Sociodemographic factors associated with anemia among women of reproductive age in Mali; Evidenced by the Mali Malaria Indicator Survey 2021: Mixed-Effects Multilevel and Spatial Model Analysis | BMC Women’s Health

Sociodemographic characteristics

This study includes a total weighted sample of 10,765 women of reproductive age from the 2021 Mali Demographic and Health Survey. 2,155 (20.0%) of the total study participants were between the ages of 15 and 19, 1959 ( 18.2%) were from the Koulikoro region, 8,103 (75.3%) were from rural areas, 6,941 (64.5%) did not attend formal classes. education, 10,196 (94.7%) were Muslim, 10,099 (93.8%) were male heads of household, 1,960 (18.2%) were the poorest, 3,567 (33.1%) had no drinking water sources improved, 5101 (47.4%) had unimproved sanitary facilities, 4094 (38.0%) were anemic. Of these, 153 (1.4%) suffered from severe anemia, while 2,530 (23.5%) and 1,410 (13.1%) suffered from moderate and mild anemia, respectively, in Mali (Table 1).

Spatial analysis results

Spatial distribution of anemia

In Mali, anemia status was analyzed geographically using 261 clusters. The number of anemia cases in each cluster corresponds to an enumeration area at each point on the map. Analysis of this study of the spatial distribution of anemia showed that a higher proportion of anemia was found in the southern and southwestern region of Mali. The northern and northeastern region of Mali had a low proportion of anemia (Fig. 1).

Spatial autocorrelation anemia

The spatial autocorrelation result reveals whether anemia in Mali is randomly distributed in the region, clustered, or scattered. The results of the spatial autocorrelation study showed a clustering effect on anemia across the country. Clustered patterns (on the right side of the red box) demonstrated a clustering effect on anemia in Mali. The exits have automatically generated keys on the right and left sides of each panel. The probability that this clustered pattern is the result of random chance is less than 1%, based on a z-score of 8.992 (p-value < 0.001). The bright red and blue colors in the trailing tails indicate a higher level of significance (Fig. 2) .

The result of the access point scan

The result of the hotspot analysis shows the high proportion (hotspot) and the low proportion (cold spot) of anemia areas in Mali. Red colors were observed in Kayes, in some parts of Koulikore, Sikasso and Segou, which are hot spot areas (high proportion of anemia). Cold spots (areas with a low percentage of women with anemia) colored green were found in the southern part of Tombouctou and in the southern part of Koulikore, Bamako, the northern part of Sikasso and Gao (Fig. 3).

Interpolation or spatial prediction

Based on the sampled region, the spatial interpolation approach predicts the proportion of anemia for unsampled areas. The map of the area was described using the standard Kriging method. The red color represents the projected low prevalence of anemia. If the color of the area changed from red to blue, it indicates that more people in the area are anemic than previously expected. The country is predicted to have anemia at a low prevalence, as shown by the red color. According to the prediction results, the southern part of Timbuktu, Gao, the eastern part of monak, the northern part of Mopti and Bamako have high prevalence of anemia. The blue color prediction showed that the regions of Kayes, Koulikore, Sikasso and Segou had the highest prevalence of anemia in the whole country (Fig. 4).

Model Comparison

Four models were built for this multi-stage investigation. The first model was built. Without independent factors, it is possible to determine how community variation affects women’s anemia status. The second model included variables at the individual level. Community-level features were incorporated into the third model. Finally, the fourth model took into account factors at both the individual and community levels. The ICC in the null model showed that among women of reproductive age, there was a variation in the state of anemia of 7.63% in the communities. The variation in the state of anemia among women of reproductive age is described by variables at the individual level in 10.62% of the occurrences. The difference in the state of anemia between women of reproductive age is explained by 9.35% of the variables at the community level. In the end, 12.47% of the variances among women of reproductive age were caused by variables at the individual and community levels. The deviance was used to assess the model’s fitness for model comparison (AIC). As a result, Model IV, which included factors at both the individual and community levels and had the lowest deviation value (AIC), was determined to provide the best fit. Variables with a significance level of p < 0.05 were considered significant predictors of anemia status among women of reproductive age (Table 2).

