Anaemia is a significant global public health issue, characterised by low blood Hb concentrations and associated with long-term cognitive and motor impairments, particularly in children(1,2) . In Brazil, the prevalence of anaemia and Fe deficiency anaemia among children aged 6 to 59 months is reported to be 10·0 % and 3·6 %, respectively(3). Fe deficiency is the leading cause of anaemia, and a substantial proportion of the world’s population has a diet that is poor in this essential mineral(Reference Kassebaum4). Apart from dietary intake, Fe is also obtained through recycling of senescent red blood cells within the body(Reference Camaschella, Nai and Silvestri5).
Fe homoeostasis is tightly regulated by hepcidin, a key peptide that plays a crucial role in communication between various sites involved in Fe absorption, utilisation and storage(Reference Camaschella, Nai and Silvestri5–Reference Muckenthaler, Rivella and Hentze7). Hepcidin acts by downregulating ferroportin, the protein responsible for releasing Fe from cells. As a result, increased levels of hepcidin are associated with impaired duodenal absorption and recycling of Fe from senescent erythrocytes. Hepcidin expression is influenced by multiple factors, including Fe overload and inflammation, as well as anaemia and hypoxemia, which can lead to decreased hepcidin levels(Reference Camaschella, Nai and Silvestri5). Understanding the intricate mechanisms involved in Fe homoeostasis is crucial for addressing the public health concern of Fe-deficiency anaemia, particularly in children under 5 years of age.
The serine transmembrane protease 6 (TMPRSS6) gene encodes the matriptase-2 protein (MT-2), which plays a crucial role in the regulation of Fe metabolism. MT-2 has been shown to suppress hepcidin, a key regulator of Fe homoeostasis, through cleavage of hemojuvelin, a co-receptor for bone morphogenetic proteins involved in hepcidin synthesis(Reference Muckenthaler, Rivella and Hentze7–Reference Du, She and Gelbart9). By inhibiting hepcidin production, MT-2 enhances the availability of Fe for erythropoiesis and other essential cellular functions(Reference Nemeth, Tuttle and Powelson8).
Emerging evidence suggests that genetic factors, including single nucleotide polymorphisms (SNP), contribute significantly to the regulation of Fe stores and the development of Fe deficiency(Reference Tanaka, Roy and Yao10). Of particular interest is the TMPRSS6 rs855791 SNP)], which has been extensively studied due to its association with various Fe-related parameters(Reference Benyamin, Ferreira and Willemsen11). This SNP involves a thymine (T) to cytosine (C) substitution at position 2321 of the gene-coding region, resulting in an alanine to valine substitution at codon 736 (A736V) of the MT-2 protein(Reference Chambers, Zhang and Li12,Reference Silvestri, Pagani and Nai13) . One in vitro study has demonstrated that the presence of valine at codon 736 is associated with increased concentrations of hepcidin(Reference Nai, Pagani and Silvestri14). This suggests that the genetic variation introduced by SNP TMPRSS6 rs855791 can influence hepcidin regulation and, consequently, impact Fe metabolism and risk for anaemia and Fe deficiency.
A study by Benyamin and cols. (2009) has shown that Caucasian and Asian individuals carrying the T allele of SNP TMPRSS6 rs855791 exhibit lower serum Fe concentrations and transferrin saturation compared with those who carry the C allele(Reference Benyamin, Ferreira and Willemsen11). Furthermore, genome-wide association studies (GWAS) have consistently demonstrated a negative association between genetic variants in TMPRSS6 (such as rs855791 and rs4820268) and reduced Fe concentrations, transferrin saturation, as well as decreased Hb concentrations, mean corpuscular volume (MCV) and mean corpuscular haemoglobin (MCH)(Reference Tanaka, Roy and Yao10,Reference Benyamin, Ferreira and Willemsen11,Reference Traglia, Girelli and Biino15–Reference Oexle, Ried and Hicks17) .
