Prevalence and factors related to anaemia in children aged 6–59 months attending a quaternary health facility in Maputo, Mozambique

ABSTRACT Globally, anaemia prevails as a public health issue, being also a concern in Mozambique, where about two-thirds of children 6–59 months of age are affected by this condition. We carried out this study to estimate anaemia prevalence and evaluate structural determinants and haematological parameters association among children aged 6–59 months attending pediatric inpatient and outpatient services in a Quaternary Health Facility in Maputo City Province, Mozambique. We collected data from 637 inpatients or outpatients who attended pediatric consultations at the Maputo Central Hospital. The overall rate of anaemia in children aged 6–59 months was 62.2% (396/637), with 30.9% moderate anaemia (197/637), 23.9% mild anaemia (152/637), and 7.4% severe anaemia (47/637). Among our study participants, critical factors for anaemia were those concerning the age group, child´s caregiver schooling, malaria and size of the liver.


Introduction
Anaemia is a condition characterised by a reduction in the red blood cell count or the concentration of haemoglobin (Sharman, 2000).It has significant adverse health consequences contributing to the increased morbidity from infectious diseases and unfavourable impacts on social and economic development (World Health Organization [WHO] & UNICEF, 2017; Sharman, 2000).Anaemia in children under five, based on the concentrations of haemoglobin in the blood, is classified as mild (10g/dL ≤ Hb ≤ 10.9g/dL), moderate (7g/dL ≤ Hb ≤ 9.9g/dL), and severe (Hb < 7.0g/dL) (Melku et al., 2018;WHO, 2011).Moderate and severe anaemia is associated with the deterioration of both physical and cognitive development in children (Ngnie-Teta et al., 2007).Severe anaemia, in particular, carries a high 'hidden' morbidity and mortality, occurring months after initial diagnosis and treatment, and is a contributing factor for overall under-five mortality (Melku et al., 2018).Microcytic iron deficiency anaemia is also a common cause of childhood anaemia and is associated with impaired cognitive and intellectual performance, motor development, coordination, language development, and academic performance (Demaret et al., 2017;Sharman, 2000;Wang, 2016).
Anaemia is far from trivial and is associated with a decreased quality of life, worse outcomes even in non-critically ill patients, and increased healthcare resource utilisation (Demaret et al., 2017).Patients admitted to hospitals are at risk of becoming anaemic from blood loss, nutritional issues, or chronic disease (Hamid et al., 2021).Despite the existence of some studies on anaemia in children aged 6-59 months in Mozambique and Maputo City and Maputo's Province, many are either outdated or mainly focused on household-level and community-level data or malaria and HIV (Cambaza, 2013;Chemane et al., 2021;Duffy et al., 2020;Moraleda et al., 2017;Muhajarine et al., 2021;Picolo et al., 2019).A study by Moraleda et al. (2017) reported that iron deficiency, undernutrition, malaria, and HIV are the main factors associated with anaemia in hospitalised Mozambican preschool children.In 2016, a study in a hospitalised cohort of HIV-infected and HIV-exposed uninfected children aged 6-59 months reported a high prevalence of anaemia (Duffy et al., 2020).
Estimating anaemia prevalence in the hospital setting can help improve clinical management, the provision of services, and intervention strategies planning for anaemia care management in children aged 6-59 months (Tesfay et al., 2021).Nonetheless, while community-level-based studies bring evidence of childhood anaemia prevalence and associated factors, hospital-level-based studies conducted until now often report only anaemia prevalence lacking possible associated risk factors (Chemane et al., 2021).Thus, eventually contributing to the continuity of in-care health facility gaps linked to poor anaemia management of cases.We carried out this study to estimate anaemia prevalence and evaluate structural determinants and haematological parameters association among children aged 6-59 months attending pediatric inpatient and outpatient services in a Quaternary Health Facility in Maputo City Province, Mozambique.Thus, evidence generated by this research will contribute to a better understanding of anaemia among this population group and help to improve and guide nutrition-specific strategies and policies, also clinical management at the health facility level.Preliminary findings of this research were presented previously as a meeting oral pitch presentation at the 15th European Public Health Conference 2022 on 12 November 2022 (Maulide Cane et al., 2022).

