Key Points
Question
Can positive effects of an early-life behavioral intervention designed for obesity prevention be sustained into middle childhood?
Findings
In an observation-only, long-term follow-up of 232 participating primiparous mother-child dyads from a randomized clinical trial conducted from 2012 to 2017, the mean body mass index from ages 3 to 9 years was significantly lower in the responsive parenting intervention group compared with controls (116 per group, 16.59 vs 16.95; absolute difference, −0.36). However, the intervention’s effect appeared to dissipate over time.
Meaning
These findings suggest that responsive parenting interventions during early life can improve weight-related outcomes, but the diminishing effect after intervention cessation suggests that a life-course approach may be required to sustain benefits for obesity prevention.
Abstract
Importance
Behavioral interventions to treat childhood obesity have had limited success. Primary prevention is desirable, but whether intervention effectiveness can be sustained is unknown.
Objective
To examine the effect of an intervention designed for the primary prevention of obesity and delivered through age 2 years on weight outcomes through age 9 years.
Design, Setting, and Participants
A longitudinal observation of a single-center randomized clinical trial comparing a responsive parenting intervention vs a home safety intervention (control) among primiparous mother-child dyads who completed the assessment at age 3 years with follow-up to age 9 years. All data were analyzed from January 21 to November 15, 2024.
Interventions
Research nurses conducted 4 home visits during infancy and research center visits at ages 1 and 2 years totaling less than 10 contact hours. The responsive parenting curriculum focused on feeding, sleep, interactive play, and emotion regulation.
Main Outcomes and Measures
The primary outcome is body mass index (BMI; calculated as weight in kilograms divided by height in meters squared) across 4 assessments from age 3 through 9 years, with the assessment of study group differences using repeated-measures analysis. A test for an interaction between sex and study group was planned. Secondary outcomes include BMI z scores and prevalence of overweight (BMI ≥85th to <95th percentile) and obesity (BMI ≥95th percentile) at 5, 6, and 9 years.
Results
Of the 232 primiparous mother-child dyads (116 per group) (7 Asian [3%], 11 Black [5%], 1 Native Hawaiian or Other Pacific Islander [0.4%], 207 White [89%], and 6 children with other race and ethnicity [including Asian, Indian, Hispanic, Dominican, and other race; 2.5%]; 121 male children [52%]), 177 (76%) had anthropometric data at age 9 years. From ages 3 to 9 years, children in the responsive parenting group had a lower mean (SE) BMI than controls (16.59 [0.18] vs 16.95 [0.18]; absolute difference, −0.36; P = .049). Sex moderated this effect; female participants in the responsive parenting group had a lower mean (SE) BMI than female participants in the control group (16.31 [0.23] vs 17.24 [0.22]; absolute difference, −0.93; P = .002), with no group differences among male participants. Cross-sectional analyses revealed no differences in BMI z scores or prevalence of overweight or obesity at ages 5, 6, and 9 years between the responsive parenting group and the control group.
Conclusions and Relevance
An early-life responsive parenting intervention resulted in lower BMI from age 3 to 9 years compared with a control intervention. This group difference was driven by effects on female participants, with differences appearing to dissipate over time. A life-course approach may be required to sustain the benefits of early-life responsive parenting interventions for obesity prevention.
Trial Registration
ClinicalTrials.gov Identifier: NCT03555331
This observation-only, long-term follow-up of a randomized clinical trial examines the effect of an intervention designed for the primary prevention of obesity and delivered through age 2 years on weight outcomes through age 9 years.
Introduction
In the US, 23% of 2- to 5-year-old children are already overweight,1 and 9% already have obesity.2 By ages 6 to 11 years, the prevalence increases to 34%1 and 17%,2 respectively, highlighting the need for early-life preventive interventions with sustainable effects. In today’s obesogenic environment, however, attempts at primary prevention of childhood obesity have been largely unsuccessful.3 Indeed, the 2017 US Preventive Services Task Force recommendations on Obesity in Children and Adolescents called for more studies assessing behavioral interventions among children aged 5 years or younger.4 Encouragingly, responsive parenting-informed interventions have shown promise for obesity prevention in early life5,6 and are endorsed in the Robert Wood Johnson Foundation’s Healthy Eating Research guidelines to prevent rapid infant growth and toddler overweight,5 although, to our knowledge, there have been few investigations reporting long-term outcomes for these interventions.7
We previously reported the positive effects of a responsive parenting intervention designed for the primary prevention of obesity on body mass index (BMI; calculated as weight in kilograms divided by height in meters squared) through age 3 years in the Intervention Nurses Start Infants Growing on Healthy Trajectories (INSIGHT) study. INSIGHT was a randomized clinical trial focused on reducing rapid weight gain in infancy and establishing healthy growth trajectories during early life.8,9 This report focuses on evaluating the long-term effects on weight status through age 9 years among children in the INSIGHT cohort.
