WU Qin, REN Jie, YANG Li Chen, LIU Jie, WEI Jun, ZHANG Wei, LI Na, WANG Yue Jiao, YAN Li Huang, MA Jian Rong, and YANG Xiao Guang

?

Cut-off Values of Diagnostic Indices to Detect Iron Deficiency in Chinese Breast-fed Infants*

WU Qin1,$, REN Jie2,$, YANG Li Chen2,#, LIU Jie3, WEI Jun3, ZHANG Wei4, LI Na4, WANG Yue Jiao3, YAN Li Huang3, MA Jian Rong4, and YANG Xiao Guang2

1. Department of chronic disease control and Prevention, Tongzhou District Center for Disease Control and Prevention, Beijing 101100, China; 2. The Key Laboratory of Trace Element Nutrition MOH, National Institute of Nutrition and Health, Chinese Center for Disease Control and Prevention, Beijing 100050, China; 3. The People's Hospital of Beijing University, Beijing 100044, China; 4. Beijing Maternity Hospital Affiliated to Capital Medical University, Beijing 100026, China

Iron deficiency anemia is one of the most prevalent nutritional deficiency worldwide. The commonly used cut-off values for identifying iron deficiency are extrapolated from older children and may not be suitable for infants. Therefore, our study aimed to establish appropriate cut-off values for the evaluation of iron status in Chinese infants. Pregnant women who delivered at ≥37 gestational weeks with normal iron status were recruited. Later, infants with normal birth weight and who were breastfed in the first 4 months were selected. Blood samples were collected to assess hemoglobin, serum ferritin, soluble transferrin receptor, mean corpuscular volume and free erythrocyte protoporphyrin. Cut-offs of all iron indices were determined as the limit of 95% confidence interval.

Iron deficiency anemia (IDA) continues to be one of the most prevalent nutritional deficiency worldwide. Infants are especially susceptible because of high iron requirement for their rapid growth[1]. It has been demonstrated that even the early stage of IDA is associated with impaired neurodevelopment in infants and children[2].

The diagnosis of iron deficiency (ID) is based mainly on a battery of laboratory tests. Early ID status in infants is difficult to evaluate because iron parameters change with age, critical values overlap at various stages of ID, and infections and inflammations can influence iron status indices[3].

WHO/CDC[4]recommends hemoglobin (Hb), serum ferritin (SF), soluble transferrin receptor (sTfR), mean corpuscular volume (MCV) and free erythrocyte protoporphyrin (FEP) as the best iron variables to assess iron status. The commonly used cut-off values to identify ID at 6-12 months of age are Hb<110 g/L[5]and SF<10-12 μg/L[6]. These values are extrapolated from older children (approximately 7 million toddlers aged 1 to 2 years)[6]and may not be suitable for infants[3]. Moreover, there is lack of data on cut-off values of other measures of iron status, such as MCV and sTfR in infants and need further research[7]. Furthermore, there are fewer studies which have focused on the iron status indices distribution ranges in Chinese breast-fed infants, even though exclusive breast-feeding is generally recommended for the first 6months of life.

Therefore, our study aimed to establish appropriate cut-offs for the evaluation of iron status or diagnosis of IDA in Chinese infants.

We recruited 800 pregnant women (>18 years old) who delivered at ≥37 gestational weeks with normal iron status(Hb>120g/L and SF>12μg/L on the last prenatal examination) and without any symptoms of anemiafrom July 2010 to May 2012 at the two hospitals in Beijing. Only infants with normal birth weight (2500-4000g) were includedin the analysis. In subsequent follow-up at 4 months, only infants who were exclusively breastfedandhad Hb level >145 g/L were considered for the inclusion. Parents were discouraged from feeding the infants complementary foods containing iron between 4 to 6 months. Signed informed consent was obtained from parents. The review and approval of ethics committee was obtained.

