Involvement of Non-Parental Caregivers in Obesity Prevention Interventions among 0–3-Year-Old Children: A Scoping Review

Introduction: We examined the scope of literature including non-parental caregiver involvement in child obesity prevention interventions. Methods: We conducted a scoping review following the Arksey and O’Malley framework, including only studies reporting the effect of an intervention on growth, weight, or early childhood obesity risk among children ages 0 to three years, published between 2000 and 2021. Interventions that did not include non-parental caregivers (adults regularly involved in childcare other than parents) were excluded. Results: Of the 14 studies that met the inclusion criteria, all were published between 2013 and 2020, and most interventions (n = 9) were implemented in the United States. Eight of the 14 interventions purposefully included other non-parental caregivers: five included both parents and non-parental caregivers, and the remaining three included only non-parental caregivers. Most interventions (n = 9) showed no significant impact on anthropometric outcomes. All interventions found improvements in at least one behavioral outcome (e.g., food groups intake (n = 5), parental feeding practices (n = 3), and screen time (n = 2)). This review can inform future interventions that plan to involve non-parental caregivers, which may be beneficial in shaping early health behaviors and preventing obesity early in life.


Introduction
Childhood obesity is a serious public health concern [1]. Globally, 38.3 million children under the age of five experience overweight or obesity; and in the United States (U.S.), 13.4% of 2-to 5-year-old children have excess weight [2], with racial/ethnic minority children at greater risk [3]. Obesity in early life contributes to a critical public health burden due to hospitalization, treatment expenses, and negative social and economic outcomes [4,5]. The disparity observed at this young age presents an early window of opportunity to develop prevention interventions that foster healthy habits that can continue into adulthood [6]. Even though the first several years of life represent a critical period for health, development, and obesity prevention [7], few prevention programs intervene during these years [8][9][10][11][12][13].
Although care of a child early in life commonly involves a parent or guardian, often a mother, as the primary caregiver, other caregivers such as grandparents, extended family and household members, friends, or childcare providers may also provide care [14][15][16]. Non-parental caregivers are involved in child feeding by controlling availability and access Int. J. Environ. Res. Public Health 2022, 19, 4910 2 of 16 of foods and can serve as active role models who can promote or hinder healthy feeding [17]. They also help shape other behaviors such as physical activity and screen use [18]. For example, evidence suggests that grandparents play an important role in child health behaviors and weight [19]. However, grandparents find it difficult to discuss childhood obesity, a situation that highlights the need to include grandparents in interventions [20]. Furthermore, parents may rely on some type of childcare (e.g., Early Head Start programs, daycare centers), and given that many children can spend a significant amount of time under the care of childcare providers early in life, it is critical to include these caregivers in prevention interventions [21][22][23][24].
Therefore, non-parental caregiver involvement has been identified as an important component in interventions intended to improve health behaviors [25,26]. However, few interventions aimed to prevent childhood obesity actively involve or target non-parental caregivers [27]. Most early-life obesity prevention interventions focus primarily on parents or a single caregiver in the immediate family and emphasize parent-child interactions, an approach that may not serve with parents in households where other non-parental caregivers are involved in feeding and care [28].
By intervening at an early age, the behaviors of parents and non-parental caregivers that can impact child diet can more easily be shaped. This is especially true for children with ages between 0 to three years, who are not yet in the school system and are often cared for by multiple caregivers. Understanding the state of research on childhood obesity prevention and the participation of non-parental caregivers in these interventions is critical and necessary to identify research gaps and opportunities. The goal of this review was to examine the scope of existing obesity prevention interventions among children 0 to three years, that involved, non-parental caregivers and to describe the non-parental caregivers' characteristics and involvement in these interventions.

Design
We conducted a scoping review of literature on obesity prevention interventions among children aged 0 to three years that involved non-parental caregivers to understand the available research given the nascent area. Specifically, we sought to examine the scope of research on interventions that reported outcomes on child growth, weight, or any other anthropometric related outcome, and to describe the non-parental caregivers and their involvement in the intervention. We defined non-parental caregivers as any person regularly involved in a child's care such as family members, neighbors, babysitters, or childcare providers, as reported in the study.
Following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines for scoping reviews [29] and the Arksey and O'Malley's (2005) fivestage framework [30], we identified a research question, examined relevant studies, selected studies for inclusion, charted, and analyzed the data, and summarized the results.

