Multiple prepubertal growth spurts in children of the Fels Longitudinal Study: comparison with results from the Edinburgh Growth Study

Abstract

Background: Contrary to the common assumption of a single midspurt, the heights of children measured at 6-monthly intervals in the Edinburgh Growth Study appear to show multiple prepubertal growth spurts of 2–3-year duration. The Edinburgh study relied on inspection of empirical accelerations to identify the spurts.Aim: The work aimed to study this phenomenon with velocity maxima and minima of kernel-smoothed heights in the more extensive 6-monthly data from the Fels Longitudinal Growth Study. Velocity maxima and minima for two sets of triplets in the Fels data were also examined for concordance of prepubertal growth patterns in the identical and fraternal members.Subjects: Subjects were 167 boys and 152 girls with complete or nearly complete 6-monthly height measurements from 2 to 18 years of age.Methods: Because the number of spurts detected by kernel smoothing was found to be sensitive to the bandwidth of the smooth, a preliminary study to determine the best choice of bandwidth was carried out. With the chosen width, 2.0 years, the number and locations of up to five maxima and preceding minima of growth velocity were determined for each case. Statistics for the distribution of age of occurrence, height, and velocity at each turning point were computed for cases having no or one to four prepubertal growth spurts.Results: Good agreement with the Edinburgh results was found. Numbers of prepubertal spurts varied from zero to four in both boys and girls. Later maturing children tended to have more spurts. Among the Fels triplets, the patterns of prepubertal growth of the identical members of each set did not appear more concordant with each other than with the fraternal member.Conclusions: The finding of similar appearing prepubertal multiple growth spurts over 2–3-year intervals by different analytical methods in two widely separated populations strongly supports the generality of this phenomenon. The timing and duration of the spurts do not suggest, however, a systematic origin. Rather, they appear to be episodes of random change in rate of growth prior to the onset of adolescence, under the constraint of the genetically determined growth potential.