Structural inequities and historical underinvestment in marginalized communities create developmental contexts that constrain access to high-quality education, healthcare, stable housing, and other critical resources. This study evaluates effects of a strengths-based, community-led intervention on young children and their families, which aims to buffer structural inequities while recognizing families' strengths. Between ages 18-36 months, English- and Spanish-speaking families consented and were randomly assigned to the intervention group (ParentChild+) or the active control group (FamilyNutrition+). Each group received 92 contacts from a specialist matched with their demographics. For the intervention group, contact focused on supporting parents and children's early learning, and families received a book or toy each week; for the control group, contact focused on supporting child nutrition, and families received a small food voucher each week. The current study evaluates whether the intervention altered parents' mental health, children's early environments, and/or children's test performance and brain development.
Supporting families with limited access to resources is essential for promoting children's health and well-being throughout the lifespan. Childhood socioeconomic status (SES), typically measured by parents' levels of education and income, is associated with a wide variety of childhood experiences including access to resources such as food and housing, exposure to stress, and access to educational opportunities. These experiences, in turn, pattern children's development. Here, the investigators partner with national nonprofit ParentChild+ to evaluate the effects of a strengths-based, community-led intervention on children's outcomes and parents' mental health and parenting behaviors. The current study compares children who completed the ParentChild+ intervention between 1.5-4 years of age to a matched, active control group who received the same amount of contact and resources related to healthy food (FamilyNutrition+). Specifically, 200 families in the Philadelphia area who were predominantly living well below the poverty line were recruited by ParentChild+ staff when children were between the ages of 18 and 36 months of age. Half of families were randomly assigned to receive the organization's 92-session home visiting intervention, and the other half were assigned to a 92-contacts active control focused on healthy nutrition. ParentChild+ Intervention. The ParentChild+ intervention consists of two 30-minute home visits per week for 46 weeks (92 visits), in which both parent and child are present. These visits may sometimes take place remotely, if best for the family. Home visitors are a family's designated Early Learning Specialist, chosen to match the demographics of the family they are assigned to (e.g., similar racial/ethnic, cultural, and/or linguistic background, living in or from the same neighborhood, etc.). A critical component of the study's design was the intentional incorporation of feedback and input from the Early Learning Specialists (ELSs) and coordinators, all of whom were selected to reflect the cultural and linguistic backgrounds of the families they served. This cultural and linguistic matching between ELSs and families is a core tenet of the ParentChild+ model, ensuring that families are able to build trust and engage meaningfully with the program. The ELSs' lived experience within these same communities informed the approach to participant outreach, engagement, and retention, both for the intervention and the control groups. Importantly, the control group, which received nutrition-related support, was also engaged by community members with similar cultural and linguistic ties, demonstrating a consistent commitment to equity and community empowerment across the study's design. An important part of the delivery model is that families and Early Learning Specialists have shared lived experiences. The same Early Learning Specialist works with the family for the duration of the 92 visits. The first session each week consists of the introduction of a new toy or book. These are chosen purposefully to reflect the culture, language, and community of the participants. The specific book or toy determined by the local ParentChild+; each is selected to be age-appropriate, about content that is interesting to the parent and/or child, and with the potential to introduce new vocabulary and/or opportunities for creativity and exploration, in addition to being free from stereotypes, representing people from diverse backgrounds, and including diverse authors, among other criteria. Each comes with an accompanying guide for engaging with it. The second session consists of working on techniques or strategies to use the introduced toy or book, or any combination of other materials in the home (at the parent's choosing). The goal of the sessions is to work from the parents' strengths to build their confidence in early literacy activities, strengthen their bonds with their children, and encourage children's explorations. Referring parents to other resources is also a primary aim of the project. ParentChild+ is well-regarded by the community, as evidenced in part by the retention of parents who completed the program as later Early Learning Specialists. FamilyNutrition+ Active Control. The control group was designed to mirror the intervention in terms of quantity of contact and resource distribution. Like ParentChild+, it consists of a 92-contact, 46-week program. However, instead of meeting with an early learning specialist, families in FamilyNutrition+ receive texts, emails, and zoom calls related to healthy nutrition and recipe ideas. In addition, families receive $25/month for groceries-a similar magnitude to the cost of the books and toys they would receive in ParentChild+. This a stringent control, as better nutrition does have the potential to nurture children's cognitive development. Design. Upon enrollment in the