Based on these findings, we hypothesise that the T allele of SNP TMPRSS6 rs855791 may be associated with Fe deficiency and anaemia in Brazilian children. Therefore, our aim was to evaluate the associations between SNP TMPRSS6 rs855791 and biomarkers of anaemia and Fe deficiency in Brazilian children attending day care centres. By examining the associations of this genetic variation with Fe-related parameters, our study will contribute to the understanding of Fe metabolism regulation and provide insights into potential avenues for personalised interventions.
Materials and methods
Study design, participants and ethics
This cross-sectional study utilised baseline data from a randomised clustered clinical trial(Reference Machado, Lopes and Schincaglia18) involving children aged 6 to 42 months who attended Early Childhood Education Centres in a Brazilian capital (Fig. 1). Early Childhood Education Centres with fewer than seven children in the age groups of interest and those that were not operating full time were excluded from the clinical trial. Children with and without anaemia, between 6 and 42 months of age, were included in the study, while those undergoing treatment for anaemia, malaria, HIV, haemoglobinopathies or haemochromatosis; with low birth weight, premature delivery (< 37 weeks), twins or allergy to any components of the fortification sachet and/or ferrous sulphate/folic acid were excluded. Children whose parents/guardians did not authorise DNA extraction, did not have enough blood for DNA extraction or had missing data from haematological and biochemical tests were also excluded.
Fig. 1. STROBE flow chart of participants’ recruitment.
This study was conducted according to the guidelines laid down in the Declaration of Helsinki, and all procedures involving human subjects were approved by the Research Ethics Committee of the Federal University of Goiás (protocol 2·641·285, 8 May 2018, CAAE: 80541717·3·0000·5083). Written informed consent was obtained from all children’s parents/guardians. The study ‘Effect of fortification with powder nutrients in prevention and treatment of nutrient deficiency’ was registered at the Brazilian Registry of Clinical Trials – ReBEC (protocol RBR-4hm7 mz; https://ensaiosclinicos.gov.br/rg/RBR-4hm7mz).
Data collection
Data were collected from March 2018 to August 2018 in ten Early Childhood Education Centres through a structured questionnaire administered to the children’s parents/guardians by trained interviewers. The questionnaire collected information on socio-economic, demographic and individual child characteristics, including sex, age, race, family income, number of family members, maternal level of education, day care attendance time, birth weight, length at birth, current weight and exclusive breast-feeding up to 6 months.
Body weight was measured in duplicate using a digital scale (SECA 877®, Seca Deutschland) with a capacity of 200 kg and an accuracy of 100 g, following a standardised protocol(Reference Lohman, Roche and Martorell19). For children younger than 2 years, weight was measured with the child on the lap of the mother/guardian or the teacher/supervisor(Reference Monego, Vieira and Menezes20).
Fe intake was determined using two methods: direct food weighing and a 24-h food recall method (R24h). Direct food weighing was conducted at the education centres to assess food consumption(Reference Cruz, Souza and Philippi21). On the same day as the direct weighing, the R24h method was applied to parents or individuals responsible for assessing food consumption at home. Home measures and photographic records were used as auxiliary tools for dietary surveys(Reference Zabotto, Viana and Gil22). Data from the R24h and food weighing were calculated using the Diet Win Professional Plus® software (Brubins Food Commerce, Brazil). Food composition data from national tables(Reference Giuntini23–Reference Philippi25) and food labels were incorporated into the software as needed. Breast-feeding was assessed by estimating the volume of milk based on the number of feeds per day(26).
Venous blood collection was performed by a trained technician. The collection was preferably done after an 8-h fast, but at least 3 h of fasting was ensured for children younger than 12 months. The blood samples were appropriately refrigerated and transported to the laboratory for further analysis.
The blood count was analysed using electronic count (Sysmex XE-2100, Sysmex Corporation). Anaemia was defined as Hb levels below 11 g/dl, and low haematocrit values were defined as below 33 %, based on the recommendations of the WHO(1). Microcytosis was defined as MCV below 70 fl for children aged 6–23 months and below 73 fl for those aged 24–59 months(1,Reference Hadler, Juliano and Sigulem27) . Hypochromia was identified as MCH values below 22 pg for children aged 6–23 months and below 25 pg for those aged 24–59 months. Altered mean corpuscular haemoglobin concentration was defined as below 32 g/dl for children aged 6–59 months(1,Reference Hadler, Juliano and Sigulem27) . Anisocytosis was defined as red cell distribution width (RDW) above 14·5 %(Reference Oski28,Reference Walters and Abelson29) .
