Prevalence and Risk Factors of Anaemia Among Children Under Five Years in Ghana: Analysis From the Ghana Demographic and Health Survey


 Background: Childhood anaemia remains a public health challenge in developing countries, mostly in children under five years in Sub-Sharan Africa. Anaemia in children is a preventable condition; however, it has serious consequences including growth retardation, low cognitive functioning, poor immune system causing an increased susceptibility to diseases, and death. ﻿This study determined the prevalence and associated risk factors of anaemia among children under five years in Ghana. Methods: Data from the most recent Ghana Demographic and Health Survey (DHS-VIII) was analysed in this study. ﻿Anaemia was examined as a binary variable- anaemic (coded as 1) and non-anaemic (coded as 0). The explanatory variables included age, child’s malaria vaccination status, mother’s number of antenatal visits and household source of drinking water. Bivariate and multivariate logistic regression model were conducted to identify the risk factors of anaemia in the defined population. Results: 2,434 children aged 6-59 months were included in this analysis. The majority were males (50.5%), aged from 24-42 months (36.1%), and had received malaria vaccine (94.3%). The prevalence of anaemia in the population was 58.35% (95%CI=52.72-63.96). The logistic regression indicated that female children under five years in Ghana are less likely to have anaemia than male children under five years [aOR= 0.74, 95% CI=0.62-0.88, p<0.001]. ﻿Also, children with malaria had about two times the odds of anaemia than those without malaria [95% CI = 1.28-1.87, p<0.001]. Further, children whose mothers had received tertiary education were less likely to have anaemia [aOR= 0.60, 95% CI=0.38-0.96, p=0.03] than those whose mothers had no formal education.Conclusion: The prevalence of anaemia is high among children under five years in Ghana, and this prevalence is determined by child’s age and malaria status, maternal education, household wealth index, and place of residence. Consequently, anaemia prevention and management strategies must prioritise these factors to reduce the anaemia prevalence in this population.

Studies have reported a high prevalence of childhood anaemia. For instance, a study based on data from Demographic and Health Survey (DHS) in Togo among children under ve years revealed 70.9% anaemia prevalence [8]. In other developing countries, the prevalence of childhood anaemia ranges from 27.2% in Mexico to 60.6% in Haiti [9,10]. In Ghana, studies have established that anaemia prevalence is 78% and 84% among children aged 2-5-years and under 2 years, respectively [2,11]. Anaemia aetiology is multifactorial and context-dependent and can include malaria, intestinal parasites, and malnutrition [7].
Commonly, it is the most vulnerable, the poorest and the least educated who are disproportionately affected by iron de ciency, and they stand to gain the most by its reduction and experience the severe forms of anaemia [12]. In addition to anaemia's adverse health consequences, the economic effect of anaemia on human capital results in the loss of billions of dollars annually which subsequently affect productivity [13,14].
This double burden of anaemia has caused both national and international up look to implement anaemia preventive programmes such as food forti cation and intermittent iron supplementation [15].
Also, in Ghana, new efforts are getting underway to improve child's health and wellbeing. It is important to characterize the baseline situation and associated factors of anaemia in the most vulnerable population groups to serve as a benchmark for monitoring progress. A recent study in Ghana using data from the DHS revealed a high prevalence of 78.4% anaemia among under ve children [16]. However, the associated determinants contributing to the higher prevalence was not established. This current study sort to bridge the gap and adds evidence to the literature on prevalence and determinants of childhood anaemia. In general, the success of community programs in Sub-Saharan Africa and particularly in Ghana require a good knowledge and understanding of the risk factors of anaemia. As a result, this study aimed to determine the prevalence and risk factors of anaemia among children under ve years (aged 6 to 59 months) using the most recent DHS dataset in Ghana.

