Abstract

Background

Despite the relatively high prevalence of Sickle cell trait (SCT) in Nigeria, there has been little research into the correlates of having children with SCT among Nigerian mothers, particularly in terms of socio-demographic differentials. This study aims to investigate the maternal socio-demographic correlates of having under-five children with SCT in Nigeria.

Method

Data from the 2018 Nigeria Demographic and Health Survey (Household Person Recode and Children Recode) were merged. Mothers with at least one under-five child whose genotype was known (n = 7,493) served as the unit of analysis. Three forms of outcome variable were explored. First was the number of children with SCT by each mother. Second, the number of children with SCT was categorized as zero, one, two or more. Lastly, each mother was categorized as either having no child(ren) with SCT or having at least one child with SCT. Subsequently, we assessed multilevel Poisson, ordinal and binary logit models to identify the best fitting model using Akaike and Bayesian Information Criterion. Multilevel binary logistic regression model was identified as best fit used to identify factors associated with having children with SCT. Adjusted Odds Ratio with 95% Confidence Interval (CI) were reported as measures of association.

Result

Nearly 62% of the mothers lived in rural areas, 38.2% had no formal education and 37.4% had ever given birth to at least five children. About 26.1% (95% CI = 25.2–26.9) of the mothers had children with SCT. By geographical variation, the Northwest region had the highest proportion of mothers of under-five children with SCT. Results from the multilevel binary logistic regression revealed that women who were traditionalists (AOR = 1.77; 95% CI = 1.04–3.02) were more likely of having children with SCT.

Conclusion

Though SCT is a genetic outcome, findings from this study suggest that important socio-demographic factors such as religion, and region of residence are significantly associated with having children with SCT in Nigeria. Sustained efforts on awareness campaigns on SCT are recommended.

Summary

This study analyzed the socio-demographic and geographic factors associated with having children who carries the sickle cell trait among Nigerian mothers, using data from the 2018 Nigeria Demographic and Health Survey. The findings provide insights that could guide public health strategies to reduce the burden of sickle cell disease in Nigeria.

Background

Sickle cell trait (SCT) is a genetic condition characterized by the presence of one abnormal hemoglobin beta gene allele (HbS) [1] and one normal hemoglobin gene (HbA). It is an inherited genetic condition that affects hemoglobin production, the protein in red blood cells responsible for carrying oxygen to the body’s tissues [2]. This contrasts with sickle cell disease (SCD), in which the affected individual has two abnormal hemoglobin beta gene alleles (SS). Unlike individuals with SCD, those with SCT do not experience symptoms related to sickling and generally enjoy a better quality of life. However, people with SCT can pass the mutated gene to their offspring, potentially leading to SCD if the other parent also carries the trait.

Sickle cell trait is found in 1 out of every 11 African Americans, though it can occur in individuals of any race or background [3]. Approximately 8% of African Americans carry the sickle cell trait, while in some African regions, more than 25% of the population are carriers [4]. West Africa has the largest number of individuals with the sickle cell trait [5].The greatest burden of SCD is in sub-Saharan Africa, where access to medical care and public health strategies to reduce mortality and morbidity are not uniformly available [6,7,8]. Findings have shown, that nearly 90% of the world’s SCD population live in three countries: Nigeria, India, and the Democratic Republic of Congo, where the disease affects up to 2% of the population [6]. Generally, people with sickle cell trait can lead normal, healthy and happy lives. However, they may experience specific health problems.

Understanding the correlates of SCT in the most affected population will be useful in preventing SCD. If effective and long-term control measures are not implemented, it is predicted that Nigeria’s SCT prevalence will increase by 100% by the year 2050 [9, 10]. Despite this, there is no comprehensive population-level or national data on sickle cell trait among Nigerian children until the conduct of the 2018 Demographic and Health Survey in the country. Similarly, no effort has been made to examine the socio-demographic correlates of having children with sickle cell traits among Nigerian mothers. Past studies have majorly focused on SCT in children [11,12,13,14,15] rather than the mothers. To fill the identified knowledge gap, this study provides empirical evidence on socio-demographic correlates of having children with sickle cell trait among childbearing women in Nigeria. The study aimed to [1] determine the percentage of childbearing women with children carrying the sickle-cell trait (SCT) in Nigeria; [2] identify the best model for factors associated with having children carrying SCT; and [3] investigate the factors associated with having children with SCT.

Methodology

Study design

A cross-sectional study, using nationally representative data from the Nigeria Demographic and Health Survey (NDHS), 2018.

