This study investigated the prevalence and factors associated with both underweight and stunting among Ugandan women of reproductive age. The prevalence of underweight was low compared to studies conducted in Nigeria (6.7%) (30) Kenya (9%) (31), and Tanzania (10%) (32). This prevalence is also within the range of 5 to 20% reported for African women (30). The prevalence of stunting is 1.3% almost similar to that in Kenya DHS (less than 1%) (31) and in Tanzania DHS (less than 3%) (32). Underweight was significantly associated with residence, wealth index and region while stunting was only significantly associated with region.
Women who belonged to the Western, Eastern and Central regions were less likely to be underweight compared to women in the Northern region. Region of residence has also been shown to be associated with undernutrition in similar low-income African settings (20, 33, 34) and in Afghanistan (35). The Northern region in Uganda is the most food insecure and the poorest (36). This could be attributed to the fact that this region experienced a long civil war which greatly affected their agricultural production and the economy compared to the other regions that have been stable without civil conflicts (37). The decreased agricultural production and poor economy which are majorly attributed to the civil war induced food insecurity by reducing own food production hence decreased food availability and access to food. (38). This leads to inadequate food in both quality and quantity risking them to underweight. Additionally, most people in the Northern region unlike the other regions are pastoral communities (some are nomadic) and this may negatively affect their consumption of the foods from agricultural origin (crops) as they mostly focus on mainly pastoral activities. As a matter of fact, pastoralism has been shown to increase risk of underweight in Ethiopian pastoral communities (20).
Women belonging to the poorest wealth quintiles were more likely to be underweight compared to those in the richest wealth quintiles. Wealth index/ socio-economic status has been shown to be significantly associated with underweight in other similar contexts (33, 39, 40). Households that are facing poverty tend to have low purchasing power for food leading to both low quality and inadequate (quantity) food intake (39, 41). The poor are usually less educated, and this negatively affects their nutritional knowledge. This limited knowledge on nutrition leads to inadequate dietary intake due to poor dietary consumption patterns like skipping meals and /or having unbalanced diets and eventually leading to underweight (4). They also tend to have poor housing and inadequate access to clean water predisposing them to various infectious diseases such as diarrhea, tuberculosis which leads to loss of nutrients from the body hence leading to undernutrition (39, 42).
Our study also shows women who belonged to rural areas were 37% less likely to be underweight compared to their counterparts in urban areas. Although most studies in the region or within similar contexts show rural areas are prone to undernutrition (39, 43), a similar reverse association was observed in Bangladesh (44). In Uganda, most agricultural production occurs in rural areas which contributes greatly to rural food availability and access (36). Uganda is experiencing rapid urbanization that has led to the poorest and highly vulnerable people to settle in poorly organized informal urban areas (20, 45). Despite the availability of regular food supply in urban areas, the urban poor households that make up the majority of the urban population have limited access to sufficient and nutritious food. This may compromise their ability to meet recommended dietary intake and therefore increasing their risk for underweight. This is supported by findings of a recent study that showed a high prevalence (88.5%) of food insecurity among women dwelling in one of the divisions of the most urbanized areas of the capital city Kampala (45). The rapid increase in the urban settlements has also led to sanitation challenge and environmental degradation (20). This has further led to people resorting to improvised unhygienic means of human excreta disposal that pose health risks such as diarrhea (46) that further increases the risk of underweight.
Stunting was only associated with region. Women in Central and Western regions were more likely to be stunted compared to those in the Northern region. This could be due to the intergenerational cycle of stunting where stunted children grow into stunted adolescents and adults. This is supported by the fact that according to the UDHS 2016 report and some studies that have looked at the trends of childhood stunting in Uganda since 1995, Western and Central regions have the highest prevalence (28, 47). Furthermore, these studies have shown slow annual stunting reduction rate of 0.45% which increases the chances of these children growing into stunted adults and this could partly explain the high proportions and increased odds of stunting in western and central regions seen in this study. An alternative explanation to these findings could be the fact that Western Uganda is partly inhabited by the Batwa tribe, an indigenous pygmy population who are relatively short (in height) compared to other Ugandans (48, 49) and the possibility of migration of stunted women from other regions to the Central and Western regions which are the most economically vibrant regions.
Strengths
We used a nationally representative sample and weighed the data for analysis and therefore our results are generalized to all Ugandan women aged 20 to 49 years. Secondly, we analyzed determinants of undernutrition considering contextual factors (household and community level factors) since also weighted data has been used. Thirdly, we used data with a large sample size which was collected, entered and cleaned by a team of trained and highly experienced scientists hence limiting mistakes in the data set.
Limitations
The cross-sectional design is limited by lack of temporality hence causality inferences cannot be made. Most data on the predictors was based on self-reporting and could not be verified through records which risks socially acceptable answers hence information bias.
Possible mis-classification of pregnant women mainly those in the first trimester as it is possible that some women are usually not aware that they are pregnant in the early stages of the pregnancies.