Menu
Menu
Menu
Menu
Background: Anaemia disproportionately affects women of reproductive age in sub-Saharan Africa including Nigeria. Yet, community-based studies on the prevalence and determinants of anaemia among women of reproductive age are scarce in Nigeria. Design: A cross-sectional community-based survey using a nationally representative sample. Objectives: This study described anaemia prevalence and its associated factors among women of reproductive age, pregnant women, and non-pregnant women in Nigeria. Methods: We analysed data from the 2018 Nigeria Demographic and Health Survey. Pregnant women with a haemoglobin level less than 11 g/dL and non-pregnant women with a haemoglobin level less than 12 g/dL were considered anaemic. Anaemia was also categorized as mild, moderate, and severe. Pearson’s chi-square test was used to evaluate the association between anaemia status and independent variables. All variables with ρ ⩽ 0.25 in bivariate analyses were further analysed using complex sample logistic regression. Results: Anaemia prevalence was 57.8%, 57.4%, and 61.1% for women of reproductive age, non-pregnant women, and pregnant women, respectively. The prevalence of severe anaemia was 1.6%, 1.5%, and 2.3% for overall women of reproductive age, non-pregnant women, and pregnant women, correspondingly. The southern regions, rural residence, low education, unemployment, low wealth index, and non-use of modern contraceptives significantly increased the likelihood of anaemia and severe anaemia among women of reproductive age and non-pregnant women. The likelihood of being anaemic was significantly increased by large family size among women of reproductive age and by being underweight among non-pregnant women. The South-East region, rural residence, low education, and unemployment were significantly associated with anaemia among pregnant women. The South-South region and unemployment increased the likelihood of severe anaemia among pregnant women. Short stature significantly reduced the odds of being anaemic and severely anaemic among pregnant women. Conclusions: Anaemia prevalence among all categories of women of reproductive age is high in Nigeria. Predictors of anaemia prevalence and severity should be considered in policies intended to reduce anaemia among women of reproductive age in Nigeria.
Nigeria had an estimated population of 195,874,683 people and annual population growth of 2.62% in 2018.34 Nigeria comprises six geopolitical regions, 36 states, and one Federal Capital territory. Each state consists of local government areas (LGAs). Each LGA is composed of wards. Approximately 50.3% of the 2018 population was urban. WRA constituted around 46% of the population.34 This study used a quantitative, cross-sectional design by analysing data from the Nigeria Demographic and Health Survey (NDHS) 2018. The sampling frame consisted of households listed in Nigeria’s 2006 Population and Housing Census (NPHC). The primary sampling unit (PSU) consisted of a distinct group of enumeration areas (EAs) from the sampling frame referred to as a cluster. An EA is usually a clearly defined geographic area which groups several households together for population and housing census. A two-stage stratified sampling technique was used to select the households. Each of the 36 states and the Federal Capital Territory was stratified into urban and rural areas, creating 74 sampling strata. In the first stage, 1400 (580 urban and 820 rural) EAs were selected from the sampling strata with probability proportional to EA size. In the second stage selection, 30 households were selected from every cluster through equal probability systematic sampling, resulting in a total sample size of about 42,000 households (Figure 1). One-third of the total sample size of households (14,000) were selected for anaemia testing. Using an estimated proportion of WRA that are anaemic (P = 0.578), design effect (Deft = 1.434), relative standard error (α = 0.01), individual response rate (Ri = 97%), household gross response rate (Rh = 95%), and the number of eligible individuals per household (d = 1.032),35 the sample size in terms of the number of households (n) was calculated using the formula36 Flowchart for the sampling procedure. The survey was successfully carried out in 1389 clusters in 36 states and Federal Capital Territory comprising 747 LGAs from August to December 2018. Eleven clusters, with deteriorating law-and-order situations, were dropped during the fieldwork. To prevent bias, no replacements and no changes to the pre-selected households were allowed in the implementing stages. Anaemia testing was conducted for WRA in one-third of sampled households selected through equal probability systematic sampling from the total sample size of 42,000 households. The inclusion criteria were all WRA, either permanent residents or visitors who stayed in the sampled household the night before the survey. Women who did not agree to provide consent and women outside the age of 15–49 years were excluded. A blood sample from a finger prick site was drawn into a microcuvette, and a haemoglobin analysis was carried out on-site with a battery-operated portable HemoCue analyser (HemoCue Hb 301 system, Sweden). Anaemia