Drivers of trimester disparities in asymptomatic malaria during pregnancy in Kintampo South District: evidence from an Oaxaca-Blinder decomposition Analysis Authors: Dennis Bardoe, Daniel Hayford, Robert Bagngmen Bio, George Ahiaka, Latifatu Alhassan Abubakar, Simon Nyarko Published: 2026 ABSTRACT AbstractBackgroundAsymptomatic malaria in pregnancy (AMiP) remains a major public health challenge with important maternal and neonatal consequences. Despite ongoing malaria control efforts, little is known about trimester-specific disparities in AMiP and the extent to which these differences are explained by socioeconomic, clinical, and environmental factors. Most existing studies have focused on overall malaria burden rather than explaining disparities across gestation. This study examined the determinants of trimester disparities in AMiP in the Kintampo South District of Ghana using Oaxaca-Blinder decomposition. MethodsA facility-based cross-sectional study was conducted among 462 pregnant women in the Kintampo South District of Ghana. Peripheral blood samples were screened for Plasmodium species using microscopy, while structured questionnaires captured sociodemographic, clinical, and environmental characteristics. Data were analysed with STATA 17 using Oaxaca-Blinder decomposition to partition trimester differences in AMiP into explained(characteristics) and unexplained (coefficients) components.ResultsThe overall prevalence of AMiP was 13.4%, with a significantly higher burden in the second trimester (χ2 = 10.96, p < 0.001). Plasmodium falciparum accounted for most infections. Oaxaca-Blinder decomposition showed that 73.7% of the trimester disparity in AMiP was explained by differences in observed characteristics (endowments). The major contributors to the explained disparity were suboptimal uptake of intermittent preventive treatment with sulfadoxine-pyrimethamine (IPTp-SP) (71.3%), frequent evening outdoor activity (58.6%), infrequent use of mosquito repellents (50.6%), proximity to overgrown vegetation (42.5%), ≤2 malaria diagnostic tests during antenatal care (ANC) (24.1%), and infrequent iron-folate supplementation (18.4%). Conversely, ownership of long-lasting insecticidal nets (LLINs) reduced the unexplained disparity by 41.9%, highlighting its protective effect. ConclusionDisparities in AMiP across trimesters are primarily driven by modifiable socioeconomic, programmatic, and environmental factors rather than inherent biological differences in gestation. These highlight the need to strengthen the prioritisation of early initiation of ANC and frequent exposure to provider-supervised IPTp-SP uptake. In addition, promoting sustained use of LLINs and repellents, alongside community-level environmental management, is critical to reducing exposure risk. Finally, embedding these targeted interventions within ANC services, supported by coordinated multisectoral efforts, will be essential to reduce trimester-specific inequities and improve maternal health outcomes.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12889-026-27837-9.Keywords: Asymptomatic malaria, Ghana, Intermittent preventive treatment with sulfadoxine-pyrimethamine (IPTp-SP), Long-lasting insecticidal nets (LLINs), Malaria in pregnancy, Maternal health, Oaxaca-Blinder decomposition, Trimester disparities Introduction Malaria in pregnancy (MiP) remains a major public health concern, particularly in malaria-endemic regions of sub-Saharan Africa [1]. Among the various forms of MiP, asymptomatic malaria in pregnancy (AMiP) presents a distinct challenge. This is because it often occurs without overt clinical symptoms despite the presence of Plasmodium parasites in the bloodstream [2]. From a biological standpoint, asymptomatic malaria becomes especially critical beyond the first trimester, when physiological and immunological adaptations increase susceptibility to malaria infection and parasitaemia [3]. This is particularly evident in primigravidae, who lack specific immunity to chondroitin sulphate A (CSA)-binding variants of Plasmodium falciparum [3]. As such, infected erythrocytes sequester in the placenta and adhere to CSA expressed on syncytiotrophoblasts. This leads to placental malaria without necessarily manifesting as clinical disease [3]. During mid-pregnancy, there is a temporary decline in cell-mediated immunity, which typically increases vulnerability to Plasmodium falciparum. This occurs because the maternal immune system prioritises fetal protection over pathogen defence [4]. In 2023, over 12.4 million pregnancies recorded in 33 moderate-to-high transmission