Trends and Determinants of Child Growth Indicators in Malawi and Implications for the Sustainable Development Goals

Sustainable development goals (SGD) 2 links malnutrition, morbidity and child mortality to stunting, wasting and overweight among children under-5 years of age. Sub-Saharan Africa still registers high nutritionally insecure people. In particular, Malawi has made modest progress in improving nutritional outcomes; and still experiences a number of structural challenges leading to negative nutritional outcomes. We describe trends of under nutrition and how the effect of selected determinants of child nutrition affect Malawian children under-5 from 1992 to 2015–16; and examine the changing patterns of the effect of selected socio-demographic characteristics on stunting and underweight using data from demographic and health surveys (DHS). The analysis included 31,630 children under-5 years from 1992, 2000, 2004, 2010, and 2015–16 DHS. Our outcome measures are stunting (height/length-for-age) and underweight (weight-for-age) less than -2 SD (Z-score). We perform logistic regression to assess the relationship between selected socio-demographic characteristics with the stunting and underweight variables. Underweight decreased by 14.0% from 24.7% (1992) to 10.7% (2015–16). Stunting decreased by 23.0% from 55.6% (1992) to 32.6% (2015–16). Underweight was more prevalent among children from central and southern regions; among male children; and children above 6 months of age or more. Later surveys were associated with reduced likelihood of underweight than the earliest surveys. Similar trends were observed between socioeconomic factors and stunting. The observed underweight and stunting prevalence is 2.2% and 1.9% lower than expected, respectively. Despite modest declines in underweight and stunting among young children in Malawi, underweight and stunting remain significant public health challenges particularly in southern and central Malawi which constitute about 85% of the total population. Interventions to address the critical malnutrition challenges in Malawi are inevitable within the context of SDG 2 on health.


Introduction
The role of nutrition as a key driver of economic development has been well recognized.
Nutrition serves as a foundation for healthy individuals and societies. By implication, malnutrition impedes human performance, health complications and human survival. Children who are well nourished are able to maximize their potential in expected growth and contribute to socio-economic development over the course of their life. The link between malnutrition, morbidity and child mortality has been highlighted in earlier studies [1] and is the focus of sustainable development goal (SGD) 2, target 2.2 calling for strategies to end all forms of malnutrition focused on stunting, wasting and overweight among children under-5 years of age [2].
Poor nutrition in the first 1,000 days of life impedes cognitive development, schooling and work performance [3] thereby lowering adult incomes and ultimately compromising national development [1]. Malnutrition is also one of the major challenges for child health in sub-Saharan Africa; and globally, more than one-third of postneonatal child deaths are attributable to undernutrition [4]. In 2015, sub-Saharan Africa was one of the two regions that accounted for almost 90% of underweight children; half were in Southern Asia and one third in sub-Saharan Africa. Despite a number of interventions to address malnutrition in sub-Saharan Africa since the 1990s, eradicating poverty and hunger remains a challenge and is at the core of the post-2015 development agenda. Many of the people suffering the most live in settings with protracted emergencies and poor access to life enhancing resources [5].
Malawi is no exception with regard to nutrition challenges, despite national and international efforts to address the problem. The country has experienced a persistent rural-urban malnutrition gap, with the prevalence of malnutrition being higher in the rural areas than in the urban areas [6].
Although Malawi was able to make economic and structural reforms and sustain its economic growth rates since 2005, poverty is still widespread and the economy remains undiversified and vulnerable to external shocks. Key challenges to growth in real gross domestic product have largely been related to severe drought and flooding in the southern part of the country; and affected agricultural production. The country registered an average 30.2% year-on-year drop in maize (key crop for food security purposes) since 2014. The Malawi Vulnerability Assessment Committee estimated that in 2016 about 7 million people would require food assistance [7].
A number of studies conducted previously have identified the magnitude and determinants of undernutrition in Malawi [6,8,9]. What lacks, however, are studies that examine trends of undernutrition and how the effect of selected determinants of child nutrition status changes over time.
We bridge this gap by describing trends in underweight and stunting among Malawian children for slightly over two decades ; and examine the changing patterns of the effect of selected socio-demographic characteristics on child nutritional status (stunting and underweight) using data from Demographic and Health Surveys (DHS). The motivation to select these two measures is two-fold: ① stunting is a better measure than underweight of the cumulative effects of undernutrition and infection during the critical 1,000-d period from pregnancy to the child's second or third birthday ② [5]; and stunting is also more common than underweight, affecting more than one in four children under-5 or 161 million children worldwide in 2013 [5]. Being underweight also puts children at greater risk of dying from common infections, increases the frequency and severity of such infections and contributes to delayed recovery. Therefore, understanding trends and determinants in stunting and underweight can inform strategies to address public health challenges of malnutrition and ensure that no child is left behind by 2030, the deadline to meet the SDGs.