Methods
Participants
The study was approved by the Human Subjects Protection Office of the Penn State College of Medicine. All participating mothers provided written informed consent. Mothers and their newborns were recruited after delivery from 1 maternity ward (Penn State Milton S. Hershey Medical Center, Hershey, Pennsylvania) (Figure 1). The enrollment period spanned from January 2012, to March 2014; 3-year assessments occurred between January 2015 and April 2017. The final 9-year assessment occurred in September 2023. Major eligibility criteria for the initial trial were full-term (≥37 weeks’ gestation), singleton newborns delivered to English-speaking, primiparous mothers 20 years of age or older. Infants born at weight less than 2500 g were excluded. Of 316 consenting mother-child dyads, 291 were randomized; 279 dyads completed the first home visit at 3 to 4 weeks after birth and were considered the study cohort for outcomes per the study protocol (Figure 1). All participants completing the 3-year assessment (N = 232) at the end of the original trial were eligible for the present observational follow-up study. More details have been published elsewhere,8,9,10 and the study protocol is available in Supplement 1.
Figure 1. Flow of Participants Through the Study.
aPotential participants may have had multiple exclusions.
bClinic visit or retrieved from medical records or parent report.
Intervention
Participant groups were randomized by research staff to the responsive parenting or home safety intervention (control) group using a secure Microsoft Excel application 10 to 14 days after delivery. The Consolidated Standards of Reporting Trials (CONSORT) reporting guideline was followed for this randomized clinical trial. The randomization scheme used permuted blocks of 6 and stratified by birth weight for gestational age11 (<50th percentile or ≥50th percentile) and intended feeding mode (breastfeeding or formula feeding). As reported previously, the responsive parenting intervention provided less than 10 contact hours of guidance that advised parents on how to respond to their child’s needs across 4 behavioral states: drowsy, sleeping, fussy, and alert (ie, interactive play and feeding).8,10 The control group received a home safety intervention that was dose matched to ensure equivalent time and intensity. The intervention was delivered via mailed instructions at 2 weeks, with nurse home visits at child’s age 3 to 4, 16, 28, and 40 weeks; research center visits at 1 and 2 years; and brief telephone contacts at 18 and 30 months. No intervention was performed during the observational follow-up.
Measures
Background Characteristics
Family demographic information was collected at enrollment. Self-identified race and ethnicity were collected using categories consistent with National Institutes of Health enrollment tables as required by the funding agency. Maternal age, prepregnancy weight, and infant gestational age were extracted from medical records.
Anthropometrics
Weight and length or height were assessed at 10 time points: 3 to 4 weeks; 16, 28, and 40 weeks; and 1, 2, 3, 5, 6, and 9 years. Measurements were completed by trained research staff through age 9 years. Staff were masked to study group beginning at the age 1–year measurement. At each time point, weight was measured in duplicate to the nearest 0.1 kg using an electronic scale (Seca 354 or 874). Recumbent length (age <2 years) or standing height (age ≥2 years) was measured to the nearest 0.1 cm using a recumbent length board (Shorr Productions) or stadiometer (Seca 216), respectively. A third measurement was taken when the 2 measurements differed by more than 0.05 kg or 1 cm, respectively. Measurements were averaged for analysis. Before age 6 years, 1% or less of measurements at each time point were retrieved directly from medical records or reported by parents from medical records when unable to conduct measurements at in-person visits; 14% (29 of 207; n = 13 from medical records, n = 16 from parent report) and 30% (53 of 176; n = 39 from medical records, n = 14 from parent report) were obtained via medical records or parent report at 6 and 9 years, respectively.
Child BMI was converted to percentiles and z scores for age and sex using the 2000 Centers for Disease Control and Prevention growth reference for children 2 years of age or older per the study protocol.12 BMI at or above the 85th percentile but below the 95th percentile was defined as overweight, and BMI at or above the 95th percentile was defined as obesity.13 Given the 2022 publication of updated growth references for those above the 95th percentile,14 post hoc analyses for BMI z scores were conducted using these references and are presented in the eResults, eFigure, and eTables 1-7 in Supplement 2. Maternal height was measured (n = 256) using a stadiometer (Seca 240) or medical records (n = 23). Maternal prepregnancy BMI was calculated as prepregnancy weight (in kilograms) divided by height (meters squared).
