Discussion
This study established reference values for RET-He in healthy children under 5 years old in Hong Kong with the reference interval of P2.5–P97.5. RET-He levels varied with age, with lower levels in infants than other age groups. An age-specific value below the lower limit of the reference range of RET-He can be used as a limit to determine ID. There were no significant gender differences in RET-He levels. RET-He was influenced by factors such as age, haemoglobin levels and reticulocyte levels, among others. These findings contribute to a better understanding of RET-He and its interpretation in the context of iron status in children.
In this study, the mean RET-He level is 30.04 pg in children aged 2–48 months, which is slightly below previously reported data with broad age ranges 30.5 pg (6 months to 5 years),18 30.8 pg (15 days to 19 years),19 30.9 pg (1–11 years).17 As the study has a higher composition of infants, who are more vulnerable to ID than older children,20 the population was further categorised into five subgroups, and RET-He reference values were detailed in each group. Based on the P2.5 level of RET-He, RET-He lower limit reference values were proposed among age groups. These values ranged from 25.92 pg to 27.15 pg in different age groups. Previously reported lower limit reference values for ID in healthy young children range from 24.9 pg to 27.5 pg, including 25.6 and 24.9 pg (4–12 months),21 25.6 pg (1–2 years),17 27.3 pg (3–5 years),17 27.5 pg (9–12 months, 6 months–5 years).16 18 Given the reference variations in different age groups, it is necessary to apply a local age-specific reference in clinical medicine when initially assessing ID status with complete blood count (CBC).
The mean RET-He fluctuated with age with a lower level in infants than children aged over 12 months old. Iron from mothers provides necessary iron for growth at birth but iron storage is depleted around 4–6 months since dietary iron becomes the sole source of iron required for continued growth.22 Breast milk, vegetables, fruits and formula milk tend to be low in iron.22 Infants have increased iron demand to support their rapid growth at this stage, making them more vulnerable to ID.23 No significant gender difference was found in RET-He, aligning with previous studies.17 18 24 Ferritin and reticulocytes were higher in females while haemoglobin did not differ by gender.
Various blood parameters were associated with RET-He. Positive correlations were found between RET-He and parameters reflecting IDA, such as haemoglobin, MCV, MCH, MCHC and reticulocytes. These findings align with a previous study,14 showing that RET-He changes are consistent with erythropoiesis-related parameters. However, no correlation was found between RET-He and ferritin. Ferritin levels showed a U-shaped curve with age, peaking at 2–6 months and gradually increasing from 6 to 48 months. Relationships between RET-He and iron metabolism biomarkers are complex and vary with age, biomarker type and ID severity. Age-specific analysis revealed no correlation between RET-He and ferritin in 12-month-old infants while a positive relationship was observed in 4-month-old infants, and a consistent negative association with transferrin saturation (TSAT) was found at both ages.21 Analysis in American children aged 6 months to 18 years revealed that RET-He significantly correlates with ferritin, TSAT and soluble transferrin receptor, making RET-He a potential marker for monitoring erythropoiesis.25 Another paediatric-population-based study did not demonstrate a linear relationship between RET-He and ferritin, indicating that red cells may reach maximum haemoglobin levels while ferritin continues to rise; RET-He may remain constant until ID is evident.26 The varying results suggest that different iron metabolism biomarkers could be used to describe different ID stages.
RET-He is an early indicator of ID in children as well as a supporting diagnostic tool in diagnosing IDA without an extra iron metabolism test, especially in inflammation status. RET-He reflects iron availability within a 4-day time frame,27 especially in predicting the absence of bone marrow iron stores.28 29 Studies focusing on children have demonstrated the RET-He in identifying ID before the onset of anaemia.30 Rescreening during the second year of life revealed a ninefold greater risk of developing IDA among children aged 9–12 months with an RET-He<27.5 pg and no initial anaemia.16 RET-He separated ID cases from the control group with comparatively high sensitivity and demonstrated improved diagnostic performance in IDA group and ID without anaemia.19 Unlike ferritin increasing in acute phase of diseases,31 RET-He has a higher specificity and a lower coefficient of variation, it is more appropriate to screen ID and IDA accompanied by inflammatory responses.32 It has been reported as a sensitive marker of body iron status to conventional tests for the detection of ID in children accompanied with inflammation.33 Combined with low haemoglobin levels, RET-He is an accurate diagnostic test for ID and IDA in ill infants and children.34 RET-He is also important in iron treatment monitoring, serving as an early response to iron therapy within days after initiating the treatment.35 Nevertheless, the use of RET-He in reflecting the effectiveness of iron supplements in IDA children warrants further study.
In our study, the results indicated that the majority of suspected thalassaemia cases had an RET-He value ≤27.8 pg, which is the cut-off determined from the ROC analysis for screening ID. The finding suggests that CHr level could serve as a diagnostic indicator for both thalassaemia and ID. Low MCV and high RCC in CBC may indicate the presence of a thalassaemia trait, requiring further investigation through a haemoglobin study. RET-He values below the cut-off suggest ID, and measuring ferritin level is advisable for confirmation and further evaluation. Another potential application of RET-He is the differentiation of haematological conditions. Thalassaemia, another microcytic, hypoanaemia, demonstrates a similar change of blood parameters as an ID in CBC.36 Reduced RET-He can also be observed in haemoglobinopathies like α-thalassaemia and β-thalassaemia, which may not be related to ID. Clinical history and genetic testing are crucial for ruling out haemoglobinopathies. Studies show that patients with β-thalassaemia have lower RET-He levels and a smaller percentage of microcytic reticulocytes compared with those with ID.28 37 RET-He testing makes it easier to differentiate between the two conditions compared with traditional assessments, allowing for more accurate and timely identification.28 To distinguish between individuals with thalassaemic traits and early ID, various factors can be considered, such as the micro/hypo ratio, Mentzer index and an algorithm using RET-He, Hb/RET-He ratio and the microratio/hyporatio.38 39
RET-He is also advantageous in clinical practice since it is cost-effective and less invasive. Though the conventional iron biomarker-TSAT is also not influenced by inflammation,33 an extra tube of blood and even a second blood draw are required. The traditional test of ferritin and TSAT parameters costs two times more than the cost of getting CBC but the RET-He can be easily obtained or measured in the same blood tube used in CBC analysis with a specific instrument.25 40 In our study, RET-He is measured by the Sysmex XN-9100/XN-1500 haematology analyser, which is widely used throughout Hong Kong, thus promoting the feasibility of adopting age-specific reference value of RET-He in initial ID screening in young children.
One notable strength of this study is a large representative sample stratified by sampling from MCHCs and kindergartens across the territory of Hong Kong. Several limitations should be considered. First, this study was cross-sectional, and causality could not be established. Second, this study only included healthy children, so the reference ranges established may not be generalisable to children with diseases that may affect iron status. As we did not measure inflammation biomarkers, participants with inflammatory status cannot be excluded. Third, this study only included children from Hong Kong so the reference ranges established may not be generalisable to other populations with different genetic and environmental factors.