Introduction

Low birth weight (LBW) is a leading determinant of neonatal mortality worldwide that particularly affects the develo** countries [1]. In the Sub-Saharan African (SSA) regions, 5 million low birth weight babies were recorded in 2015, of those 20% were secondary to malaria in pregnancy [2, 3] and the other causes included the maternal socio-demographic, gyneco-obstetrical characteristics and the fetus genetic disorders [4, 5]. Despite that LBW association with neonatal death is well established in SSA, half of all birth weights is still not recorded due weaknesses in the routine systems, which understates the magnitude of the problem and delays decision making [1, 6]. In some selected countries, the reported prevalence varied between 8 and 17% [7]. In Burkina Faso, the prevalence was 15.8% in 2005 and decreased to 13.4% in 2011 [7, 8].

To reduced malaria related adverse birth outcomes, the intermittent preventive treatment of malaria in pregnancy using sulfadoxine-pyrimethamine (IPTp-SP) in combination with vector control approaches such as free distribution of insecticide treated bed nets (ITNs) were adopted [9]. From an initial two-doses policy in the 2000s, the approach was revised in 2012, and a minimum of three doses of SP were recommended [9, 10]. The strategy consists of the administration of a full treatment course of sulfadoxine-pyrimethamine (SP) to pregnant women from the second trimester, monthly until delivery [9], and was proven efficacious [11]. The increasing trends of P. falciparum resistance to SP, particularly in western and central African regions are however raising concerns regarding the efficacy of the strategy [12]. Indeed, high prevalence of triple and quadruple mutations in the dihydropteroate synthase (dhps) and dihydrofolate reductase (dhfr) genes of P. falciparum are widely reported [13]. Also, the magnitude of the quintuple and sextuple mutations in the dhps and dhfr genes along with the emergence of septuple even octuple mutations reported in several countries is a further threat for the efficacy of the strategy and is reported to cause low birth weight in sub-Saharan Africa [14,15,16,1).

Table 1 List and definition of main variables
Full size table

Laboratory procedures

Thick and thin blood smears were prepared, dried and stained with 5% Giemsa for 30 min. Two independent microscopists examined the slides at 1000x magnification using light microscopy. Parasite densities were calculated by counting the number of asexual parasites per 200 white blood cells (WBC), and parasites per μl calculated assuming WBC count of 8000 cells per μl of whole blood. When the number of asexual parasites was less than 100 per 200 WBC, counting was done against at least 500 WBC. A slide was considered negative if no parasite was found after a review of 1000 WBC or 100 fields containing at least 10 WBC per field. In case of discrepant results (discrepant species or count difference of at least 50% between the two microscopists), a third microscopist assessed the slides and the average of the two closest reads was used.

Data processing and analysis

Data were collected on electronic questionnaires using REDcap (Research Electronic Data Capture) data collection tool and exported onto Stata version 15 (StataCorp. 2017, TX, USA) for cleaning and analysis. Categorical variables were summarized on frequency tables. Mean or median with respective standard deviations or quartiles were used to summarize numerical variables. T-test was used to compare means and odds ratios (OR) with 95% confidence intervals (95% CI) calculated by univariable logistic regression. Adjusted OR (aOR) were derived by backward multivariable logistic regression of factors which p-values were < 0.1 at univariable analysis and kee** those with p-values < 0.1 in the final model. Maternal age was not included in the model because of its strong correlation with the number of pregnancies. In addition, gravidity was dichotomized (1 or ≥ 2 deliveries) for birth weight factors assessment. The significance level was set at 5% (two-sided p-value).

Results

Study participants background characteristics

A total of 684 participants were assessed, 600 (87.7%) were included and 84 (12.3%) not eligible. Reasons for non-eligibility included 12 preterm births, 18 non-singleton births, 12 stillbirths, 2 miscarriages, 5 neonatal deaths, 15 data losses due to electronic pad breakdown, 16 concurrent participations to another research and 4 very ill neonates referred with missing birth weights.

Majority of women were unemployed (69.2%) with none or primary level of education (67.8%). Mean maternal age was 25 ± 6 (range; 15–42) years and 23% were aged less than 20 years. The median number of pregnancies stood at two (interquartile range, IQR; 1–4). Regarding malaria prevention measures, 92% stated using ITN the night before admission to delivery clinic and the IPTp-SP uptake was 77, 17.0, 5.0 and 1% for 3 or more doses, 2 doses, one dose and no dose respectively. Malaria infection was diagnosed in 17.5% (105/600) of participants, among them 83.8% were asymptomatic (Table 2).