countries in the WHO African Region were infected with malaria [5]. West and Central Africa recorded the highest burden [5]. Despite increasing awareness and interventions, asymptomatic cases often go undetected and untreated, especially in resource-limited settings where routine screening is uncommon or poorly implemented [6]. This is because the asymptomatic nature of the infection fosters complacency, delays in detection, and a false sense of health security among patients and providers [7]. As part of the response, the World Health Organization (WHO) has championed the use of intermittent preventive treatment with sulfadoxine-pyrimethamine (IPTp-SP), along with the promotion of long-lasting insecticidal nets (LLINs) and prompt diagnostic testing as part of antenatal care (ANC) protocols [5]. By 2021, over 90% of pregnant women and young children had received free LLINs, surpassing the 80% target [8]. In addition, about 80% of pregnant women had at least two IPTp-SP doses [9]. In spite of extensive research on MiP in the Bono East Region [10–12], there is a lack of evidence explicitly quantifying and explaining trimester-specific disparities in AMiP within the Kintampo South District. Existing studies [10, 13–15] have largely examined overall prevalence or focused on symptomatic infections, with limited attention to how the burden of asymptomatic infection varies between the second and third trimesters. More importantly, to the best of our knowledge, no study has assessed what proportion of these differences could be attributed to observable socioeconomic, clinical, and environmental factors versus unobserved influences. This distinction is critical because without disentangling the drivers of trimester disparities, intervention strategies remain broadly targeted and may fail to address the underlying sources of inequity. Against this background, this study examines trimester disparities in asymptomatic malaria in pregnancy and quantifies the contribution of socioeconomic, clinical, and environmental factors using Oaxaca-Blinder decomposition. The central research question guiding this study is “What factors account for the observed differences in AMiP between the second and third trimesters, and to what extent are these disparities driven by differences in characteristics and variations in the effects of these characteristics?” By addressing this question, the study contributes to a deeper understanding of host-parasite dynamics during gestational immunomodulation and the persistence of subclinical malaria beyond the first trimester. The findings are also expected to inform more targeted ANC strategies, strengthen surveillance for asymptomatic malaria, and support efforts to reduce inequities in malaria burden. Methods Study area and context The study was conducted in the Kintampo South District of Ghana (Fig. 1). It covers a total land area of approximately 1,774.85 km² with an estimated population of 97,690 [16]. The district shares boundaries with several malaria-endemic areas [10, 16] and continues to experience high malaria transmission [17]. Ecologically, it lies within the Voltaian Basin and Southern Voltaian Plateau, with a woodland savannah vegetation that supports diverse mosquito breeding habitats [16]. Climatically, the district has a modified tropical continental climate with two rainy seasons (March-July and August-November), annual rainfall ranging from 1,150 to 1,250 mm, high relative humidity (up to 90–95%), and temperatures between 24 °C and 30 °C [16]. These conditions favour year-round mosquito breeding and malaria transmission. Besides, the District is served by one district hospital, two health centres, one clinic, 17 CHPS compounds, and a private maternity home [16]. However, coverage of key malaria preventive interventions, including intermittent preventive treatment in pregnancy using sulfadoxine-pyrimethamine (IPTp-SP) and insecticide-treated net (ITN) use, remains sub-optimal [13, 15]. This gap, in the context of high and continuous malaria transmission, contributes to a higher likelihood of asymptomatic malaria infections among pregnant women. Bono east regional map showing the study area Study design and population This study employed a facility-based cross-sectional analytical study design among pregnant women attending routine antenatal care (ANC) at Kintampo South District Hospital. It spanned between July and October 2024, coinciding with the peak malaria transmission season. The target population consisted of apparently healthy pregnant women in their second and third trimesters without symptoms of malaria