Increasing Fruit and Vegetable Consumption Among Preschool-Age Children and Their Families: Protocol for the Project V.E.G.G.I.E. Pilot/Feasibility Study

Introduction

Background

Food insecurity (FI), the inability to acquire enough food to support a healthy, active lifestyle [1], is a major health concern in the United States [2]. Recent prevalence estimates indicate that 17.9% of households with children experienced FI at some point in 2023 [1]. This is problematic as FI is related to poor mental (depression [3], anxiety [4]) and physical (obesity [5], diabetes [6]) health outcomes among adults. Children living in households experiencing FI are at an increased risk of behavioral, academic, and socioemotional problems [7] as well as poor physical health [8]. Further, individuals experiencing FI report lower diet quality than those classified as food secure. Adults living in households experiencing FI consume fewer fruits and vegetables (FVs) [9] and a higher amount of energy-dense foods with added sugars and high-fat dairy products [10]. Children in these same households have similar dietary patterns, consuming diets high in calories, sugar, and fat [11] and low in FVs [12], compared to their food-secure peers.

Within the context of FI, nutrition security refers to the healthfulness or nutritional adequacy of available foods [13]. FV consumption is one aspect of nutrition security as it is known to be relevant for the prevention of chronic disease [14,15]. Disparities in FV consumption among those experiencing FI exist. Among children aged 1 to 5 years living in households experiencing FI, only 41% ate a daily vegetable and 54% ate a daily fruit compared with those living in food-secure households (53% vegetable; 70% fruit) [16].

An underexplored reason for disparities in FV consumption among this age group is the importance of repeated exposure to novel/disliked foods (such as vegetables). Repeated exposure is consistently identified as necessary for children to accept a variety of foods; children may need up to 15 exposures before they “like” a new food [17]. However, it is common for parents to only provide children with a novel food 3 to 5 times before determining that their child does not like it [18] and no longer offering it. Lack of repeated exposure to novel foods may be especially problematic for households experiencing FI because these families cannot afford to endure costs associated with wasted food [19]. As such, families experiencing FI may rely on highly palatable foods that they know their children will accept, leading to the disparities in FV consumption. The importance of repeated exposure to novel/disliked food is highly relevant for preschool-age (2‐5 years old) children because this developmental period often coincides with an increase in food neophobia (fear of new foods) and picky eating. The onset of picky eating typically peaks around 2 years of age and persists until age 4 to 6 years [20,21], thus creating a vulnerable developmental period when parents may decrease offering foods their child is likely to reject [19].

Children are not only influenced by their parents’ feeding practices, but also the environment around them. Many families experiencing FI may send their children to home-based childcare settings as they are normally more affordable and flexible for parents [22]. These home-based childcare providers may be in smaller settings with fewer children, focusing on a family-like environment [22]. In these settings, children also lack exposure to new FVs as the families attending may not be able to afford that food waste in or out of the house.

One strategy to increase FV consumption among families experiencing FI is to provide them with access to FVs at no cost. Such support may enable parents to provide their preschool-age children with the repeated exposures necessary for them to learn to like and eat FVs. However, the efficacy of these initiatives is mixed [23,24]. One highly relevant study which provided discounted FVs detected changes in parents’ but not children’s diets [25], indicating that access alone may not be enough to influence children’s diets.

Many programs to increase FV access involve barriers to participation such as advanced purchasing/ordering and inconvenient pick-up times/locations [25]. We attempt to address this issue by delivering FVs to childcare providers, a convenient location for parents. Families can pick up their FVs when they pick up their child. Much of the literature leveraging childcare settings has focused on center-based care providers and failed to include home-based childcare settings [26]. However, many families with low resources may not work typical hours and use home-based childcare providers due to scheduling flexibility; they are the most common form of nonparental care for young children in the United States [27]. These providers have fewer resources than those in center-based childcare settings [27], making home-based childcare providers an ideal and underutilized location to provide resources to families with preschool-age children at risk of experiencing FI.