program, families are randomly assigned to a condition by ParentChild+ staff and fill out a questionnaire (T1 questionnaire). Upon completion of the program, families fill out another questionnaire (T2 questionnaire). If families dropped out during the 92-session period, they are still invited to fill out a T2 questionnaire at the time they would have completed the program. Next, when children turn 4 years old, they are invited into the lab to complete a behavioral session and a scan session. The current study asks whether families that were assigned to the intervention, relative to the control group, have differences in parental mental health, parenting, and child outcomes. To test the question about parental mental health, the investigators will run a linear mixed effects model that predicts the latent factor of parental mental health (from a factor analysis) from the interaction between timepoint (T1 and T2) and treatment (ParentChild+ or FamilyNutrition+), with a random intercept for participant. The investigators predict a significant interaction, such that the slope of change in parental mental health between T1 and T2 is more positive for families in the ParentChild+ condition than the FamilyNutrition+ condition. For the other questions, the investigators will repeat this analysis with measures of parenting and child outcomes in place of parental mental health. The current study will retain as many responses as possible even when families skip questions. For the measure of parenting (MAPS, described below), analyses with T1 data show that randomly removing 50% of responses to questions on the MAPS produced scores that were highly correlated with the full score (Pearson correlation \> .9). Thus, to preserve as much data as possible, the investigators will compute average scores from all questionnaires in which families have at least 50% of questions completed. For all analyses, the investigators will conduct the main analyses with an intent-to-treat sample: anyone randomized to the intervention or control group who was not excluded. The investigators will conduct sensitivity analyses with only those who completed at least 50% and 80% of the program, respectively, in line with ParentChild+ recommended dosage. Measures and exclusion criteria are described in more detail in the following sections.
Study Type
INTERVENTIONAL
Allocation
RANDOMIZED
Purpose
BASIC_SCIENCE
Masking
DOUBLE
Enrollment
199
92-session early literacy focused intervention
92-session active control focused on child nutrition
University of Pennsylvania
Philadelphia, Pennsylvania, United States
Parental mental health: anxiety
Parents will fill out the Generalized Anxiety Disorder 7 item scale (GAD-7), which asks seven questions about the extent to which parents have felt various symptoms of anxiety over the last two weeks (e.g., "not being able to stop or control worrying"). These are summed to create a score ranging from 0-21, with higher scores indicating more anxiety. This will be submitted to a factor analysis with the two other mental health scales to create a composite of mental health.
Time frame: Collected at age 1-2 and 3-4
Parental mental health: depression
The Patient Health Questionnaire Scale (PHQ) asks eight questions about the extent to which the parent has felt various symptoms over depression over the last two weeks (e.g., "little interest or pleasure in doing things"); responses are summed to create a score of depressive symptoms. These are summed to create a score ranging from 0-27, with higher scores indicating more depression. This will be submitted to a factor analysis with the two other mental health scales to create a composite of mental health.
Time frame: Collected at age 1-2 and 3-4
Parental mental health: perceived stress
The Perceived Stress Scale (PSS-10) asks 10 questions about stress parents experienced over the last month (e.g., "In the last month, how often have you felt that you were unable to control the important things in your life?"). Parents rate each question on a 5-point scale from "never" to "always;" responses are summed to create a score of perceived stress, ranging from 0-40, with higher scores indicating more perceived stress. This will be submitted to a factor analysis with the two other mental health scales to create a composite of mental health.
Time frame: Collected at age 1-2 and 3-4
Parenting behaviors
The multidimensional assessment of parenting scale (MAPS) asks parents to think about their parenting over the last two months and rate 34 behaviors on a 5-point scale from "never" to "always" (e.g., "I have warm and intimate times together with my child"). These are summed and averaged to produce parenting style factor scores for dimensions of hostility, physical control, lax control, positive parenting, positive reinforcement, warmth, supportiveness, broadband positive parenting, and broadband negative parenting. We will use broadband positive and negative parenting as our primary variables of interest; scores range between 1-5, with higher scores indicating more positive parenting and more negative parenting, respectively.
Time frame: Collected at age 1-2 and 3-4
Parent involvement in early learning
The Parent Involvement in Early Learning Scale (PIEL), was validated with a low-income sample and asks 25 questions about how frequently family members engage in various activities (e.g., "Bring the child on errands, like to the laundromat or grocery store") on a 4-point scale from "rarely" to "always." Responses are summed to scores of up to 100, to form a summary score of the family's involvement in learning activities for children, with higher scores indicating more parental involvement in early learning.
Time frame: Collected at age 3-4
Child vocabulary
The Macarthur Bates Communication Development Inventory (MCDI) asks about the words children says in any language, from a list of common words. Scores from all 100 words are summed to produce a score ranging from 0-100, with higher scores indicating higher vocabularies.
