Preterm children have clinically important and statistically significant differences in all growth measures, including weight, length/height and head circumference through 36 months of CA when their growth is plotted according to their chronological age compared to their corrected ages. At 4 months of age, about 40% of extremely and very preterm infants [21] who were growing appropriately (>−2 z-scores at post term ages) [17] would be misclassified as growth faltering for weight and length if their chronological age was used for assessing their growth.

The common belief that most preterm infants are growth faltering could be at least partially attributable to the lack of proper correction for prematurity when assessing preterm growth [22,23,24,25]. Our findings show that infants born very preterm make consistent progress in catching up, with only 9.9% had height and 4% had BMI measures <−2 z scores by 36 months CA. When chronological age was used, more than twice as many children (66% vs 28%, misclassifying 38%) would have been considered stunted at 4 months of age. With regards to growth charts, the plotted growth patterns of weight-for-age, length-for-age, weight-for-length, head circumference-for-age, and age-specific BMI showed considerable differences in growth patterns according to the type of age used. When chronological age was used, the children plotted lower on the growth charts for all growth indicators, suggesting poorer growth patterns. Using corrected age provided a better representation of the plotted growth measures on the growth chart curves for weight, length and head circumference. These results confirm the need for age correction through 36 months of CA for children born extremely and very preterm to achieve accurate growth assessments.

While some very and extremely preterm infants in this study were small relative to the chart curves median at 0 months CA followed by steady catch up in length and weight to 36 months CA, their WFL categorizations suggested the opposite direction; 19.7% were categorized as overweight at 0 months CA, then at the older ages only 2.3% remained in that category. The categorization difference is likely since their lengths (27.1% stunted) at 0 months CA were lagging their weights (10.2% underweight). One could assume this data at 0 months CA suggests risk of overweight in later life. Our longitudinal analysis shows that this effect is not a prediction of later overweight but is rather a problem using a WFL metric for preterm infants prior to length catch-up, a problem also been seen from using BMI for preterm infants [26].

Preterm head circumferences catch up earlier than weight and length [27], which has been referred to as head sparing [3, 28] and has been observed to be an indicator for good development [28]. Neurodevelopmental outcomes of this cohort have been examined elsewhere [27].

The current findings are in line with those obtained by Wang and Sauve over two decades ago, who using a similar approach also observed large significant differences in growth classifications in preterm growth based on chronological versus corrected ages [6]. The need for age correction was also highlighted by literature assessing developmental outcomes in preterm populations [1, 10, 29, 30]. A recent study by Aylward indicated the need for age correction for cognitive assessments up until 2 years, and for 3 years when assessing language and motor composite scores, for all degrees of prematurity [1] to provide accurate age-appropriate assessments.

Given the complexities surrounding preterm birth, including low birthweights, immature organs and NICU exposures, children born prematurely should not be expected to mature and achieve developmental milestones faster than expected by their term-born peers [31]. While some factors can be influenced by postnatal age such as body fat deposition [32] and gastrointestinal adaptions (which is influenced by feeding) [33, 34], most aspects of childhood growth and development require correction for prematurity. Their chronological age disregards the fact that these infants are born at earlier postmenstrual ages than their term-born counterparts of the same chronological age. Age correction is grounded in the assumption that “early development proceeds as a function of time since conception” [2]. For instance, when chronological age is used to assess the growth of an infant at six months who was born four months early, the infant would be expected to have achieved the size and development of a six-month infant. When the CA is used however, the infant is considered two months of age and thus can be expected to have grown and developed accordingly. These considerable age differences can result in varying and possibly contrasting conclusions regarding the infant’s growth and developmental status, as seen in this study. We agree with D’Agostino that age correction may improve clinical capacity “to accurately recognize genuine delays as opposed to perceived delays related to a child’s gestational age at birth” [5].

In the context of childhood obesity epidemic as well as prevalent food insecurity, it is paramount that assessments of children’s growth are accurate. Misclassified infant and child growth could result in implementing additional clinical services that are not needed. When based on flawed assessments, nutrition interventions aimed at accelerating infant growth could have serious negative consequences on an infant’s health, wellbeing [35] and the feeding relationship within the family [36]. Misguided parenting practices may include over/force feeding that can disrupt children’s capacity to self-regulate their food intake [37,38,39], negatively influencing the child’s ability to maintain their healthy weights and have a healthy relationship with food [37,38,39].

Our study has a few limitations. First, this sample of extremely and very preterm infants had a variety of prenatal and neonatal morbidities and feedings that may have contributed to their growth patterns; however, our mix of patients with morbidities, a range of sizes at birth and a mix of breast and formula feeding are not unique. Accordingly, our results may be similar to results of other NICUs. Second, the anthropometric measurements were made by several healthcare providers; however, they were likely reliable since they were made by trained staff in the Follow-up Clinics [40] using regularly calibrated scales. Third, while it would be valuable to link the present findings to cardiometabolic outcomes, the required data were not available. Lastly, it was not possible to assess variations in growth classifications by age for moderate to late preterm infants as their data were not included in the PreM Growth Study dataset.

To the best of our knowledge, the study findings provide the most up-to date evidence addressing the issue of age correction in preterm growth assessments based on recent very preterm data and widely used WHO growth charts. The study sample size was sufficient to capture important statistical differences. Additionally, the plotted growth patterns on WHO growth charts provide visual evidence for the dramatic differences in growth distributions by chronological and corrected ages.

In conclusion, considering the substantial and statistically significant differences observed in preterm growth classifications according to the age used, available evidence supports the practice of age correction through 36 months of CA to avoid misclassifications of growth among extremely and very preterm children. Future research should explore age correction effects in moderate and late preterm children, and clinical guidelines should be formulated to streamline age correction in routine care to ensure optimal support for children born very prematurely.