Outcomes
The primary outcome was BMI measured across 4 assessments from 3 through 9 years. BMI was selected as the primary outcome given statistical concerns with using standardized scores, such as BMI z scores over longitudinal time. BMI measured across study assessments from early infancy (age 3-4 weeks) to 9 years was a secondary outcome, as were BMI z scores and overweight (BMI ≥85th percentile and <95th percentile) and obesity (BMI ≥95th percentile) status at 5, 6, and 9 years. In addition, blood pressure and waist circumference were determined at ages 6 and 9 years. Blood pressure was measured using automated oscillometric monitors (Omron Model HEM-705CPN) on the upper arm.15 Children’s blood pressure was categorized as elevated if systolic or diastolic blood pressures were in the range from in the 90th percentile or higher to less than the 95th percentile and as stage 1 or 2 hypertensive if in the 95th percentile or higher.16 Waist circumference was measured to the nearest 0.1 cm and was compared with established norms for sex and age to calculate the z score.17 At age 9 years, body composition and metabolic health indicators were also assessed. Body composition (fat mass percentage and fat-free mass percentage) was determined by whole-body air displacement plethysmography (BOD POD; COSMED USA, Inc).18 A fasting laboratory evaluation for metabolic health included glucose, hemoglobin A1c, alanine aminotransferase, total cholesterol, low-density lipoprotein cholesterol, non–high-density lipoprotein (HDL) cholesterol, HDL cholesterol, and triglyceride levels. Abnormal values were determined from consensus guidelines.13,19,20
Sample Size and Power Calculation
The power calculation for the primary outcome of BMI across 3 to 9 years was based on the effect sizes reported through age 3 years.8 Assuming 90 participants per group, we anticipated having greater than 80% power to test the treatment effect across 4 measurements (3, 5, 6, and 9 years). Additionally, using the recommendations for growth curve power simulations in Gelman and Hill,21 we also expected 85% power to detect an interaction effect of 0.45 between study groups and within-person slopes.
Statistical Analysis
Analyses were determined a priori and included all randomized participants who had completed the assessment at age 3 years. Repeated-measures analysis of variance was used to evaluate the main effect of study group on BMI across 4 assessment time points from ages 3 to 9 years. In the context of this model, we also examined the main effect of time (exact age in years) with a random slope estimated for child age on BMI and the potential group-by-time interaction. Next, sex was examined as a moderator of intervention effects. Variables included as covariates were maternal age, marital status, and household income at enrollment, as well as prepregnancy BMI and child weight for length at birth, which were selected based on previously reported associations with BMI and/or were related to missingness.7 The same approach was used for the secondary outcome of treatment effect across 10 assessment time points from ages 3 to 4 weeks to age 9 years. To examine cross-sectional differences by study group and sex at ages 5, 6, and 9 years, fixed-effects models were used for the BMI z score, and ordinal logistic regression was used for the prevalence of overweight and obesity.
Similar models were estimated for additional outcomes: ordinal logistic regressions conducted for blood pressure categories and a fixed-effects model for body composition and waist circumference. Because of small cell sizes for metabolic outcomes, the Fisher exact test was used to examine study group differences. Multiple imputation allowed for all mixed models to include all available data from the observational follow-up study’s analytic sample. In addition, as a sensitivity analysis, all analyses for primary and secondary outcomes were conducted with multiple imputation via a fully conditional specification method as described in the eMethods in Supplement 2. The multiple imputation analyses were used to compare effects among the original study cohort (N = 279 [ie, if attrition had not occurred between the first study assessment at age 3-4 weeks and age 3 years]) with those eligible for the longitudinal follow-up evaluation (N = 232) that is the focus of this report. All analyses were conducted in SAS, version 9.4 (SAS Institute Inc) from January 21 to November 15, 2024. Statistical significance was defined as P < .05, and all tests of intervention effects were 2-sided. A Tukey adjustment was applied for post hoc comparisons of significant main effects and interactions; a comparison of least-squares mean values was used for mixed-effects or fixed-effects models, and odds ratios were used for logistic regressions.
Results
Of the 232 primiparous mother-child dyads eligible for the longitudinal follow-up study (116 per group) (7 Asian [3%], 11 Black [5%], 1 Native Hawaiian or Other Pacific Islander [0.4%], 207 White [89%], and 6 children with other race and ethnicity [including Asian, Indian, Hispanic, Dominican, and other race; 2.5%]) (Table 1), 96% (n = 222), 88% (n = 203), and 76% (n = 177) provided height and weight data at 5, 6, and 9 years, respectively. Attrition did not differ by study group. Sociodemographic data did not differ between those with and those without anthropometric data with 1 exception. Children with unmarried parents were more likely to having missing anthropometric data between ages 5 and 9 years (P = .01).
Table 1. Mother and Child Characteristics of Participants With Assessments at Age 3 Years (N = 232).
| Characteristic | Responsive parenting group (n = 116) | Control group (n = 116) |
|---|---|---|
| Children | ||
| Male sex, No. (%) | 62 (53) | 59 (51) |
| Gestational age at birth, mean (SD), wk | 39.2 (1.3) | 39.2 (1.2) |
| Birth BMI, mean (SD) | 13.2 (1.2) | 13.3 (1.3) |
| Hispanic ethnicity, No. (%)a | 9 (8) | 4 (3) |
| Race, No. (%)a | ||
| Asian | 3 (3) | 4 (3) |
| Black | 6 (5) | 5 (4) |
| Native Hawaiian or Other Pacific Islander | 1 (1) | 0 |
| White | 102 (88) | 105 (91) |
| Otherb | 4 (4) | 2 (2) |
| Mothers | ||
| Age, mean (SD), y | 32.3 (4.2) | 32.4 (4.7) |
| BMI, mean (SD) | 27.6 (6.3) | 27.6 (6.6) |
| Marital status, No. (%)a | ||
| Married | 94 (81) | 94 (81) |
| Not married, living with partner | 16 (14) | 16 (14) |
| Single | 5 (4) | 6 (5) |
| Divorced, separated, or widowed | 1 (1) | 0 |
| Education, No. (%)a | ||
| ≤High school graduate | 6 (5) | 7 (6) |
| Some college | 27 (23) | 27 (23) |
| College graduate | 47 (41) | 50 (43) |
| ≥Graduate degree | 36 (31) | 32 (28) |
Abbreviation: BMI, body mass index (calculated as weight in kilograms divided by height in meters squared).