Pearson’s Chi-Square Analysis of Factors Associated with Anemia

Pearson’s Chi-square analysis was used for age, place of residence, region, religion, educational level, sex of household head, wealth index, sources of drinking water, and types of facilities. health among women of reproductive age. The result of Pearson’s Chi-square analysis showed that anemia had a significant association with place of residence, region, religion, educational level, household sex, wealth index, drinking water sources, and households. types of sanitary facilities among women of reproductive age (Table 3).

Bivariate and multivariate logistic regression

Bivariate logistic regression was used for age, place of residence, religion, educational level, household sex, wealth index, drinking water sources, and type of bathroom among women of reproductive age. The result of the bivariate analysis showed that anemia had significant relationships with age, place of residence, region, religion, educational level, gender of the head of household, wealth index, sources of drinking water and income. type of health service among women of reproductive age. Variables that had a p value less than 0.05 were considered in the multivariate analysis. According to the result of the multivariable regression, the key variables related to anemia among women of reproductive age were the age of the woman, the place of residence, the educational level, the sex of the household head, the wealth index , the sources of drinking water and the type of sanitary service. The probability of anemia among reproductive women who were between the age range of 20 to 24 years was 0.8 times lower. [ AOR = 0.817; 95% CI = (0.638,1.047); P = 0.000] female relatives who were between the age range of 15 to 19 years. The probability of anemia among women of reproductive age living in rural areas was once more likely [ AOR = 1.053; 95% CI = (0.880,1.260); P = 0.000] in relation to women living in the urban area. The likelihood of anemia among women of reproductive age who attended higher education were 0.4 times less likely to have anemia [AOR = 0.401; 95% CI= (0.278,0.579); P = 0.000] compared to women who did not attend formal education. The likelihood of anemia among women of reproductive age who were animists were 3.1 times more likely to be anemic [AOR = 3.10; 95% CI= (0.763,12.623) P = 0.04] in relation to women who were Muslims. The probability of anemia among male-headed households was 0.65 times less likely to be anemic [AOR = 0.653; 95% CI= (0.536,0.794); P = 0.000] compared to a female-headed household. The probability of anemia among the richest was 0.62 times less likely to have anemia [ AOR = 0.629; 95% CI= (0.524,0.754) P = 0.000] relative to women who were poorer. The probability of anemia among women who used not improved drinking water sources was 1.1 times more likely[aor=1117;Ic=(10171228);P=0021}EncomparaciónConthewomenwhoareusedtobeinfectedAor=1018;Ic=(09171130);P=0041}comparedtowomenwhousedimproveddrinkingwatersources(Table[AOR=1117;CI=(10171228);P=0021}comparedtowomenwhousedimproveddrinkingwatersourcesTheoddofanemiaamongwomenwhowereusedunimprovedtoiletfacilitywere1timemorelikely[AOR=1018;CI=(09171130);P=0041}comparedtowomenwhousedimprovedtoiletfacility(Table[AOR=1117;IC=(10171228);P=0021}encomparaciónconlasmujeresqueutilizaronfuentesmejoradasdeaguapotableLaprobabilidaddeanemiaentrelasmujeresqueusaronbañosnomejoradosfue1vezmásprobable[AOR=1018;IC=(09171130);P=0041}encomparaciónconlasmujeresqueusaroninstalacionessanitariasmejoradas(Tabla[AOR = 1117;CI=(10171228);P = 0021}comparedtowomenwhousedimproveddrinkingwatersourcesTheoddofanemiaamongwomenwhowereusedunimprovedtoiletfacilitywere1timemorelikely[AOR = 1018;CI=(09171130);P = 0041}comparedtowomenwhousedimprovedtoiletfacility(Table4).