With our new protocol, the mean weight z-score decreased much less birth to discharge, and the head circumference z-score decreased almost not at all, but the length z-score still decreased. Very few babies actually fell enough off their curves enough to satisfy definitions of post-natal growth failure, again, if you define by weight. It is clear, despite all that, that the babies are all a bit short and fat when they go home. It may also be that some of the improvement in head circumference is due to subcutaneous fat, so even though I am very happy that head circumference is much better, I don’t know for sure that brain growth is as good as it should be. Making the babies longer as well as heavier is something we would like to be able to do, I think by improving mineral supply, as well as protein intakes, we could probably improve skeletal growth some more.

Which brings us to the thorny question “What should be our goals for nutritional outcomes in the neonatal unit, of very preterm babies?”

There are recent publications with new growth standards, for fetal growth, growth of infants, and growth of late preterm babies, but the data for the very preterm/extreme preterm was quite limited, they are based on “healthy” babies, so of course the numbers of extreme preterms is extremely small.

One of the comments supporting the development of these new standards appears to be that, since we fail to achieve intra-uterine growth, it is not attainable, and we should stop trying, and accept less growth and create new curves. I think that is mistaken,

Until we have other reliable prospective data showing impacts on long-term outcomes of extreme preterm babies, I think the goal of our nutritional interventions should be to aim for the weight, length, head circumference and body composition by 42 weeks PMA that the baby would have had if they had remained in utero with a placenta that functioned well, and then delivered at 39.5 week, and undergone the usual postnatal water and weight loss.

That may be an approach that poses unanswerable questions, but it is the approach taken in a new publication that I only saw after writing that previous paragraph! (Landau-Crangle E, et al. Individualized Postnatal Growth Trajectories for Preterm Infants. JPEN. 2018;42(6):1084-92.) One of the problems with the current best growth charts for preterm babies is that our patients normally cross percentile lines in the downward direction for several days, or even 3 weeks after birth. That is something that we want to happen, as babies who lose less weight have more complications. We can manipulate that weight loss to an extent by starting sodium supplementation too early, for example, but how much free water we give has very little effect.

I have worried for a while about how we could incorporate a curve reflecting postnatal weight loss into a growth chart, and how you would then know which percentile the baby should be on.

This is an extract from the figure legend:

…C) Fetal‐Median‐Growth and Growth‐Velocity Approach. Application of Fenton day‐specific median growth velocities or day‐specific median growth velocities adjusted by a factor from day of life 21 until 42+0/7 weeks (pink). ΔW = difference between target WHOGS weight at 42+0/7 weeks PMA and predicted individual growth trajectory weight…

The authors of this paper have made a number of assumptions, and tested different models to see how they fit. They assumed that by 42 weeks post-menstrual age a preterm baby should be on the WHO growth chart at the percentile that they would have been on had they stayed in utero with a functioning placenta, and then delivered at term and lost their postnatal water, which they term “Contraction of the Extracellular Space” or CES which can be at term, TeCES, or preterm, PreCES. Their preferred model uses day specific corrected median growth velocities from Tanis Fenton’s work, starting after postnatal weight loss, and found that they routinely lined up with the WHO standards at 42 weeks.

They have now set up a website that you can visit https://www.growthcalculator.org that will plot an individualized growth chart if you put in the gestation, sex and birthweight, you can then print it out if you wish. Here is an example showing the curve generated using the criteria shown, and the redlines being the personalized percentiles for that baby.

Our nutritional outcomes that I mentioned above are similar to those of some other groups such as a recent paper from Southampton (Andrews ET, et al. Early postnatal growth failure in preterm infants is not inevitable. Arch dis childh FN. 2018). This article showed very good nutritional outcomes, in terms of all the things that we usually measure, and similar discharge Standard Deviation Scores to our publication, including a similarly worse outcome in terms of length than the outcomes for weight. An editorial accompanying this article (Menon G, et al. Is preterm nutrition a trade-off between head and heart? Archives of disease in childhood Fetal and neonatal edition. 2018) noted some of the limitations with this approach, and they note the lack of large prospective trials that have proven that improvements in short-term growth and particularly weight growth, are beneficial for functional outcomes, or the risk of metabolic disorders in later life.

Using these new weight charts is definitely worth investigating, and I think measures of body composition are important also, even though direct analysis of body composition is tricky, at least an accurate measure of length would be a good idea. Measuring with a tape measure is practically useless; when I was at the Royal Victoria Hospital in Montreal we used a stadiometer, but that requires a fair amount of handling of the baby. Recent studies have used a caliper type of device, and found it quite accurate, and feasible with limited handling. (Pavageau L, et al. Valid serial length measurements in preterm infants permit characterization of growth patterns. J Perinatol. 2018;38(12):1694-701). Of note, the new percentiles that I have been mentioning are not accompanied by length or head growth percentile charts. They also are based an assumption that a baby who starts out at the 1st percentile was destined to end up around the 1st percentile, by 42 weeks; and I am not sure about that, the extremely preterm babies in NICUs are often the result of an intrauterine environment that was poor, and they then are growth restricted. Are babies like that better of if we let them stay on a very low percentile or if we enhance their nutrition and obtain some catch-up growth? Menon and his co-authors in the above-mentioned editorial address briefly some of those questions, and note that the only way to answer them is with prospective trials.