing standardized measurement tools for assessing socioemotional development and behavioral problems that can be used reliably and cross-culturally in the context of interventions with young children.

Second, we analyzed the impact of parenting interventions on parents themselves and found improvements in parenting knowledge, parenting practices, and parent–child interactions. However, 40% of studies aimed at improving ECD did not measure a single parenting outcome. Given that the primary focus of parenting interventions is to support parents and improve the caregiving environment, it is critical to measure parenting as a way of understanding how interventions impact ECD. More mediation analyses are needed to empirically substantiate program theories of change and identify the key caregiving mechanisms by which parenting programs lead to improved ECD outcomes.

We found little evidence to support that parenting interventions reduced parental depressive symptoms. By covering a broader set of parenting interventions and including studies in HICs, our results extend the findings of a prior meta-analysis that found that stimulation interventions did not significantly reduce maternal depressive symptoms in LMICs. This overall null effect on depressive symptoms is likely due to the fact that few of the parenting interventions had an explicit psychological component to address parental mental health. However, one notable exception was a parenting group intervention in Uganda that integrated sessions on not only stimulation and responsive caregiving but also parental emotion regulation, coping strategies, and principles of cognitive behavioral therapy. This intervention found medium-sized reductions in maternal depression. While bundling multiple components together (e.g., mental health and parenting) to create intervention packages is recommended in the literature, additional research is needed to determine the independent and additive benefits of particular components (e.g., those that target mental health versus parenting) in order to optimize intervention effects on outcomes.

In spite of the observational evidence that fathers are critically important for ECD, only 7 parenting interventions in this review (7%) engaged fathers to some degree, and only 1 study measured paternal outcomes from fathers directly (who only comprised 5% of the respondents in this study). A prior review has drawn attention to the limited number of parenting interventions with fathers and the program implementation barriers that limit fathers’ participation in parenting interventions across global contexts. More research is needed to determine how to optimize the design and delivery of parenting programs so that they are inclusive of fathers and can achieve benefits for maternal and paternal parenting and ECD.

Third, moderator analyses showed that parenting interventions had greater effects in LMICs compared to HICs for improving child cognitive, language, and motor development and parenting practices. Parents and children in LMICs are more likely to be exposed to additional risk factors that constrain nurturing care and ECD (e.g., low parental education, malnutrition, fewer early learning opportunities), such that support for parenting may be more beneficial in low-resource contexts. More research is needed to understand whether and how variation in risk profiles within populations may influence the degree of benefits from a parenting intervention. It is worth noting that none of the included studies directly compared the same intervention program in a LMIC and HIC. As a result, it is possible that these subgroup results may also, in part, be confounded by other differences in intervention design and implementation between LMICs and HICs.

Regarding intervention content, we found that those with responsive caregiving components had significantly greater impacts on child cognitive development, parenting practices, and parent–child interactions, compared to interventions that did not include responsive caregiving content. These results underscore the importance of not only increasing early play and learning materials (e.g., provision of homemade toys), which have traditionally been the focus of parenting intervention for young children especially in LMICs, but also incorporating program components that directly support parental sensitivity and responsiveness.

At the same time, we did not find evidence to suggest that child age, intervention duration, delivery modality, or setting moderated intervention effects on more than one outcome. The lack of statistically significant subgroup results by child age are consistent with the few prior meta-analyses that have similarly shown no age effects among parenting interventions focused during early childhood or spanning broader age ranges from early to middle childhood or birth to age 18. Despite strong neuroscience and economic arguments for intervening as early in life as possible, our review suggests that parental engagement and support programs may indeed have universal benefits on child and parent outcomes irrespective of child age or the timing of introduction.

Furthermore, the lack of statistically significant subgroup results by program duration, delivery modality, and setting are also consistent with the few prior meta-analyses that have conducted similar analyses and shown no associations between these program aspects and intervention effects on ECD and parent-level outcomes. One recently published cluster RCT specifically compared the effectiveness of weekly home visits to weekly mother–child group sessions over a 2-year period in rural India. Trial results revealed that group sessions were as effective as home visits in improving child cognition and language, and both delivery models similarly had null impacts on child motor, socioemotional, behavior development, and parenting knowledge and practices. Of note, cost analysis revealed that the group-based model required a quarter of the costs compared to home visiting. Nevertheless, our findings suggest that there are various effective implementation strategies and program models that can be used to deliver parenting interventions and achieve positive impacts on ECD and parent outcomes. Decisions regarding program duration, delivery modality, and setting should be determined based on the existing resources and systems, community needs, risk profiles of the population, and cultural context.