Cut-off iron status indices were calculated as the limit (±2 SD) of 95% confidence interval () by setting a ‘healthy’ group sample with normal iron nutrition. Three methods were used to determine the ‘iron-normative group’. First, we enrolled study participants not likely to have a high prevalence of ID, but without any further selection. It was assumed that the infants born at full term with normal body weight who are exclusively breastfed by healthy mothers would have a replete iron store at 4 months of age. We called this the ‘unselected’ normative group.

It is known that infants with borderline ID at 4 months would be expected to become gradually more iron deficient with time (such as at 6 months) if they are not supplemented with iron. Thus, we excluded such 6-months-old infants as per the conventional cut-off values for iron status variables. We called this ‘iron-screened’ normative group. To avoid using adult reference values in infants, conventional cut-offs for infants at 6 months were established by calculating the 95% confidence interval of infants at 4 months. We excluded infants with any abnormal iron status variables [Hb, MCV, FEP, SF, or sTfR] as per 2 SD cut-off values at 4 months.

On the other hand, individuals with slight IDA were identified by analyzing the Hb response after a period of iron supplementation. The infants at 4 months were randomly assigned to either control group or ‘iron-supplement’ group. Infants in the control group were considered as the ‘unselected’ group while infants in the iron-supplemented group were given an iron supplement (liquid amino acid chelated iron, Singcom Pharma) at a dose of 1 mg/kg per day (standard treatment doses is 6 mg/kg per day) for 2 months, from the age of 4 months to 6 months. Hb response to iron supplementation was evaluated as an alternative definition of ID, and we called this ‘iron-supplemented’ group. After 2 months, infants with an increase in Hb level of >10 g/L[8]were regarded as possibly iron-deficient and excluded from the ‘iron-supplemented’ normative group at 4 months to further minimize the risk of including ID individuals (Figure 1).

Figure 1. showing the study design flowchart.

Venous blood samples were collected from infants at birth, 4 months and 6 months. At each stage, iron status indices including Hb, SF, sTfR, MCV, and FEP levels were tested. The presence of inflammation or infection was assessed by measuring the levels of acute-phase proteins: C-reactive protein (CRP) and α-glycoprotein (AGP). Infants with CRP>5 mg/L or AGP>1 g/L were excluded. Changes in Hb level were assessed following iron supplementation.

Automated analyzers (Sysmex XE 2100, Kobe, Japan) were used to assay Hb and MCV while CRP, SF, sTfR, and AGP levels were analyzed on an automatic analyzer (7080, Hitachi, Japan) using particle enhanced transmission immunoturbidimetric passed assay. FEP level was tested using afluorescence spectrometer (SB-046, PerkinElmer, USA). All reagents were purchased from Roche, USA.

Our preliminary results which included 40 infants at 4 months at the two hospitals showed that Hb level change in infants with ID in the iron supplement group was 8 g/L greater than that in the control group. Hence, permissible error in the estimated standard deviation () was 8 g/L with power (1-)=80%, significance l=0.05,meansvalue, using the formula:

N=2[(Z+Z)/]2(1)

:Sample size.Z: the standard normal difference oflevel.: the standard normal difference oflevel.: estimated standard deviation.: permissible error.

The minimum sample required was 21 infants in each group at 4 months. Considering the prevailing breastfeeding rates of about 50%-60% and the estimated subsequent loss of participants to follow-up, the minimum sample was 130 at the beginning.

The distribution of iron nutritional indices in the three groups were compared. The 2 SD cut-off values were used to diagnose ID in 4- and 6-months-old infants with nearly exclusive breastfeeding (although some water could be fed to the infants). The ratio of sTfR/SF was calculated by dividing sTfR (in μg/L) by SF (in μg/L).

All statistical analyses were performed using SPSS 17.0. Kolmogorov-Smirnov test was used to verify if iron status indices were consistent with normality. Statistical methods used to compare means weretest and analysis of variance (ANOVA) and significance was defined as<0.05.

