Public health prevention continuum

The public health prevention continuum provides a useful framework for researchers and interventionists considering how to best direct their efforts to prevent or mitigate the impact of ACEs on lifelong health [8]. Primary prevention efforts are designed to either prevent or reduce the likelihood of ACEs occurring in the first place. These efforts acknowledge that ACEs can develop due to structural and Social Determinants of Health (SDOH) factors such as historic, local, and political contexts where there are multiple stressors on families–racism, discrimination, lack of affordable housing and barriers to childcare–and can include efforts to act on these ‘upstream’ factors [9–15]. Secondary prevention initiatives have an early detection focus, where, after ACEs have occurred but prior to the onset of identifiable symptoms, those who have been exposed to ACEs are identified and supported to access interventions designed to reduce the risk of any sequelae [16, 17]. Tertiary prevention focuses on treatment for families, children and young people who are identified as having already experienced ACEs and their impacts, with the goal to mitigate worsening consequences [18]. While all three levels of preventive interventions have a place in a robust multi-pronged approach that can truly impact the prevalence and downstream effects of ACEs, understanding which interventions have evidence of effectiveness on each level of the prevention continuum will help determine which interventions to apply at each level, and which levels of prevention should be an intervention priority. In addition, life course frameworks suggest other useful considerations for the design of effective ACEs prevention interventions.

Perinatal interventions

Although preconception is a logical stage of the life course to consider ACEs prevention for the next generation, inconsistent contact with the health system and lack of an obvious delivery platform, make interventions in that phase of life challenging [23]. Instead, researchers interested in ACEs prevention have often focused on pregnancy and the perinatal period as the earliest time when expectant parents have frequent contact with the health care system, are engaged with efforts to achieve the best outcomes for their child, and may be most adaptable and motivated to change their own behaviors [24]. Because of this they may be open to exploring events or experiences in their own personal history or circumstances that could pose a risk for their child (i.e. through their parenting or mental or physical health), and they may be motivated to participate in interventions designed to reduce or even eliminate those risks. Moreover, intervention programs that act early in the prenatal period, might offer the greatest potential for both effectiveness and future cost savings over the entire life course [25, 26].

A significant body of literature exists examining perinatal interventions to improve broad child health outcomes. However, to our knowledge, there are few reviews that specifically consider ACEs preventive interventions during this life stage, and none that apply the public health prevention and life course frameworks or focus on rigorously designed experimental studies. To address this gap in the literature, we undertook a scoping review of perinatal interventions designed to address at least one risk or exposure related to child ACEs. Our aims were (1) to conduct a systemic search of published perinatal interventions tested in RCTs with the purpose of preventing or addressing ACEs and their risk factors and had reported results for mother and child; (2) to assess the degree of potential bias, transparency, and reproducibility of the reported study designs and evaluation procedures; (3) to perform content analysis using (a) the public health prevention continuum and (b) life course intervention characteristics framework to describe where the interventions lie along the prevention continuum and the degree to which they incorporate life course characteristics; and (4) to summarize which interventions demonstrate effectiveness at reducing the risk of future exposure of the child to ACEs and its sequalae.

Inclusion criteria

Given that the term ‘ACEs’ has only recently become widely used, and few interventions directly addressed ‘ACEs prevention’, studies were included in this analysis if they aimed to address risk factors or potential exposure for the future experience of an ACE by the infant. Search terms included all the named adversities in the landmark Felitti study (physical, sexual, or emotional abuse, physical or emotional neglect, parental mental illness, parental substance abuse, intimate partner violence, parental divorce, having a relative in jail or prison), as well as adversities such as experience of racism, stigma, discrimination, or prejudice; experiences of community violence; peer victimization and bullying; family conflict; and chronic social vulnerability including poverty and food insecurity [1, 27–30]. These broader adversities were included as they are increasingly recognized as stressors and traumatic experiences relevant to traditionally disadvantaged and marginalized groups and likewise associated with poor health outcomes [31–33].

Interventions were categorised as perinatal if they were begun during pregnancy and were substantially delivered during the time between conception and 4 months post-partum, the latter period often referred to as the “fourth trimester of pregnancy”, acknowledging it as a challenging time of adaptation for parents and child [34]. Interventions that initially engaged and intervened with mothers, infants, and families only after birth were excluded from this review. To focus on interventions tested with the strongest study design, this review was further limited to Randomized Controlled Trails (RCTs). To be most relevant to contemporary family and community conditions, we included only articles published between January 2000 and November 2023 in English. As we wished to focus on interventions that adopted at least a dyadic approach, this review was further limited to studies that reported outcomes for both mother and child.