Providing parents with evidence-based strategies to improve their own feeding practices for their child(ren) alongside increased access to FVs may be necessary to impact children’s FV consumption. In line with the Family Systems Theory [28], educating parents on evidence-based parent feeding practices may enable them to endure children’s resistance when offered novel or disliked FVs and ultimately lead to greater child FV consumption. Parents are considered “nutritional gatekeepers” for their children because they have such an important influence on children’s diets. Parents determine not just what children eat, but also when, where, and how much food children eat [29]. Such education may be particularly relevant for families experiencing FI as FI may have detrimental impacts on parent feeding practices. Parents experiencing FI may restrict access to food when resources are scarce and pressure children to eat in the absence of hunger when food is available [30]. These practices disrupt children’s innate ability to self-regulate the amount they eat; they are related to problematic eating behaviors and obesity [30]. The Satter Division of Responsibility in Feeding is considered the gold standard for evidence-based parent feeding practices. According to this framework, both the parent and child play important roles within a healthy eating environment [31]. Parents are responsible for what food is served as well as the timing and location of meals and snacks; the child is responsible for how much (and whether) they eat from the options provided by the parent [31]. Unfortunately, when a family deviates from these responsibilities (such as the child determining what is served, or the parent determining how much the child must eat), unhealthy child eating behaviors, like picky eating, may occur [31]. In addition to the distinct roles of both parent and child, the Satter Division of Responsibility in Feeding recommends the following strategies to promote a healthy diet for children: repeated exposure to novel/rejected foods, structured meals and snacks with consistent rules/routines, and the importance of parental modeling of desired eating behaviors (eg, FV consumption) [31]. When these strategies are consistently implemented, child food preferences and consumption can be improved. However, families with low resources, such as those experiencing FI, may struggle to implement such education without additional material support. Therefore, to increase FV consumption, it is necessary to provide families with preschool-age children experiencing FI with both education on evidence-based parent feeding practices as well as additional access to FVs.

Objective

The primary objective of Project V.E.G.G.I.E. (Vegetable Eating Gets Going by Increasing Exposure) is to increase FV consumption among preschool-age children and their families at risk of experiencing FI. To achieve these goals, we propose providing families with increased access to FVs alongside education on evidence-based parent feeding practices (based on the Satter Division of Responsibility in Feeding). The goal of this study is to determine the feasibility and acceptability of the study protocol. Preliminary data will inform the format and contents of future iterations of the Project V.E.G.G.I.E. intervention as well as demonstrate feasibility, acceptability, and a preliminary signal of program efficacy. Figure 1 depicts the Project V.E.G.G.I.E. conceptual model and intervention targets.

‎

Methods

Design Overview

The Project V.E.G.G.I.E. Pilot/Feasibility Study was a cluster nonrandomized controlled trial (non-RCT). We originally planned for it to be a proof-of-concept study where all participants received the intervention. However, due to changes in the federal administration and potential changes in policy due to Department of Government Efficiency cuts [32], a control group was added to account for potential confounding variables (eg, potential cuts to social safety net programs such as the Supplemental Nutrition Assistance Program, the Special Supplemental Nutrition Program for Women, Infants, and Children, and childcare subsidies). Participants in both conditions completed assessments at 3 time points: baseline (prior to the start of the program), posttest (weeks 10‐11), and follow-up (4 weeks after the final box delivery). See Table 1 for the assessment schedule.

Ethical Considerations

The study protocol was approved by the institutional review board of the University of Florida (study ID: IRB202401344).

Participants

Recruitment of Childcare Providers

Parents and their preschool-age children were recruited from childcare providers (N=5) located in Orange County, Florida, in February/March 2025. The research team partnered with the Early Learning Coalition of Orange County (ELCOC) to identify partner childcare providers. The ELCOC’s mission is to provide all children in the Orlando area with fair access to quality learning environments. Childcare providers in the ELCOC network are Supplemental Nutrition Assistance Program eligible and accept childcare vouchers. ELCOC staff invited childcare providers they had successfully collaborated with on previous initiatives (eg, art programs) to participate in Project V.E.G.G.I.E. All contacted childcare providers agreed to participate in the research study (n=5, 100%). Originally, 3 childcare providers (n=2 home-based; n=1 center-based) were contacted to participate in the Project V.E.G.G.I.E. Pilot/Feasibility Study, with all participants receiving the program (proof-of-concept study). An additional 2 home-based childcare providers were later invited to serve as control sites due to concerns that potential changes in federal policy could impact key intervention outcomes (eg, food/nutrition security). Because of this, recruitment of control participants and their baseline assessments occurred a few weeks later. Interested individuals were informed of group allocation during informed consent.

Recruitment of Participants

Partner childcare providers shared the opportunity to participate in the research study with families enrolled at their site via study flyers (hard copies) featuring a QR code that linked to the informational website, which included informed consent documents and an online screener. Additionally, childcare providers directly shared the link to the study website via their normal communication methods (apps). The recruitment process for the control and intervention participants was identical; both groups were informed of their group allocation (intervention vs control) based on their childcare provider during the online informed consent process. Participants were informed of their eligibility/enrollment status and next steps by the research team (not the childcare providers) via text message.