Serum concentrations of α-1-acid glycoprotein (AGP) were analysed using immunoturbidimetry (AU480 Chemistry Analyzer, Beckman Coulter, Brea), while serum concentrations of C-reactive protein (CRP) were determined by turbidimetry (Cobas 8000 c502, Roche Diagnostics). Children with serum AGP levels exceeding 100 mg/dl and/or CRP levels exceeding 0·5 mg/dl were considered to have inflammation(Reference Lundeen, Lind and Clarke30,31) . Serum ferritin concentrations were analysed by chemiluminescence (UniCel DxI 800, Beckman Coulter), and Fe deficiency assessed by ferritin was defined as ferritin levels below 12 ng/dl in the absence of inflammation and below 30 ng/dl in the presence of inflammation(31). Serum soluble transferrin receptor (sTfR) concentrations were analysed by nephelometry (BN II System, Siemens Healthineers). Fe deficiency was defined as sTfR levels exceeding 1·76 mg/l, which is the reference value from the analysis kit.
Genomic DNA was extracted from leukocytes in whole blood using the PureLink Genomic DNA Mini Kit® (Thermo Fisher Scientific). Quantification was performed using a Qubit fluorimeter® (Invitrogen), and DNA purity was verified using a NanoDrop spectrophotometer® (Thermo Fisher Scientific). Samples were considered of good quality when the A260/280 ratio fell between 1·7 and 2·0.
Genotyping was conducted using real-time polymerase chain reaction (RT-PCR) with an Applied Biosystems TaqMan® SNP Genotyping assay (Thermo Fisher Scientific) for SNP TMPRSS6 rs855791. Amplification reactions for SNP genotyping were performed in a StepOne ® thermocycler (Thermo Fisher Scientific). As a methodological control, 10 % of the samples were double genotyped. The analysis and processing of fluorescence data for allelic discrimination were performed using the StepOne® v. 2·1 software (Thermo Fisher Scientific).
Statistical analyses and justification of sample size
The data were entered twice using Epi Info version 6·04d (CDC) and Microsoft Excel 2010 (Microsoft) to ensure consistency. To assess the goodness of fit to the normal distribution, the Shapiro–Wilk test was used. Normally distributed continuous variables were presented as mean ± sd, while non-normally distributed variables were presented as median with interquartile range. Categorical variables were presented as absolute (n) and relative (%) frequencies.
The sample size calculation was based on an absolute error of 5 %, a power of 80 %, an allocation rate of 1:1 and an effect size of 0·38. The effect size was determined based on the prevalence of anaemia found in a previous Brazilian study(Reference Cardoso, Augusto and Bortolini32). The calculated sample size was 164 children, and an additional 22 % was added to account for potential losses during the data collection process, resulting in a total sample size of 200 children.
Multiple logistic regression models with hierarchical selection of variables were utilised to estimate the effects of various independent variables, including children’s sex and age, maternal level of education, family income, number of family members, day care time, weight and length at birth, breast-feeding up to 6 months, dietary Fe intake, CRP, AGP, current weight, as well as additive and dominance effects of the T allele, on categorised outcomes such as anaemia, Fe deficiency and haematological biomarkers. Linear regression models with hierarchical selection of variables were employed to estimate the effects on blood markers including Hb, ferritin, sTfR and other haematological biomarkers.
In both regression models, the variables were organised into blocks and prioritised based on their presumed influence on the outcome. The distal block encompassed variables such as family income, maternal level of education, number of dependents and children’s sex and age. The medial block included day care time, while the proximal block included weight and length at birth, current weight, exclusive breast-feeding up to the sixth month, dietary Fe intake, CRP and AGP concentrations, as well as additive and dominance effects of the T allele of SNP TMPRSS6 rs855791.
