Menu
Menu
Menu
Menu
N/A
The prevalence of underweight among children below 60 months old in Nigeria remains a significant public health challenge, especially in northern geopolitical zones (NGZ), ranging from 15% to 35%. This study investigates time-based trends in underweight prevalence and its related characteristics among NGZ children below 60 months old. Extracted NGZ representative dataset of 33,776 live births from the Nigeria Demographic and Health Survey between 2008 and 2018 was used to assess the characteristics related to underweight prevalence in children aged 0–23, 24–59, and 0–59 months using multilevel logistics regression. Findings showed that 11,313 NGZ children below 60 months old were underweight, and 24–59-month-old children recorded the highest prevalence (34.8%; 95% confidence interval: 33.5–36.2). Four factors were consistently significantly related to underweight prevalence in children across the three age groups: poor or average-income households, maternal height, children who had diarrhoea episodes, and children living in the northeast or northwest. Intervention initiatives that include poverty alleviation through cash transfer, timely health checks of offspring of short mothers, and adequate clean water and sanitation infrastructure to reduce the incidence of diarrhoea can substantially reduce underweight prevalence among children in NGZ in Nigeria.
The NDHS 2008, 2013, and 2018 standardised national representative surveys were combined, and data related to the NGZ were extracted for analysis. Structured questionnaires were used to gather data regarding the health and demographic characteristics of the child and mother, including anthropometry data during the surveys. Women aged between 15 and 49 years who were interviewed during the surveys detailed the live births of their children. In all the surveys considered, approximately 62,169 live births of children less than 60 months old occurred in the NGZ; of these, 17,184 were from the 2008 NDHS, 21,693 were from the 2013 NDHS, and 23,292 were from the 2018 NDHS. A digital display scale, particularly that of SECA 878U, was used to measure the children’s weight. The weight of children was documented using the standardised weight-for-age measurement procedure as described by the WHO [6]. In the pooled surveys, 33,776 children below 60 months old who had comprehensive and valid data concerning the birth date and weight measurements in the NGZ were used for the study analysis. The statistical methodology used to record live births and the anthropometric guidelines regarding the weight measurement of children below 60 months old have been detailed elsewhere [9,10,11]. The dependent variables were underweight in children below 60 months old and disaggregated into three age categories in months, that is, 0–23, 24–59, and 0–59. Underweight cases arising within the study age groups were measured twofold: the case of underweight was estimated as the WAZ score < −2 SD and coded as 1, and non-case underweight with WAZ ≥ −2 SD was coded as 0. A recent nutritional framework described by the UNICEF [7], and past studies conducted in developing countries [12,13,14] were used to identify the possible confounding factors to be examined in the current study. The dependent variables were investigated against all 29 potential confounding variables selected, which were categorised into seven distinct classes (Table 1). Classification and description of likely independent confounding related characteristics to underweight in children below 60 months old for the study analysis. Notes: ‡, maternal height measured in centimetres; MBMI, maternal body mass index (estimated in kilograms/square meters). The UNICEF framework entails direct immediate characteristics, which include child nutrition and disease occurrence. It has been previously suggested that poor child nutrition and recurrent child illness increase the association with underweight nutrition. Dietary diversity score (DDS) mirrors the prevalence of eight possible food categories taken by a child in the last 24 h in children younger than five years [15,16,17]. Feeding practices and DDS were parameters used to measure adequate child nutrition prior to the survey interview, and the food categories were classified into two classes in the study analysis (child consumed ≥ five food categories and child consumed < five food categories). These eight food groups have been reported elsewhere [11]. Recurrent child illnesses (e.g., diarrhoea and fever) were possible disease occurrences in the last 14 days before the survey interview date. According to the earlier literature [12,14,18,20], socioeconomic characteristics (e.g., educational attainment by mothers/fathers, economic status of a household, mother’s work status, and the number of wives or women living in a household) are associated with underweight prevalence in children below 60 months old. Household income or expenditure data were not used to quantify the economic status of households due to unavailability of data; however, a self-reported household asset-based factor score was utilised as a wealth proxy measure to categorise the household economic status using principal component analysis [21]. This means weights were assigned to the self-identified assets, and these assets have been listed in the NDHS report [11]. In the three combined surveys, the household wealth index factor scores were classified into three groups: poor, middle, and rich households. Individual child and maternal characteristics are increasingly associated with underweight prevalence in children below 60 months old [12,18]. The child and maternal characteristics incorporated are presented in Table 1. Maternal autonomies [22], such as having healthcare, earning/financial, and movement autonomies, were considered, and these autonomies were grouped as household decision-related characteristics. Additionally, included in the study analysis was maternal access to electronic or print media classified as knowledge of healthcare through media. Healthcare service-related characteristics (e.g., mode of birth, birth assistance, and birthplace) and community-level characteristics (type of residence and geopolitical zone) were also included in this study. Table 1 depicts the classification of all the independent characteristics used in the study analysis. For each wave of the surveys, the frequency distribution and underweight prevalence of children below 60 months old for all potential associated independent variables listed in Table S1 were estimated with a confidence interval (CI) of 95%. Data from all surveys considered were pooled to identify the odds of the relationship between the potential independent variables and the study dependent variables. Multilevel logistics regression was used to conduct the multivariable analyses that estimated the adjusted odds ratios (AORs), which measured the strength of association with the dependent variables. Survey clusters and weights were adjusted using the STATA/MP 14.1 version ‘SVY’ command. A stage modelling approach was adopted for the multivariable analyses, entailing that each of the seven-level factors presented in Table 1 was assessed independently. Firstly, the community-level characteristics were entered as a baseline first-stage model and those characteristics that met the 5% significance level criteria were retained (model 1). In the second-stage modelling, the significant variables in model 1 were added to the socioeconomic characteristics, and again, those variables that were significantly significant were retained in model 2. This procedure was repetitively used for the inclusion of individual maternal and child-related, knowledge of health services through (media), household decision autonomy, healthcare-related service, and immediate feeding practices in the third, fourth, fifth, sixth, and seventh stages, respectively. Variables that were statistically significant in stage model 7 are reported in the study (Tables S2–S4). This procedure permits factors that indirectly affect child health to be satisfactorily examined without interfering with direct factors that impact child health (e.g., child’s nutritional intake and disease incidence).
