pediatrics
February 2016, VOLUME137 /ISSUE 2

# Promotion of Positive Parenting and Prevention of Socioemotional Disparities

2. Carolyn Brockmeyer Cates, PhDa,
3. Benard P. Dreyer, MDa,
4. Samantha Berkule Johnson, PhDa,b,
5. Harris S. Huberman, MD, MPHa,c,
6. Anne M. Seery, PhDa,
7. Caitlin F. Canfield, PhDa, and
8. Alan L. Mendelsohn, MDa
1. aDepartment of Pediatrics, New York University School of Medicine/Bellevue Hospital Center, New York, New York;
2. bDepartment of Psychology, Marymount Manhattan College, New York, New York; and
3. cDepartment of Pediatrics, SUNY Downstate Medical Center, Brooklyn, New York
1. Dr Weisleder drafted the initial manuscript, conceptualized and designed the study, conducted statistical analyses, and reviewed and revised the manuscript; Dr Mendelsohn conceptualized and designed the study, conducted statistical analyses, designed the data collection instruments, and reviewed and revised the manuscript; Drs Seery and Canfield conducted statistical analyses and critically reviewed the manuscript; Drs Brockmeyer Cates and Berkule Johnson conceptualized and designed the study, designed the data collection instruments, coordinated and supervised data collection, and critically reviewed the manuscript; and Drs Dreyer and Huberman conceptualized and designed the study and critically reviewed the manuscript. All authors approved the final manuscript as submitted.

## Abstract

OBJECTIVE: The goal of this study was to determine what effects pediatric primary care interventions, focused on promotion of positive parenting through reading aloud and play, have on the socioemotional development of toddlers from low-income, primarily immigrant households.

METHODS: This randomized controlled trial included random assignment to 1 of 2 interventions (Video Interaction Project [VIP] or Building Blocks [BB]) or to a control group. Mother–newborn dyads were enrolled postpartum in an urban public hospital. In VIP, dyads met with an interventionist on days of well-child visits; the interventionist facilitated interactions in play and shared reading through provision of learning materials and review of videotaped parent–child interactions. In BB, parents were mailed parenting pamphlets and learning materials. This article analyzes socioemotional outcomes from 14 to 36 months for children in VIP and BB versus control.

RESULTS: A total of 463 dyads (69%) contributed data. Children in VIP scored higher than control on imitation/play and attention, and lower on separation distress, hyperactivity, and externalizing problems, with effect sizes ∼0.25 SD for the sample as a whole and ∼0.50 SD for families with additional psychosocial risks . Children in BB made greater gains in imitation/play compared with control.

CONCLUSIONS: These findings support the efficacy of VIP, a preventive intervention targeting parentchild interactions, for enhancing socioemotional outcomes in low-income toddlers. Given the low cost and potential for scalability of primary care interventions, findings support expansion of pediatric-based parenting programs such as VIP for the primary prevention of socioemotional problems before school entry.

• Abbreviations:
ANOVA
analysis of variance
BASC-2
Behavior Assessment System for Children–Second Edition
BB
Building Blocks
CI
confidence interval
ES
effect size
ITSEA
Infant–Toddler Social and Emotional Assessment
RCT
randomized controlled trial
ROR
VIP
Video Interaction Project
• #### What’s Known on This Subject:

Differences in socioemotional development are an important contributor to poverty-related disparities in school readiness. It is unknown whether pediatric interventions seeking primary prevention of disparities by promoting parent–child interactions through reading aloud and play can enhance socioemotional outcomes.

This randomized controlled trial demonstrated the effects of pediatric-based interventions focused on promoting positive parenting through reading and play on socioemotional outcomes of toddlers from low-income families. Findings support the efficacy of a primary preventive intervention with high potential for scalability.

Poverty-related disparities in child development and school readiness emerge early in childhood and widen over time,13 placing disadvantaged children at risk for lower educational attainment.4,5 Addressing these disparities is a national priority, as evidenced by American Academy of Pediatrics policy and recent federal initiatives.6,7 The present article examines the potential for pediatric-based primary preventive interventions to reduce poverty-related disparities in school readiness by enhancing children’s socioemotional development.