Multiple models were created by testing variables from the distal to the proximal block using an automated stepwise approach. Variables that were associated with the presence of the T allele were identified based on a significance level of P value < 0·05 after adjusting for potential confounders within the same block and superior hierarchical block(s). In cases where significant deviations from normality were detected, the models were adjusted using a boxcox transformation. However, as the results obtained with and without transformation were congruent, the results were presented without any transformation.
The adherence of SNP TMPRSS6 rs855791 to the Hardy–Weinberg equilibrium was assessed using the Pearson χ2 test, conditioned on the allele frequencies estimated from the study children. All statistical analyses were conducted using R version 4.2.2(33). The significance level for all analyses was set at 5 %.
Results
Sample characteristics
The final sample consisted of 163 children, of whom 53·4 % were female, with a median age of 24·0 (14·5–34·5) months. The children had been attending the Early Childhood Education Centre facility for a median duration of 13·0 (4·0–24·0) months. The frequency of the minor allele was 0·399. The prevalence of anaemia was 14·4 % (n 22), while the prevalence of Fe deficiency, assessed by serum ferritin and sTfR concentration, was 21·6 % (n 35) and 65·4 % (n 106), respectively. Other socio-economic, demographic, genotype and biochemical data are provided in Table 1.
Table 1. Baseline characteristics of the sample (n 163)
Continuous variables are presented as median (interquartile range) and categorical variables are presented as absolute number [percentage]. MAF, minor allele frequency; CRP, C-reactive protein; AGP, alpha-1-acid glycoprotein; sTfR, soluble transferrin receptor; MCV, mean corpuscular volume; MCH, mean corpuscular haemoglobin, MCHC, mean corpuscular haemoglobin concentration; RDW, red cell distribution width.
* Microcytosis (< 70 fl–6 to 23 months; < 73 fl–24 to 59 months).
† Hypochromia (< 22 pg–6 to 23 months; < 25 pg–24 to 59 months).
‡ Altered (< 32 g/dl–6 to 59 months).
§ Anisocytosis (> 14·5 %). Sample sizes for sex, age, skin colour, maternal education level, family members, day care time, Hb, SNP TMPRSS6 rs855791, haematocrit, MCV, MCH, MCHC and RDW: n 163; family income, current weight, ferritin, CRP, AGP, sTfR and Fe deficiency–ferritin and sTfR: n 162; birth weight: n 160; birth length: n 156.
Association of anaemia, iron deficiency and biochemical and haematological markers with SNP TMPRSS6 rs855791
The distribution of genotypes observed in the study did not deviate significantly from the expected frequencies under the Hardy–Weinberg equilibrium (χ2 = 0·0076; P = 0·9305). In the additive model analysis of SNP TMPRSS6 rs855791 (CC × CT × TT), the T allele was significantly associated with anaemia and Fe deficiency assessed by sTfR concentrations. For each copy of the T allele, there was a 4·5-fold increase in the odds of presenting anaemia (OR = 4·49, 95 % CI = 1·56, 12·91, P = 0·005) and a 4·2-fold increase in the odds of presenting Fe deficiency assessed by sTfR concentrations (OR = 4·23, 95 % CI = 2·07, 8·65, P < 0·001). No associations between the dominant model and anaemia were observed, and this model was not included in the final model for the other outcomes.
Anaemia and Fe deficiency assessed by sTfR concentrations were significantly associated with maternal level of education, family income, AGP and CRP (P < 0·05). In addition, male sex was associated with a 3·8-fold increase in the odds of developing Fe deficiency assessed by sTfR concentrations compared with female sex (OR = 3·80, 95 % CI = 1·57, 9·20, P = 0·022, Table 2). Separate models for males and females were analysed, and differences were observed only for Fe deficiency assessed by sTfR concentrations. For each copy of the T allele, female children had a 6·2-fold increase in the odds of developing Fe deficiency (OR = 6·2, 95 % CI = 2·3, 16·6, P < 0·001), and only male children showed a 2·6-fold increase in the odds of presenting high RDW (OR = 2·6, 95 % CI = 1·04, 6·3, P = 0·041) for each copy of the T allele (data not shown).