Search method

A research team was formed, and with the assistance of research librarians, a literature search was conducted in February 2022 with a supplemental search in November 2023 in Psychinfo and PubMed databases. ACEs did not exist as a MESH term until 2019, therefore, we used MESH terms and key words based on the ‘Improve Child Outcomes Associated with Adverse Childhood Experiences’ systemic review, to guide our search strategy [18]. The terms included were child maltreatment, early life trauma, child abuse, childhood trauma, abuse, neglect maltreatment or sexual violence, battered child mistreatment, partner violence, domestic violence, various mental health diagnoses, substance use disorders, incarcerated, divorced, separated parents, social discrimination, racial biases, and racism. We included primary, secondary, and tertiary prevention interventions regardless of their setting. We also reviewed reference lists of relevant evidence reviews for primary studies that fit the inclusion criteria.

We used an adaptation of the PRISMA-ScR process to determine which studies would be included in our full analysis [35]. (Refer to S1 Appendix, PRISMA-ScR-Checklist.) First, a single reviewer independently screened titles and abstracts for the 2148 studies identified through the initial search. Excluding duplicate articles, prevalence studies, opinion pieces, descriptive association studies, those that did not report results, and those not relevant to the topic excluded 1586 articles, leaving 562 under review. Review of reference lists led to identification of 205 additional candidate articles, yielding 767 for further review. Of these, 195 were deemed out of scope. (see “Fig 1”).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. PRISMA chart outlining the flow of studies at each stage of the review.

https://doi.org/10.1371/journal.pone.0307441.g001

Once narrowed to 572 studies, three research team members independently read and classified the studies using the following categories: title, author, year of publication, type of study, the strength of the study (design, bias), description of population, description of intervention, and reported outcomes. At this step, articles were further excluded if they were not RCTs, not implemented during the perinatal period, and did not report outcome measures for both the mother and child. This left 57 articles which on further review yielded 53 separate interventions (4 were articles about interventions already included) for our final analytic sample of studies for deeper analysis.

Data extraction

Three members of the research team reviewed the included studies and extracted data and descriptions according to the categories listed below. A summary of the key features of included articles were captured in a table. (Refer to S2 Appendix, Features of articles). Categories were selected based on relevance to the study aims and on the PRISMA-ScR checklist [36]. Discrepancies in observations between reviewers were resolved through dialogue to reach consensus.

1. Study description: author, year, title, source, abstract, study design, randomization method, blinding method of participants and researchers, allocation method, setting of intervention, geographical location, level of detail about the intervention provided.
2. Type of study: name of intervention, type of intervention, description of the intervention, the aim of the intervention, target of intervention, the intervenors, dose of the intervention
3. Participants: the number of participants and their race, ethnicity, language spoken, income level, educational level, relational involvement in intervention,
4. Impact and outcome: measures used, impact and outcome, length of follow-up, consideration of bias risk, and strength of study overall.

Considerations of bias

The degree of study bias was evaluated by adopting the Cochrane Collaboration’s tool for assessing the risk of bias in randomized trials by Higgins et al. [37]. This tool incorporates six principles for assessing bias in randomized controlled trials. The bias domains included: Selection bias (1. Random sequence generation, 2. Allocation concealment), Performance bias (3. Participant and 4. Personnel blinding), Detection bias (5. Outcome blinding), and Attrition bias (6. Incomplete outcome data). Studies were classed as having a low risk of bias if each domain was classified as low risk and the presence of bias was unlikely to alter the results seriously; a high risk of bias if at least one of the domains was classified as having a high risk of bias and the bias may alter the results seriously, and an unclear risk if the authors were not able to gather enough information from the written reported process about the details of the bias control process. Two evaluators worked together to assign categories and resolve any discrepancies, with a final assignment based on consensus. For a summary of this process refer to S3 Appendix, Critical Appraisal.

Transparency and reproducibility of intervention content

We also evaluated the level of detail provided in each study publication to assess the study’s transparency about the content of the intervention, and to determine whether sufficient detail was provided, or available, to allow for reproduction, or spread and scale of the intervention. Papers with sufficient detail to understand what occurred at each step of the intervention were classed as ‘details were provided’. If the team were unsure about what occurred, or unable to reproduce an intervention, ‘limited detail reported’ was noted. These categories were based on current intervention reporting guidelines [38].