Assessed during the delivery or birth hospitalization.
Including Asian, Indian, Hispanic, Dominican, and other race.
Primary Outcome of BMI From 3 Through 9 Years
There was a significant main effect of study group on BMI across ages 3 to 9 years; children in the responsive parenting group had a lower mean (SE) BMI than controls (16.59 [0.18] vs 16.95 [0.18]; absolute difference, −0.36; P = .049) (Figure 2) with BMI increasing for both groups over time (P < .001) (Table 2). There was no significant study group by time (ie, age) interaction. Alternatively, sex was a significant moderator of the main effect of study group (ie, study group by sex interaction) such that female participants in the responsive parenting group had a lower mean (SE) BMI compared with female participants in the control group (16.31 [0.23] vs 17.24 [0.22]; absolute difference, −0.93; P = .002) (eResults in Supplement 2), with no significant study group difference for male participants (16.81 [0.21] vs 16.66 [0.21]; absolute difference, 0.15; P = .92). Although BMI was lower for female participants in the responsive parenting group, post hoc analyses of BMI change between 3 and 9 years revealed significantly greater increases for female participants in the responsive parenting group than for those in the control group (3.35 vs 1.89; absolute difference, 1.46; P = .03). The greater increase in BMI for female participants in the responsive parenting group compared with those in the control group was most profound between ages 6 and 9 years (2.78 vs 1.80; absolute difference, 0.98; P = .04). For male participants, in addition to no main effect difference in BMI between groups across ages 3 to 9 years, there was no significant difference in BMI change (study group by time interaction) over this interval by study group.
Figure 2. Mean Difference in BMI Between Children Who Received Responsive Parenting vs Home Safety (Control) Interventions at Each Study Time Point.
BMI indicates body mass index (calculated as weight in kilograms divided by height in meters squared). BMI difference was calculated as responsive parenting minus control. Error bars indicate 95% CIs. Data are presented as BMI units. The circles indicate data among the sample completing the age 3–year assessment eligible for the longitudinal follow-up (N = 232).
Table 2. Weight Outcomes by Study Group, Age, and Sex.
| Outcome | Study group | |||||||
|---|---|---|---|---|---|---|---|---|
| Age 3 y | Age 5 y | Age 6 y | Age 9 y | |||||
| Responsive parenting | Control | Responsive parenting | Control | Responsive parenting | Control | Responsive parenting | Control | |
| Overall | ||||||||
| No. | 116 | 116 | 111 | 111 | 99 | 104 | 87 | 90 |
| BMI, mean (SD) | 15.83 (1.19) | 16.20 (1.32)a | 16.05 (2.05) | 16.04 (1.77) | 16.34 (2.29) | 16.39 (2.18) | 18.49 (3.54) | 18.16 (3.37) |
| BMI z score, mean (SD) | −0.13 (1.00) | 0.15 (1.01)a | 0.28 (1.11) | 0.31 (1.04) | 0.32 (1.05) | 0.38 (1.03) | 0.50 (1.08) | 0.36 (1.16) |
| BMI percentile, mean (SD) | 46.9 (29.6) | 53.8 (28.9) | 56.1 (29.9) | 59.3 (28.8) | 58.7 (28.5) | 61.3 (28.6) | 63.5 (29.3) | 60.1 (32.3) |
| Overweight, No. (%)b | 10 (9) | 14 (12) | 14 (13) | 19 (17) | 14 (14) | 16 (15) | 12 (14) | 11 (12) |
| Obesity, No. (%)c | 3 (3) | 9 (8) | 11 (10) | 9 (8) | 9 (9) | 11 (11) | 16 (18) | 15 (17) |
| Female participants | ||||||||
| No. | 54 | 57 | 52 | 56 | 49 | 51 | 41 | 46 |
| BMI, mean (SD) | 15.67 (1.20) | 16.43 (1.48)d | 16.08 (2.02) | 16.45 (2.09) | 16.54 (2.39) | 16.69 (2.56) | 19.06 (3.94) | 18.24 (3.08) |
| BMI z score, mean (SD) | −0.14 (0.96) | 0.41 (1.02)d | 0.36 (1.07) | 0.59 (0.99) | 0.43 (1.03) | 0.51 (0.94) | 0.58 (1.14) | 0.43 (0.96) |
| BMI percentile, mean (SD) | 46.1 (29.8) | 61.4 (28.0)d | 59.2 (30.0) | 67.2 (26.3) | 61.4 (29.1) | 65.5 (25.2) | 65.4 (29.9) | 63.5 (27.3) |
| Overweight, No. (%)b | 3 (6) | 10 (18)a | 9 (17) | 13 (23) | 7 (14) | 8 (16) | 4 (10) | 5 (11) |
| Obesity, No. (%)c | 2 (4) | 6 (11) | 6 (12) | 6 (11) | 6 (12) | 4 (8) | 10 (24) | 4 (9) |
| Male participants | ||||||||