Socioemotional development is a core component of school readiness, affecting children’s school adjustment and academic performance.811 Critical socioemotional factors associated with children’s transition to school and long-term academic trajectories include attention, social competence, and behavior.8,11,12 Although positive parenting plays a critical role in supporting children’s socioemotional development,13,14 poverty is associated with reductions in responsive parent–child interactions.1517 Promotion of positive parenting during early childhood is a promising avenue for enhancing socioemotional development in low-income children in general, and it may be especially important for children at increased risk due to family psychosocial factors.18,19

Pediatric primary health care provides a significant opportunity for population-level, primary prevention of school readiness disparities through promotion of positive parenting (eg, using the Reach Out and Read [ROR] and Healthy Steps programs),2022 given the frequency and near universality of well-child visits and the opportunity to leverage existing health care infrastructure. Previous studies have shown that this platform can be effectively used to enhance socioemotional development through promotion of positive parenting in children with already-identified behavioral problems (eg, using the Incredible Years program).2325 However, it is unknown whether pediatric interventions seeking the primary prevention of disparities through promotion of parent–child interactions in reading aloud and play can also enhance socioemotional development.

We addressed this gap through a study of 2 pediatric, primary care–based preventive interventions: the Video Interaction Project (VIP) and Building Blocks (BB). VIP was designed as an enhancement to ROR, adding an interventionist who uses videorecordings of the parent and child interacting to promote parents' self-reflection and encourage positive parenting behaviors.26 Previous studies of VIP have shown beneficial effects on parent–child interactions and reduced family psychosocial stressors, including maternal depression and parenting stress.2730 BB is a lower intensity intervention that uses mailed newsletters highlighting positive parenting strategies. BB has also been shown to enhance responsive parenting and reduce maternal depressive symptoms, although the effects are smaller.27,28

The present article describes findings from a randomized controlled trial (RCT) examining the effects of VIP and BB on parenting, child development, and school readiness. The present analyses examined effects of VIP and BB on children’s socioemotional development from 14 to 36 months, a primary outcome of the RCT. We hypothesized that: (1) socioemotional outcomes would be improved in VIP and BB compared with the control group, with stronger effects of VIP given its higher intensity and previous evidence of greater impacts; and (2) intervention effects would be larger in families with increased psychosocial risk.

## Methods

### Study Design

A single-blind, 3-way RCT entitled the Bellevue Project for Early Language, Literacy and Education Success (BELLE) was conducted at an urban public hospital serving low-income families (Bellevue Hospital Center).27 Consecutive enrollment of mother–child dyads occurred in the postpartum unit, with 675 dyads enrolled. Approval was obtained from the Institutional Review Board at the New York University School of Medicine, Bellevue Hospital Center, and the New York City Health and Hospitals Corporation. The initial study design included follow-up of all groups at 6, 14, 24, and 36 months. Financial constraints, however, led to limited follow-up of BB families at 24 months and no follow-up at 36 months.

### Randomization Groups

After enrollment, dyads were randomized to VIP, BB, or control by using a random number generated by the project director by using Microsoft Excel 2003 (Microsoft Corporation, Redmond, WA). Group assignments were concealed from staff and study participants until enrollment was completed.