Study design and setting
From August 2020 to August 2022, we collected secondary data from inpatients or outpatients who attended pediatric consultations at the 'Maputo Central Hospital (HCM)'a quaternary health facility in Maputo City Province.Mozambique's health facilities are hierarchically organised into primary, secondary, tertiary, and quaternary levels to ensure a comprehensive referral system and continuum of care.Typically, tertiary and quaternary health facilities are provincial or regional referral hospitals (generally located in densely populated urban areas) (Fernandes et al., 2023).The Pediatrics Department of Maputo Central Hospital is a reference at the national level that integrates ten services: intensive care, pediatric emergency, surgery, infants, hemato-oncology, general illnesses, infectious diseases, pneumology, neonatology, outpatient consultations, and has the following subspecialties: cardiology, neurology, nephrology, hemato-oncology, and pulmonology (Departamento de Pediatria [DP], 2023).

Selection and description of participants
The sample universe was the Pediatrics Department of Maputo Central Hospital.All medical records of children aged 6-59 months who attended pediatric consultations during the period´s study were included (census approach).In total, 637 children attending these pediatric consultations and that complied with the study criteria were selected (see Figure 1, Supplementary Table 1).Inclusion criteria were: a child aged between six and 59 months.Exclusion criteria were: a child who does not have a haemoglobin (Hb) determination or who has carried a haemoglobin determination out of the period covered by the study; a child younger than six months; a child older than 59 months.Children under five have anaemia when the haemoglobin concentration is below 11.0 g/dL at sea level (see Supplementary Table 1).The cut-off values for anaemia in children aged 6-59 months, based on haemoglobin levels, were: mild (10g/dL ≤ Hb ≤ 10.9g/dL), moderate (7g/dL ≤ Hb ≤ 9.9g/dL), and severe (Hb < 7.0g/dL) (WHO, 2011).Children were categorised by age as follows: 6-11 months, 12-23 months, and 24-59 months (MISAU et al., 2015;MISAU & INS, 2015;UNICEF, 2021b).

Data collection
We collected data from clinical records, child's health cards, consultation registration books, and monthly summary records.A pre-tested electronic questionnaire (Kobo Toolbox [Kobo], 2023) was used to compile information on sociodemographics, health status, and nutritional variables.The questionnaire, prepared in Portuguese, included four sections that dealt with (i) sociodemographic characteristics, (ii) anthropometric characteristics, (iii) feeding practices and therapeutic diet, and (iv) health status.Data collectors received training on study objectives, ethics, and data collection proceduresincluding a pilot day.Data collection was closely monitored and supervised by investigators.To ensure good data quality, we performed double data entry and compared the information in physical records (paper-based form) and the electronic questionnaire.Data cleaning was performed to verify frequency, consistency, and missed values, and any errors identified were corrected.

Data analysis
Data was analysed using SPSS 28.0 software (International Business Machine Corporation [IBM Corp], We performed descriptive statistics 2021), (frequencies, cross-tabulations) and measures of association (Cramer's V, Phi, and contingency coefficient) to describe study participants and determine anaemia prevalence.We estimated the odds ratio and associated confidence intervals (CI 95%) using logistic regression models.Primarily, we examined the association between the independent variables and anaemia using bivariate logistic regression.Subsequently, we included variables with a p-value less than 0.05 in the adjusted multiple logistic regression model (Hsieh et al., 1998;Ranganathan et al., 2017).The anaemia (has anaemia = Hb < 11.0g/dL; without anaemia = Hb ≥ 11.0g/dL) was considered a dependent variable in the model.Gender, age, residence area, province, child's caregiver schooling, breastfeeding, complementary feeding, type of breastfeeding, porridge intake, family meals, eating practices, mid-upper arm circumference (MUAC), malaria, human immunodeficiency virus (HIV), COVID-19, glucose level, urine testing, vomiting, cough, pale conjunctiva, and size of the liver were analysed as independent variables in the model.