Further content analysis

For each of the 57 articles, the team performed further analyses utilizing two frameworks: the Public Health Prevention continuum and the Life Course Intervention Research Framework [8, 20]. Consideration of the intervention for place on the public health prevention continuum was undertaken to determine the focus and emphasis of existing ACEs prevention research and to identify gaps. When applying the public health continuum to this review, primary prevention- focused interventions were defined as those designed for healthy women with no identified risk factors that may contribute to their child being exposed to ACEs; secondary as those offered to families at risk of exposing their child to ACEs e.g., a prior history of maternal depression, or history of substance use in remission at risk of relapse; and tertiary were those where an ACE exposure was identified (e.g. ongoing maternal substance use) and interventions were focused on eliminating, reducing the impact of and/or mitigating against the adverse exposure to the child. Each intervention was categorized according to its place along the continuum.

Consideration of Life Course Intervention Research (LCIR) Framework was also undertaken to determine to what degree existing ACEs prevention efforts incorporated a life course perspective. This framework is a useful tool to use since SDOH and the Socio Ecological Model are both built into this framework, because they are recognised as strong contributors to perinatal ACEs due to their impact on maternal physical and mental health. The LCIR framework is also helpful, given that ACEs themselves have impacts across the whole life course, and potential for intergenerational effects. The Life Course Intervention Research Network (LCIRN) recently undertook a modified Delphi process to delineate characteristics of life course interventions [20]. This process yielded an initial list of 12 characteristics. Life course interventions are (1) aimed at optimizing health trajectories; (2) developmentally focused, (3) longitudinally focused, and (4) strategically timed; and are (5) designed to address multiple levels of the ecosystem where children are born, live, learn, and grow and (6) vertically, horizontally, and longitudinally integrated to produce a seamless, forward-leaning, health optimizing system. Interventions are designed to (7) support emerging health development capabilities; are (8) collaboratively codesigned by transdisciplinary research teams, including stakeholders; and incorporate (9) family-centered, (10) strengths based, and (11) antiracist approaches with (12) a focus on health equity. These characteristics can be incorporated at all stages of an intervention, from conception and planning, to design, implementation, evaluation, and translation, with the whole comprising a Life Course Intervention Research (LCIR) Framework [39].

It has been suggested, but not proven, that interventions that adopt these characteristics may be more effective in having a positive impact on health. As a preliminary test of this theory, we analyzed to what extent interventions incorporated these characteristics, and whether interventions that incorporated a greater number of characteristics were more likely to show positive impacts on maternal and/or child outcomes. The number of life course characteristics included in the intervention design were determined for each intervention by two research team members, blinded to each other, and discrepancies resolved with a discussion between three team members until consensus was reached. To guide this scoring, the research team met and discussed the discrepancies and refined operational definitions of how each characteristic can be applied to interventions in the perinatal period. If the member’s score differed by three or less, the average was taken. For those that differed by more than three, the operational definitions were re-visited and refined, and then all three members re-scored the study independently. The re-score resulted in all studies having a less than 3-point difference among the three-member team. The operational definitions of the LCIR characteristics developed for this review were adapted by the authors, from the descriptions provided in Russ et al, 2021 [20]. (Refer to S4 Appendix, LCIR Characteristics.)

To understand the relationship between the number of incorporated life course characteristics and the outcomes reported for the mother and/or child, we performed a two-tailed t test on the equality of means, assuming unequal variances between the number of incorporated life course characteristics and whether the study reported a positive outcome for the mother and/or child. To evaluate for a threshold effect, we dichotomized the number of characteristics documented using a threshold of seven characteristics (i.e., 6 or less characteristics = low level of LCIR characteristics; 7 or more characteristics = high level of LCIR characteristics) based on the empiric distribution. We used a Fischer’s exact chi square to test the association between seven or more LCIR characteristics and whether the study reported a positive outcome for the mother and/or child. As robustness checks, we also used a threshold of 6 and 8 characteristics.

Quality of studies

No studies were identified with a high risk of bias, so none were excluded due to bias concerns.

Study location

The interventions analyzed were from a wide range of countries. Less than half (40%) studies were from the United States, 15% each from the United Kingdom and Australia, 8% from other European countries, 6% from South Africa and China, and one each from Iran, Pakistan, Northern Ghana, Malaysia, India, and Nigeria. Most interventions were conducted in higher income countries (HIC) (83%), with 17% from a subset of Low- and Middle-Income Countries (LMICs) in the Middle East, Africa, or Asia.