Search Strategy
We searched the Cochrane Library and PubMed databases to identify studies using a standardized search string, which followed the patient/population, intervention, comparison, and outcomes (PICO) framework. These terms included medical subject headings and keywords. A medical education and clinical outreach librarian at a public research institute provided advice and direction during the search strategy development and literature review. Table 1 provides the string of search terms used.

Study Selection
Studies were included if they were: (1) peer-reviewed; (2) published between 2000 and 2021, as there has been a significant growth of scientific publication related to infant feeding and growth after the year 2000; (3) conducted in English or Spanish; (4) reported on growth, weight, or any other anthropometric related outcomes such as body mass index (BMI) as primary or secondary outcomes; (5) included children in the age range 0 to three years; and (6) included non-parental caregivers. Domestic (i.e., U.S.) and international studies were eligible. Excluded literature included dissertations, protocol papers, published abstracts, unpublished work, literature reviews, and meta-analyses.
We identified 3641 references in the databases. The first author exported the references to Endnote, removed duplicates, and screened article titles and abstracts for inclusion; and then performed a hand search. Three authors (AR, AMC, SH) assessed the full text of the remaining articles against the eligibility criteria, and independently reviewed and extracted the data. AR and AMC reviewed the final set of articles, confirmed their inclusion, discussed uncertainties and ambiguities, and reviewed for non-parental caregiver participation in the intervention.

Data Analysis
During the data extraction, eligible studies were imported into an Excel spreadsheet and organized by article citation, research question (s), involvement of other caregivers (yes/no), approach or theoretical model, use of community-engaged approaches, research design, type of research, data collection methods, key variables, scales or measures, sample size, participant age range, participant inclusion and exclusion criteria, main results, and limitations. We excluded articles that did not include non-parental caregivers as well as those articles that did not clearly or explicitly indicate non-parental caregiver status.
Following recommendations for conducting scoping reviews, our analysis focused on describing the interventions, and non-parental caregivers' characteristics. Based on this information, we identified gaps in the research and next steps for the involvement of non-parental caregivers in obesity prevention interventions [29][30][31].

Study Selection
The search from the Cochrane Library and PubMed yielded 3641 articles. After removing duplicates, 2920 titles and abstracts were screened, of which 2754 were excluded from full review for not meeting the inclusion criteria (n = 1647) including not focusing on the relevant outcomes for this analysis (i.e., infant growth, weight, BMI), or the children not being within the age range.
The search from the Cochrane Library and PubMed yielded 3641 articles. After removing duplicates, 2920 titles and abstracts were screened, of which 2754 were excluded from full review for not meeting the inclusion criteria (n = 1647) including not focusing on the relevant outcomes for this analysis (i.e., infant growth, weight, BMI), or the children not being within the age range.
After the full-text review of 178 articles, 164 articles did not meet the inclusion criteria and were excluded (125 of which did not include non-parental caregivers). A final sample of 14 articles were included in the scoping review. The complete screening process is presented in Figure 1.

2-5 years
Cluster RCT 42 adult-child dyads Low-income parents, with a child between 2-5 years-old, with high levels of perceived stress.

Hammersley et al. (2019) Australia
Does a parent-focused, internet-based lifestyle program aimed to overweight children or at risk to becoming overweight, improve child BMI, obesity-related behaviors, parent modeling and self-efficacy?

RCT 86 adult-child dyads
Children were at or above the WHO fiftieth percentile for BMI for their age and sex; they were excluded if they had a medical condition that affected weight.  Sample sizes ranged from 42 adult-child dyads [35] to 450 adult-child dyads [36], with other studies including more than 500 children from childcare centers [37,39,41,43]. Nine studies targeted children two to five years [32,33,[35][36][37][38][40][41][42], two studies extended this range to six years [43,44], one study included children between two to four years [45], and another study included children between 1.5 to four years [39]. One intervention was initiated during pregnancy (n = 1) [34]. No study included children within the specific age range 0 to three years.