Data
We analysed publicly de-identified data from five DHSs

Undernutrition measurements and socio-demographic characteristics
Our measures were stunting and underweight, defined as height/length-for-age and weight-for-age less than -2 SD (Z-score) defined by the WHO growth reference standards for children of the same age and sex [10]. Simply put, stunting is defined as inadequate height for age whereas underweight is body weight considered too low to be healthy. We pool all the DHS data (n=31,630 children) and use the two dependent variables: stunting (yes, no) and underweight (yes, no) in a logistic regression to assess the relationship between selected socio-demographic characteristics with the stunting and underweight variables. Table 1 provides a description of the selected measures in the study and how the characteristics were measured. Analyses were conducted via survey procedures including weights to account for the multistage survey design used in the DHS. The distribution of children by stunting and underweight were calculated using chi-square contingency tables and applying weighted frequencies. Statistical significance was assessed at P < 0.05.

Results
The study included children aged 0-59 months with a mean age of 27.   At least 85% of children in each of the survey period lived in urban areas. Except for 2010, nearly half of children were sampled from Southern region followed by Central region and Northern region ( Table 2). About 43% of all the households were below the middle wealth group and did not   Size at birth showed that children who were born with an average size declined from 63.6% in 1992 to 49.6% in 2015-16; with corresponding increases in the percentage of children whose size at birth was large and decreases in the percentage of children whose size at birth was small. Across all the surveys, about one in five children had diarrhoea two weeks prior to the surveys whereas almost 4 in 10 children (each) had a fever and cough two weeks prior to the survey.
Unadjusted bivariate analysis of pooled data using the survey as a variable and 1992 as a reference category showed a significant decline in the odds of being underweight with time ( Figure   3). Children were less likely to be underweight
The present study also documented significant levels of low size at birth of babies across all surveys; indicating a greater likelihood of mothers' poor nutritional status. Alderman and Behrman [14] highlighted the relationship between low birth weight and cognitive as well as neurological impairment among low birth weight babies; including the risk of stunting. The observed low size at birth in this study was also significantly associated with underweight and stunting. Poor cognitive and neurological development among children is associated with lower productivity and poor educational and economic outcomes, among others. Similar findings on the association between maternal under-nutrition and fetal growth restriction were also reported by Black et al. [15] in a global assessment of maternal and child under-nutrition.
The current findings support the literature on the negative correlation between under nutrition among children and wealth [12]. Also notable in the study are significant levels of underweight and stunting by region of residence where underweight was more pronounced in the central and the southern regions than in the northern region. Stunting was also more pronounced in the central region than the northern region. Some of the reasons for the health and socioeconomic advantage of the northern region compared with the central and the southern regions have been documented elsewhere (e.g., [16]). In brief, the relative advantage of the northern region reflect intergenerational effects of education compared with the southern and central regions of Malawi whose cumulated education effects may be lower than those of the north.
The link between socio economic factors and under-nutrition has been observed in this study with the poorest households registering almost twice as much stunting and underweight compared with the richest household in all the five surveys. However, all surveys indicate reduction of both stunting and low birth weight by half. This finding emphasizes the importance of interventions which target reduction of food insecurity at the household level. Food insecurity is associated with unavailability of food, insufficient purchasing power, inappropriate distribution, or inadequate use of food which are factors that are related to poverty levels. Apart from lack of nutritious food, child malnutrition is also caused by other factors such as by frequent illness, poor care practices and lack of access to health; these factors can well be tackled within the domain of improved economic status.
