Growth chart comparison – Incidental Findings

The 2025 Fenton third-generation growth charts for preterm infants was released recently and I thought it would be interesting to examine how various preterm and term growth charts compared and what effects they had on the concept of “small”, “large”, and “appropriate” size for gestational age.

A total of 6 weight-for-age growth charts (Fenton 2003/2013/2025, Olsen 2010, WHO infant & expanded) and 5 percentile curves (3, 10, 50, 90, 97) are represented – full set at the bottom.

Evolution of preterm growth charts

I am not an expert in growth charts, but my understanding is that preterm growth charts attempt to describe normal growth trajectories of fetuses in utero, but are based on the weight of preterm infants after they are born. One challenge is that growth faltering of a fetus in utero can be an indication for elective preterm delivery, thereby biasing the sample population used to generate preterm growth charts to smaller sizes.

The Fenton third-generation growth charts attempt to reduce this bias by excluding fetuses at higher risk of having abnormal growth trajectories. As expected, compared to the second generation Fenton 2013 growth charts, the largest differences were in the lower percentile weight curves, with slopes of growth velocity improved to better match expected growth assessed by fetal ultrasound. The concrete result of this is that more preterm infants will be classified as small for gestation.

The maximum differences in the 10th percentiles between the second and third-generation charts were at 28 weeks of gestation, where the third- generation’s 10th percentiles placed on the second-generation charts at the 29th percentile for girls and the 27th percentile for boys. (Fenton 2025)

Small for gestational age

Above is an plot overlaying the Fenton 2013 and Fenton 2025 male weight-for-age charts at the 10th, 50th, and 90th percentiles (as well as two versions of the WHO growth standards). The plot is interactive – you can choose which lines to select, zoom in / out, and hover to see actual values. The distance between the red 10th percentile lines demonstrates the difference between the Fenton 2013 and 2025 cutoff values for the usual definition of “small for gestational age” (SGA).

Hamonization with WHO growth standards

After the original Fenton 2003 preterm growth chart, the 2013 and 2025 charts were smoothed to harmonize well with the WHO growth standards. The plot above also overlays two versions of the WHO growth standard. The who_2006_infant infant growth standard has monthly values from 0 to 24 months (fairly coarse), while the who_expanded child growth standard has daily values from 0 to 1,856 days (approximately birth to 5 years age). The who_expanded chart has the interesting property of showing longitudinal changes in growth, where zooming in, it is possible to see the result of expected postnatal weight loss in the first days of life.

One unexpected finding was that while the Fenton 2013 growth chart harmonizes very precisely with both WHO standards, practically perfectly superimposed by 50 weeks post-menstrual age, the new third generation Fenton 2025 weight-for-age charts remains significantly below the WHO growth standards. This does not appear to be due to the WHO origin age at 40 weeks gestation (Fenton 2025 reported a weighted average of birth at 39.3 weeks), as shifting the curves did not particularly improve the harmonization.

Potential consequences of growth chart selection

“Full term” newborns and hypoglycemia

One of the more common use of growth charts at the time of delivery is to identify full term newborns at increased risk of low blood sugar (hypoglycemia). Commonly, small and large for gestational age (SGA, less than the 10th percentile; LGA, greater than the 90th percentile) are used as discrete decision points for blood sugar monitoring. (Preterm newborns are universally screened.)

Some institutions recommend using the 2006 WHO growth standard for all “full term” babies and using an uncorrected chronologic age. For a male infant at age 0 with the WHO 2006 growth standard, the 10th percentile weight would be 2,750 grams and 90th percentile would be 4,000 grams.

However, full term is defined as birth from 37 0/7 to 41 6/7 weeks gestation.

I do not agree with categorizing, for example, a baby born “full-term” at 37 0/7 weeks with a birth weight of 3,990 grams as being appropriately-sized (per WHO 2006); on the Fenton 2013 chart, this would be 99%ile (Z = +2.26). Likewise, a male infant born 41 6/7 weeks and birth weight 2,770 grams would be AGA per WHO 2006, but severely growth restricted (<0.1%ile, Z = -2.64) on Fenton 2013.

My concern is that using the WHO growth standard for all deliveries between 37 0/7 and 41 6/7 weeks could miss severely SGA and LGA babies at significant risk for hypoglycemia. Because of this, my personal practice is to use the Fenton 2013 chart at birth even for “term”, although that is not its recommended use. The design of the Fenton 2013 charts to precisely harmonize with the WHO 2006 growth standard by 50 weeks postmenstrual age also gives me confidence in this practice.

Preterm newborns and growth monitoring considerations

Unlike the previous discussion about potentially missing babies at risk for hypoglycemia, preterm newborns are universally screened – the categorization of preterm newborns as “SGA”, “AGA”, and “LGA” has less practical impact on clinical care decisions.

Many institutions now track preterm growth outcomes over the course of a hospitalization by change in Z-scores (e.g., for weight) from birth to discharge, with a common expectation of a ∆Z from birth to discharge -0.8 as being acceptable.

However, the new third-generation Fenton 2025 growth charts will now assign more negative weight Z-scores at birth compared to Fenton 2013. As these babies approach discharge closer to 40 weeks, the differences between Fenton 2013 and 2025 disappear. For identical babies and growth trajectories (particularly those born small), the ∆Z weight will be less negative when using the Fenton 2025 charts.

Overall, there will be fewer babies being categorized as poor growth (∆Z from birth to discharge less than -0.8) than previously. This “∆Z -0.8” is a fairly arbitrary cutoff and to my knowledge, not linked to long-term neurodevelopmental or other health outcomes.

But it’s worth keeping in mind that changing the growth chart used for quality improvement monitoring will change change the categorization of growth outcomes. Ideally, future studies will help determine what growth trajectories yield optimal long-term outcomes (and potentially redefining what a “good” ∆Z score is).

Comparison of more growth charts and percentiles

For further exploration, below is an interactive plot that includes more growth charts and percentile curves:

percentiles: 3, 10, 50, 90, 97
charts: Fenton 2003, Olsen 2010, Fenton 2013, Fenton 2025, WHO 2006 infant, WHO expanded child growth standard

(Click an entry in the legend to hide it; draw a box on the plot to zoom in; double-click the plot to zoom back out; hover to get more information.)