Natale et al. (2014) Healthy Inside-Healthy
Outside (HI-HO) SEM 6-month, culturally appropriate intervention that implemented a nutrition and PA curriculum. CCs teachers received two trainings about healthy menus, and child nutrition; parents received information about nutrition and PA, and at home activities once a month by a registered dietitian; the centers incorporated policies.
Majority of families identified as Hispanic, and 35% were Spanish-speaking. Providers were also ethnically diverse.
No significant changes in PA, weight z-score, height z-score; however, BMI z-score was negatively correlated with their participation in home activities, and 97% of the children with normal weight remained with normal weight. Children in the intervention consumed more F/V, 1% milk, and less juice and junk food.

Yilmaz et al. (2015) Name was not specified SCT
Conducted by health care practitioners during maintenance visits in a hospital; non-parental caregivers received home visits. Consisted of four intervention components, with a total duration of 6 weeks. Included printed materials, interactive CDs, and a counseling call, which addressed consequences of increased screen time, and alternatives to watching TV.
Ethnicity and race were not presented in the results. Majority of families had an annual income between $10,000-20,000.
In the intervention group vs. control, there was a significant reduction in meals consumed in front of a screen, screen time of children, parents, and non-parental caregivers, and ins aggressive behaviors. There were no significant differences in BMI z-scores. Parents in the intervention vs. control decreased restrictive feeding practices relative. Similar changes in children's weight-related behaviors were observed in the intervention and control parents. Providers at CCs that serve low-income children.
Race-ethnic data were not available.
In the HAP + CCHP program, there was a significant increase in the proportion of children exposed to nutrition and PA best practices, and a significant reduction in child BMI. 3.5% of the children in the study were Australian-Aboriginal. No significant differences in the BMI, PA, screen time, or sleep outcomes.
Intervention vs. control group showed a reduced intake of discretionary food, and parents improved self-efficacy and child feeding pressure to eat.  Web-based, 12-month intervention designed to address barriers to guideline implementation. It included the use of a web-based menu-planning program, educational resources reminder to increase compliance), training, and support by health promotion officer).
CCs from different economic backgrounds. Ethnicity/race was not presented.
In the intervention group vs. control, there was a significant increase in mean child consumption of fruit and dairy and a reduction in consumption of discretionary foods.
No significant differences were observed in diet quality, BMI z-scores, or child health-related quality of life.

Jastreboff et al. (2018) Name was not specified X
Intervention was developed for parents as the primary participants, but other adults participated: 95% (n = 59) were biological mothers, one was the biological father, one was an adoptive mother, and one was a grandmother.

Hammersley et al. (2019) Time2bHealthy
X X Intervention was developed for parents as the primary participants, but other adults participated: 93% of the adults were the children's mother, 5% were the father, and 2% were other. Intervention included a study partner selected by the mother. Approximately, half of the study partners (54.6%) were the infant's father, 27.5% were the infant's grandmother, 11.5% were another type of relative (infant's aunt, cousin, grandfather, sister or unspecified), and 6.4% were not non-relatives (mother's roommate, infant's godmother, or unspecified).

Interventions That Included Non-Parental Caregivers Post-Hoc
Six interventions [33,35,36,38,42,44] did not initially aim their intervention activities toward non-parental caregivers; however, these caregivers were included because they were involved in the participating child's care.
Two interventions of the six that involved non-parental caregivers post-hoc, two were conducted with AI participants [33,36] and included any caregivers as primary adult participants. In one of these two studies, approximately 85% of the enrolled caregivers were the children's mother, and the rest were the children's grandparent/other (12.7%) and father (2%). In the other study, the caregiver-child relationship was not described, though most of the caregivers were women (94.7%) [36].
In the remaining intervention of the six that involved non-parental caregivers posthoc [44], non-parental caregivers were included after the intervention started, since the investigators recognized that participating parents dealt with time limitations. The investigators included home visits to caregivers who were involved in the children's care, along with parents receiving the intervention at doctors' offices.
The remaining intervention included pregnant woman and study partner dyads [34]. The former was asked to identify a person who was involved in the child's care and was important in their decision-making about infant care. The study partners received their own set of educational material and participated in home sessions with the mothers.