Table 2. Association of SNP TMPRSS6 rs855791 genotypes with biochemical and haematological markers of anaemia and iron deficiency (n 153)*
(a), Additive model (CC × CT × TT); d, dominant model (CC × CT + TT); AGP, alpha1-acid glycoprotein; sTfR, soluble transferrin receptor; CRP, C-reactive protein; MCV, mean corpuscular volume; MCH, mean corpuscular haemoglobin; MCHC, mean corpuscular haemoglobin concentration; RDW, red cell distribution width.
* Only children with complete data for all evaluated markers were included.
† Adjusted for age, number of dependents, birth weight, exclusive breast-feeding up to sixth month and mean weight (0·05 > P ≤ 0·20).
‡ Adjusted for age (0·05 > P ≤ 0·20).
§ Adjusted for exclusive breast-feeding up to sixth month (0·05 > P ≤ 0·20).
|| Adjusted for AGP (0·05 > P ≤ 0·20).
¶ Adjusted for sex (0·05 > P ≤ 0·20).
** Adjusted for income, number of dependents and CRP (0·05 > P ≤ 0·20).
The results of the linear regression analyses revealed significant associations between genetic and demographic factors and various haematological parameters. Each copy of the T allele was found to be associated with an increase of 0·18 mg/l in sTfR concentrations (P < 0·001) and reductions of 1·41 fl in MCV (P = 0·006) and 0·52 pg in MCH (P = 0·013), as shown in Table 3. Moreover, an increase of 0·01 mg/l in sTfR concentration (P < 0·001) was observed for each unit increase in AGP. Additionally, with each month increase in the child’s age, there was a reduction of 0·01 in sTfR concentrations (P = 0·004) and 0·06 in RDW (P < 0·001), accompanied by increases of 0·13, 0·07 and 0·04 in MCV, MCH and mean corpuscular haemoglobin concentration, respectively (P < 0·001). Furthermore, each additional year of maternal education was associated with an increase of 0·34 in MCV (P = 0·001). Notably, male sex was found to be positively associated with sTfR concentrations (P = 0·013) and RDW (P = 0·003), while being negatively associated with MCV (P < 0·001) and MCH (P < 0·001), as indicated in Table 3.
Table 3. Association of the additive model of SNP TMPRSS6 rs855791 with haematological and biochemical markers – final model (n 153)*
(a), Additive model (CC × CT × TT); AGP, alpha1-acid glycoprotein; sTfR, soluble transferrin receptor; CRP, C-reactive protein; MCV, mean corpuscular volume; MCH, mean corpuscular haemoglobin; MCHC, mean corpuscular haemoglobin concentration; RDW, red cell distribution width.
* Only children with complete data for all evaluated markers were included.
† Adjusted for maternal level of education and birth length (0·05 > P ≤ 0·20).
‡ Adjusted for sex, number of family members and AGP (0·05 > P ≤ 0·20).
§ Adjusted for income, exclusive breast-feeding up to sixth month and CPR (0·05 > P ≤ 0·20).
|| Adjusted for birth weight (0·05 > P ≤ 0·20).
¶ Adjusted for income, birth weight and AGP (0·05 > P ≤ 0·20).
** Adjusted for income (0·05 > P ≤ 0·20).
Discussion
This is the first study to evaluate associations between SNP TMPRSS6 rs855791 and various biochemical, haematological and immunological markers related to anaemia and Fe deficiency in Brazilian children. The T allele was associated with an increased odds of developing anaemia and Fe deficiency, as assessed by sTfR concentrations. Additionally, each copy of the T allele was associated with an increase of 0·18 mg/l in sTfR concentrations and reductions of 1·41 fl and 0·52 pg in MCV and MCH, respectively.
Comparing our results with previous studies, the frequency of the T allele in Brazilian children was similar to that found in Polish children with coeliac disease(Reference Urbaszek, Drabińska and Szaflarska-Popławska34) and other Caucasian, European and Asian populations(Reference Benyamin, Ferreira and Willemsen11,Reference Chambers, Zhang and Li12,Reference Wanjiku, Towers and Swinkels35) , but higher than South African populations(Reference Gichohi-Wainaina, Melse-Boonstra and Swinkels36). These differences between populations may potentially impact the outcomes of various diseases, including anaemia and Fe deficiency.