Sociodemographic, nutritional, and health parameters associated with anaemia
Anaemia in children aged 6-59 months was associated with age, child's caregiver schooling, complementary feeding, malaria or HIV, cough and size of the liver (p < 0.05) in the bivariate logistic regression analysis.At the same time, age, child´s caregiver schooling, malaria and size of the liver remained significant in the multivariable logistic regression analysis (p < 0.05).Children aged 6-11 months are more prone to have anaemia than children from other age groups (OR = 2.39; CI 95% = 1.37-4.16).Children whose caregivers have no schooling (OR = 3.07; CI 95% = 1.37-6.89)or with a primary (OR = 2.71; CI 95% = 1.23-5.98)or secondary (OR = 1.96;CI 95% = 1.07-3.60)level of education are also more likely to have anaemia.Malaria (OR = 3.56; CI 95% = 1.63-7.77)and the size of the liver (OR = 10.57;CI 95% = 1.36-81.95)are also factors with a crucial role in anaemia (see Table 3).

Discussion
In this study, we described the magnitude of anaemia in children aged 6-59 months who attended Maputo Central Hospital and evaluated its association with structural and haematological parameters.We observed a high prevalence of anaemia among children aged 6-59 months of 62.2% (CI 95% = 2.6-3.0%) in children aged 6-59 months.This prevalence is higher than 40% and is considered a severe public health problem (WHO, 2011(WHO, , 2023)).Our findings are similar to those of  Mavale et al. (2000), who reported a high prevalence of anaemia (68.7%) in infants attending Alto-Mae and Xipamanine health centres in Maputo city.
Although addressing micronutrient deficiencies in children under 59 months of age is one of the highest priorities in the Mozambican health sector (Boletim da República [BR], 2021; Amaro, 2019), current strategies may still be inadequate, and more efforts are needed to address anaemia.´Homemade food fortification with multi-nutrient powder´(MNP) is one of the responses to micronutrient deficiencies, including iron, adopted by the Mozambican Ministry of Health in 2015 (MISAU, 2022).However, it faces systematic stock-outs limiting the supply of all required doses to children.As such, in 2021 was reported that only 9.0% of children (7.021 out of 410.684) had received the first dose of MNP and failed to receive the second and third doses of MNP (MISAU, 2022).Ensuring a strong and resilient health system able for timely delivery of MNP supplementation to children under five through the reinforcement of human resources logistic management skills can contribute to reducing the anaemia burden.
We found an association between anaemia and age, caregiver schooling, malaria, and liver size.This association of anaemia with malaria was also observed in previous studies (Duffy et al., 2020;Moraleda et al., 2017;Muhajarine et al., 2021), highlighting the need for more monitoring.Foreseeable, the association between anaemia and the liver size reported by us is also in good agreement with previous findings in the literature (Divya et al., 2020;Muhe et al., 2000;White, 2018;Yurdakök et al., 2008).Anaemia commonly develops rapidly in acute malaria, and the liver and spleen enlarge rapidly (White, 2018), explaining the results found in our study.
The current study shows that children aged 6-11 months are more prone to suffer anaemia than their peersthis is consistent with previous findings in Ethiopia and China (Fentaw et al., 2022;Li et al., 2020).Possible explanations for these results may be related to theage at which complementary foods are given to the children.As suggested by several authors (Csölle et al., 2022;Miniello et al., 2021Miniello et al., , 2017)), the age of introduction of complementary feeding might also be critically important if infants are not fed properly, with an external source of iron, which might be associated with childhood anaemia.In our study, the exact time age´s introduction of complementary foods to the children evaluation couldn´t be carried out due to the unavailability of information (not featured in the clinical data collection charts/tools).Such limitations may be considered in further research, ensuring a better understanding of the age-time of complementary foods introduction and its relationship with anaemia in this setting.
In line with our findings, children whose caregivers have a lower level of education may be more exposed to the risk of anaemia than those whose caregivers have a higher level of education, which is well documented in studies conducted in Ethiopia (Gebereselassie et al., 2020;Melku et al., 2018;Regassa et al., 2023) and Tanzania (Khatib & Joho, 2022).Some authors suggested that health education strengthening caregivers' scientific nutrition knowledge can be carried on during the prevention and treatment of maternal anaemia during pregnancy (Li et al., 2020).In addition, our results highlight the importance of increasing educational interventions for caregivers whose children are intruded to complementary foods-helping to improve feeding practices and contributing to the management of anaemia.
Although we observed a slightly higher prevalence of anaemia among boys comparing girls and those children living in rural areas compared to urban areas, we found no association between anaemia and gender or residence area.Even so, that does not imply that gender may not play a role in anaemia´s occurrence.Some authors argue that marked urban-rural anaemia prevalence variation may indicate the importance of targeting specific areas or districts (Ncogo et al., 2017).Several studies reported the higher prevalence of anaemia in children who are males is linked to a higher growth rate in boys as they have higher iron body requirements, often not supplied by the diet (Elmardi et al., 2020;Tesema et al., 2021;Zuffo et al., 2016).However, evidence for the role of these factors in anaemia in Mozambican children remains weak and warrants further research.