Also noticeable in this analysis is the association between stunting and underweight and maternal education. All the surveys register significant proportions of stunting and underweight with the least educated mothers registering the worst outcomes, emphasizing the need for interventions to improve maternal education. In our study, we observed that underweight and stunting was less likely to occur among girls than boys; a finding that may be linked to biological differences and the role of environmental factors in influencing health and mortality outcomes [17]. The prevalence of both underweight and stunting rises significantly after >6 months of age. This could be an indication of inappropriate complementary feeding practices, cumulative effects of illness and lack of access to health care which seem to last even at 60 mo. A systematic review utilizing data from 147 countries concluded that poor breastfeeding and complementary feeding practices are some factors associated with stunting and underweight among children under-5 [18]. Maleta et al. [8] also argued that inappropriate complementary feeding was one significant contributor to stunting among children in rural Malawi.
Consistent with earlier studies, we found significant association between stunting, and diarrhoea and non-improved water sources across all the surveys. Although the prevalence of both stunting and underweight declined during the last two survey rounds (2010 and 2015-16), the proportions are still significantly high. The high prevalence of diarrhoea could be due to the use of un-chlorinated or contaminated water. More improved sanitary and water sources as well as better health seeking behaviour would result in improved nutritional outcomes, especially in reductions in stunting.
Similar findings were also reported in Uganda [19] and Tanzania [20]. Again, improved water sources and sanitary facilities could be associated with reduced transmission of diarrhoea, hence the lower prevalence of underweight among children who had no diarrhoea [20].
Preventing stunting requires interventions that address, among other factors, the complex array of factors such as the quality and quantity of foods, reductions in the incidence and prevalence of diseases, educational outcomes, and maternal factors linked to birthweight and preterm delivery.
While the evidence base on some of the potential interventions is limited [21], there are nutrition-specific interventions whose evidence in reducing the prevalence of stunting is proven and robust particularly among pregnant women and mothers of infants and young children. These include: ① antenatal micronutrient supplementation, which are low-cost and easily administered to improve birth outcomes and the risk of stunting [22]; ② Promoting consumption of healthy, diversified diets, including high-quality, nutrient-rich food in the complementary feeding period (6-24 mo); ③ strengthened and concerted efforts targeted towards improved nutrition for pregnant mothers to reduce intra-uterine growth retardation; and ④ intermittent presumptive treatment of malaria in malaria endemic countries like Malawi [23].
For infants and young children, proven interventions include 100,000 international units (IU) of vitamin A for infants 6-11 months of age and 200,000 IU of vitamin A every four to six months for children 12-59 months in settings with a night-blindness prevalence of 1% or higher among children aged 24-59 months or where vitamin A deficiency is at least 20% infants and children aged 6-59 months [23]. Prophylactic zinc supplementation is an ideal intervention associated with increased mean height and decreased diarrheal incidence in children [24]. Other key interventions include public provision of complementary food for children which has the potential to reduce stunting by 67% in food-insecure settings [25].
Interventions to improve nutritional and health outcomes among children under-5 can only be effective if the Government and stakeholders make more investments towards addressing gender gaps through women empowerment and efforts such as increased education access to ensure adequate knowledge and ability to access food, provide better child feeding practices and care.
Government and stakeholders should also enhance efforts to implement evidence-based policies on food nutritional supplementary interventions. Partnership between the government and the research community is inevitable in order to identify new strategies and technologies to address complex nutritional problems such as stunting and anaemia.

Conclusion
This study confirms that stunting and underweight remain significant public health problems in Malawi. We note that the high levels of stunting call for measures directed towards long term investments in interventions that lead to improved energy intake, weight gain, improved nutritional outcomes, and overall health outcomes. Community-based interventions that promote access to better water, sanitation and hygiene facilities remain vital and are the pillars of hope to achieve SGD 2, target 2.2 (ending all forms of malnutrition focused on stunting, wasting and overweight among children <5 yr of age) by 2030.