at the time of enrolment. The focus on women beyond the first trimester was intentional. This was because AMiP are more reliably detectable during this period due to increased placental sequestration of parasites compared to the first trimester [18]. In addition, many women initiate ANC after the first trimester in this setting [19]. This makes the second and third trimester populations more representative of routine ANC attendees. Furthermore, preventive interventions such as IPTp-SP are administered starting from the second trimester [15]. These factors make the second and third trimesters particularly relevant for assessing programmatic and behavioural influences on AMiP. Eligibility criteria Inclusion Pregnant women were eligible for inclusion if they were in their second or third trimester of pregnancy, resided within the district at the time of the study, were enrolled in ANC services at the hospital, and provided informed consent. Exclusion Pregnant women were excluded if they presented with acute febrile illness requiring immediate malaria treatment. Sample size estimation The sample size was estimated using the Cochran formula for proportions (Eq. 1), assuming a 95% confidence level, a 5% margin of error, and a conservative prevalence of 50% in the absence of prior local estimates. To account for the multistage sampling design and potential heterogeneity in the study population [20], a 20% adjustment was applied for design effect and non-response. The final minimum sample size was 462 participants. Sampling and enrolment procedure First, the Kintampo South Government Hospital in Jema was purposively chosen. As the district’s main hospital, it provides a wide range of essential health services, including ANC, interventions for malaria prevention and treatment, and other maternal health services. Within the hospital, systematic random sampling was applied to recruit participants. On each clinic day, the ANC register provided the sampling frame. The number of eligible women present (N) was divided by the daily recruitment target (n) to obtain the sampling interval (k) using the notation (Eq. 2): The first participant was chosen randomly by selecting a number between 1 and k, after which every kth woman on the register was recruited until the target was reached. Eligible women were approached after ANC registration. The study objectives, risks, and rights to voluntary withdrawal were explained and those who agreed provided written informed consent (signature or thumbprint) to affirm their engagement. Study variables Outcome variable The primary outcome variable was the asymptomatic malaria status of the consented pregnant women (AMiP). This was defined as the laboratory-confirmed presence of malaria parasites detected by microscopy in the absence of fever (axillary temperature < 37.5 °C) or malaria-related symptoms. Main explanatory variable The main explanatory variable was the trimester of pregnancy, with the second and third trimesters treated as distinct categories. In this study, the second trimester was defined as the period between 14 and 26 weeks of gestation, and the third trimester was defined as the period between 27 weeks and delivery. These two trimesters were analysed as separate categories to assess trimester-specific differences in the prevalence of AMiP. Covariates The covariates included sociodemographic characteristics, clinical (obstetric) and programmatic factors, and environmental exposure and household practices. Wealth index, as part of the sociodemographic characteristics, was computed as a proxy for household socioeconomic status using principal component analysis (PCA). The computation was based on participants’ self-reported information about household assets and amenities. Each participant was assigned a wealth score derived from the first principal component, and these scores were ranked to generate wealth categories. For this study, the original quintiles were collapsed into three groups (poor, middle, and wealthy) to ensure adequate sample sizes and clearer comparisons across groups. This approach is consistent with methods applied in previous malaria-related studies [21]. Although IPTp-SP is initiated from 14 weeks of gestation, the number of doses recorded in this study mirrors cumulative uptake up to the time of assessment. This was classified as sub-optimal (≤ 2) and optimal (≥ 3). With early and consistent ANC attendance, participants could have received three or more IPTp-SP doses by the late second or early third trimester at the time of data collection. Therefore, the classification of ≥ 3 doses