Fast growth in the first years raises the risk of later obesity, and interventions that bend the growth curve have been thin on the ground. The INSIGHT trial randomized first-time mothers to one of two home visiting curricula: one on answering a baby’s cues around feeding, sleep, play and emotion, one on home safety. Nurses visited four times in infancy. What did the scales say at three?
The intervention group came out lower, on one measure and not the other. Mean at three years was -0.13 against 0.15 in the control group. Mean was 47th against 54th, which did not reach significance, and percentile had been the trial’s originally designated . Overweight ran at 11.2 percent of the intervention children against 19.8 percent of the controls.
Rapid growth and high weight in early childhood raise the risk of obesity later, and the list of things that reliably change an early growth trajectory is short. The INSIGHT trial tried a parenting curriculum rather than a diet.
Two hundred ninety-one first-time mother and child pairs at one center were randomized, and 279 received the first home visit. Half drew , a curriculum on reading and answering a baby’s cues around feeding, sleep, interactive play and emotion regulation. The others drew a control curriculum about home safety. Research nurses made four home visits during infancy, with annual center visits out to age three, and 232 pairs finished.
At three years the responsive parenting children had a lower mean BMI z score than the controls, -0.13 against 0.15. The absolute difference was about -0.28, and its confidence interval only just clears zero. Overweight was less common too, 11.2 percent of the intervention children against 19.8 percent of the controls.
Now the complication, and it is a real one. BMI percentile did not differ significantly, 47th against 54th, and percentile had been designated previously as the trial’s primary outcome before the z score took that place. A reader is entitled to weigh that. Add a single center, and mothers who were 86 percent white and 86 percent privately insured, and what remains is a small, real and narrowly generalizable effect.
wat. Four home visits about reading a baby’s cues left a quarter of a standard deviation on the growth curve three years later. The percentile measure did not follow.
wat wat
A quarter of a standard deviation in a preschooler is small in itself. Interventions that move early growth at all are rare enough that a small one earns the attention.
How do we know?
Paul and colleagues randomized 291 first-time mother and child pairs at a single center to a responsive parenting curriculum or a home safety control. Research nurses made four home visits in infancy, with annual visits to age three.
What is still unknown
One center, mothers who were 86 percent white and 86 percent privately insured, and the outcome that reached significance is not the one the trial originally named as primary.
Learn something?
Where to next
Does processing itself make you eat more, or is it the sugar and fat?satiety