#### VIP

VIP sessions occur at regularly scheduled pediatric visits beginning in the first month, totaling 15 possible sessions through age 3 years. An interventionist meets with families in 1-on-1 sessions for 25 to 30 minutes. Parent–child dyads are videorecorded during 5- to 7-minute play/shared reading interactions by using a developmentally appropriate toy and/or book provided by the program. The interventionist reviews the video along with the parent to identify and reinforce responsive interactions and promote parent self-reflection. The video is given to the parent to promote generalization of identified behaviors in the home; pamphlets provide suggestions related to positive parenting during play, reading, and daily routines, as well as opportunities for parents to develop their own plans for interacting with their child. When possible, the same interventionist meets with each family at each session. Interventionists typically have bachelor’s degrees in fields related to young children and receive training and supervision by Bellevue Project for Early Language, Literacy, and Education Success Project leadership. VIP has an estimated cost of $150 to$200 per child per year.26

#### BB

BB utilizes mailed information and learning materials to improve parent self-efficacy and promote positive parenting. Each month, parents are mailed a toy or book, along with a newsletter that provides information on encouraging learning and ideas for interactions around a specific developmental goal. Parents are also asked to complete Ages and Stages developmental questionnaires31 every 4 to 6 months. BB has an estimated cost of $75 to$100 per child per year.

#### Control

Control families received standard pediatric care, including recommended anticipatory guidance and monitoring. Standard pediatric care included ROR for all groups.

### Participants

Enrollment occurred between November 2005 and October 2008, as previously described.27 Consecutive mother–infant dyads meeting inclusion criteria and providing informed consent were enrolled. Inclusion criteria were: plans to receive pediatric care at Bellevue Hospital Center; term birth; no significant medical complications or eligibility for early intervention at birth; and a mother who was the primary caregiver, aged ≥18 years, spoke primarily English or Spanish, and was able to maintain contact with the program.

### Measures

#### Sociodemographic Characteristics

Information on sociodemographic characteristics was collected during the postpartum period through interviews with the mother, as previously described.27 At the 6-month assessment, maternal literacy was assessed in the mother’s preferred language by using the Woodcock-Johnson III/Woodcock-Muñoz Bateria III Letter-Word Identification Test32; for cases with missing literacy data, educational level was used as a proxy.

Families were considered to be at increased psychosocial risk if the mother reported ≥1 of the following: being a victim of violence, homelessness, involvement with child protective services, significant financial hardship, food insecurity, cigarette smoking or alcohol use during pregnancy, or history of previous mental illness.

#### Dependent Variables

At 14 and 24 months, socioemotional outcomes were assessed via mother interviews by using 3 subscales from the Infant–Toddler Social and Emotional Assessment-Revised (ITSEA)33: imitation/play, attention, and separation distress. These subscales were selected to assess key dimensions of socioemotional development at this age, including social skills, attention, and behavior. The ITSEA includes items describing behaviors that the parent rates as “not true/rarely,” “somewhat true/sometimes,” or “very true/often,” with scores ranging from 0 to 2. It is available in English and Spanish and has been validated in children aged 12 to 36 months. At the 14-month assessment, the ITSEA was not included in the protocol until ∼4 months into the process; thus, although the 14-month assessment was performed for 327 families, the ITSEA was collected for 242 families at this age.

At the 36-month assessment, interviews were conducted with mothers by using 4 subscales from the Parent Rating Scales of the Behavior Assessment System for Children–Second Edition (BASC-2)34: social skills, attention problems, hyperactivity, and aggression. The BASC-2 has been normed for use in English and Spanish. T scores (mean ± SD: 50 ± 10) were calculated for each subscale, and a composite score was computed for externalizing problems (hyperactivity and aggression). Clinically at-risk behavior was defined based on the manual as a T score ≥60 for attention problems, hyperactivity, aggression, and externalizing problems; a T score ≤40 was used for social skills.

### Statistical Analysis

A total of 225 families were enrolled per group, providing 80% power to detect a minimum effect size (ES) of 0.33 SD, assuming 33.3% attrition by age 3 years. Statistical analyses were performed based on an intention-to-treat model. Analysis of variance (ANOVA) and χ2 tests were used to compare sociodemographic characteristics between randomization groups and to compare families who contributed data on at least 1 time point versus those who did not.