In line with our study, previous research has also explored the role of TMPRSS6 gene variants in Fe-related haematological parameters in different populations. Our findings are consistent with previous GWAS that identified associations between SNP TMPRSS6 rs855791 and lower values of Hb, Fe and MCV in European, North American, Asian and Australian populations(Reference Benyamin, Ferreira and Willemsen11,Reference Chambers, Zhang and Li12,Reference Read, Schlauch and Elhanan37) . Consistent with our findings, Batar et al. (2018) reported significant associations between SNP TMPRSS6 rs855791 and several haematological parameters in Turkish patients with Fe-deficiency anaemia(Reference Batar, Bavunoglu and Hacioglu38). However, our results differ from a study in Indonesian children, where no associations were observed between this SNP and sTfR and Hb concentrations(Reference Shinta, Adhiyanto and Htet39).
In healthy individuals, hepcidin binds to ferroportin, forming a complex that prevents Fe outflow and increases cellular Fe stores as ferritin(Reference Camaschella, Nai and Silvestri5,Reference Ginzburg6) . However, in individuals carrying the SNP TMPRSS6 rs855791, which causes a change in amino acids in MT-2, the Fe flow in enterocytes is altered, leading to reduced absorption(Reference Chambers, Zhang and Li12). Carriers of the risk allele exhibit inefficient regulation of hepcidin synthesis, resulting in consistently high concentrations of hepcidin, even during Fe deficiency. This sustained elevation of hepcidin levels prevents the restoration of Fe homoeostasis at both intestinal and macrophage levels, leading to recurrent Fe deficiency. As a result, oral supplementation may not be fully effective in these individuals, as its effectiveness depends on intestinal absorption(Reference Benyamin, Ferreira and Willemsen11,Reference Nai, Pagani and Silvestri14,Reference Urbaszek, Drabińska and Szaflarska-Popławska34) .
The presence of the T allele of SNP TMPRSS6 rs85579 in carriers is typically associated with low Fe concentrations. This genetic variant promotes the formation of a complex between responsive elements and Fe-regulating proteins in the mRNA of cellular transferrin receptor, leading to an increase in transferrin receptor synthesis as a compensatory mechanism for Fe uptake(Reference Wang and Babitt40). The concentration of transferrin receptor is directly correlated with sTfR concentrations(Reference Read, Schlauch and Elhanan37). A meta-analysis conducted in European populations estimated that the T allele is associated with a 0·20 mg/l increase in sTfR concentrations(Reference Wanjiku, Towers and Swinkels35), which supports our findings.
While sTfR concentrations could provide valuable information to complement the diagnosis of anaemia, only Hb concentrations have been prioritised. However, a definitive diagnosis of anaemia also considers MCV and MCH values(Reference Cappellini, Russo and Andolfo41,Reference Dallman, Reeves and Stekel42) , which indicate microcytosis and hypochromia associated with Fe deficiency anaemia(Reference Buttarello43). Interestingly, our findings revealed associations between SNP TMPRSS6 rs85579 and MCV and MCH, showing significant reductions in these markers with each copy of the T allele, suggesting the need for further investigation.
Apart from the significant associations with SNP TMPRSS6 rs85579, it appears that other genetic variations may also be involved in the regulation of sTfR concentrations. For instance, in African individuals, SNP TF rs1799852 was associated with lower transferrin levels and higher sTfR concentrations(Reference Gichohi-Wainaina, Melse-Boonstra and Swinkels36). In Australian individuals, SNP HFE rs1800562 and TF rs1799852 and rs3811647 together explained 40 % of the variation in sTfR concentrations(Reference Benyamin, Ferreira and Willemsen11). Similarly, in African women, SNP TF rs1799852 and rs3811647 explained 13 % of the variations in sTfR concentrations(Reference Gichohi-Wainaina, Melse-Boonstra and Swinkels36).
