Effect of a Behavioral Intervention for Underserved Preschool-Age Children on Change in Body Mass Index: A Randomized Clinical Trial

Abstract

Importance: Prevention of obesity during childhood is critical for children in underserved populations, for whom obesity prevalence and risk of chronic disease are highest.

Objective: To test the effect of a multicomponent behavioral intervention on child body mass index (BMI, calculated as weight in kilograms divided by height in meters squared) growth trajectories over 36 months among preschool-age children at risk for obesity.

Design, setting, and participants: A randomized clinical trial assigned 610 parent-child pairs from underserved communities in Nashville, Tennessee, to a 36-month intervention targeting health behaviors or a school-readiness control. Eligible children were between ages 3 and 5 years and at risk for obesity but not yet obese. Enrollment occurred from August 2012 to May 2014; 36-month follow-up occurred from October 2015 to June 2017.

Interventions: The intervention (n = 304 pairs) was a 36-month family-based, community-centered program, consisting of 12 weekly skills-building sessions, followed by monthly coaching telephone calls for 9 months, and a 24-month sustainability phase providing cues to action. The control (n = 306 pairs) consisted of 6 school-readiness sessions delivered over the 36-month study, conducted by the Nashville Public Library.

Main outcomes and measures: The primary outcome was child BMI trajectory over 36 months. Seven prespecified secondary outcomes included parent-reported child dietary intake and community center use. The Benjamini-Hochberg procedure corrected for multiple comparisons.

Results: Participants were predominantly Latino (91.4%). At baseline, the mean (SD) child age was 4.3 (0.9) years; 51.9% were female. Household income was below $25 000 for 56.7% of families. Retention was 90.2%. At 36 months, the mean (SD) child BMI was 17.8 (2.2) in the intervention group and 17.8 (2.1) in the control group. No significant difference existed in the primary outcome of BMI trajectory over 36 months (P = .39). The intervention group children had a lower mean caloric intake (1227 kcal/d) compared with control group children (1323 kcal/d) (adjusted difference, -99.4 kcal [95% CI, -160.7 to -38.0]; corrected P = .003). Intervention group parents used community centers with their children more than control group parents (56.8% in intervention; 44.4% in control) (risk ratio, 1.29 [95% CI, 1.08 to 1.53]; corrected P = .006).

Conclusions and relevance: A 36-month multicomponent behavioral intervention did not change BMI trajectory among underserved preschool-age children in Nashville, Tennessee, compared with a control program. Whether there would be effectiveness for other types of behavioral interventions or implementation in other cities would require further research.

Trial registration: ClinicalTrials.gov Identifier: NCT01316653.

Conflict of interest statement

Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Mr Escarfuller and Dr Stevens reported receiving grants from the National Institutes of Health (NIH). No other disclosures were reported.

Figures

Figure 1.. CONSORT Flow Diagram of the Parent-Child Pairs

Of the 610 parent-child pairs randomized, 304 were randomized to the intervention group and 306 were randomized to the control group. At each time point, the number retained represents the number of children for whom BMI was collected. Missing body mass index (BMI) falls into 2 categories: BMI measure missing but BMI was collected at a later time point vs permanently lost to follow-up with no further BMI measures collected. The cumulative number of parent-child pairs permanently lost to follow-up is indicated at each time point above, and was 26 for the intervention group and 34 for the control group at year 3 follow-up. aBecause the primary analysis used an intention-to-treat approach, all participants were analyzed in the group to which they were randomized, regardless of missing data.

Figure 2.. Child BMI by Group at Each Follow-up Time Point

Box plots are shown where the middle line represents the median observed child body mass index (calculated as weight in kilograms divided by height in meters squared), boxes represent the interquartile range, whiskers extend to the most extreme observed values with 1.5*IQR of the nearer quartile, and dots represent observed values outside that range.

Figure 3.. Intervention Effect on Child Body Mass Index (BMI)

A, Model-estimated child BMI trajectory for the intervention (n = 304) and control (n = 306) groups. Shaded regions represent 95% CIs around each trajectory. Model-based estimates indicate the BMI linear intervention effect (BMI difference/year) was −0.082 (95% CI, −0.246 to 0.082; P = .33) and the BMI quadratic intervention effect (BMI difference/yearsquared) was 0.032 (95% CI, −0.014 to 0.078; P = .18). The joint likelihood ratio test failed to reject the null hypothesis that the linear and quadratic terms were jointly different from zero (P = .39). B, Model-estimated difference in the mean child BMI trajectories between intervention and control groups, where a value of zero indicates no difference. Shaded region represents 95% CIs around this difference. BMI was calculated as weight in kilograms divided by height in meters squared.