For 14- and 24-month outcomes, ANOVA was used, followed by Tukey post hoc testing, to compare ITSEA mean scores between randomization groups. Socioemotional development was then compared over time by using multilevel modeling with Stata SE 12 (Stata Corp, College Station, TX); this approach allowed us to test group differences while accounting for repeated measures. Separate multilevel modeling was performed for each subscale. We first calculated a model with age and group predicting ITSEA mean scores; a second model was then computed that included group-by-age interaction terms. Multilevel models adjusted for all baseline sociodemographic characteristics; child age was included as a random effect. In the main effects models, coefficients represent the difference in scores for VIP or BB compared with control across the entire 14- to 24-month period; in interaction models, coefficients represent the difference in slope (change from 14 to 24 months) for VIP or BB compared with the control.

For 36-month outcomes, t tests were used to compare BASC-2 T scores between VIP and the control. Findings for ES were calculated by using Cohen’s d, which reflects mean differences between groups in SD units. We performed χ2 analyses to compare proportions of children with clinically at-risk behaviors. To assess whether intervention effects were more pronounced for higher risk families, subgroup analyses were performed comparing VIP and control children of mothers at increased psychosocial risk. To examine dose–response effects, piecewise regressions were performed with the number of VIP sessions completed as predictor variables.

## Results

### Study Sample

A total of 463 (69%) of 675 families contributed data on socioemotional outcomes at ≥1 assessment point (analytic sample) (Fig 1), including 78% of VIP and control participants and 49% of BB participants (BB follow-up was lower due to financial constraints, as described earlier). Descriptive statistics for dyads in the analytic sample are provided in Table 1. Mothers were primarily Hispanic/Latino and born outside the United States; 34% of families were at increased psychosocial risk. VIP families completed a median of 9.5 sessions; 67% of families completed one-half or more sessions, whereas 15% completed fewer than one-third (5 sessions).

FIGURE 1

Participant flowchart.

TABLE 1

Demographic Characteristics of Families in the Analytic Sample

Dyads contributing data did not significantly differ from those who did not contribute data for maternal age, literacy, psychosocial risk, or birth order. However, mothers contributing data were less likely to have graduated high school or be living with a partner, and they were more likely to be Hispanic/Latina, immigrants, of lower socioeconomic status, and speak Spanish as their primary language (all P < .01); children in the analytic sample were more likely to be female (P < .05). Importantly, there were no differences in sociodemographic characteristics according to randomization group for participants in the analytic sample (Table 1).

### Primary Analyses

Table 2 displays the socioemotional outcomes as a function of group at the 14- and 24-month assessments. At 14 months, ANOVA tests showed significant differences between groups for imitation/play (P < .01) but not for attention (P = .75) or separation distress (P = .20). Tukey’s honestly significant difference post hoc testing at 14 months found that VIP children had higher imitation/play scores than BB children (P < .01), with an ES based on Cohen’s d of 0.49; post hoc comparison of VIP and the control did not reach statistical significance (ES: 0.31; P < .1). At 24 months, ANOVA tests showed differences between groups for attention (P < .05); differences in imitation/play (P = .30) and separation distress (P = .07) did not reach statistical significance. Post hoc testing at 24 months showed that VIP children had significantly higher attention scores than control subjects (ES: 0.30; P < .05). There were no significant effects of BB at 14 or 24 months.

TABLE 2

Effects of VIP and BB on Socioemotional Outcomes at 14 and 24 Months: Mean ± SD Scores for ITSEA Subscales

Figure 2 displays the trajectories of intervention effects on socioemotional development from 14 to 24 months. Results of multilevel modeling (Table 3) found a significant main effect of VIP for imitation/play (P < .05) and separation distress (P < .05), indicating that VIP children were more advanced in imitation/play and had lower separation distress than control children throughout this period. The main effect of VIP on attention did not reach significance (P = .08); however, there was a significant positive interaction between VIP and age (P < .05), indicating that VIP children made greater gains in attention between 14 and 24 months than control children. There were no significant main effects of BB. However, there was a significant positive interaction between BB and age for imitation/play (P < .05), indicating that BB children made greater gains in imitation/play between 14 and 24 months than children in the control group.