Eligibility

Eligibility criteria for primary caregivers included being at least 18 years of age and the primary caregiver (at least 50% custody) of a 2- to 5-year-old child enrolled at a childcare provider participating in the research study. Primary caregivers had to be able and willing to complete surveys in either English or Spanish with smartphone access (to complete study assessments). Exclusion criteria for primary caregivers included not having a child within the target age range (eg, <2 or >5 years old) and having a child with a medical condition that significantly impacts their diet (parent reported).

Procedures

Study Timeline and Data Collection Overview

Parents completed assessments at 3 time points: baseline (prior to the start of the program), posttest (weeks 10‐11), and follow-up (4 wk after the end of the program). Parent surveys and daily diaries were distributed via text message with personalized links and completed by parents on their smartphones.

Intervention

The intervention group received the Project V.E.G.G.I.E. program, which included 6 boxes with 10 to 14 pounds of culturally familiar FVs biweekly delivered to their childcare provider for 10 weeks. Due to the university spring break schedule and corresponding lack of availability of undergraduate research assistants, 2 FV boxes were delivered in 2 consecutive weeks. During this time, parents also received education on evidence-based parent feeding practices.

FV Boxes

We partnered with Second Harvest Food Bank of Orange County to provide the FV boxes using their existing infrastructure. Undergraduate research assistants collected the FV boxes from Second Harvest and delivered them to the childcare providers. Research assistants remained at the childcare providers to facilitate distribution and collect process evaluation data (eg, number of boxes distributed). To reduce stigmatization, all families enrolled at intervention childcare providers received an FV box during each distribution, regardless of study enrollment or eligibility.

Educational Materials

Parent

Educational materials were provided in both English and Spanish. They focused on evidence-based parent feeding practices and included strategies to increase children’s FV consumption tailored to the unique needs of preschool-age children based on the Satter Division of Responsibility in Feeding [31]. Educational materials covered the following topics: normalizing child eating behaviors that are often frustrating for parents (such as inconsistent eating volume, refusing previously liked foods, and preferring snacks over meals), the importance of repeated exposure to FVs and role modeling, not pressuring children to eat certain foods, the distinct roles of parents and children at mealtimes (according to the Satter Division of Responsibility in Feeding), and noneating exposure to foods such as getting their child involved in food preparation. Based on findings from a needs assessment, education was delivered through print materials included with the FV boxes, as well as text messages (infographics) sent to parents’ smartphones.

Child

As an additional method of exposure [33], children also received print materials with activities and simple information about FVs. Children’s materials included topics like how various fruits and vegetables grow (carrots in the ground, apples in trees) and what animals eat vegetables, as well as a checklist to mark what FVs they received in their box that week. Children’s materials were not a primary intervention target; they were added to the research study because they were specifically requested by the ELCOC.

Control/Comparison Group

Participants in the control group completed the same study assessments (eg, parent surveys and daily diaries) as the intervention group. However, they did not receive any intervention materials until after the follow-up assessments were completed. Families in the control group received one FV box and all educational materials delivered at once via hard copies in early July 2025.

Assessments

Parent Surveys

The parent survey assessed the following domains: parent and child diet [34], FI [35,36], nutrition security [13,37], parent feeding practices [38], child eating behaviors [39,40], as well as contextual factors that may influence the main outcomes of interest. These included parent mental health [41], parenting stress [42], household disorganization [43,44], parenting skills [45], competence [46], and satisfaction [47]. All survey measures are described in Table 2. On average, participants completed the parent survey in 36.35 (SD 16.74) minutes. This information was extracted from Qualtrics survey metadata, which includes duration in seconds. Parents were texted up to 3 reminders to complete their parent survey at each time point.

Daily Diaries

At each time point, parents completed electronic daily diaries for a 7-day period to assess relevant day-level behaviors/contextual factors. Using modified/abbreviated versions of the cited scales, parents reported on day-level parent and child diet [48], access to food [51], parent feeding practices [38], and child eating behaviors [39,40]. Daily diary items and survey flow are included in Table 3. Each daily diary took parents an average of 5.33 (SD 2.93) minutes to complete (survey duration included in Qualtrics metadata). Automated text messages were sent at 8 PM with a personalized link specific to that day as a reminder to complete the daily diary. Daily diaries expired after 24 hours so that parents were only able to complete one daily diary per day. Parents who completed at least 5/7 daily diaries were considered compliant with the study protocol. Up to an additional 7 days of daily diaries were sent to participants who requested additional survey opportunities.

Exit Survey

The exit survey was embedded in the posttest parent survey for those who received the Project V.E.G.G.I.E. intervention. It was designed to assess acceptability and utility of the FV boxes and educational materials via the following domains: quantity, variety, quality, consumption, waste, familiarity, enjoyment, helpfulness, and program impacts. See Table 4 for exit survey measures and questions.