Study setting

Just over half (53%) of the interventions were home-based, 21% took place in a clinic setting, 17% in a community setting, and 4% were a combination of home and clinic based. Six studies incorporated a virtual component (phone or internet-based).

Subject of intervention

The focus of the intervention varied; 43% worked directly with the individual pregnant woman, 38% had a focus on the mother-infant/child dyad, and 23% involved more than the mother/child dyad. Three interventions (19%) reported they involved the family unit, with seven studies (13%) who had a focus on involving co-parents / couples.

Participants

Adversities

Studies aimed to address a wide variety of traumas, adversities, or risk conditions, frequently addressing more than one risk. The adversities were grouped into eight broad categories which are not mutually exclusive. “Table 1” lists the number of interventions that targeted these risks or adversities.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Adversities addressed across interventions.

https://doi.org/10.1371/journal.pone.0307441.t001

Outcomes

Various outcome measures were used across studies, 76 for mothers and 41 for children. The most common maternal measure was the EPDS (Edinburgh Postnatal Depression Scale), used in 39% of the studies, with each version validated for its country of study. “Table 2” lists the tools used across the studies to measure maternal outcomes and “Table 3” lists tools used to measure child outcomes.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Tools used to measure maternal outcomes.

https://doi.org/10.1371/journal.pone.0307441.t002

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Tools used to measure child outcomes.

https://doi.org/10.1371/journal.pone.0307441.t003

Impact

Overall, from the 57 articles 42 (74%) reported a positive impact on mother and/or child.

• 37 (65%) reported positive outcomes for mothers
• 37 (65%) reported positive outcomes for children,
• 32 (56%) reported positive outcomes for both mother and child
• 15 (26%) interventions reported no significant impact on either the mother or child

“Table 4” indicates the different intervention types and “Table 5” indicates the discipline of the person who delivered the intervention.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Reported impact on the mother and/or child by the intervention type.

https://doi.org/10.1371/journal.pone.0307441.t004

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 5. Reported impact on mother and / or child by discipline of the intervenors.

https://doi.org/10.1371/journal.pone.0307441.t005

“Table 6” below categorizes interventions based on content and the ACEs risk factor they were designed to address.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 6. A summary of the type of interventions that were provided to address ACE risk factors.

https://doi.org/10.1371/journal.pone.0307441.t006

“Table 7” lists the number of studies with certain follow up periods within their study and also identifies whether they had a positive or negative impact on the mother and/or the child.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 7. Follow up period.

https://doi.org/10.1371/journal.pone.0307441.t007

In terms of longitudinal follow-up, there were 42 studies (74%) that measured outcomes beyond the 4-month post-partum period. Nine of the fourteen interventions that reported follow up for two years or more reported a sustained positive impact on the mother and/or child. Almost all of these were home visiting programs, apart from one participatory intervention community-based woman’s group. Most interventions (63%) had periods of follow-up of 12 months or less, therefore that their long-term or life course impacts are unknown.

Content analysis used in studies

See “Table 8” for the distribution of interventions across primary, secondary, and tertiary prevention categories.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 8. Interventions by public health prevention continuum domain.

https://doi.org/10.1371/journal.pone.0307441.t008

“Table 9” describes the number and percent of interventions that incorporated each of the LCIR characteristics.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 9. Incorporation of LCIR characteristics into interventions.

https://doi.org/10.1371/journal.pone.0307441.t009

“Fig 2” demonstrates the number of interventions that incorporated various levels of LCIR characteristics into their design. All 53 interventions were developmentally focused and strategically timed to a sensitive period of brain development, as they were all implemented during the peri-natal period. Over half incorporated a longitudinal focus, targeted emergent health development capabilities, and were multi-level. No interventions incorporated all LCIR characteristics into their design. Characteristics that were less frequently incorporated included a focus on health equity, being family-centered, explicitly anti-racist (only 1(2%)), integrated with other services, and collaboratively co-designed with family and community representatives and with potential intervention recipients (Only 2 (4%)). Fathers were actively involved in the intervention in only 13/53 (25%) studies.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. The number of LCIR characteristics incorporated in an intervention and reported impact.

https://doi.org/10.1371/journal.pone.0307441.g002

Of the 14 interventions which reported no significant impact for mother or child, all 14 lacked collaborative co-design and specific anti-racist approaches, 10 did not incorporate family-centered design or service integration characteristics, nine did not include emerging health development capabilities, seven did not include holistic or multi-level characteristics, six did not have an equity characteristic and two did not report longitudinally focused characteristics.

