J. Van Den Broeck et al., Length velocity acceleration at 9 months of age in a representative birth cohort of Dutch infants, J PED END M, 13(1), 2000, pp. 45-54
According to the ICP (infancy-childhood-puberty) growth model, statural gro
wth can be divided into three partially superimposed components assumed to
represent different physiologic mechanisms, This model predicts a sudden ac
celeration of length velocity (LV) at the onset of the childhood component
around 9 months. The existence of such an infancy-childhood growth spurt ha
s not yet been firmly corroborated by epidemiological studies. In the prese
nt study length measurements were made at the target ages of 1, 3, 6, 9, 12
, 15, 18 and 24 months in a birth cohort of 2034 infants. In order to check
whether length growth showed a continuous smooth pattern, different mathem
atical models were fitted to the individual growth curves. The models inclu
ded Count and Guo functions, 5th order polynomial and combinations of 5th o
rder polynomial with the logarithmic term of the Count function and the squ
are root term of the Guo function. We showed that in boys and girls there i
s a small but systematic lack of fit of the mathematical modeling, due to a
sudden acceleration of LV around 9 months. In addition there was an increa
se in variation of attained length at this age. Comparison of unbalanced AN
OVA models with and without addition of dummy variables for the target ages
confirmed that there was an acceleration around 9 months that, if correcte
d for, leads to a significantly improved model fit (likelihood ratio test p
<0.0001). In absolute terms of LV, the misfit at 9 months was not greater t
han 0.5 cm/year on average. We conclude that the results of this study supp
ort the existence of a late infancy growth spurt. In our opinion, however,
the magnitude of the phenomenon does not legitimate construction and use of
discontinuous growth references such as the ICP reference.