represents optimal uptake relative to gestational age. Data collection methods Data were collected using a structured, pre-tested questionnaire (Supplementary File 1) adapted from the Ghana Demographic and Health Survey 2022, complemented by laboratory screening for malaria infection. The questionnaires were administered through face-to-face interviews, while blood samples were collected for parasitological examination. Validity and reliability of the questionnaire The questionnaire (Supplementary File 1) consisted of four sections and was self-administered through face-to-face interaction. Section A captured sociodemographic information. Section B collected clinical and obstetric data. Section C covered programmatic factors. Section D focused on environmental exposure and household variables. It was pre-tested among 10 pregnant women at a nearby facility to identify ambiguities, after which necessary modifications were made. Internal consistency of the questionnaire was assessed using McDonald’s omega (ω). This measures how well the items capture the same underlying concept. The obtained value (ω = 0.961) indicates a high level of reliability. This suggests that the items were highly consistent in measuring the constructs of interest. Serological screening, identification, and estimation of parasitaemia of Plasmodium species In the laboratory, blood samples were obtained by finger prick after cleaning the site with an alcohol swab by the available medical laboratory technician. Approximately 10 µL of blood was used to prepare thick smears on glass slides measuring 15 × 15 mm. The smears were air-dried, fixed in 70% methanol for 30 s, and labelled with the participant’s identification number. Each slide was then stained with 10% Giemsa solution for 15 min and examined under oil immersion (×100) using a binocular light microscope. Parasite density was determined by counting the number of asexual parasites per 200 white blood cells (WBCs) in the thick film. A smear was classified as negative when no parasites were detected after examining 200 high-power fields [22]. In cases where the readings differed (positive versus negative, or a discrepancy of ≥ 50% in parasite density), the senior microscopist’s report was taken as final. Parasite counts were expressed as parasites per microliter of blood, assuming a standard WBC count of 8,000/µL [23], using the formula (Eq. 3): Data management and statistical analysis Data were double-entered, cleaned, and analysed with STATA 17 (StataCorp LLC, College Station, USA). Preliminary data checks were performed to identify and address missing values, inconsistencies, and outliers. Descriptive statistics, including frequencies, percentages, means, and standard deviations, were computed to summarise the characteristics of the study population. Pearson’s chi-square tests were used to assess the difference in proportions between categorical variables. To decompose differences in asymptomatic malaria prevalence between trimesters, the Oaxaca-Blinder decomposition was performed using the “xi: oaxaca outcomevar varlist, by(groupvar) logit” command. This method partitions the overall disparity into the “endowment effect (E)” (differences in observed characteristics) and the “coefficient effect (C)” (differences in the effects of these characteristics). The general form of the decomposition model was expressed as (Eq. 4): Where is the difference in mean prevalence between the second (A) and third trimester (B), is the mean outcome (prevalence of AMiP) in each group, is average covariates (endowments), is the pooled reference coefficients, is group-specific coefficients, and is the interaction due to simultaneous differences in endowments and coefficients. Components Explained (E) = portion of the gap due to group differences in characteristics (endowments). Expressed as: Unexplained (C) = portion of the gap due to differences in coefficients (returns to characteristics, including potential unobserved factors). Expressed as: Interaction (I) = portion due to simultaneous differences in endowments and coefficients. Expressed as: Contribution of individual covariates () = contribution of a single variable j to the explained disparity. Expressed as: The percentage contribution of variable to total disparity is expressed as: Results Sociodemographic characteristics of pregnant women, stratified by trimester The sociodemographic and clinical characteristics of the 462 pregnant women included in the study are presented in (Table 1). Their mean age was 29.3 ± 3.61 years (95% CI: 28.97–29.63), with most participants (71.2%) aged