We started with 800 pregnant women at the beginning. As the study progressed, some pregnant women or infants were dropped-out for not meeting our inclusion criteria including not exclusively breastfeeding the infants during their first 4 months and feeding the infants with iron-enriched supplementary foods while some were lost due to relocation. Out of 332 infants at 4 months, 23 were further excluded including 10 infants with CRP>5 mg/L or AGP>1 g/L, 5 infants who were fed iron-enriched supplementary foods, and 8 infants who were not exclusively breastfed. Therefore, the number of eligible newborn participants was 309. They had mean birth weight of 3.3±0.3 kg, were born at gestational week of 40±1, and had mean Hb of 175.0±18.0 g/L. The latest test before the delivery showed that Hb of mothers was 125.0±4.0 g/L and SF was 46.5±26.6 μg/L. At 4 months, the average weight of infants was 7.6±0.9 kg, height was 65.5±4.6 cm, and head circumference was 41.3±1.2 cm. During 4 to 6 months, 86 infants were excluded which included 63 infants who were lost, 8 infants who could not tolerate iron supplements and 2 infants had CRP>5 mg/L or AGP>1 g/L. In addition, 13 infants in the iron-supplement group were excluded for having an Hb response level >10 g/L after 2 months of iron supplementation. Thereby, 223 infants remained in the study at 6 months. Figure S1 (see in www.besjournal.com) showed the change in the sample size during the study.

By normality test, Hb, sTfR, MCV, and FEP were consistent with arithmetic normality, so arithmetic means were used. Since SF and sTfR/SF were consistent with a log-normal distribution, geometric mean was used.

To develop the cut-off values to evaluate the iron status of 4- and 6-months-old infants,test and ANOVAwere used to compare means of different iron variables in different groups (Table 1). As iron status did not differ significantly between the unselected group and the iron-supplemented group at 4 months (>0.05), the mean of the two groups were used to establish the cut-off values at 4 months. A similar method was used to develop the cut-off values of other iron indices such as sTfR, Hb, sTfR/SF ratio, and FEP for 6-month-old infants (Table 1). SF showed significant differences between unselected group, iron-supplemented group, and iron-screened group at 6 months (=0.01). After pairwise comparisons, the ‘unselected’ group had a significantly lower SF level than the other two groups which was possibly a reflection of the fact that there was a small proportion of iron-deficient infants in the unselected group (<0.05). So, the mean SF of iron-supplemented group and iron-screened group was used to establish the cut-off values.

Iron status variable distribution of infants in different groups is shown in Table 1 and the ±2 SD cut-off values for each variable in different groups are shown in Table 2. Our study suggested the following 2 SD cut-off values for the diagnosis of IDA in 4- and 6-months-old Chinese infants: SF<16 μg/L, sTfR>6 mg/L, Hb<101 g/L, MCV<72 fL, and FEP>205 μg/dL at 4 months; SF<11 μg/L, sTfR>6 mg/L, Hb<100 g/L, MCV<69 fL, and FEP>209 μg/dL at 6 months (Table 2). Hb is the commonly used index for the diagnosis of ID and we established the cut-off Hb levels for full-term breastfed infants as 101 g/L and 100 g/L at 4 and 6 months, respectively. As per a WHO report, the cut-off Hb levels for 6-12-month-old infants was 110 g/L[9], which is higher than our result. One of the reason for this difference may be that the WHO data was extrapolated from older age groups[7]. A study conducted on 253 healthy Swedish and Honduran infants suggested 105 g/L as the cut-off value of Hb for both 4- and 6-month-old infants[3]. Another study in Britain suggested 95 g/L as the cut-off value of Hb for 8-month-old infants[7]. Considering that Hb concentration shows a physiological descent during the age of 6-9 months, it is reasonable that cut-off values for the 4- and 6-month-old infants in our study are higher than 95 g/L.