FIGURE 2

Trajectories of mean scores on (A) imitation/play, (B) attention, and (C) separation distress from 14 to 24 months for children in each of the 3 randomization groups; the y-axis displays the predicted values based on multilevel models.

TABLE 3

Multilevel Model Results: Predictors of ITSEA Mean Scores at 14 to 24 Months

Table 4 displays socioemotional outcomes at 36 months for the sample as a whole and for families at increased psychosocial risk. VIP children had significantly reduced hyperactivity (ES: 0.27; P < .05) and externalizing problems (ES: 0.27; P < .05) compared with control children. Although VIP children also tended to have lower aggression scores, this finding did not reach statistical significance (ES: 0.20; P < .1) for the sample as a whole. Among higher risk families, VIP children had lower aggression scores than those in the control group (ES: 0.48; P < .05), as well as reduced hyperactivity (ES: 0.55; P < .01) and externalizing problems (ES: 0.57; P < .01).

TABLE 4

Effects of VIP at 36 Months for the Entire Sample and for the Subgroup of Families at Increased Psychosocial Risk: Mean ± SD BASC-2 T Scores

Table 5 presents the proportion of children that reached thresholds for clinically at-risk behaviors on the BASC-2 at 36 months. No statistically significant differences were found in the proportion of children with at-risk scores for the sample as a whole. Among families at increased psychosocial risk, VIP children were significantly less likely than control children to be at risk for hyperactivity (15% vs 41%; P < .01) and externalizing problems (11% vs 30%; P < .05); these findings corresponded to an absolute risk reduction for VIP of 26% (95% confidence interval [CI]: 8%–43%) for hyperactivity and 18% (95% CI: 2%–34%) for externalizing problems, and a relative risk reduction for VIP of 63% (95% CI: 23%–82%) for hyperactivity and 62% (95% CI: 8%–84%) for externalizing problems. Among families at higher psychosocial risk, 4 VIP families (95% CI: 2–12) would need to receive VIP to prevent 1 child from being in the at-risk category for hyperactivity, and 5 (95% CI: 3–44) would need to receive VIP to prevent 1 child from being in the at-risk category for externalizing problems. Among higher risk families, VIP children tended to be less likely than control children to be in the at-risk category for aggression (8% vs 20%) and social skills (0% vs 7%), but these findings did not reach statistical significance (P < .1)

TABLE 5

Effects of VIP at 36 Months for the Entire Sample and for the Subgroup of Families at Increased Psychosocial Risk: Proportion of Children Reaching Threshold for Clinically At-Risk Behaviors on BASC-2 Subscales

Within-group analyses were conducted of dose effects for families randomized to receive VIP. In piecewise regressions, a dose–response was seen for each additional visit completed beyond the first 5 visits for hyperactivity (B = 1.00, β = –0.26, P < .05) and externalizing problems (B = 0.80, β = –0.24, P < .05). This outcome is equivalent to a reduction in T score of 1 unit for every visit beyond the fifth visit for hyperactivity and of 0.8 U per visit for externalizing problems. There was no significant change in T scores related to completion of the first 4 visits (hyperactivity: β = –0.10, P = .69; externalizing: β = –0.14, P = .58).

## Discussion

The present study found that low-cost, pediatric parenting interventions beginning in early infancy and delivered before the identification of behavior problems can have meaningful effects on socioemotional development during the toddler period. These findings are important because of increasing evidence that socioemotional capacities are critical to school readiness and predict a range of adult outcomes across domains of education, employment, criminal activity, substance use, and mental health.8,10,11,35,36

Overall, we found strong effects of VIP on children’s socioemotional outcomes, with more limited impacts of BB. In particular, VIP produced positive effects on key aspects of socioemotional development beginning at 14 months and extending through 36 months, including enhanced imitation and attention, and reduced separation distress, hyperactivity, and externalizing problems. The presence of effects across multiple time points and domains of socioemotional development provides strong support for VIP’s impacts during the toddler period. We did not document effects of BB at any single time point; however, BB infants made greater gains in imitation/play from 14 to 24 months compared with control infants, suggesting that there may be some positive effects of this very low intensity intervention. Stronger effects for VIP compared with BB are consistent with the idea that specific characteristics of VIP, such as the relationship with the interventionist and use of videotaping to promote self-reflection, are important for achieving these outcomes. An important implication of our work is that relationship-based interventions should be strongly considered as policy makers seek to address school readiness disparities.

