Postpartum moderate to severe anemia in a low-income country, Burkina Faso

Background In Burkina Faso, Sahelian country with limited resources, the maternal mortality ratio is still high and anemia is one of the major contributing factors. To identify epidemiological aspects of postpartum anemia in order to better guide public policy in its prevention and management. Methods This was a secondary analysis of data from the 2010 Demographic and Health Survey in Burkina Faso. The classification of anemia in women according to the World Health Organization was considered. For the research of associated factors, moderate or severe anemia was considered. Results A total of 1616 women were included. Hemoglobin levels were less than 12 g/dl in 839 women (51.9%). It was less than 10 g/dl in 251 cases (15.5%). There has been a significant increase in the risk of postpartum anemia for women in the Sahel region (OR=2.9; IC [1.2-7.4]; p=0.024). The adjusted effect of the woman's status about postpartum amenorrhea also showed an increased risk in women still having no return from menstruation. Conclusion Anemia is common during postpartum. It is of multifactorial origin. Further studies are needed to better adapt its current prevention strategies and investigate the relationship between Sahel region and anemia.


Abstract
Background In Burkina Faso, Sahelian country with limited resources, the maternal mortality ratio is still high and anemia is one of the major contributing factors. To identify epidemiological aspects of postpartum anemia in order to better guide public policy in its prevention and management.
Methods This was a secondary analysis of data from the 2010 Demographic and Health Survey in Burkina Faso. The classification of anemia in women according to the World Health Organization was considered. For the research of associated factors, moderate or severe anemia was considered.
Results A total of 1616 women were included. Hemoglobin levels were less than 12 g/dl in 839 women (51.9%). It was less than 10 g/dl in 251 cases (15.5%). There has been a significant increase in the risk of postpartum anemia for women in the Sahel region (OR=2.9; IC [1. 2-7.4]; p=0.024). The adjusted effect of the woman's status about postpartum amenorrhea also showed an increased risk in women still having no return from menstruation.
Conclusion Anemia is common during postpartum. It is of multifactorial origin. Further studies are needed to better adapt its current prevention strategies and investigate the relationship between Sahel region and anemia.

Background
Maternal mortality remains a scourge, particularly in West African countries including Burkina Faso [1]. It is a silent burden described as a neglected public health problem [2]. It depends on multiple factors. Among the indirect obstetric causes of maternal death, anemia plays an important role [3].
Although anemia has been recognized as a public health problem for many years, little progress has been reported in this area and its high global prevalence remains a problem [4]. The World Health Organization and the United Nations Children's Fund also reiterate the urgency of combating anemia in general and stress the importance of recognizing its multifactorial etiology, a prerequisite for effective control programs [5]. Postpartum anemia is therefore a common complication with prevalence ranging from 10-80% depending on the geographic area [6.7]. It can be a result of pregnancy anemia that can be aggravated by blood loss during childbirth [8]. In addition, in the postpartum period, iron deficiency anemia can occur in breastfeeding mothers who are sensitive to iron depletion if energy and nutrient intake in their diet is inadequate [9]. In Burkina Faso, Sahelian country with limited resources [10], the maternal mortality ratio is still high and anemia is one of the major contributing factors [11]. There are no national data on prevalence or associated factors of postpartum anemia. The objective of this study was to identify these epidemiological aspects of postpartum anemia in order to better guide public policy in its prevention and management with the aim of contributing to the reduction of maternal mortality.