| No. | 62 | 59 | 59 | 55 | 50 | 53 | 46 | 44 |
| BMI, mean (SD) | 15.98 (1.18) | 15.97 (1.10) | 16.01 (2.09) | 15.63 (1.27) | 16.15 (2.19) | 16.11 (1.72) | 17.99 (3.11) | 18.08 (3.68) |
| BMI z score, mean (SD) | −0.12 (1.04) | −0.11 (0.93) | 0.20 (1.16) | 0.03 (1.02) | 0.21 (1.07) | 0.25 (1.10) | 0.43 (1.03) | 0.29 (1.34) |
| BMI percentile, mean (SD) | 47.6 (29.6) | 46.6 (28.1) | 53.4 (29.8) | 51.2 (29.3) | 56.1 (28.0) | 57.4 (31.2) | 61.8 (28.9) | 56.5 (36.8) |
| Overweight, No. (%)b | 7 (11) | 4 (7) | 5 (8) | 6 (11) | 7 (14) | 8 (15) | 8 (17) | 6 (14) |
| Obesity, No. (%)c | 1 (2) | 3 (5) | 5 (8) | 3 (5) | 3 (6) | 7 (13) | 6 (13) | 11 (25)a |
Abbreviation: BMI, body mass index (calculated as weight in kilograms divided by height in meters squared).
P < .05.
Overweight: BMI from in the 85th percentile or higher to less than the 95th percentile.
Obesity: BMI in the 95th percentile or higher.
P < .01.
Secondary Outcomes
When examining the entire study period (3-4 weeks to 9 years), there was a significant main effect of study group such that children in the responsive parenting group had a lower mean (SE) BMI than children in the control group (16.47 [0.13] vs 16.86 [0.13]; absolute difference, −0.39; P = .01) (Figure 2). Although there was a main effect of time on BMI (P < .001), there was no significant study group by time interaction. Sex moderated the study group effect; female participants in the responsive parenting group had a lower mean (SE) BMI compared with female participants in the control group (16.18 [0.18] vs 16.84 [0.17]; absolute difference, −0.66; P = .01) (eResults in Supplement 2), with no significant group difference for male participants. In cross-sectional comparisons at ages 5, 6, and 9 years, there were no significant differences between study groups in BMI z score (Figure 3) or in the prevalence of overweight or obesity (Table 2). When these cross-sectional comparisons were stratified by sex, there was 1 significant difference between study groups: male participants in the responsive parenting group were less likely to have obesity at age 9 years than male participants in the control group (odds ratio, 0.24 [95% CI, 0.06-0.92]).
Figure 3. Distribution of BMI z Score at Ages 5, 6, and 9 Years Among All Children and Male and Female Participants Who Received Responsive Parenting vs Home Safety (Control) Interventions.
BMI indicates body mass index (calculated as weight in kilograms divided by height in meters squared). The BMI z score was calculated using the 2000 Centers for Disease Control and Prevention growth reference.
Additional Outcomes
Among the children with recorded blood pressure measurements, 31 of 153 (20.3%) and 33 of 154 (21.4%) had elevated blood pressure or were in the hypertensive range at ages 6 and 9 years, respectively, with no observed differences between study groups (eResults in Supplement 2). Similarly, there were no differences between groups for waist circumference at either time point. For body composition at age 9 years, the responsive parenting group and the control group had a mean (SD) of 22.9% (9.5%) and 22.9% (9.4%) fat mass, respectively. Among the 81 participants who completed the laboratory evaluation, no between-group differences were detected for any test conducted.
Discussion
A responsive parenting intervention designed for the primary prevention of obesity that spanned the first 2 years after birth resulted in lower BMIs than controls from ages 3 to 9 years. BMI was also lower for children in the responsive parenting group vs children in the control group when examined from the first month after birth through 9 years. The difference in BMI between groups was driven by the lower BMI among female participants in the responsive parenting group compared with those in the control group, an effect that appeared to dissipate over time after cessation of the intervention. By age 9 years, there were no overall study group differences in any weight-related outcome, including BMI and proportion of children with overweight or obesity.