There were 33 interventions (62%) that incorporated up to half (six or less characteristics), and of these studies, 67% had a positive impact on the mother or child. Of the twenty interventions that incorporated seven or more characteristics, 75% reported a positive impact and 25% reported insignificant impact on mother and or child.

The mean number of LCIR characteristics was significantly higher in the studies that reported a positive impact on the mother and/or child (p = 0.003). We found a significant association between the number of life course characteristics incorporated in an intervention and whether the study reported a positive impact on the mother and/or child at a threshold of six (p = 0.004) and eight (0.04), but not seven (p = 0.07) life course characteristics.

Longitudinal impact

Expectant parents can be highly motivated to make changes to benefit their unborn child, providing an ideal time during the life course for community programs to embrace this opportunity [22, 40, 41]. Yet, this potential is easily squandered if the interventions offered to them are not informed by the evidence base, are too narrowly defined, or fail to meet family and community needs. Ideally, ACEs preventive interventions would have long-lasting impacts that go beyond simple time-limited risk reductions and instead improve life course health trajectories. Given that only approximately one-third (37%) of studies followed up for more than one year, and only 26% for more than two, it is not possible to determine the long-term impacts of most of these interventions. However, of the interventions that did follow up for two years or more, 67% reported sustained positive impact, an encouraging finding. While this may reflect some bias whereby better-funded studies with more intense interventions were also more likely to be able to follow participants for longer time periods, and that long-term follow up studies with positive findings may be more likely to be submitted and published than those with negative findings it could also suggest that these more comprehensive interventions may indeed have longer-lasting impacts.

Home visiting interventions

Nurse-family partnership home visiting programs are by far the most intensively studied perinatal intervention related to ACEs prevention. While many home visiting interventions are based on the Olds model [42, 43] there is considerable variation in content and delivery, including whether the intervention is led by nurses, paraprofessionals, or volunteers. A large body of literature has yielded mixed results on their effectiveness, efficacy, and potential for spread and scale [44]. The Home Visiting Evidence of Effectiveness program maintains a website summarizing the state of the evidence for different models (https://homvee.acf.hhs.gov/). In turn, there have been tensions about whether to prioritize fidelity to original models, or to emphasize the need for sometimes significant cultural and local adaptations [45]. These difficulties have been compounded by the challenges inherent in determining which components of the program are essential vs flexible, and the need for greater detail in reporting on the content and activities undertaken during visits for different populations [46, 47]. In our review, 63% home visiting interventions reported positive impacts, but more than one-third did not, a high proportion when one considers the resources applied to home visiting. Only 20 (35%) incorporated 7 or more life course characteristics into their design, only 13 (26%) involved family members other than the child’s mother, 2 showed any evidence of any co-design, and only 1 study reported incorporation of anti-racist principles. Adaptations to home visiting programs that incorporate these elements could improve impact and are worthy of further study.

The higher likelihood of a reported positive perinatal intervention outcome when more life course characteristics were reflected in the intervention design suggests that greater incorporation of life course characteristics in general, beyond just home visiting programs, merits further study. For example, it would be of great interest to know whether incorporation of all 12 characteristics could result in interventions having even greater positive dyadic impacts that were sustained across the life course. One intervention that incorporated 10 characteristics was the secondary prevention-focused indigenous Family Spirit program, with strong culturally led equity-focused co-design, that reported sustained (3 years) positive impacts on mother and child [48, 49]. Characteristics least likely to be incorporated in current interventions include some identified by stakeholders as most important for life course orientation, including strengths-based approaches, a focus on health optimization, collaborative co-design and incorporation of specific anti-racist components [20].

Interventions in the context of the developmental ecosystem

Life course approaches acknowledge that ACEs occur and exert their effects in the context of the child’s family, neighborhood, social and community environment, yet very few studies in this review either reported on or considered in depth this broader developmental ecosystem. Most studies focused on the mother alone, some included the child or infant and only 13% included the child’s father or second parent or any other family members. This approach represents a missed opportunity for wider family and community engagement in ACEs prevention efforts that may have limited intervention effectiveness [50]. Co-parenting interventions such as Family Foundations could prove particularly impactful for ACEs prevention yet are not in widespread use [51]. Similarly, a lack of co-design, present in only 4% interventions, represents a missed opportunity for individuals and families to contribute their own lived experience and knowledge of local context to the development of culturally and contextually relative interventions created in equal partnership. Only one quarter of interventions considered individual, family and community strengths, representing a missed opportunity, and also pointing to a need for more detailed research on definitions, principles and use of different types of strengths of character, skill, and resources [52]. These omissions also point to a gap in research for new approaches to family and community engagement in intervention studies to better understand priority needs.

