Childhood Malnutrition: Determinants Among Under Five Children

Objective

The aim of the paper was to investigate the determinants of malnutrition in children under 5 years of age in Ghana using multilevel regression analysis among a total of 2083 children within 1641 households where the data was extracted from the 2008 Ghana demographic and Health Survey(GDHS). The article also highlights childhood malnutrition as a multi-dimensional issue (Wong et al, 2014) that needs immediate intervention specially in developing countries. Clusters of malnourished children are mostly found in sub Saharan Africa and South Asia (Poel et al, 2008) additionally stressing the need of strategies directed towards recognizing risk factors related to the varied household conditions and formulating and implementing policies to house these burning issues.

Background

According to WHO, malnutrition can be defined as “deficiencies, excesses or imbalances in a person’s intake of energy or nutrients (World Health Organization, 2016)”. Among its two groups of conditions, three WHO derived indicators were specified to obtain height for age (HAZ), weight for age (WAZ) and weight for height(WHZ) which depicted stunting, underweight and wasting subsequently. Malnutrition is a major issue all over the world but especially can be seen in different countries in South Asia like India, Nepal and Afghanistan and sub Saharan African countries like Ghana and Jamaica. According to the World Bank, Ghana has a population of 28. 8 million with an annual Gross Domestic product of 8. 5% and Gross national income of 1490 $ (World Bank, 2018). Here, 40% of the deaths in children is caused owing to improper nourishment (Aheto et. Al, 2015). This article examines the agents of malnutrition in children under the age of five to direct the stakeholders and policy makers to change the health outcomes related to malnutrition.

Study Design and Participants

The authors have used a cross sectional study based on population of Ghana and most of the data on maternal and child health issues like nutritional status, awareness on family planning methods, fertility preferences, domestic violence, childhood mortality rate, breastfeeding practises were extracted from the Ghana Statistical Service in the year 2008 by GDHS. Initially a total of 12,323 households were selected out of a countrywide 411 sampling units. The data was further divided into: - Rural and Urban communities (5175 and 6603) - Women and men according to ages where there were 4916 women of 15-49 years of age and 4568 men of 15-59 age group. - Children: Total of 2992 children were examined out of which 756 children were excluded on account to missing exposure data. Additionally, 153 children were further excluded as their Z scores were outside the plausible biological range that were stated by the World Health Organization. Therefore, a cumulative of 2083 children under the age of 5 were examined who were living in 1641 (400 communities), of which 25. 4% (416) had two or more children under the age of 5years. Variables: The Primary Outcome of this study was to determine malnutrition among the children below 5 years of age.

However, the authors have not described why only the children under 5 years of age were selected. According to a study conducted by Liu et. al, malnourished children at the age of 3years showed prominence and activeness that resulted in discrepancies with motor and conduct disorder later when they were 17 years of age (Liu et al,2004). On this ground, examining the children at the age of 3 would be more conclusive even though the study population would be smaller in comparison. Nonetheless, most of the studies on malnutrition have examined children below 5 years of age.

Multiple exposures: The study demonstrates multiple exposures divided into two aspects that led to malnutrition. However, the demarcation between these two aspects were not clear in the article from the very beginning. Individual factors such as age of the child, breastfeeding time, number of births, diarrhoea, poor households, unstandardized toilet systems, mothers having no insurance coverage were considered whereas the elements that did not influence each child but were common to all members of the family like the quantity of food available were categorized as socioeconomic factors. Moreover, factors like educated/uneducated mothers and overall BMI of the mother seems to be the potential confounders. Although the writers have categorised the effects in both an individual level and a more generalized socioeconomic (household related) level, there is a lack of proper exploration of the economic aspect which it is vital in impacting the health status of people in Ghana. Poverty plays a crucial role as a determinant related to disease and malnourishment. Scientific evidences demonstrate that an economic state within or between countries directs the adverse effects of health where the relationship between an illness or condition and socio-economic status is recognized to be inverse (Zere et al, 2003). Due to difficulty in measuring “wealth” in this case, the use of asset-based index as a proxy has proved to be a thoughtful and effective solution for the article.

Method A random intercept multilevel regression analysis was used to calculate the probabilities of malnutrition i. e stunting, underweight and wasting using Z scores. Any child was sorted as malnourished if his/her Z score was less than minus two of the standard deviation taken from the median of the population(Aheto et al, 2015). Statistical Analysis Even though there are no hierarchical structure of the data mentioned, the use of a multilevel regression analysis is fitting for the evaluation of both aspects namely individual respondents and social contexts (Snijders,2012). However, this could also result in errors of standard deviation for regression coefficients which can eventually produce incorrect data. For further investigation, to analyse the relation between outcomes of malnutrition and risk factors, each of the significant factors were pitched against a computed probability for that outcome (Aheto et al, 2015). The additional use of the software R for statistical analysis has made the results even more full proof.

Results

We can observe from the above table that the investigators have appropriately differentiated the variables for each of the three outcomes. Additionally, having toilet facilities inside the household, fever history, age of the mother and insurance cover related to the mother were the factors that did not make any difference to WAZ. Similarly, breast feeding periods, mother’s age, history of fever and increased number of births were not relevant to WHZ. On the other hand, we can conclude that mother’s BMI has a huge role to play in all 3 of the outcomes. Even so, the authors have failed to provide suitable data or a number to determine the appropriate weight during pregnancy or childbirth.