Intervention Settings
The interventions that purposely included non-parental caregivers were primarily implemented at childcare centers or early care education (ECE) settings (n = 7) [32,37,[39][40][41]43,45] whereas most interventions that purposely did not include non-caregivers were implemented at home (n = 3) [33,36,42]. Only two of the 14 interventions included in this review were multi-setting: one was delivered in doctors' offices together with home visits [44], and the other targeted the preschool, home, and community settings [45]. The intervention settings are described in Table 3.

Intervention Results
Most of the interventions (n = 9) showed non-significant improvements in children's anthropometric outcomes. Of the five interventions that found positive anthropometric outcomes, three were childcare center based interventions with teacher, parent, and policy implementation components [32,40,41], one involved childcare staff and consisted of program integration into existing public health infrastructure [37], and the last one included a healthy lifestyle toolkit delivered by in-home mentoring (intervention) compared to receiving the material in the postal mail (control) for any adult caregivers that lived with the child [33]. The latter found results only when the control and the intervention participants were combined. Nonetheless, all 14 interventions had a positive impact on one or more behavioral outcomes, independent of the non-caregivers being primarily targeted.

Discussion
This scoping review found that there are very few obesity prevention interventions for children 0 to three years that include non-parental caregivers such as other family members or childcare providers. Although we found 125 articles that focused on interventions addressing child growth, height, BMI, or early childhood obesity risk, only 14 included non-parental caregivers. Despite the recognition of non-parental caregivers' involvement in childcare, we found that interventions continue to target primarily parents (specifically mothers). Of the 14 interventions that were analyzed in this review, only eight interventions included non-parental caregivers as part of the intervention design, whereas the remaining six interventions included non-parental caregivers later in the study in a post-hoc manner, as recruitment was flexible to include non-parental caregivers actively involved in the child's care. These findings show that when interventions are being developed, nonparental caregivers, who might have an important influence on children's health, are being overlooked.
Given the important role non-parental caregivers play in shaping health behaviors early in life, there is a need to develop interventions that actively involve them. By not doing so, prevention efforts may fail to capture the experiences of families with caregiving arrangements that do not rely solely or primarily on mothers. Families involving single parents, multi-generation, low-income, or same-sex parents as well as families that reside in multi-family households and low-income families could benefit from prevention efforts that address non-parental caregivers. Additionally, including non-parental caregivers in interventions along with parents can provide peer support; this can be particularly important among low-income families that bear the burden of social and economic factors on health outcomes [34,46]. Wasser et al. (2020) suggest that the definition of non-parental caregivers should extend beyond traditional notions of childcare involvement (e.g., grandmothers) and that interventions should consider the mothers' circumstances (e.g., single parent, full-time employee) when considering other non-parental caregivers in infant care and feeding [34].
Childcare providers were also identified as non-parent caregivers in four of the studies presented in this analysis. Childcare has been identified as an important setting for intervention programs due to the considerable amount of time that children spend in childcare and the association between the feeding behaviors of childcare providers and childhood obesity [21][22][23][47][48][49].
Most of the interventions included in this review did not find significant changes in anthropometric outcomes. This finding is similar to previous literature among children ages 0 to five whereby interventions that focus on diet or physical activity have not significantly impacted these outcomes [8]. These results may be due to the length of time between the intervention and the measurement of the outcomes [45], food insecurity [50], or short duration of the interventions [45], which would suggest that longer-term intervention may be needed. In this scoping review, almost half of the studies lasted less than three months [32,34,35,38,42,44], even though it is recommended that interventions for obesity prevention range between two to 12 months in length [51]. Interventions can also address responsive parental feeding practices, since they have been associated with reduced growthrelated indicators of obesity risk; this marks the importance of not only what is being fed, but also how and when feeding occurs [52,53].
Nonetheless, all interventions in this review were effective in improving other behavioral outcomes. These findings suggest that interventions may be effective in parent and non-parent caregiver behavior change in the context of childcare specific to child screen time, dietary intake, physical activity, sleep, stress, and parental feeding practices to improve anthropometric outcomes.