Methods
This was a secondary analysis of data from the 2010 Demographic and Health Survey (DHS) in Burkina Faso. A database request was made on the DHS website (https://dhsprogram.com/What-We-Do/Survey-Types/DHS.cfm). The various databases were obtained electronically in SPSS and Stata format. The collection of DHS data was carried out according to a rigorous methodology and complied with the ethical procedures of research on human subjects as well as the legislation in force in Burkina Faso. This was a cross-sectional study with mixed sampling (stratified on the place of residence and in clusters). It concerned the population of individuals who were living in ordinary households throughout the country. The sample was stratified to provide adequate representation of urban and rural residence in the thirteen regions of the country. The survey was conducted in 14424 households. The data were collected by direct interview with individuals and by measuring biological parameters including hemoglobin (Hb) levels and anthropometric parameters such as weight and height. The blood samples were obtained by the "Dried Blood Spots" (DBS) technique using blotting paper and centralized in the laboratory of the Regional Blood Transfusion Centre in Ouagadougou where tests were conducted to determine hemoglobin levels. Hemoglobin levels were measured in g/dl. The classification of anemia in women according to the World Health Organization was considered: no anemia for hemoglobin levels greater than or equal to 12 g/dl; mild anemia for hemoglobin levels less than 12 g/dl but above 10 g/dl; moderate anemia for hemoglobin levels below 10 g/dl but greater than or equal to 8 g/dl; severe anemia for hemoglobin levels below 8 g/dl [12].
From the DHS database obtained, only data on women who were in postpartum at the time of the survey were extracted and analyzed. The period from delivery to one year after birth was considered.
On this basis, 1616 women were included (Figure 1). The dependent variable was represented by hemoglobin levels. For the research of associated factors, moderate or severe anemia was considered. A link between a hemoglobin level under 10 g/dl and the socio-demographic characteristics, clinical characteristics of women was explored from pregnancy, childbirth and postpartum data. The STATA 14.2 software was used for analysis. After a description of the sample, multilevel logistic regression was used for univariate and multivariate analysis to identify factors associated with postpartum anemia. The first level was represented by the woman and the second level by the household and the third level was the clusters (enumeration areas of the National Institute of Statistics and Demography). We appreciated the relevance of multilevel analysis by judging the values of the intraclass correlation coefficient from one level to another. We also adjusted the data using the weighting performed on the statistical analysis unit. The odds ratio (OR) was the measure of association used in studying the sense of association, strength, precision and significance.
All variables that had a p-value of less than 0.2 during the univariate analysis were progressively included in the multivariate model with evaluation of the Akaike information criterion (AIC). The significance level used in multivariate was p <0.05.

Prevalence of postpartum anemia
Hemoglobin level was less than 12g/dl in 839 women (51.9%). It was less than 10g/dl in 251 women (15.5%). The distribution of women by hemoglobin level is summarized in Table 1.

2.2.
Factors associated with moderate or severe anemia in postpartum Univariate analysis the odds of being admitted for males is 5.44 times that of females.
The odds of being anemic (moderate or severe) among postpartum women in the Sahel region were 3.7 times that of postpartum women in the Central Region (OR = 3.7, CI [1. 5-8.9]; = 0.004).
The odds of being anemic (moderate or severe) among women of the second quintile were significantly high compared to odds of women of the fourth quintile (OR = 1.9, CI [1.0-3.5]) with a lightly degree of significance (p = 0.041). Table 2 summarizes the relationship between the sociodemographic characteristics of patients and the presence of moderate to severe postpartum anemia in univariate analysis.
There was no significant association between moderate or severe postpartum anemia and clinical characteristics (pregnancy's characteristics) in univariate analysis (Tables 3 and 4).
Finally, we found a significant association between severe to moderate anemia and being in lactational amenorrhea or being in the first trimester of the postpartum. In fact, compared with those who were not in lactational amenorrhea, those in amenorrhea were 1.9 times more likely to have postpartum anemia than the others (OR = 1.9, CI: [1.2 -2.9], p = 0.005). Women in the first trimester of the postpartum compared to women in the fourth trimester of the postpartum were 2.1 times more likely to be anemic (p = 0.001). The associations between moderate to severe anemia and childbirth and postpartum characteristics in univariate analysis are shown in Table 5.

Multivariate analysis
Following the multivariate analysis, the administrative region of woman, the woman's status about postpartum amenorrhea, and woman's marital status were lightly significantly associated with moderate to severe postpartum anemia. There was a significant increase in risk of nearly 200% to present postpartum anemia for women in the Sahel region combined with a decrease of 70% of risk of postpartum anemia for women in the Hauts-Bassins region in comparison with the women of the Central region. The adjusted effect of the woman's status about postpartum amenorrhea also showed an increased risk in women still having no return from menstruation. The same effect was found for marital status where women who are in union have been found to be less at risk of developing postpartum anemia than non-union women. Table 6 presents the factors associated with postpartum anemia in multivariate analysis.