Our suggested cut-off value of MCV of 72 fL at 4 months and 70 fL at 6 months are consistent with those in the literature. Cut-off values in a study on Finnish infants who were fed iron-fortified formula were 76 fL at 4 months and 68-70 fL at 6-9 months[10]while a study in Swedish and Honduran infants suggested 73 fL at 4 months and 71 fL at 6 months[3].

Our suggested cut-off SF levels of 16 μg/L at 4 months and 11 μg/L at 6 months were similar to those reported in a research conducted in Sweden and Honduras infants (=197)[3]in which -2SD cut-offvalues were 20 μg/L and 9 μg/L at 4 and 6 months of age, respectively. Another study of Finnish infants who were fed with iron-fortified formula (=36-43) suggested -2 SD cut-off values of 37 μg/L and 19 μg/L at 4 and 6 months of age, respectively[11]. The higher SF values in that study could be a result of iron-fortified formula feeding. All the above-mentioned results differ from the 12 μg/L cut-off value at 6-12 months of age proposed by the World Health Organization (WHO) in 2003[12].

Table 1. Comparing of the Means of Iron Status Variables Between Infants in the Unselected, Iron-supplemented, and Iron-screened Groups

a: Geometric mean±SDThe arithmeticis added (subtracted) to the mean, whereas the geometricis multiplied (divided) by the geometric. Geometric=expe,elnx.b: SF has significant differences between the unselected group, iron-supplemented group and iron-screed group (=0.01).c: Hb: hemoglobin, SF: serum ferritin, sTfR: soluble transferrin receptor, MCV: mean corpuscular volume, FEP: free erythrocyte protoporphyrin.d:test between unselected and iron-supplemented group.e: ANOVA test among the unselected group, iron-supplemented group and iron-screened group.

Table 2.±2 SD Cut-off Values for Chinese Infants at 4 and 6 Months of Age

a: Geometric mean±SD. The arithmetic SD is added (subtracted) to the mean, whereas the geometric SD is multiplied (divided) by the geometric mean.b: Hb: hemoglobin, SF: serum ferritin, sTfR: soluble transferrin receptor, MCV: mean corpuscular volume, FEP: free erythrocyte protoporphyrin.C:of screened: When discussing about Hb, infants with abnormal iron status variables of MCV, FEP, SF and sTfR were excluded, 53 infants remained, so=53. SF:=87, sTfR:=76, MCV:=82, FEP:=73, sTfR/SF:=83.d: The 2 SD cut-off values for the diagnosis of IDA in 4- and 6-month-old Chinese infants: SF<16 μg/L, sTfR>6 mg/L, Hb<101 g/L, MCV<72 fL, and FEP>205 μg/dL at 4 months; SF<11 μg/L, sTfR>6 mg/L, Hb<100 g/L, MCV<69 fL, and FEP>209 μg/dL at 6 months.

Our study suggested a cut-off sTfR level of 6 mg/L both at 4 months and 6 months. These values are consistent with those in two Canadian studies: one study of 389 Canadian ID infants suggested a cut-off sTfR level of 7 mg/L for male infants and 6 mg/L for female infants at 4-6 months of age[13]; and another study of 485 Canadian healthy infants suggested a cut-off value of 6.6 mg/L at 9-15 months[14]. These two studies also used enzyme immunoassay methods to analyze sTfR, but the reagents used were from a different company. One major problem in comparing the results of sTfR from different studies is the wide distribution range among laboratories and different commercial kits; however, it has been reported that even though the values obtained from different sTfR detection kit may differ, they have good correlation and comparable validity to identify ID if the reference values for that particular assay were used[1].

Our study suggested the sTfR/SF cut-off values of 257 at 4 months and 559 at 6 months. In one study of 515 Chilean infants, aged 8-15 months, the suggested sTfR/SF cut-off value was 975[15]. A possible reason for this difference is the higher age of the infants in the Chilean study as SF level drops after 4 months of age, while sTfR level remains steady.