In summary, our study had several strengths. These include the comprehensive review of RCTs conducted in LMICs and HICs, the investigation of multiple child development and caregiver outcomes, the large number of studies and participants represented, and the various moderator analyses conducted to explore potential features associated with intervention effectiveness. At the same time, there are a number of limitations that should be noted. First, there was considerable unexplained heterogeneity in the pooled effects across outcomes. Future research should further explore heterogeneity in treatment effects by other program factors (e.g., content, theoretical goals, behavior change techniques) and intervention implementation features (e.g., dosage, fidelity, population risk profiles, supervision and training provided to delivery agents). Considering the generally inadequate reporting of intervention details that we identified across the majority of studies, future trials should use standardized guidelines such as CARE and TIDieR for common reporting of intervention content and implementation to facilitate more nuanced comparisons and subgroup analyses. Second, our review focused on parenting interventions for children during the first 3 years of life. As a result, other relevant parenting interventions with preschool-aged children or those covering a broader age range of children (e.g., 2- to 6-year-olds) were excluded, which limits the generalizability of our results. Third, the reporting of the psychometric properties of measures (e.g., reliability and validity evidence, adaptation procedures) was highly variable across studies, with many studies not reporting any such details. Measures of ECD and parenting ranged widely from validated and adapted assessments to unstandardized measures, which may affect comparability across studies and robustness of findings. Fourth, our meta-analysis involved 6 ECD outcomes and 4 parent-level outcomes and 7 moderator variables, which raises the issue of statistical multiplicity and the increased risk for false positive findings in our estimates of pooled effect sizes and subgroup analyses. Although we used robust variance estimation to address multiplicity and statistical dependencies, this method might raise issues of disproportionate influence from studies that contribute a large number of effect sizes relative to those that contribute fewer, and particularly when the overall number of studies and effect sizes is relatively small. Future reviews should apply additional methods to address issues of multiplicity. Fifth, we found evidence suggesting small-sample bias for 2 outcomes (child language and parent–child interactions) based on the results of Egger’s tests. Future reviews should include searches of grey literature, regional databases, and in multiple languages to improve the global coverage and reduce publication bias in the evidence base.

Finally, our review also highlighted other broader issues with regard to risk of bias in the design, conduct, and reporting of parenting intervention trials. Specifically, the majority of trials had potential concerns of selection bias due to inadequate details about what methods were used to conceal the allocation sequence before assignment to determine whether intervention allocations could have been foreseen before or during enrolment. Additionally, the majority of trials were not preregistered and did not have a study protocol of prespecified outcomes to assess whether selective outcome reporting bias or selective analysis reporting bias was present. Selective reporting bias, especially of null results, may additionally contribute to the publication bias detected for some of the outcomes. In light of our findings, future trials of parenting interventions should preregister primary and secondary outcomes and their scoring methods in study protocols, report all analyzed outcomes regardless of statistical significance levels, and provide adequate descriptions in publications not only regarding random sequence generation but also the allocation concealment process to especially improve the quality of this body of evidence.

Conclusions

Parenting interventions during the first 3 years of childhood are effective for improving ECD and caregiving. In the last decade, there has been considerable policy attention on ECD globally. The United Nation’s Sustainable Development Goal Target 4.2 is to ensure that all girls and boys have access to quality early childhood development, care, and education services. However, the current policy landscape for ECD has not yet resulted in greater investments and implementation of large-scale national parenting programs. More implementation science research regarding how to adapt and deliver parenting programs across diverse local contexts, coordinate within existing systems and services, finance programs so that they are cost-effective and sustainable, and scale parenting programs while maintaining quality and effectiveness is needed to accelerate political will and action. Finally, while the present review supports robust immediate benefits of parenting interventions during early childhood, more follow-up studies of individual trials are needed to determine whether there are sustained, longer-term effects for children and caregivers over the life course. Greater knowledge regarding the sustainability or fadeout of effects can guide the design and packaging of improved parenting programs, such as the integration of subsequent “booster” sessions that can potentially further support children and caregivers’ evolving needs and sustain intervention benefits over the life course.

이 글은 위 출처의 원문을 그 라이선스(CC BY 4.0)에 따라 옮겼습니다. 참고문헌 목록과 인용 번호, 그림 · 표 · 각주 · 수식 · 부록은 빼고 절별로 나누어 실었으며, 그 밖의 문장은 원문 그대로입니다. 원 저작자와 발행 기관은 이 화면을 만들거나 검토하지 않았고, 이곳의 해석과 낱말 풀이는 공식 번역이 아닙니다.