Data sources

The 2018 NDHS, for the first time in a DHS survey, collected information on sickle cell disease (SCD) and sickle cell trait (SCT). Data from the Person Recode file (PR) and the Children Recode file (KR) of the 2018 DHS were merged and analyzed for the purpose of this study. A valid genotype was obtained from 11,210 children (unweighted counts) from households that were surveyed, after informed consent for the testing was obtained from a parent or other individual responsible for the child. The unit of analysis in the present study is mothers in a household who have at least one child with a known genotype result.

Dependent variable

The dependent variable was the number of children with SCT for each woman. We explored three forms of this variable. First, the number of children with SCT was used as a count variable and as an ordered polytomous variable. Thereafter, the count of the number of children with SCT each woman had was dichotomized into zero (i.e. women with no SCT children) and at least one (i.e. women with at least one child that has SCT).

Independent variables

The present study made use of variables that are similar to other studies [16,17,18] as well other available socio-demographic variables in the DHS data. These factors are considered to possibly influence the genotype of the child through the decision of the parents. Such factors include domicile (rural/urban), parity, mother’s history of child death, mother’s marital status, wealth index, maternal education, age at first birth, and religion.

Data management

For this study, two different recodes (Household Person Recode - PR, and Children’s Recode - KR) of the 2018 DHS dataset were merged. These data are publicly available on the Demographic and Health Survey (DHS) website (dhsprogram.com). While the KR file contains information relating to characteristics of the mother and child, the PR file contains the genotype of each child. Of the 28,973 records that were uniquely matched from the two datasets, 10,198 (35.2%) of the children had genotype results (Appendix B). Furthermore, these children were from 7,493 women. Since the unit of analysis was women with a known genotype result of their child(ren), only these 7,493 were eventually used for analysis.

Statistical analysis

Demographic and biological characteristics of the study participants were first described using mean, standard deviation and percentages. In addition, hotspot analysis was used to describe the distribution of mothers of under-five children who had children with SCT using ArcMap (v.10.9). For hotspot analysis, a shape file which consist of names of states in Nigeria, their coordinates, and prevalence of SCT was imported into ArcMap for analysis. Bivariate analysis was done to examine the patterns of the study outcome. We explored three forms of the outcome variable to identify the best fitting model for multivariable analysis. First, the number of a children with SCT that a woman had was modelled using Multilevel Poisson regression. Secondly, an Ordinal logistic regression model was fitted. Thereafter, the outcome variable was dichotomized as women having children with or without the sickle cell trait and a multilevel logit model fitted. The Maximum likelihood ratio test, the Akaike’s Information Criteria (AIC) and the Bayesian Information Criteria (BIC) were used to determine the model of best fit. The multilevel models accounted for hierarchical structure by adjusting the standard errors of the regression coefficients so that characteristics shared by women selected from the same cluster during the survey do not bias the estimates.

Finally at the multivariate analysis, multilevel binary logistic regression analysis (being the model of best fit) was used to examine the correlates of having children with sickle cell trait among mothers of under-five children in Nigeria. Two separate models were fitted. The first model (Model I) is the univariable model which measures the relationship between the outcome and each predictor variable. While the second model (Model II), is the adjusted model where all other socio-demographic, socio-economic and health-related factors that were significant (p < 0.10) in Model I were included. Although, the DHS data that was used for the analysis is a publicly available secondary data, approval was sought and obtained (Appendix A) prior to conducting the study. All analyses were conducted at 5% level of significance using Stata MP Version16.

Results

Socio-demographic characteristics of study participants

A total of seven thousand, four hundred and ninety-three (7,493) records were analysed. Of these, nearly three-third (59.5%) were from the Northern region with 24.0% specifically from the Northwest (TableÌý1). Also, the majority (61.1%) of these women were domiciled in rural areas, (26.7%) between the ages 25–29 years old, and 54.9% who had not more than primary education. Furthermore, majority (51.9%) of these women practiced Islam, and 40.2% belong to the poor/poorest household wealth quintile. Meanwhile, a high proportion (37.4%) of the women had given birth to at least five children, and 28.3% of these women had experienced child mortality.

Table 1 Socio-demographics of mothers of under-five children who had data on SCT among under-five children in Nigeria, DHS 2018

While the Northwest remains the region with the highest proportion of mothers of under-five children with sickle cell trait, states such as Jigawa, Yobe, Borno, Taraba (from the North), and Edo, Bayelsa and Enugu (from the South) had very high proportion (Fig.Ìý1).