Any positive effect on weight-related outcomes in childhood could be considered encouraging and even surprising for an early-life behavioral intervention with a limited number of contacts.22 INSIGHT’s total number of contact hours for intervention was fewer than 10, on average, and occurred largely in the first year after birth, and the intervention was completed when children were 2 years of age. Although there was hope that early-life effects might persist consistent with the developmental origins of health and disease framework, the finding that these positive effects dissipated by age 9 years may reflect the need for a life-course approach to obesity prevention in today’s obesogenic environment.
In the first few years after birth, children have limited independence and are reliant on parents for choices related to feeding, diet, sleep, and play. With increasing independence and the wider variety of exposures that occur during the toddler and early school age periods, additional challenges to obesity prevention occur. As summarized by the World Health Organization,23 a life-course approach that recognizes that all stages of life are related to each other, the lives of those around us, and to past and future generations may help maintain the benefits of early-life intervention. Such an approach reflects an understanding that there are risk factors and protective factors for disease throughout an individual’s life that must be addressed by that individual as well as by society. Continued parenting support provided to families at developmentally relevant time points could be combined with policy and systems approaches to decrease the extent to which the environment operates in opposition to individual-level behavioral intervention efforts.
During the longitudinal follow-up period of the study, the lack of continued, developmentally relevant guidance was associated with “catch-up” growth for the dyads that were most affected by the early-life intervention, the female participants in the responsive parenting group. This negated the early difference in the effect of the responsive parenting intervention between sexes, favoring girls, which can be interpreted in the context of literature describing less responsiveness among parents of boys. Instead, boys are exposed to more directive and controlling parenting than girls, particularly during the first years after birth.24,25 As such, parents of girls may have found the INSIGHT intervention content to be more relevant and applied its content more effectively than parents of boys. This hypothesis is supported by previous findings from INSIGHT that showed that girls fared better on some outcomes that have been associated with responsive parenting. For example, we found that mothers of boys reported their children had more problem behaviors at ages 3 and 6 years,26 and they were more likely to use restrictive feeding practices with boys than girls.27 In contrast, boys in the responsive parenting group showed greater intervention effects on cognitive aspects of self-regulation in childhood,28 and the intervention improved the emotional aspects of self-regulation overall across the sexes.
Limitations
As we previously reported, the INSIGHT study has several limitations. The relatively homogeneous sample was recruited from a single center with a high proportion of White and native English-speaking mothers with middle income. Although it is possible that responsive parenting interventions could have an even greater effect on populations at greater risk for obesity, we are unable to answer that question with our sample, although a similar responsive parenting intervention was tested in such a population in rural Georgia.29 Additionally, the INSIGHT sample was limited to firstborn children as we sought to establish rather than change parenting behaviors. Fortunately, we have previously reported positive carryover effects of the INSIGHT intervention to second-born children despite the fact that families received no additional intervention.30,31 As acknowledged previously, the 2-group randomized design precludes us from answering the question of what intervention components are required to affect weight outcomes. We acknowledge the attrition of our sample through age 9 years. Despite extensive attempts to retain participants, 63% of the original INSIGHT study cohort and 76% of those eligible for the longitudinal follow-up contributed data to the primary outcome at age 9 years. Yet, our multiple imputation analyses suggest that it is unlikely that a higher rate of retention would result in different outcomes.
Conclusions
In this single-center randomized clinical trial for primiparous mother-child dyads, a responsive parenting intervention initiated in early infancy resulted in a reduction in BMI across ages 3 to 9 years compared with a control intervention. The intervention effect was stronger for girls and appeared to dissipate over time after intervention cessation, particularly for girls, with no significant intervention effect on most weight-related outcomes by age 9 years. Further research is needed on how to incorporate a life-course approach to obesity prevention, which may help to sustain the benefits that responsive parenting interventions have demonstrated in early life. Additional research on improving responsiveness for parents of boys may also be warranted.
Trial Protocol
eMethods.
eResults.