According to McCall et al, women with >/= 60 kg/m2 of BMI had greater risks of thromboprophylaxis in both antenatal and postpartum periods and had double the chances of pre-eclampsia or eclampsia (McCall et. al, 2018). More importantly, the education of a mother played a positive role in the increasing the weight of the child but did not make any difference to wasting and stunting in a baby (Aheto et al, 2015). The chances of stunting went up with the child’s age whereas conversely, wasting probabilities went down with the child’s age. Likewise, multiple births by mothers put the risk of underweight and stunting in children whereas having a smaller size during birth makes the child more prone to all three of the factors. Studies have shown that smaller size puts the child into elevated risks of death by infectious disease (Johanne et. al, 2014) and about 14% of infants in developing countries weigh less than 2. 5kg at birth, out of which many are born pre-term (O’Leary et. al,2017).

The authors of the article have investigated and covered a large range of population which makes the study and the results more plausible and meaningful but have included the lack of data for all of the variables as both strength and a limitation. The use of mother’s word of mouth for size of the child seems like a quick fix but creates a reporting bias. To solve this issue, a logistic regression analysis was used to obtain the odds ratio. This method allowed avoiding confounding effects by examining all the factors together(Sperandei,2014). Furthermore, under the socioeconomic heading, the authors have stressed upon childhood nutritional differences that further implies that the results have a level of variation even after accommodating for household characteristics and the child. Additionally, unobserved household factors could affect about 32%,23% and 20% in WAZ, HAZ and WHZ respectively (Aheto et al, 2015). These factors could be environmental (location of the houses or geographic variations) or social but the author has not gone into much detail regarding this matter. Undoubdtedly, this article supports other developing countries regarding nourishment needs and requiring better health outcomes but does not describe details of other countries or continents who are in the same vicious cycle of malnutrition for comparision.

For instance, malnutrition is still prevalent in children under 5 years of age in developing countries. According to Kramer et al, in the year 2013, 51 million children suffered from moderate wasting (8% globally) and 17 million had severe wasting (3% globally) with highest occurrence in Asia (71%) and Africa (28%). Furthermore, 161 million children were stunted with the highest prevalence in Southern Asia (56%) and in Africa(36%) (Kramer et al,2015) which meant almost half of the pre-school children in Bangladesh are malnourished (Akhtar,2016). It does however briefly talk about Jamaica to describe the difference in association of breastfeeding time between low and middle-income countries. The article further needs to elaborate on how the determinants affect malnutrition in Ghana for the stakeholders to better formulate policies. As mentioned by the UN food and Agriculture Organization, only 50% of children below 6 months are breastfed (Aheto et. al,2015).

The immunological advantages of breast milk can help dramatically reduce infections in children. Furthermore, studies have proved that consumption of breastmilk can diminish the risk of diarrhoeal infections among children (Black et al,2010). Concurrently, improving the education standards of mothers and caregivers in effective weaning periods in relation to weight gain (Akhtar, 2015) and in building up spaces between two births (Basit et. al, 2012) should be reinforced in order to reduce malnutrition in children. The unaffordability of nutritionally rich food after breastfeeding owing to poverty is another huge issue that needs to be corrected. Many of the families in Ghana belong to a low-income group. In such conditions, separating funds for healthcare automatically reduces the budget for healthy food. Admittedly poverty and shortage of food hastens the process of food insecurity and also grants access to disease epidemics ( Akhtar,2016). Besides that, malnutrition achieved during the first two years of life can even prolong into adulthood(Bradshaw, 2006). Undoubtedly, the first few years of life is a critical time to upgrade immunity to avoid a more permanent damage.

Compared to the uninsured Ghanaians the members of NHIS are more inclined to use the prenatal, delivery and postpartum services. In addition, the members also get to take advantage of the preventive checkups which can automatically reduce the fetal and maternal death rates. Empirical Analysts confirm that the NHIS is generating positive health outcomes for the people who are insured (Mensah et al,2010). That being said, the uninsured members are in a much worser condition than when NHIS was not implemented. Since cost was the major reason for people to go uninsured, systems which identifies the core poor citizens in the most extreme geographical areas should be developed to allow exemptions and discounts (Mensah et. al,2010) to mothers who require it the most.

On the other hand, any form of change requires time and patience to be implemented. So schemes should be put in place that focuses on short term solutions for the uninsured groups (Mensah et. al, 2010). Along with that, mothers should be informed about the benefits of having NHIS insurance which will in turn create better health outcomes for both mothers and their children. Hence a collaborative approach by capable healthcare providers, responsible administrators, stakeholders and policy makers is required to bring about the desired change in Ghana. To recapitulate, the article has aptly described the concept of malnutrition and its determinants but stands in necessity more information on prevalence of malnutrition in Ghana and comparison with other countries.

Likewise, the article also provides the implications of malnutrition and policies that can reduce and even eradicate malnutrition in Ghana. Inclusion of policies on eliminating poverty, clean sanitation, empowering women education, reduction of maternal and childhood mortality rates, providing secondary and basic education to people in a distributed manner all coincide with the central argument of the article. While the study methods are described clearly, further addition of statistical evidences of other studies regarding malnutrition could have made this article much stronger.

Overall, the article is relevant to the present status in Ghana and can be used to support potential policies in health and nutrition. While it cannot be denied that it showed variation in describing the risk factors and its reasonings, addition of the economic aspect separately could have clearly marked off where the policymakers could start working on since poverty is a major issue in Ghana. The data errors have been tried to be corrected with the use of proxies and logistic regression analysis for known cases but there is still the need to account for the unidentified household level variations in nutritional outcomes in children below five years of age. On a different context, the author has used recent articles which were above the year 2008 either from journals and from official government sites validating the article’s authenticity and used suitable lengths of paragraphs to engage the reader’s interest.