Fig. 1

Geographical clustering of mothers with under-five children who had SCT in Nigeria

Distribution of mother with children carrying SCT in Nigeria, 2018 DHS

Among the different age groups, mothers who were aged 20–24 years had the highest proportion (25.3%) of children with the sickle cell trait. This proportion is similar t (25.1%), among those who were aged 25–29 years and (24.8%) among those of 40–44 years old. Also, women who had primary education (24.6%) had the highest proportion of children with SCT, followed by those with no formal education (24.5%). In addition, women who were traditionalist/practiced other religion beside Christianity and Islam (33.8%), who had given birth to three or four children (25.0%), from the poorest wealth quintile (25.7%), and had experienced multiple child mortality (26.8%) had the highest proportion of children with sickle cell trait. As indicated in TableÌý2, the mother’s religion was significantly associated (p < 0.05) with having at least one child with the sickle cell trait.

Table 2 Pattern of SCT distribution among mothers of under-five children in Nigeria, 2018 NDHS

Model selection

As presented in TableÌý2, the Multilevel Poisson regression had the least log-likelihood estimate. However, the multi-level binary logistic regression had smaller estimates of the Akaike’s Information Criteria (AIC) and Bayesian Information Criteria (BIC). Given the limitation of log-likelihood in terms of how it is affected by large parameters, and the consistency of the other two tests (AIC and BIC), it was concluded that the multilevel binary logistic regression was the best model to fit the present data. Thus, multilevel logistic regression model was used for further analysis of the correlates of having children with sickle cell trait among Nigerian mothers.

Table 3 Tests for model of best fit

Correlates of having children with sickle cell traits in Nigeria, 2018 DHS

Religion is one of the identified key factors that was associated with having SCT children among mothers of under-five in Nigeria. According to TableÌý3, Muslim women were more likely (OR = 1.12, 95% CI = 1.01–1.25) more likely than Christian women to have children with sickle cell trait. Also, traditionalist women had a higher likelihood of having children with SCT (OR = 1.75 (95% CI = 1.03–2.96). The results showed that mothers who had had multiple child deaths were 1.16 times (95% CI = 0.98–1.36) as likely to have children with sickle cell trait than women who had not experienced any child deaths. In addition, women from the North West of the country were 1.18 (95% CI = 0.99–1.41) times more likely to bear children with the sickle cell trait than women from the North Central.

Table 4 Multilevel binary logistic regression model of having children with Sickle Cell traits in Nigeria, 2018 DHS

In Model II, being a traditionalist woman persisted as a risk factor for having children with the sickle cell trait even after accounting for certain potential confounders (Model II). The findings showed that women who were traditionalist were more likely than women who were Christians to have SCT children, even when other factors (region of residence, father’s living situation, and mother’s experience with child mortality) were controlled (AOR = 1.77 (95% CI = 1.04–3.02).

Discussion

This study examined the socio-demographic differentials of having children with sickle cell trait among mothers of under-five children in Nigeria using a derivative dataset from the children recode and household recode file of the Demographic and Health Survey, 2018. Nigeria accounts for the highest number of under-five and maternal deaths globally and is currently off track to achieve key health-related Sustainable Development Goals [19]. Thus, the study investigated possible socio-demographic and socio-economic factors that associated with having children carrying SCT and the variation of such factors across the different regions of the country. Considering that the country faces significant challenges due to a shortage of healthcare professionals [8], findings from this study will be helpful to prevent the spread SCT in the country.

Distribution of sickle cell trait in Nigeria

The prevalence of sickle cell trait among under-five children is still unacceptably high in tandem with findings from other studies that were conducted in Nigeria [20,21,22,23] and Africa as a whole [21, 24]. Additionally, it was found that a high proportion of the mothers who had children with SCT were living in urban areas, from the poorest households (according to wealth index), and had no formal education. This finding is a cause for concern as individuals with HbAS have an increased risk of giving birth to children with the sickle cell trait or disease. It further highlights the importance of socio-economic factors in the distribution of SCT in Nigeria. Notably, this aligns with findings of another study that was conducted in Brazil where socio-demographic characteristics was found to play a significant role in the trajectory of sickle cell disease [25]. It could be that poverty, and lack of formal education are drivers of SCT in the country. These factors have been found to be associated with SCD [26]. Hence, this finding may serve as an impetus for further studies to investigate the role of these socio-economic factors on SCT in the country.