eTable 1. Summary Statistics for Observed and Imputed Primary Variables From 0 to 9 Years
eTable 2. Mixed Effects Repeated Measures Model Pooled Estimates Examining BMI Trajectories From 3 to 9 Years (Model A) and 0 to 9 Years (Model B) (N = 279)
eTable 3. Mixed Effects Models Pooled Estimates Examining Differences in BMI z Scores at 5, 6, and 9 Years With Imputed Data (N = 279)
eTable 4. Ordinal Logistic Regressions Pooled Estimates Predicting Weight Status at 5, 6, and 9 Years Using Imputed Data (N = 279)
eTable 5. Descriptive Statistics Comparing 2000 and Extended Growth References for BMI z Score
eTable 6. Least Square Means Results From Mixed Effects Models Predicting BMI From 3 to 9 Years and 3 4 Weeks to 9 Years
eTable 7. Additional Outcomes at Ages 6 and 9 Years
eFigure. Mean Difference in BMI by Study Group Membership and Child Sex at Each Study Timepoint: a) Females b) Males
Data Sharing Statement
References
- 1.Ogden CL, Carroll MD, Kit BK, Flegal KM. Prevalence of childhood and adult obesity in the United States, 2011-2012. JAMA. 2014;311(8):806-814. doi: 10.1001/jama.2014.732 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Ogden CL, Carroll MD, Lawman HG, et al. Trends in obesity prevalence among children and adolescents in the United States, 1988-1994 through 2013-2014. JAMA. 2016;315(21):2292-2299. doi: 10.1001/jama.2016.6361 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Blake-Lamb TL, Locks LM, Perkins ME, Woo Baidal JA, Cheng ER, Taveras EM. Interventions for childhood obesity in the first 1,000 days a systematic review. Am J Prev Med. 2016;50(6):780-789. doi: 10.1016/j.amepre.2015.11.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Grossman DC, Bibbins-Domingo K, Curry SJ, et al. ; US Preventive Services Task Force . Screening for obesity in children and adolescents: US Preventive Services Task Force recommendation statement. JAMA. 2017;317(23):2417-2426. doi: 10.1001/jama.2017.6803 [DOI] [PubMed] [Google Scholar]
- 5.Perez-Escamilla R, Segura-Perez S, Lott M. Feeding guidelines for infants and young toddlers: a responsive parenting approach. Healthy Eating Research. February 2017. Accessed February 2, 2025. https://healthyeatingresearch.org/wp-content/uploads/2017/02/her_feeding_guidelines_report_021416-1.pdf
- 6.Spill MK, Callahan EH, Shapiro MJ, et al. Caregiver feeding practices and child weight outcomes: a systematic review. Am J Clin Nutr. 2019;109(suppl 7):990S-1002S. doi: 10.1093/ajcn/nqy276 [DOI] [PubMed] [Google Scholar]
- 7.Taylor RW, Gray AR, Heath AM, et al. Sleep, nutrition, and physical activity interventions to prevent obesity in infancy: follow-up of the Prevention of Overweight in Infancy (POI) randomized controlled trial at ages 3.5 and 5 y. Am J Clin Nutr. 2018;108(2):228-236. doi: 10.1093/ajcn/nqy090 [DOI] [PubMed] [Google Scholar]
- 8.Paul IM, Savage JS, Anzman-Frasca S, et al. Effect of a responsive parenting educational intervention on childhood weight outcomes at 3 years of age: the INSIGHT randomized clinical trial. JAMA. 2018;320(5):461-468. doi: 10.1001/jama.2018.9432 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Savage JS, Birch LL, Marini M, Anzman-Frasca S, Paul IM. Effect of the INSIGHT responsive parenting intervention on rapid infant weight gain and overweight status at age 1 year: a randomized clinical trial. JAMA Pediatr. 2016;170(8):742-749. doi: 10.1001/jamapediatrics.2016.0445 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Paul IM, Williams JS, Anzman-Frasca S, et al. The Intervention Nurses Start Infants Growing on Healthy Trajectories (INSIGHT) study. BMC Pediatr. 2014;14:184. doi: 10.1186/1471-2431-14-184 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Oken E, Kleinman KP, Rich-Edwards J, Gillman MW. A nearly continuous measure of birth weight for gestational age using a United States national reference. BMC Pediatr. 2003;3:6. doi: 10.1186/1471-2431-3-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Centers for Disease Control and Prevention; National Center for Health Statistics. Growth charts. Centers for Disease Control and Prevention. September 2, 2024. Accessed September 2, 2024. https://www.cdc.gov/growthcharts/
- 13.Hampl SE, Hassink SG, Skinner AC, et al. Clinical practice guideline for the evaluation and treatment of children and adolescents with obesity. Pediatrics. 2023;151(2):e2022060640. doi: 10.1542/peds.2022-060640 [DOI] [PubMed] [Google Scholar]
- 14.Hales CM, Freedman DS, Akinbami L, Wei R, Ogden CL. Evaluation of alternative body mass index (BMI) metrics to monitor weight status in children and adolescents with extremely high BMI using CDC BMI-for-age growth charts. Vital Health Stat 1. 2022;(197):1-42. doi: 10.15620/cdc:121711 [DOI] [PubMed] [Google Scholar]
- 15.Gillman MW, Cook NR. Blood pressure measurement in childhood epidemiological studies. Circulation. 1995;92(4):1049-1057. https://www.ncbi.nlm.nih.gov/pubmed/7641339. doi: 10.1161/01.CIR.92.4.1049 [DOI] [PubMed] [Google Scholar]