Our study obtained the appropriate cut-off values for the assessment of iron status in Chinese infants aged 4-6 months using a ‘healthy-group approach’. Although, the sample size of our study was smaller than in most other studies on reference values for other age groups, the advantage of our dataset is the strict definition of breast-feeding, presence of an iron-supplemented group of infants as well as an un-supplemented group. In addition, CRP and AGP were used as the inflammation markers and the two-hospital design provided the wide range in iron status, all of which is essential for evaluating iron status cut-off values during infancy.

We will continue to coordinate with clinicians to validate the cut-off values determined in this study.

We wish to express our gratitude to HAN Chi and ZHANG Xin for their assistance with the experiment.

1. Domell?f M1, Braegger C, Campoy C, et al. Iron requirements of infants and toddlers. J Pediatr Gastroenterol Nutr, 2014; 58, 119-29.

2. Lozoff B, Smith JB, Kaciroti N, et al. Functional significance of early-life iron deficiency: outcomes at 25 years. J Ped, 2013; 163, 260-6.

3. Domell?f M1, Dewey KG, L?nnerdal B, et al. The diagnostic criteria for iron deficiency in infants should be reevaluated. J Nutr, 2002; 132, 3680-6.

4. WHO/CDC. Assessing the iron status of population. In. 2nd ed. Geneva, Switzerland: World Health Organization, 2004.

5. World Health Organization (WHO), United Nations Children's Fund (UNICEF), United Nations University (UNU). Iron deficiency anaemia: assessment, prevention and control, a guide for programme managers. Geneva: World Health Organization, 2001.

6. Looker AC, Dallman PR, Carroll MD, et al. Prevalence of iron deficiency in the United States. JAMA: the journal of the American Medical Association, 1997; 277, 973-6.

7. Aggett P, Agostoni C, Axelsson I, et al. Iron metabolism and requirements in early childhood: do we know enough? A commentary by the ESPGHAN Committee on Nutrition. J Pediatr Gastroenterol Nutr, 2002; 34, 337-45.

8. Baker RD and Greer FR. Diagnosis and prevention of iron deficiency and iron-deficiency anemia in infants and young children (0-3 years of age). Pediatrics, 2010; 126, 1040-50.

9. Stoltzfus RJ, Dreyfuss ML, LSI. International, Guidelines for the use of iron supplements to prevent and treat iron deficiency anemia. 1998: ILSI Press Washington.

10.Saarinen UM, Siimes MA. Developmental changes in red blood cell counts and indices of infants after exclusion of iron deficiency by laboratory criteria and continuous iron supplementation. J Pediatr, 1978; 92, 412-6.

11.Saarinen UM, Siimes MA. Serum ferritin in assessment of iron nutrition in healthy infants. Acta Paediatr Scand, 1978; 67, 745-51.

12.World Health Organization (WHO), United Nations Children's Fund (UNICEF). Global Strategy for Infant and Young Child Feeding.In, 2003.

13.Yeung GS, Zlotkin SH. Percentile estimates for transferrin receptor in normal infants 9-15 mo of age. Am J Clin Nutr, 1997; 66, 342-6.

14.Punnonen K, Irjala K, Rajamaki A. Serum transferrin receptor and its ratio to serum ferritin in the diagnosis of iron deficiency. Blood, 1997; 89, 1052-7.

15.Pisacane A. Neonatal prevention of iron deficiency. BMJ, 1996; 312, 1102-3.

Figure S1. The change in the sample size durting the study.

$These two authors contributed equally to this work.

#Correspondence should be addressed to: YANG Li Chen, E-mail: yanglichen28@126.com, Tel: 86-10- 83132906, Fax: 86-10-67711813.

Biographical notes of the first authors: WU Qin,female,born in 1985,Master graduate,majoring in trace iron; REN Jie, female, born in 1991, Master graduate, majoring in trace iron.

Accepted: November 1, 2016

10.3967/bes2016.111

August 4, 2016 ;

*This project was supported by Natural Science Foundation of China (Grant No. 30972483); The Chinese clinical trial registry number is ChiCTR-TRC-12002838.