Only 1 of the 53 interventions in this review specifically addressed racism as a root cause contributor to adversity. This critical omission suggests that current intervention approaches are not considering the role of structural factors and biases in society and institutions that can contribute to early adversity and reinforce inequity. Greater incorporation of equity aims and anti-racist approaches into intervention design and implementation could go some way to ensuring root causes of adversity and trauma are acknowledged [53].

Finally, despite good evidence for short and medium-term positive impacts of a range of perinatal interventions, each intervention on its own addressed the prevention of only certain adversities. One option for future research is to integrate or “bundle” together multiple interventions, or core components of evidence-based interventions into larger, or stacked multi-level interventions, linking them more overtly with existing programs to address multiple co-existing needs [54, 55]. This approach has proven promising in early childhood interventions to promote early learning, but requires rigorous study in relation to ACEs prevention [56].

Limitations

Limitations of this scoping review include our choice of search terms and inclusion criteria which inevitably impact the articles under review, and our limitations on detailed understanding of the intervention content, intensity, and activities to the detail reported in the manuscripts. This review did cover a select population due to the fact that we included only articles published in English, and most reported. most interventions were delivered in English and 1 or at most 2 other languages. Therefore, it is uncertain to what degree our findings might be relevant for non-English speaking countries, groups and cultures. All review elements were limited to manuscript content–for example if authors chose not to publish negative impact or lack of impact of interventions this could not be considered in our review. Due to the imbalance in the number High Income Countries (HIC) compared with Low and Middle Income Countries (LMIC) comparisons between them were not conducted. This means our findings cannot be generalized to non-English speaking countries and to LMICs.

The two conceptual frameworks driving this review, the public health prevention continuum, and the Life Course Intervention Research framework, gave useful insights, but different conclusions may have been drawn from use of other frameworks. Moreover, the finding regarding life course characteristic incorporation and parent/child outcomes should not be interpreted as causal as there is a potential for unmeasured confounding.

Interventions were classed as having positive impact if the statistical analysis reported in the manuscript supported this conclusion, however few authors commented on the degree to which significant findings were clinically meaningful or were classed as meaningful in a practical sense by the intervention recipients themselves. Incorporation of standardized measures across research on ACEs prevention could facilitate cross study comparisons. The authors encourage greater attention to the sharing of intervention content and design processes globally to increase learnings across the field.

Future directions

There were a limited number of longitudinal studies identified for this review therefore there was limited ability to understand the impact on adolescents and adults across the lifecourse from the interventions. The authors support further longitudinal research, both to better understand the pathways from ACEs to health outcomes later in the life course, and to evaluate the impact of perinatal prevention efforts beyond childhood through adolescence, adulthood and later motherhood and fatherhood.

The imbalance between the number of studies conducted in HIC, compared with LMIC made comparisons difficult. The authors encourage further research to gain an understanding of the effects of WEIRD (Western, Educated, Industrialized, Rich and Democratic) contexts on intervention design and effectiveness.

The authors suggest that the utilization of the LCHD framework by researchers, practitioners and policy makers would be helpful to guide implementation efforts and help to shift the focus from individual to more multi-level interventions and can help to guide investment further upstream to prevent perinatal ACEs. Although the LCHD Framework was used to give consideration of the impact of SDOH by the incorporation of the Socio Ecological Model, more research is needed to better understand the evidence base for interventions that are multi-level and target broader structural issues.

Public health implications

The perinatal period holds promise as a life course stage when public health programs targeting ACEs preventive interventions can have a positive impact on the future health of both mother and child. Existing service delivery programs that already work with perinatal populations e.g. home visiting programs could incorporate additional evidence-based ACEs prevention interventions identified in this review, including co-parenting and integrative health interventions.

Bundled multi-level evidence-based interventions that address social and structural determinants of health in their design and delivery could form a more comprehensive perinatal ACEs prevention strategy. Incorporation of life course characteristics including co-deign of interventions with potential recipients, and incorporation of trauma-informed elements into interventions is a strategy worthy of future public health research.