Strengths and limitations
As far as we know, this was the first study with a large sample size that attempted to evaluate the determinants associated with the prevalence of anaemia in children aged 6-59 months in this specific setting.Therefore, our study can contribute to filling the gap in the evidence/literature on this topic and improve ongoing interventions related to the prevention and management of anaemia in children aged 6-59 months.We are aware that our study may have some limitations.The first limitation is that we could not identify the exact age at which complementary foods were introduced to childrenas this information is not available in clinical records, child health cards, consultation registers, and monthly summary records.The second limitation is that information on the child's intestinal parasitosis is unavailable, remaining poorly captured by the data collection tools used in this study.Therefore, we could not compare the risk of anaemia in children concerning the introduction of complementary feeding at different ages and intestinal parasitosis.These limitations highlight the difficulty of collecting data in this quaternary settingwhere there are still no electronic records and the conditions for archiving physical charts/records are often inadequate or inappropriate.The third limitation is that a few variables presented a high proportion of missing values (above 15%), namely, porridge intake, family meals intake, eating practices, MUAC, COVID-19, urine testing and glucose.However, such variables did not enter into the multivariate logistic regression model.Thus, the model obtained was not affected.

Conclusion
Our findings provide evidence of factors associated with the prevalence of anaemia in children aged 6-59 months of age.More than half of the children were anaemic, demonstrating that anaemia prevails as a severe public health problem, particularly in this setting.Among our study participants, keyfactors for anaemia were age, caregiver schooling, malaria and size´s liver.To tackle anaemia, a set of multisectoral efforts should take place targeting the improvement of management for timely delivery of homemade fortification by the Ministry of Health, the continuous monitoring of children who have malaria by the health practitioners, the reinforcement of educational interventions among caregivers whose children are introducing complementary foods by nutritionists and health practitioners, as well as, the strengthening of strategic community interventions adapted to the local context and aimed at reducing food insecurity, particularly during economic, environmental and public health emergencies.

Figure 1 .
Figure1.Flowchart of the study population.

Table 2 .
Continued.MUACmid-upper arm circunference; Hbhemoglobin; RCB-red blood cell; MCV-mean corpuscular volume; HIV-Human immunodeficiency virus; COVID-19 -Coronavirus disease 2019; BGLblood glucose level; CI-confidence interval.(*) Children that have meals/eat food with their family members.(**) It refers to whether the child eats or not family meals (if the child consumes meals with other family members, such as breakfast, lunch, afternoon snack, and dinner).Family meals include consuming food from the household´s usual diet (other family members, children older than five years, adolescents, and adults).(***) Food groups include a. Cereals, tubers, and derivates; b.Meat, fish and eggs; c.Legumes; d.Vegetables; e. Dairy products; F. Fruits; g.Fats and oils.

Table 3 .
Analysis of factors associated with anaemia in children aged 6-59 months at Maputo Central Hospital, PAMC, August 2020-August 2022.

Table 3 .
Continued.Notes: Bivariate analysis: (*) Crude Odds Ratio (COR) and p values; (***) COR and p-value were not performed for variable pale conjunctiva for presenting no values for 'Without anaemia'/'Yes category'; ( a ) global p value, for categories with more than 2 levels.Multivariate analysis: included 403 participants.(**)AOR and p values for variables of multivariate analysis; ( b ) global p value, for categories with more than 2 levels.Only variables with a p-value less than 0.05 (5% of significance level) were included in the adjusted multiple logistic regression model; (****)According to the multivariate regression model (backward stepwise likelihood method), the variables complementary feeding, HIV, and cough are no longer significant (p > 0.05), being automatically excluded from the model.Thus, odds ratios are not presented for these variables.Abbreviations: MUACmid-upper arm circunference; HIV-Human immunodeficiency virus; COVID-19 -Coronavirus disease 2019; Hbhemoglobin; BGLblood glucose level; CIconfidence interval.