- 16.National High Blood Pressure Education Program Working Group on High Blood Pressure in Children and Adolescents. The fourth report on the diagnosis, evaluation, and treatment of high blood pressure in children and adolescents. Pediatrics. 2004;114(2 suppl 4th report):555-576. [PubMed] [Google Scholar]
- 17.Fryar CD, Gu Q, Ogden CL. Anthropometric reference data for children and adults: United States, 2007-2010. Vital Health Stat. 2012;11(252):1-48. [PubMed] [Google Scholar]
- 18.Fields DA, Goran MI. Body composition techniques and the four-compartment model in children. J Appl Physiol (1985). 2000;89(2):613-620. [DOI] [PubMed] [Google Scholar]
- 19.Expert Panel on Integrated Guidelines for Cardiovascular Health and Risk Reduction in Children and Adolescents; National Heart, Lung, and Blood Institute. Expert panel on integrated guidelines for cardiovascular health and risk reduction in children and adolescents: summary report. Pediatrics. 2011;128(suppl 5):S213-S256. doi: 10.1542/peds.2009-2107C [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Vos MB, Abrams SH, Barlow SE, et al. NASPGHAN clinical practice guideline for the diagnosis and treatment of nonalcoholic fatty liver disease in children: recommendations from the Expert Committee on NAFLD (ECON) and the North American Society of Pediatric Gastroenterology, Hepatology and Nutrition (NASPGHAN). J Pediatr Gastroenterol Nutr. 2017;64(2):319-334. doi: 10.1097/MPG.0000000000001482 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Gelman A, Hill J. Data Analysis Using Regression and Multilevel/Hierarchical Models. Cambridge Unversity Press; 2007. [Google Scholar]
- 22.Brown T, Moore TH, Hooper L, et al. Interventions for preventing obesity in children. Cochrane Database Syst Rev. 2019;7(7):CD001871. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Müller A, ed. The Life-Course Approach: From Theory to Practice: Case Stories From Two Small Countries in Europe. World Health Organization. Regional Office for Europe; 2018. [Google Scholar]
- 24.Tamis-Lemonda CS, Briggs RD, McClowry SG, Snow DL. Maternal control and sensitivity, child gender, and maternal education in relation to children’s behavioral outcomes in African American families. J Appl Dev Psychol. 2009;30(3):321-331. doi: 10.1016/j.appdev.2008.12.018 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Endendijk JJ, Groeneveld MG, Bakermans-Kranenburg MJ, Mesman J. Gender-differentiated parenting revisited: meta-analysis reveals very few differences in parental control of boys and girls. PLoS One. 2016;11(7):e0159193. doi: 10.1371/journal.pone.0159193 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Tauriello S, Savage JS, Goldsmith J, Kubiniec E, Paul IM, Anzman-Frasca S. Effect of the INSIGHT responsive parenting intervention on parenting and child behavior at ages 3 and 6 years. J Pediatr. 2023;255:72-79. doi: 10.1016/j.jpeds.2022.10.025 [DOI] [PubMed] [Google Scholar]
- 27.Hyczko AV, Ruggiero CF, Hohman EE, et al. Sex differences in maternal restrictive feeding practices in the intervention nurses start infants growing on healthy trajectories study. Acad Pediatr. 2021;21(6):1070-1076. doi: 10.1016/j.acap.2021.05.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Hernandez Acton E, Kubiniec E, Bhargava S, et al. INSIGHT responsive parenting intervention effects on child self-regulation at ages 3 and 6 years. Dev Psychol. 2024;2024. doi: 10.1037/dev0001839 [DOI] [PubMed] [Google Scholar]
- 29.Lavner JA, Hohman EE, Beach SRH, Stansfield BK, Savage JS. Effects of a responsive parenting intervention among black families on infant sleep: a secondary analysis of the sleep SAAF randomized clinical trial. JAMA Netw Open. 2023;6(3):e236276. doi: 10.1001/jamanetworkopen.2023.6276 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Savage JS, Hochgraf AK, Loken E, et al. INSIGHT responsive parenting educational intervention for firstborns is associated with growth of second-born siblings. Obesity (Silver Spring). 2022;30(1):183-190. doi: 10.1002/oby.23301 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Hohman EE, Savage JS, Marini ME, et al. Effect of the INSIGHT firstborn parenting intervention on secondborn sleep. Pediatrics. 2022;150(1):e2021055244. doi: 10.1542/peds.2021-055244 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Trial Protocol
eMethods.
eResults.
eTable 1. Summary Statistics for Observed and Imputed Primary Variables From 0 to 9 Years
eTable 2. Mixed Effects Repeated Measures Model Pooled Estimates Examining BMI Trajectories From 3 to 9 Years (Model A) and 0 to 9 Years (Model B) (N = 279)
eTable 3. Mixed Effects Models Pooled Estimates Examining Differences in BMI z Scores at 5, 6, and 9 Years With Imputed Data (N = 279)
eTable 4. Ordinal Logistic Regressions Pooled Estimates Predicting Weight Status at 5, 6, and 9 Years Using Imputed Data (N = 279)
eTable 5. Descriptive Statistics Comparing 2000 and Extended Growth References for BMI z Score
eTable 6. Least Square Means Results From Mixed Effects Models Predicting BMI From 3 to 9 Years and 3 4 Weeks to 9 Years
eTable 7. Additional Outcomes at Ages 6 and 9 Years
eFigure. Mean Difference in BMI by Study Group Membership and Child Sex at Each Study Timepoint: a) Females b) Males
Data Sharing Statement