Exit Interviews

Both intervention and control participants were invited via text message to complete exit interviews about their experience with the study. Exit interviews were conducted in the participant’s preferred language (either English or Spanish) over the phone using a semistructured interview guide. Interviews took approximately 30 minutes, and questions assessed participants’ satisfaction with the Project V.E.G.G.I.E. program, including the ease of picking up the FV boxes at the childcare centers, the topics covered in the educational materials, as well as the quality and quantity of produce in the boxes. Participants in the control group were asked questions about their satisfaction with the study protocol and their one box of FVs. The semistructured interview guides for both groups are included as Multimedia Appendices 1 and 2. All exit interviews were recorded and transcribed using Otter.ai, an online transcription software. Transcriptions were imported into NVivo (version 15.3; Lumivero) software for data analysis, conducted by trained research assistants using an immersion crystallization approach (inductive) and constant comparison methodology.

Compensation

Parents received up to US $175 in Amazon e-gift card codes via text message. Gift card codes were distributed after completion of each assessment. Table 5 presents participant compensation by assessment and time point.

Statistical Analysis

Feasibility

Descriptive statistics of the feasibility outcomes will be presented as: recruitment (number of families who complete the screener), enrollment rate (proportion of families enrolled relative to the number of eligible families), retention rate (proportion of families who remain in the study across time points relative to the total number of participants), and adherence to the study protocol/missing data (proportion of parent surveys and daily diaries completed at each time point). We will evaluate intervention fidelity as delivery of FV boxes and educational materials. Delivery of FV boxes and printed educational materials will be directly assessed by onsite undergraduate research assistants. Educational materials sent via text message will be documented by a trained graduate research assistant. Participant barriers and facilitators to receiving the intervention (FV boxes; education) will be included in the exit interviews. In line with previous pilot/feasibility studies among similar populations (families with low resources), the protocol and intervention will be deemed feasible if enrollment, retention, and protocol compliance rates of at least 70% are achieved [58,61-64].

Acceptability

The proportion of participants who indicate they are satisfied with each facet of the program will be reported; the program will be considered acceptable to families if at least 70% of participants indicate they are satisfied/very satisfied across measures.

Progression Criteria

Progression to larger-scale trials will be determined jointly with community partners based on feasibility and acceptability metrics. It is expected that the intervention will be refined and the protocol modified prior to future iterations of the Project V.E.G.G.I.E. study in order to enhance recruitment, retention, acceptability, and intervention fidelity.

Efficacy

To assess an initial signal of program efficacy, we will assess whether those who received the intervention changed differently than those in the control group over time (time-by-treatment interaction). We will conduct a series of mixed-effects models, accounting for the nesting of repeated measures within participants with one analysis per outcome of interest: diet, FI, nutrition security, parent feeding practices, and child eating behaviors. Full information maximum likelihood estimation will be used to account for missing data and allow all available data to be included in the models. Standard effect size estimates (Cohen d) will be calculated as baseline to posttest and baseline to follow-up. Results will be interpreted with caution due to small sample size and limited statistical power. Due to the small number of childcare providers (n=5), analyses will not account for nesting of participants within sites. Demographic characteristics will be presented separately for those in the treatment and control conditions, and differences will be evaluated (independent samples t tests). We will consider and discuss detected differences between groups and their potential impacts on study outcomes. Analyses will be conducted in Stata 18.0 statistical software (StataCorp LLC).

Sample Size

This pilot/feasibility study consisted of a sample size of 20 participants recruited from 5 childcare providers. This sample size is sufficient to provide an initial signal of the feasibility and acceptability of the Project V.E.G.G.I.E. intervention [65]. A well-powered RCT is needed to evaluate study outcomes; as such, a power analysis was not necessary for this pilot/feasibility nonrandomized trial.

Results

The Project V.E.G.G.I.E. Pilot/Feasibility Study was internally funded by the Department of Health Education and Behavior in the College of Health and Human Performance at the University of Florida. Recruitment of participants (n=20) began in February and ended in March 2025. The study protocol and data collection took place from March to July 2025. Data cleaning and analyses are underway (January 2026), and the outcome paper for this pilot/feasibility study is expected to be submitted for publication in fall 2026.