The distribution of SCT among children of Nigerian women provides valuable insights into the burden of sickle cell in the country. A high proportion of mothers of under-five with SCT children was found in seven states of the country, with four of these states located in the Northern region. This is a cause for concern and highlights the need for targeted interventions to reduce the future burden of SCD in the region.

Correlates of having children with SCT in Nigeria

The study identified factors associated with higher odds of having children carrying the SCT in Nigeria. These factors include religious affiliation, geographic region, and the mother’s history of child mortality. Although the latter two did not attain statistical significance.

Regarding religion, this study found that Muslim and traditionalist have a higher likelihood of bearing children with sickle cell trait. This may be due to differences in cultural practices, attitudes and beliefs about genotypes among these women. The Northern region of Nigeria is predominantly occupied by those who practice Islam with common beliefs and practices. For instance, though cultural and religious beliefs of the Northern residents do not prohibit them from seeking genetic counselling prior to marriage, most of them do not patronise any genetic counselling and testing centre [17]. The higher odds of having children with SCT among Muslim and Traditionalist women may be an indication of negative attitude towards sickle cell trait. Such negative attitude has been noticed among secondary school students in the North where majority of them reported that genotype of their partner will not influence their decision of partner before marriage [18]. Meanwhile, religion has been regarded as important within African cultures, emphasizing its roles in health-seeking behaviour [27].

The mother’s experience of child mortality is another factor that increased, though not significantly, the likelihood of having children with sickle cell trait in Nigeria. This suggests that women who have had multiple child deaths were more likely to have children with sickle cell trait compared to women who have not experienced any child deaths. This may be related to the fact that the mother’s experience of child mortality most probably could have resulted from undiagnosed / diagnosed sickle cell anemia (disease) in the deceased siblings, since most parents are not aware of their own genotype until the birth of a child with sickle cell disease [28].

Overall, these findings have important implications for public health efforts aimed at reducing the burden of sickle cell disease in Nigeria. Distribution of childbearing with SCT differ across geo-political regions and certain socio-demographic variables. This suggest that enlightenment and public health education can be more effective if tailored to peculiarities of regions and religious affiliation. The study findings also suggested that the prevalent high under-five mortality in Nigeria may have been driven covertly by sickle cell disease. This imply that there is need to strengthen health education around the subject of sickle-cell among intending and potential parents so that preventive measures such as pre-marital genetic counseling can be given due attention, since Nigeria and Democratic Republic of Congo are predicted to remain the countries most in need of policies for the prevention and management of sickle cell anemia in 2050 [6].

In summary, the findings suggest that regional and socio-economic factors are important determinants of SCT prevalence in Nigeria, and that targeted public health interventions aimed at addressing these factors could help to reduce the burden of sickle cell disease in the country.

Limitation of study

Sickle cell trait is associated with sudden death in individuals during periods of extreme physical exertion [29]. However, the fact that SCT may be associated with mortality cannot be investigated in this present study. This is because it was only the genotype of living children that were readily available in the dataset. Hence, it was not possible to associate mortality with genotype result in the present study. Also, there was no measure of the women’s awareness and knowledge on SCT in the dataset. This could have been a useful factor to be considered in the correlates of having children with sickle cell trait in Nigeria.

Implication for practice and policy recommendations

Based on the importance of geographical region, religion, and experience of child mortality in reducing the prevalence of SCT, the following recommendations were made: Focus on targeted interventions for women from specific regions, such as the North West. There may be misconceptions associated with certain health issues, including sickle cell. Efforts to improve religious beliefs and attitudes towards sickle cell could involve targeted campaigns and education initiatives aimed at women of religious groups that are most vulnerable to having children with SCT.

Conclusion

In conclusion, the study identified factors associated with having children carrying the sickle cell trait among reproductive-age women in Nigeria. These factors include religious affiliation, the living status of the husband, and the mother’s history of child mortality. These findings suggest that cultural, socio-demographic, and geographic variations are potential game changers in the distribution of SCT across populations in Nigeria.

Religious beliefs may influence perception and attitude towards choice of partner and reproductive decisions before marriage. Additionally, geographic disparities in healthcare services and education likely contribute to regional differences in proportion of sickle cell trait. Meanwhile, the mother’s experience of child mortality further highlights the potential danger of having more than one child with sickle cell trait in the family. On the other hand, having more than one child with SCT possibly emphasizes beliefs that influence attitude of the mothers towards SCT.