In addition to the associations between SNP TMPRSS6 rs855791 and markers of anaemia and Fe deficiency, we identified associations between these markers and other variables, including age, income, maternal education level, current weight, sex and inflammatory status. These findings highlight the importance of considering socio-demographic variables when evaluating anaemia and Fe deficiency, as they are often intrinsic determinants of poor nutrition. Our results are supported by other studies that have reported similar relationships between markers of anaemia and Fe deficiency with maternal education level, family income, child’s age, sex and concentrations of CRP and AGP(Reference Engle-Stone, Aaron and Huang44–Reference Zuffo, Osório and Taconeli46).
It is noteworthy that respiratory and inflammatory diseases are common in children attending day care centres and educational institutions, likely due to close contact with other children and the immaturity of their immune system(Reference Lambrecht, Bridges and Wilson47,Reference Pedraza, Queiroz and Sales48) . Fe deficiency and inflammation are closely related, as elevated concentrations of hepcidin are often found in children with pre-existing inflammatory conditions(Reference Prentice, Bah and Jallow49). Therefore, our findings emphasise the importance of further investigating inflammatory markers, anaemia, Fe deficiency and their associations with SNP TMPRSS6 rs85579.
Our study has several limitations, including the absence of assessment for hepcidin levels. Additionally, we only investigated one SNP in the TMPRSS6 gene, although it is well established that other genetic variants may also be associated with the outcomes we investigated. Despite the significance of our findings, we acknowledge the need for further investigations in other regions of Brazil, particularly due to the high level of population miscegenation. Nevertheless, this study is the first to analyse the associations between SNP TMPRSS6 rs855791 and a wide range of parameters related to anaemia and Fe deficiency in Brazilian children, including inflammatory markers, which expands our understanding of the genetic aspects of Fe metabolism.
In conclusion, our study revealed significant associations between the T-risk allele of SNP TMPRSS6 rs855791 and anaemia and Fe deficiency in Brazilian children attending day care centres. Children carrying the T-risk allele had an increased risk of developing anaemia and Fe deficiency, as evidenced by higher serum concentrations of sTfR. Moreover, the dose of the T allele was associated with lower MCV and MCH values. These findings have implications for directing intervention studies and developing precision nutrition-based strategies to mitigate the adverse effects of these health problems.
It is important to note that the development of effective precision nutrition strategies requires a multifactorial understanding of Fe homoeostasis, including the influence of other genetic variants, dietary factors and environmental influences. Therefore, it is crucial to conduct additional research to identify other relevant genetic markers and their interactions, as well as to investigate gene–environment interactions to refine the selection process for precision nutrition strategies.
Acknowledgements
*Members of FORNUTRI (Fortification with NutriSUS) Working Group: Paulo Sérgio Sucasas da Costa, Ana Paula Viana de Siqueira, Bárbara Pimenta, Vanessa Farias Franco and Nágila Alencar Afyoni. The authors would like to thank all the professionals of the education centres, parents/guardians and children who trusted our work and to the team that helped in data collection.
Financial support
Funding was provided by Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) [public call CNPq/MS/SCTIE/DECIT/SAS/DAB/CGAN Nº 13/2017, grant Nº 408786/2017–5] and by Fundação de Amparo à Pesquisa do Estado de Goiás (FAPEG) [public call Nº 04/2017, Programa Pesquisa para o SUS, FAPEG/SES-GO/CNPq/MS/DECIT/2017 PPSUS-GO, grant Nº 201710267001247]. The funders had no role in the design, analysis or writing of this article.
Authorship
The authors’ contributions are as follows: N.M.S. and M.P.L. participated in data acquisition, analysis, interpretation, discussion and wrote the first version of the manuscript; R.M.S. and A.S.G.C. participated in data analysis, interpretation, discussion and reviewed the manuscript. C.C. was the co-supervisor, contributed to data interpretation and wrote and reviewed the manuscript; M.C.C.M.H. was the supervisor responsible for the study conception, data interpretation and reviewed the manuscript. Paulo Sérgio Sucasas da Costa, Ana Paula Viana de Siqueira, Bárbara Pimenta, Vanessa Farias Franco and Nágila Alencar Afyoni for pediatric care, prescription of medications and data acquisition. Maria Aderuza Hortz who contributed to conceptualization and methodology. All authors read and approved the final version of the manuscript.
Conflict of interest
The authors declare no conflict of interest.