status at the time of the survey is the dependent variable. Pregnant women with a haemoglobin level less than 11 g/dL and non-pregnant women with a haemoglobin level less than 12 g/dL were considered anaemic.35,37 Anaemia was categorized as mild (haemoglobin (Hb) of 10.0–10.9 g/dL for pregnant women and 11.0–11.9 g/dL for non-pregnant women), moderate (Hb of 7.0–9.9 g/dL for pregnant women and 8.0–10.9 g/dL for non-pregnant women), and severe (Hb < 7.0 g/dL for pregnant women and 0)’ where ‘adjust’ is the amount of the adjustment, ‘alt’ is the altitude in 1000 feet (converted from metres by dividing by 1000 and multiplying by 3.3), ‘adjHb’ is the adjusted haemoglobin level, and ‘Hb’ is the measured haemoglobin level in grammes per decilitre. Regarding smoking adjustment, no adjustment for women who smoked less than 10 sticks per day, while the haemoglobin of women who smoked 10–19, 20–39, and 40 or more sticks of cigarette per day were adjusted by –0.3, –0.5, and –0.7 g/dL, correspondingly. The variables were grouped into individual maternal characteristics, socio-economic and household characteristics, and health service–related factors based on the conceptual framework for maternal anaemia determinants.2 The individual characteristics included the age of the respondent, marital status (never in a union, married/living with a partner, and divorced/separated/widowed), family size (<5 and ⩾5), sex of household head (female and male), ever had a termination of pregnancy (yes and no), breastfeeding status (yes and no), body mass index (BMI) (underweight, normal, overweight, and obese), and modern contraceptive use (yes and no). The total children ever born (0, 1, 2–4, and ⩾5) were regrouped into four categories of parity (nulliparity, primiparity, multiparity, and grand multiparity), correspondingly.39 BMI was converted from a numeric to a categorical variable based on the World Health Organization (WHO) BMI.35 As BMI is not appropriate for pregnant women, we used stature (height) for all categories of WRA categorized as short stature (<145 cm) and normal (⩾145 cm).35 The socio-economic and household characteristics included region (North-Central, North-East, North-West, South-East, South-South, and South-West), type of residence (urban and rural), highest education (no education, primary, secondary, and higher), employment (unemployed and employed), wealth index (poorest, poor, moderate, rich, richest), access to sanitation (unimproved and improved), the main source of drinking water (unimproved and improved), ownership of a mosquito bed net for sleeping (yes and no), respondent having slept under a mosquito bed net the night before the survey (yes and no), and media exposure (none and any form). Based on the consumption of 10 food groups in the 24 h preceding the survey, women were categorized into low (<5) and high diversity (⩾5) groups.35 The health service–related factor is the extent to which respondents considered the distance to a health facility as a problem (not a problem, not a big problem). Data were analysed using SPSS 20 (IBM Corp., Armonk, NY). We adjusted the data for sampling weights, stratification, and multistage sampling before analysis to account for the non-proportional allocation of the sample to the different states and provide representative population estimates. The basic characteristics of the respondents were presented using frequencies, population estimates, and percentages (weighted). Pearson’s chi-square test was used to evaluate the association between anaemia prevalence and independent variables. Multicollinearity was assessed using the variable inflation factor (VIF). The independent variables showed no multicollinearity (minimum VIF = 1.00, maximum VIF = 3.80). All variables with a p value ⩽ 0.25 in bivariate analyses were further analysed using multivariable complex samples logistic regression. In addition, we included age, stature, and parity in the model for pregnant women based on clinical significance. The results of regression analysis were presented by crude/unadjusted odds ratio (COR) and adjusted odds ratio (AOR) with 95% confidence intervals (CIs), F statistics, and p values. The McFadden test statistic for overall WRA, non-pregnant women, and pregnant women ranged from 0.02 to 0.04. Since values ranging from 0.2 to 0.4 indicate good model fit and values beyond 0.4 indicate excellent fit, our models might not be the best fit.40 However, McFadden test, a log-likelihood-based pseudo-R2 that represents the improvement in model likelihood over a null model, is influenced by sample size (the smaller the sample size, the higher the value), number of predictor variables, and number of categories of the dependent variable and its distribution asymmetry.40 Statistical significance for the multivariable complex sample logistic regression analyses was set at p < 0.05. The 2018 NDHS protocol was reviewed and approved by the National Health Research Ethics Committee of Nigeria (NHREC) and the ICF Institutional Review Board. Informed consent was obtained from participants before interviews or biomarker tests were conducted. Consequently, our study, being a secondary analysis, did not require further ethical approval.
N/A