Data source
The data used in the study was extracted from the 2019 Ghana Malaria Indicator Survey (GMIS) which is the eighth phase of the Ghana Demographic and Health Survey (GDHS) programme. GDHS has been effective since 1988 and is part of the international Demographic and Health Survey (DHS) programme which seeks to collect household surveys in developing countries with the goal to provide governments with recent and reliable information on population health indicators to allow for adaptation of health policies. The DHS led by the United States Agency for International Development (USAID) deals with many health issues, including anaemia, family planning and reproduction, child health (breastfeeding and nutrition, vaccinations, acute respiratory infections, and diarrhoea), malaria and the Human Immunode ciency Virus (HIV) [17]. Financial support for the survey was provided by the United States Agency for International Development (USAID) and the Government of Ghana. The survey was designed to provide estimates at the national level, urban and rural areas, and for each of the 10 administrative regions. In 2019, Ghana created six new regions; however, the new administrative boundaries were not available during survey design of the 2019 GMIS. The survey used a strati ed, two-stage cluster design, where a total of 200 clusters, which consisted of 97 clusters in urban areas and 103 in rural areas were chosen. Details about the survey can be found in the 2019 MIS report [18]. Since the main objective of the study was to determine the prevalence and risk factors of childhood anaemia among under ve years, the children's data recode le part of 2019 GMIS datasets was used for the analysis [18]. Outcome variable During the 2019 GMIS, anaemia was tested with a single-use retractable, spring-loaded, sterile lancet for the nger or heel prick. Haemoglobin (Hb) analysis was performed on site with a battery-operated portable HemoCue 201+ analyser, which could give result in less than a minute (haemoglobin level) in grams per decilitre (g/dl) of blood. The World Health Organisation (WHO) estimates that children aged 6-59 months with Hb concentration level below 11.0 g/dl corresponds to anaemia, a level below 7.0 g/dl is considered as severe anaemia, a level between 7.0 g/dl and 9.9 g/dl is considered as moderate anaemia and a level between 10.0 g/dl and 10.9 g/dl is considered as mild anaemia [2]. In regard to the WHO haemoglobin levels of classi cation, the dependent variable was coded into two levels (anaemia and not anaemia). A child was classi ed as anaemic if the Hb level was below the WHO threshold of 11.0 g/dl.
Weighting the sample In order to adjust for the unequal probability of selection between the geographically de ned strata, and to compensate for sample variability as well as for nonresponse during the survey, a sample weight was applied during the analysis to get statistics that are representative of Ghana. A detailed explanation of the weighting procedure can be found in the methodology section of the survey report [18].

Independent variables
The independent variables were group into child related characteristics, Maternal characteristics, and Household characteristics. The child's related variables were sex, child's age, malaria status, and whether a child took the mosquirix (malaria vaccine). Variables related to the child's mother were educational status, mother's age, ethnicity, number of antenatal clinic (ANC) visit, and religion. Household related characteristics were wealth index, place of residence, household source of drinking water, and region located.

Statistical Analysis
Data analysis was conducted with STATA/SE version 17. The characteristics of the sample was summarised with frequencies and proportions in a descriptive analysis. Then, statistical signi cant differences between the independent variables and anaemia were determined using the Pearson Chi-Square test (χ 2 ). Further, logistic regression analysis was tted to estimate the odds of anaemia odds ratios (ORs). The ndings were reported with a 95% Con dence Interval (95% CI) and a p-value less than 0.05 was considered statistically signi cant.

Ethical approval
Ethical approval was received from the Brunel Research Ethics Committee [Reference: 31563-NER-Jul/2021-33647-1]. Also, written consent was obtained from the Demographic and Health Survey (DHS) programme data archivist before the dataset was accessed and used for the analysis.