Table 2 Socio-demographic, gyneco-obstetric and medical characteristics of study participants in Yako health district, Burkina Faso (n = 600)
Full size table

Factors associated with neonate low birth weight

Among the 600 neonates examined 51.7% were females. The prevalence of low birth weight was 11%, and the mean birth weight was 2942 ± 459 g. Mean birth weight was significantly lower in female neonates compare to males (2893.5 versus 2993.9 g, p < 0.01), in neonates from mothers with less than three SP doses than those with three or more doses (2821.1 versus 2978.7 g, p < 0.001), and if the mothers had malaria infection (2858.7 versus 2959.7 g, p = 0.02). By multivariate analysis, first pregnancy, malaria infection, uptake of less than three SP doses, maternal fever and being a female neonate were independently associated with an increased risk of LBW (Table 3).

Table 3 Prevalence of low birth weight and associated factors in rural Burkina Faso 2019–20 (n = 600)
Full size table

Discussion

This paper examined the prevalence and associated factors of low birth weight in rural health centers in Burkina Faso 5 years after the start of the three doses IPTp-SP policy. Importantly, the study was conducted in the rural area where the prevalence of low birth weight secondary to malaria was the highest [23]. To conduct this study, we hypothesized that low birth weight prevalence would be reduced from 13.4 to 10% 5 years after the IPTp-SP policy change [7]. In this study, we reported a prevalence higher than that expected (11.0%) and this advocates that more efficacious interventions are needed to reduce LBW prevalence. In the literature, low neonate birth weight is usually related to situations where fetal intrauterine malnutrition is produced due to alterations in the placental blood flow which originates from several factors including the mother socio-economic status, inadequate pregnancy care, or infections [24].

We found that the uptake of less than three doses of SP was significantly associated with a higher prevalence of low birth weight and a significant lower mean birth weight. This finding supports that the new IPTp-SP policy remains efficacious in the study settings to reduce low birth weight prevalence, and advocates more efforts to reduced non optimal uptake of SP. The reason behind the efficacy of the new policy could be related to the reduction of the placental inflammation caused by infectious pathogens [25], which reduces the dysregulation of the placental angiogenesis secondary to these infections, source of intrauterine growth retardation [26]. Also, the number of SP doses administered to women during pregnancy could be considered as a surrogate of the quality of care provided to pregnant women. Indeed, as the anti-malarial is administered under health care workers’ supervision, each contact represents an opportunity to detect and address maternal conditions that may affect the birth weight at delivery. Thus, increasing the number of SP doses, increases the number of contacts with the health care workers and improve the quality of care.

As expected [3], malaria was associated with an increased risk of LBW in the study area. This indicates that despite widespread use of malaria prevention methods, it remains a major factor of low birth weight in the rural area of Burkina Faso. Plasmodium resistance to sulfadoxine pyrimethamine reported in sub-Saharan Africa and Burkina Faso, could be the main reason of this reduced efficacy [14,15,16,17]. Indeed, the high number of triple and the presence of quintuple mutations (triple dhfr and double dhps mutations) reported in Burkina Faso decreased the efficacy of this antimalarial to reduce adverse birth outcomes [27]. Therefore, as well as the need to update resistance level grows [27, 28], it becomes also necessary to assess alternative interventions. Mefloquine and azithromycin-based combinations were tested and are considered equivalent to the current approach although more evidence is needed [29].

We noted that first pregnancy infants were at particular higher risk of LBW [30,31,32]. Several factors including the physiological [33,34,35], structural and immunological [36] patterns of the uterus could impact the placental blood flow and induced inadequate fetal growth. In addition, the higher predisposition of first pregnancy placenta to malaria parasites’ sequestration put the first pregnancy fetus at higher risk of LBW [37]. A special attention should be paid to those women, and tight supervision of these pregnancies are required in the rural area.

As, already reported [38], female neonates were at higher risk of low birth weight. The difficulty in this study was that neonate sex was determined only after delivery which limits the utility of the information for decision making. However, in settings where ultrasounds are available, it would be useful to tailor pregnancy care to the fetus sex.

In this study we described the prevalence of low birth weight and tentatively identified some risk factors in women from Yako health district. However, some limitations are worth noting. The consideration of term pregnancy neonates and the exclusion of twin deliveries and stillbirths somehow understate the actual magnitude of low birth weight at the site. Gestational age was collected using the mother knowledge of last menstrual period or the Ballard score due to the absence of ultrasounds, and this is frequently prone to errors. Also, the cross-sectional design of the study does not guaranty causality between exposures and outcomes. However, the study is worth underlying the extent of some factors influencing adequate maternal and child health and help identify areas with the most need of interventions.

Conclusion

Low birth weight prevalence was high in the study area. Reducing the proportion of pregnant women with low IPTp-SP uptake and malaria infection could reduce low birth weight occurrence and improve neonatal survival particularly among infants born from young mothers.