Discussion

Overview

Ultimately, Project V.E.G.G.I.E. aims to increase FV consumption among preschool-age children and their families at risk of experiencing FI through enhanced nutrition security (increased FV access) and education tailored to the needs of parents of preschoolers. The underlying premise of Project V.E.G.G.I.E. is that children from families at risk of experiencing FI lack the necessary exposure to FVs, an unexplored factor contributing to inadequate FV consumption. Our study protocol overcomes the weaknesses of previous interventions using increased FV access. Programs that provide individuals with reduced- or no-cost FVs often involve many challenges, such as advanced purchasing/ordering and inconvenient pick-up times/locations. We aim to create a more convenient program by delivering FVs to childcare providers, an ideal time and location for parents of preschool-age children, thus increasing program uptake and use. In addition to increased access, we provided parents with educational materials focused on evidence-based parent feeding practices aimed at increasing child FV consumption. By targeting both FV access and parent feeding practices, we hypothesize that Project V.E.G.G.I.E. will result in improved program acceptability, treatment fidelity, and ultimately a more efficacious approach to increasing FV consumption among preschool-age children at risk of experiencing FI.

Strengths and Limitations

A key strength of this research study is the capable and experienced community partners who assisted with participant recruitment and intervention implementation: ELCOC and Second Harvest Food Bank of Central Florida. By leveraging the existing ELCOC network of childcare providers serving families with low resources, we were able to connect with 5 childcare providers eager to be involved with the pilot/feasibility study, as demonstrated by our 100% recruitment rate. Their large network (632 childcare centers and 31,761 children) [66] will enable us to recruit additional sites/participants for future iterations of the Project V.E.G.G.I.E. intervention.

The Second Harvest Food Bank of Central Florida is a private, nonprofit organization that has served the community for over 40 years. They collect, store, and distribute food to more than 750 feeding partners in seven Central Florida counties. In 2024, they distributed enough food for 82 million meals to partner programs [67]. Their “Bring Hope Home” initiative delivers boxes of food (including fresh FVs) to the homes of senior citizens and individuals with disabilities [68,69]. As such, Second Harvest was able to leverage this existing infrastructure to build the FV boxes for Project V.E.G.G.I.E. They have the resources and expertise to support Project V.E.G.G.I.E. as we scale up to a well-powered cluster RCT and provide support to additional families throughout central Florida.

This study is not without limitations. As a pilot/feasibility study, we are not adequately powered to evaluate intervention impacts or account for clustering within childcare providers. As such, between-group differences may be due to differences at the site level that cannot be disentangled from the intervention effects, limiting our ability to determine causation. A well-powered cluster RCT is necessary to evaluate the efficacy of the Project V.E.G.G.I.E. intervention. Because the intervention is implemented at childcare providers, it is necessary for all participants enrolled at a given site to receive the same treatment. This study design reduces the risk of contamination and is essential for our community partners. It would be unethical to supply food to only some families while withholding it from others who witness its distribution.

Our sample (n=5 childcare providers; n=20 parent/child dyads) recruited from Orange County, Florida, is not generalizable to the entire population of US families at risk of experiencing FI. However, there are a number of reasons why Central Florida is an ideal location to conduct this work. Orange County, Florida, is racially/ethnically diverse, with approximately 37% of the population categorized as Non-Hispanic White, 19% Non-Hispanic Black, and 33% Hispanic [70]. In 2023, 23% of residents were foreign born [70], and 13% of residents and 15% of residents with children were classified as below the Federal Poverty Line, slightly higher than the national average [70]. Further, although the Orlando area is classified as urban/suburban, the county also includes “Rural Residential Enclaves,” single-family neighborhoods with low residential densities similar to rural settings. Thus, this location will enable future Project V.E.G.G.I.E. studies to assess intervention efficacy and unique barriers/facilitators across varying levels of urbanicity/rurality.

As previously mentioned, the control group was not initially included in the study. It was added later due to potential changes in federal policy that could impact the target population. Because of this late addition, recruitment and baseline dates differed for the intervention and control, resulting in potential site-selection bias and temporal confounding.

Due to limited funding, this pilot/feasibility study relies on self/parent-report to evaluate all outcomes of interest, including diet, food and nutrition security, parent feeding practices, and child eating behaviors. Although there is error inherent in self/parent-reported measures, we attempt to improve data quality by collecting day-level data (improved recall) and using validated measures. In the future, we hope to address these weaknesses by conducting a well-powered cluster RCT that includes an objective measure of FV consumption (eg, VeggieMeter) [71-73].

Conclusions

This study will provide insight into the feasibility and acceptability of the intervention and study protocol. Feedback from participants will enable us to refine the Project V.E.G.G.I.E. intervention prior to scaling up to a well-powered cluster RCT. Ultimately, policy and research implications of this work include facilitating programs that provide families with preschool-age children with increased access to FVs, as well as education around evidence-based parent feeding practices. Integrating such programs into childcare settings may overcome barriers typically associated with initiatives to increase FV access.