Characteristics of the study population
The dataset consisted of three thousand and four (n=3004) children under ve years of age. However, this analysis only included two thousand four hundred and thirty-four (n=2434) children aged 6-59 months whose anaemia status was determined and were available. The proportion of male and female populations in the sample was 1228 (50.5%) and 1206 (49.6%), respectively, and the most represented age group were children aged 24-42 months (n=879, 36.1%). The majority (n=2296, 94.3%) had received malaria vaccine (Mosquirix); however, more than one-third (n=857, 35.2%) reported fever or malaria. Furthermore, more than 50% of the children (n=1335) lived in poor households, and nearly 62% (n=1508) were from rural areas. Also, the source of the most common household drinking water was dug well (41.2%).
Most of the children were from the Northern region (15.9%), were from poor households (54.85%) and had mothers who had no formal education (46.1%). Also, while nearly 27% of the mothers did not visit the Antenatal Clinic (ANC), about 47% of them visited the ANC 6-10 times.
Moreover, anaemia prevalence was signi cantly higher among children from poor households (66.67%, p<0.001) and those living in rural areas (62.74%, p<0.001). The distribution of anaemia prevalence among children in the regions of Ghana is Upper East (72.84%), Northern (67.87%), Central (66.20%), Upper West (62.57%), Brong Ahafo (59.36%), Volta (53.55%), Western (50.51%), Greater Accra (47.43%), Eastern (47.08%), and Ashanti (42.91%). The descriptive characteristics of the children, their mothers, households, and the distribution of anaemia prevalence across the child-related factors, maternal and household factors with respective p-values and 95% CI are shown in Table 1 Page 6/20   [1]. This nding is of public health importance, and a continued effort is imperative to create anaemia awareness in Ghana to save lives. The prevalence of anaemia identi ed in this study is consistent with similar studies conducted in other lower-and-middle-income countries (LMICs), such as Uganda (58.8%) and India (59%) [4,19]. However, it is considerably higher than the prevalence reported by studies in Ethiopia, another LMIC [20,21]. The differences in anaemia prevalence between Ghana and Ethiopia could be due to variations in geographic characteristics, like climate changes, in the two countries [22]. However, it could also emanate from differences in this study's research approach and that of the studies from Ethiopia.
This study also identi ed varied anaemia prevalence among the different geographic areas in Ghana. For example, the Upper East (72.84%), Northern (67.88%), Central (66.20%) and Upper West (62.57%) regions in Ghana have a relatively higher anaemia prevalence compared to the other regions. Also, the prevalence of anaemia was higher in rural (62%) than urban (51.2%). Again, these ndings are consistent with similar studies on anaemia prevalence in under ve years old in Ghana [16,23]. The rural and urban anaemia prevalence variation could be attributed to poverty and low propitious living conditions, poor sanitation, and di cult access to good drinking water in the rural areas, resulting in diseases that subsequently increase anaemia risk [22]. For emphasis, this study identi ed that children under ve years from households with dug wells, rivers, and rainwater as the source of drinking water recorded are more likely to be anaemic, and these drinking water sources are more characterised in rural areas. These inherent rural characteristics further explain the signi cant relationship between rural dwellings and the risk of anaemia in children under ve years in Ghana.
Also, the ndings suggested that as the children aged, their odds of anaemia become progressively lower.
This observation may be due to the iron supplements given to the older children (≥6years old) in Ghana. Thus, to address this anaemia burden in children under ve years, the Ghana government could include this population in their nationwide iron supplemental for children intervention, especially as younger children are more susceptible to diseases and infections resulting in anaemia [16,25]. The nding on age and anaemia is similar to a study conducted in Togo on prevalence and risk factors of anaemia in children aged from 6 to 59 months [8]. Again, a lower prevalence of anaemia as a child age was reported in another study [24]. Hence, health systems must prioritise younger children in their anaemia prevention strategies.
The mother's educational status also showed a statistically signi cant association with childhood anaemia in this study. Children whose mothers had primary, secondary and tertiary education were 22%, 35% and 40% times less likely to have anaemia. Additionally, children whose mothers had frequent ANC visits were less likely to be anaemic under ve years old. Similar ndings have been reported in several studies [26,27]. This nding could be since educated mothers are more likely to know better and observe safe and hygienic practices to prevent diseases that could trigger anaemia in children. Also, FOCUSED antenatal care intervention, a agship ANC policy implemented in Ghana, could account for the low anaemic risk in children whose mothers frequently visit ANC centres. This intervention, spearheaded by the Ghana government, provides tailor-made and individualistic interventions to expectant mothers in government hospitals across the country. Based on this study's ndings, this intervention is recommended to be sustained and extended to hard-to-reach areas in Ghana to ensure comprehensive coverage and reduction in nationwide anaemia prevalence. Finally, even though the prevalence of anaemia in this study is a tad lower than that reported by WHO in 2020 (59.5%) for children under ve years old in Ghana [1], there is still the crucial need for critical stakeholders in Ghana to review and improve their anaemia prevention and management strategy to signi cantly reduce the burden of anaemia in Ghana.
To the best of the researchers' knowledge, this is the rst study to estimate anaemia's prevalence and risk factors in Ghana using the most recent DHS. Therefore, the ndings could inform contemporary interventions to address the anaemia prevalence in Ghana. Additionally, the relatively larger sample size in this study due to the use of the DHS dataset ensured the external validity of the study's ndings.
Nonetheless, because of the retrospective nature of the DHS dataset, there is a tendency of recall bias; thus, the ndings of this study must be interpreted with caution. Finally, because of the observational nature of this study, no causal association could be inferred from the explanatory variables and anaemia.

Conclusion
This study identi ed a high prevalence of anaemia among under ve years children in Ghana. This prevalence was determined by factors such as child's age and malaria status, maternal education, household wealth index and place of residence. The ndings suggest the need for Ghana to target these factors in their anaemia prevention and control strategies. Policies like continuous distribution of mosquito nets, especially in the northern region of Ghana, could help address the risk of malaria and by extension anaemia. Also, mass media campaigns on anaemia prevention should be sustained and trickled down to sub-district and regional levels in Ghana. FOCUS antenatal practice in Ghana should be continued with increased emphasis on anaemia education.