TORONTO – Teaching parents of newborns to respond to eating and satiety cues in ways that promote self-regulation was associated with improvements in some weight outcomes at 3 years in a randomized clinical trial.
For the primary outcome of body mass index (BMI) z score at 3 years, a significant difference favoring the responsive parenting (RP) intervention was seen (–0.13 vs. 0.15 for controls; absolute difference, –0.28; P = .04). A longitudinal analysis examining the entire intervention period confirmed that the mean BMI group differences across seven study visits confirmed the effect of the RP intervention on BMI (P less than .001).
“We felt that the BMI z score and longitudinal growth analysis are probably the most sustained effects for an early-life intervention that have been recorded to date,” reported Ian M. Paul, MD, MSc, of Penn State University, Hershey. “While the differences between study groups were modest and not all achieved statistical significance, all favored the responsive-parenting intervention.”
Mean BMI percentile, a secondary outcome, was 47th for the RP group and 54th for controls, narrowly missing statistical significance (P = .07). Similarly, the percent of children deemed overweight at 3 years was 11.2% for the RP group and 19.8% for controls (P = .07), while 2.6% and 7.8%, respectively, were obese (P = .08).
No significant differences were seen in growth-related adverse events, such a weight-for-age less than the 5th percentile. The issue of “inducing” failure-to-thrive with a feeding intervention is a concern, said Dr. Paul, but there was no evidence for it in their study.
“One could question whether [the small differences seen between groups] are clinically significant, but if we look at how small differences have changed in the population over time and how those equate as far as longitudinal risk for cardiovascular outcomes and metabolic syndrome, etc., the small differences [we saw] might be important on a population level,” said Dr. Paul at the Pediatric Academic Societies meeting.