Discussion Strengths and limitations of this study
This study is one of the few population studies witch are investigate postpartum anemia. Indeed, most of the data in the literature on the subject concern samples in a hospital environment. As a result, we filled a major gap that was the lack of postpartum anemia prevalence data in the similar countries.
However, important limitations are to be taken into account when analyzing the DHS data: first, the data are from 2010, so the current situation may have changed. Next, we did not have any variables that indicated the clinical status of women during pregnancy, particularly in terms of general illnesses or health events that could explain the subsequent occurrence of anemia in the postpartum (placenta previa, bleeding during pregnancy, anemia during pregnancy, etc.). In addition, the DHS database

Prevalence of postpartum anemia
Hemoglobin levels were less than 12 g/dl in 51.9% of women. Moderate to severe anemia (Hb levels below 10 g/dl) were found in 15.5% of cases. Heterogeneity of thresholds for defining postpartum anemia has been noted in the literature, making comparisons difficult. Indeed, while some studies have considered mild or severe anemia, others have just chosen to address the issue in women whose anemia was moderate or severe (case of our study). Also, most of these studies were conducted in hospitals. The prevalence of moderate to severe anemia found in our series testifies to the problem of anemia in our country. This is all the more so since this high prevalence could contribute to maintaining already significant maternal mortality in our context. Lakew et al reported an anemia prevalence of 22.1% in 2015, considering a threshold of 12g/dl in a population of 7332 lactating women in Ethiopia. [9]. Bergman et al noted a higher prevalence (22%) of moderate or severe anemia (Hb level below 10 g/dl) in a large retrospective study of 43,807 women who gave birth between 1993 and 2008 [13]. However, this latest study only involved women in the immediate postpartum (48 hours of postpartum). Our study concerned the expanded postpartum to twelve months after delivery. This longer period would help reduce this prevalence of moderate or severe anemia in our study. Indeed, physiologically, blood losses are compensated in a few weeks or even months, following their occurrence if all physiological mechanisms are not altered and also if these losses are not severe. In 2009, Somdatta et al in a cross-sectional study in a village in northern India reported a 70% prevalence of postpartum anemia by setting a threshold for hemoglobin levels below 11g/dl [14]. In 2014, Rakesh et al in India found a prevalence of 47.3% by the sixth week of postpartum with a threshold value of 12g/dl [15]. Other proportions, still high, were found as 60. professional activity [19], wealth quintile [4], age [15,19], ethnicity [16], educational attainment [4,16,18] and race [8,17,18,20,21]. The same was true for variables such as multiple pregnancies [13], caesarean delivery [17] and postnatal consultation within two months of delivery [22], obesity [8], and parity [22]. The administrative region was associated with moderate or severe postpartum anemia, with two regions acting in the opposite way. Indeed, while belonging to the Sahel region was a risk factor for postpartum anemia, belonging to the Hauts-Bassins region was a protective factor (reducing the risk of postpartum anemia) when comparing these two regions to Central Region.
Geographical factors combined with other factors such as cultural practices coupled with religious practices and eating habits may explain these differences. However, in our series, some key variables were not included in the analysis model because of their sensitivity: religion and ethnicity.

Conclusion
Anemia in postpartum is common in our country despite the policy of pre-and post-natal supplementation in iron and folic acid. Given its importance and also its variability according to the geographical regions, it is necessary to investigate the quality of the conduct of the anti-anemic prophylaxis at the pregnant woman and the postpartum and to carry out additional research to better understand the supposed link between postpartum anemia and geographical areas. This could guide the fight against this maternal mortality factor in order to be more effective     Table 6. Factors associated with postpartum anemia in multivariate analysis Figure 1 flowchart of selected women