The ES findings for VIP were within the range of those found for interventions using other platforms (eg, home visiting, community-based programs) that tend to have higher costs.37,38 These averaged 0.25 SD for the sample as a whole and were even greater (0.50 SD) for children in families with additional psychosocial risks. For higher risk families, the relative risk reduction for clinically at-risk hyperactivity and externalizing behaviors due to VIP was >60%, which, together with the low number needed to treat (4 and 5, respectively), suggest potential for substantial benefits.

One of the most important scientific and policy implications of the present study is that interventions focused on promotion of activities such as reading aloud and play may be effective in enhancing socioemotional development, in addition to their well-documented role in enhancing cognitive-language outcomes.39,40 Indeed, most existing parenting interventions shown to affect socioemotional outcomes in young children include direct teaching of discipline strategies.23,41 Notably, VIP affects socioemotional development without teaching these strategies, instead focusing on promotion of responsive parent–child interactions in the context of activities such as reading aloud and play that are typically considered to be cognitively oriented.

This study had 3 main limitations. First, because of resource limitations, there was no follow-up of the BB group past 24 months; findings related to BB should therefore be interpreted with caution. Second, socioemotional outcomes were based on parent report, which, while obtained by using reliable and standardized instruments, can be subject to bias. Third, participating mothers were primarily first-generation Hispanic/Latino immigrants; results may not be generalizable to families with other sociodemographic characteristics.

## Conclusions

This study showed that pediatric primary care interventions focused on promotion of positive parenting through reading aloud and play can enhance socioemotional development among children in poverty. Given the potential for low cost and population-level reach of primary care interventions, these findings suggest that the pediatric platform should play an important role in primary prevention of poverty-related disparities in school readiness.

## Acknowledgments

We are grateful to the many individuals who contributed to this project, including Melissa Acevedo, Jenny Arevalo, Nina Burtchen, Diego Catalan Molina, Laura Curiel Duran, Aida Custode, Hannah Goldman, Pamela Kim, Jennifer Ledesma, Jennifer Elizabeth Lee, Maya Matalon, Andrea Paloian, Caroline Raak, Daniela Romero, Casilda Suárez-Hesketh, Melissa Tunik, Jessica Urgelles, Kristina Vlahovicova, Linda Votruba, Caroline Wilkes, Lisa White, Margaret Wolff, and Brenda Woodford. We would especially like to thank the parents and children who participated.

## Footnotes

• Accepted November 16, 2015.
• Address correspondence to Adriana Weisleder, PhD, Department of Pediatrics, New York University School of Medicine, 550 First Ave, OBV A529, New York, NY 10016. E-mail: adriana.weisleder{at}nyumc.org
• This trial has been registered at www.clinicaltrials.gov (identifier NCT00212576).

• FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.

• FUNDING: Supported by grants R01 HD047740 (Principal Investigator, Dr Mendelsohn) and R01 HD40388 (Principal Investigator, Dr Huberman) from the National Institutes of Health/National Institute of Child Health and Human Development; the Tiger Foundation; the Marks Family Foundation; Children of Bellevue, Inc; KiDS of NYU Foundation, Inc; the Rhodebeck Charitable Trust; and the Academic Pediatric Association Young Investigator Award Program. Dr Weisleder was supported in part by a National Research Service Award from the Health Resources and Services Administration (T32 HD047740), with training supported in part by a New York University Clinical and Translational Science Award (UL1TR000038) from the National Institutes of Health National Center for the Advancement of Translational Science. Funded by the National Institutes of Health (NIH).

• POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.