Time frame: Collected at age 1-2 and 3-4
Child behavioral problems
The Early Childhood Behavior Questionnaire (ECBQ) asks parents to rate how often children behaved a certain way in the last 2 weeks, on a 7-point scale from "never" to "always" (e.g., "When you were busy, how often did your child find another activity to do when asked?"). This scale produces three scores indicating attentional focusing, attentional shifting, and inhibition, ranging from 1-7, with higher scores indicating more behaviors on each trait.
Time frame: Collected at age 1-2 and 3-4
Child emotional problems
The Child Behavior Checklist (CBCL) asks parents to rate how well items describe their children now or over the last six months, on a 3-point scale from "Not true" to "Very true or often true." This produces a raw score and a t-score of children's internalizing and externalizing symptoms; t-scores typically range from 50-70, with the following clinical cut-offs: ≤ 59 indicates non-clinical symptoms, 60-64 indicates risk for problem behaviors, ≥ 65 indicates clinical symptoms.
Time frame: Collected at age 1-2 and 3-4
Child language performance
Children will complete the QUick Interactive Language Screener (QUILS) in their parents' primary language (English or Spanish). If Spanish, they will complete only the Spanish version of the bilingual QUILS. This measure calculates percentile ranks based on normative data, where higher scores indicate higher performance relative to norms.
Time frame: Collected at age 4-5
Child reasoning performance
Children will complete the matrix reasoning subtest of the Wechsler Preschool and Primary Scale of Intelligence (WPPSI-IV). This creates scaled scores based on a child's age, ranging from 1-19, where higher scores indicate higher performance on matrix reasoning.
Time frame: Collected at age 4-5
Child science learning performance
Children will complete the science subscale of Woodcock Johnson test (age: Preschool-Grade 1). This produces percentile ranks based on normative data, with higher scores indicating higher performance relative to norms.
Time frame: Collected at age 4-5
Child fluctuation amplitude
We will characterize the fluctuation amplitude of children's resting fMRI signal (obtained through pure rest, or through a task with its signal regressed out, depending on children's ability to complete the scan). We will compute the average power of low-frequency fluctuations (0.01-0.08 Hz) in the rs-fMRI signal, after preprocessing the data with fMRIprep, including susceptibility distortion correction, slice-timing correction, coregistration to the anatomical scan, resampling to standard space, and correction for 36 confounds including those related to physiological noise, motion, and their temporal derivatives.
Time frame: Collected at age 4-5
Child cortical thickness
Children will complete a T1-weighted MRI scan, from which cortical thickness will be derived using Freesurfer surface-based analysis tools. The distance between the white and pial surface is defined as the cortical thickness at each cortex location using Freesurfer. Surfaces will be checked and edited as needed. Surfaces of individual participants will next be resampled to a standard brain (fsaverage) and smoothed with a 15-mm full-width half-maximum kernel. We will extract the mean cortical thickness for each child. We will additionally extract regional-level cortical thickness using the Schaefer-100 parcellation.
This platform is for informational purposes only and does not constitute medical advice. Always consult a qualified healthcare professional.
Time frame: Collected at age 4-5
Child cortical surface area
Children will complete a T1-weighted MRI scan, from which cortical surface area will be derived using Freesurfer surface-based analysis tools. Surfaces will be checked and edited as needed. Surfaces of individual participants will next be resampled to a standard brain (fsaverage) and smoothed with a 15-mm full-width half-maximum kernel. We will extract the mean surface area for each child. We will additionally extract regional-level surface area using the Schaefer-100 parcellation.
Time frame: Collected at age 4-5
Child hippocampal volume
Children will complete a T1-weighted MRI scan, from which hippocampal volume will be derived using Freesurfer volume-based analysis tools. Segmentation of individual participants will be checked and edited as needed.
Time frame: Collected at age 4-5
Child molar eruption
We will measure molar eruption from T2-weighted scans. Molar eruption for the first four molars will be rated by a dental student, using the 4-point scale developed for our past work (1: unerupted; 4: fully erupted). Molar eruption rating will be averaged across the four molars for each participant.
Time frame: Collected at age 4-5
Child language-related brain activation
Children complete a block-design task in the scanner, in which they listen to blocks of Sesame Street characters speaking. Half of the blocks are normal, forward speech and half of the blocks have the speech reversed. Children complete this task in their parent's native language (English or Spanish). Language-related brain activation will be defined as activation to the forward relative to backwards speech, within six ROIs constituting the search space.
Time frame: Collected at age 4-5