This review also points to the multiple environmental and social influences on childcare and the need for obesity prevention interventions to include community or environment-based strategies [54], especially among low-income communities [55]. Most studies in the review included racial/ethnic minority participants, which is incredibly valuable given the disproportionate burden of childhood obesity among racial/ethnic minority child populations [56]. However, attention to environmental and social influences on caregiving was limited; only one intervention in our review included community support for healthy behavior change among families [45]. In addition, interventions tend to focus on single settings and lack comprehensiveness [45]. Multi-setting interventions (e.g., the combination of childcare center and home environments) have shown more significant and beneficial results on weight-related outcomes compared to single-setting interventions [49]; however, in this scoping review, only two interventions included more than one setting [44,45]. As one study demonstrated, including both the preschool and home setting decreased sedentary behavior more than focusing only on the preschool setting, which implies that children benefit if there are also changes in the home setting [45]. This underscores the importance of targeting multiple environments, in particular, the home and preschool, as previously suggested [57].
One way this can be conducted is by employing principles of community-based participatory research (CBPR) in health research, which privileges the needs of patients, community members, and diverse stakeholders in research [58]. As we found in our review, two studies in engaged American Indian community members and other key stakeholders in decision making about the study design, incorporation of American Indian values and knowledge-based approaches, and development of intervention material [33,36].
This approach informed a more inclusive study design permitting non-parent caregivers to be recipients of the intervention. Another study [34] employed CBPR approaches and designed their research to be inclusive of the diversity of adults involved in infant care, which prompted mothers to invite the infant's father, grandmother, other relatives (e.g., aunts, siblings, cousins), or nonrelative adults. Similarly, another Such approaches demonstrate the value of CBPR and the importance of community and key stakeholder input into studies designed to address health disparities in early childhood obesity [59].
Several limitations of our review should be considered. First, articles in languages other than English or Spanish were not included in our analysis; this exclusion limits the generalizability of our findings to other countries. Second, although we conducted a systematic search in the database for broad coverage, our search strategy may not have identified all existing childhood obesity prevention interventions. Third, we selected the age range of 0 to three to capture early life childcare and feeding prior to the start of preschool, according to the definition of the Center for Disease Control [60]. Because of this focus, we included studies involving preschool age children only if the intervention targeted early life (i.e., 0 to three years), which overlaps with preschool years. It is possible that some articles that include interventions with two-to three-year-old children may have been excluded from our search. Fourth, studies in which non-parental caregivers' participation was unclear or unspecified may have been excluded from the analysis. Fifth, in some of the studies included in this review [33,35,36,38,42], it was not clear whether the non-parental caregivers were substitutes for the parents as the primary caregiving parent was unavailable, or if the non-parental caregivers were the primary caregiver. This distinction could help explore the implications for interventions where multiple caregivers belong to the same family or live in the same household as well as the dynamics between multiple caregivers and how care is managed among them. In addition, information about the non-parental caregivers was not described extensively in some studies, which was a limitation to describing their characteristics. Sixth, we only included interventions that included anthropometric measures as outcomes; some obesity prevention programs therefore might have been overlooked. Finally, the included studies were heterogeneous, and we did not evaluate the quality of the studies; therefore, our results are subject to bias due to the internal and external validity of the study results.

Conclusions
The scoping review provided a comprehensive overview of the state of research on early life obesity prevention interventions involving non-parental caregivers. Even though non-parental caregivers are commonly involved in the care of infants and toddlers, we found in this scoping review that only a few obesity prevention interventions for children between 0 and three years involved non-parental caregivers. The findings from this review highlight the need to target non-parental caregivers as participants in childhood obesity prevention interventions to better capture the multiple perspectives and competing interests that inform infant feeding practices and risk for early childhood obesity. In addition, this review can help inform the development of childhood obesity prevention interventions. The findings highlight the value of community-engaged approaches that can incorporate cultural values and knowledge-based approaches into interventions designed to address health disparities in early childhood obesity. For interventions to respond to the characteristics and needs of families, future research needs to better understand the role of non-parental caregivers and their dynamic with parents in infant and toddler feeding and early childhood obesity risk.