between 26 and 30 years. A higher proportion of the pregnant women were married (69.3%). In terms of education, formal education was more common (64.7%), and the vast majority were employed (96.5%). A greater proportion of pregnant women lived in urban areas (60.4%), and most (89.4%) lived in block houses roofed with iron sheets. Wealth distribution showed that 43.5% of women were in the middle tertile. With respect to obstetric history, 45.7% were secundigravidae, and the majority were multiparous (74.2%). In addition, more than half of the pregnant women (54.8%) had a normal Body Mass Index (BMI). The mean haemoglobin level was 9.78 ± 1.89 g/dl (95% CI: 9.60–9.94), with a considerable proportion of women classified as anaemic (39.4% had moderate anaemia, 19.3% mild anaemia, and 14.1% severe anaemia). Table 1Baseline sociodemographic characteristics stratified by trimesterVariablen [%]2nd trimester (337 [72.9%])3rd trimester (125 [27.1%])Age Mean ± SD [29.30 ± 3.61 (95% CI: 28.97–29.63)] 18–2549 [10.6]44 [9.5]5 [1.1] 26–30329 [71.2]242 [52.4]87 [18.8] 31–4084 [18.2]51 [11.0]33 [7.1]Marital status Not married142 [30.7]111 [24.0]31 [6.7] Married320 [69.3]226 [48.9]94 [20.3]Education No formal education163 [35.3]114 [24.7]49 [10.6] Formal education299 [64.7]223 [48.3]76 [16.5]Labour force Unemployed16 [3.5]14 [3.0]2 [0.4] Employed446 [96.5]323 [69.9]123 [26.6]Place of residence Rural183 [39.6]143 [31.0]40 [8.7] Urban279 [60.4]194 [42.0]85 [18.4]Household category Mud with thatch27 [5.8]24 [5.2]3 [0.6] Mud with iron sheets22 [4.8]18 [3.9]4 [0.9] Blocks with iron sheets413 [89.4]295 [63.9]118 [25.5]Wealth index Poor105 [22.7]80 [17.3]25 [5.4] Middle201 [43.5]136 [29.4]65 [14.1] Wealthy156 [33.8]121 [26.2]35 [7.6]Gravidity Multigravida106 [22.9]54 [11.7]52 [11.3] Secundigravida211 [45.7]163 [35.3]48 [10.4] Primigravida145 [31.4]120 [26.0]25 [5.4]Parity Multiparous343 [74.2]307 [66.5]36 [7.8] Nulliparous119 [25.8]30 [6.5]89 [19.3]Body Mass Index (BMI) Obese (≥ 30 kg/m2)66 [14.3]47 [10.2]19 [4.1] Overweight (≥ 25.0 - <30 kg/m2)117 [25.3]80 [17.3]37 [8.0] Underweight (< 18.5 kg/m2)26 [5.6]22 [4.8]4 [0.9] Normal weight (≥ 18.5 - <25.0 kg/m2)253 [54.8]188 [40.7]65 [14.1]Anaemia statusMean haemoglobin ± SD [9.78 ± 1.89 (95% CI: 9.60–9.94)] Severe Anaemia (< 7 g/dl)65 [14.1]50 [10.8]15 [3.2] Moderate Anaemia (7–9.9 g/dl)182 [39.4]138 [29.9]44 [9.5] Mild Anaemia (10–10.9 g/dl)89 [19.3]58 [12.6]31 [6.7] Non-anaemic (> 11 g/dl)126 [27.3]91 [19.7]35 [7.6]Frequency of ANC visits 1–3191 [41.3]139 [30.1]52 [11.3] 4–7193 [41.8]138 [29.9]55 [11.9] ≥ 878 [16.9]60 [13.0]18 [3.9]Media exposure (Frequency of listening to the radio and watching television) Less than twice a week122 [26.4]85 [18.4]37 [8.0] At least twice a week161 [34.8]114 [24.7]47 [10.2] Almost every day179 [38.7]138 [29.9]41 [8.9]Distance from household to health facility Major problem330 [71.4]251 [54.3]79 [17.1] Not a problem132 [28.6]86 [18.6]46 [10.0]Ownership of long-lasting insecticidal nets (LLINs) Yes276 [59.7]182 [39.4]94 [20.3] No186 [40.3]155 [33.5]31 [6.7]Number of IPTp-SP taken ≤ 2 (Sub-optimal)297 [64.3]280 [60.6]17 [3.7] ≥ 3 (Optimal)165 [35.7]57 [12.3]108 [23.4]Frequent uptake of iron folate Not frequent191 [41.3]154 [33.3]37 [8.0] Frequent271 [58.7]183 [39.6]88 [19.0]Malaria diagnosis at ANC ≤ 2319 [69.0]234 [50.6]85 [18.4] ≥ 3143 [31.0]103 [22.3]40 [8.7]Frequent use of mosquito repellents Not frequent114 [24.7]89 [19.3]25 [5.4] Frequent348 [75.3]248 [53.7]100 [21.6]Frequent evening outdoor activity Yes204 [44.2]150 [32.5]54 [11.7] No258 [55.8]187 [40.5]71 [15.4]Proximity of the household to the drainage tunnel Close (≤ 50 m)158 [34.2]128 [27.7]30 [6.5] Far (> 50 m)304 [65.8]209 [45.2]95 [20.6]Proximity of the household to the refuse dumping site Close (≤ 50 m)157 [34.0]130 [28.1]27 [5.8] Far (> 50 m)305 [66.0]207 [44.8]98 [21.2]Proximity of the household to overgrown vegetation Close (≤ 50 m)180 [39.0]144 [31.2]36 [7.8] Far (> 50 m)282 [61.0]193 [41.8]89 [19.3]Proximity of the household to livestock shelters Close (≤ 50 m)130 [28.1]93 [20.1]37 [8.0] Far (> 50 m)332 [71.9]244 [52.8]88 [19.0]Frequent draining of household gutters as larval control practices Yes304 [65.8]221 [47.8]83 [18.0] No158 [34.2]116 [25.1]42 [9.1]Household water storage practices Storage media without a seal221 [47.8]167 [36.1]54 [11.7] Storage media with a seal241 [52.2]170 [36.8]71 [15.4]Total462 [100]337 [72.9]125 [27.1]Abbreviations: ANC Antenatal care, BMI Body Mass Index, n Frequency, % Percentage, SD Standard Deviation, ≤ Less than or equal to, ≥ Greater than or equal to, < Less than, > Greater than, kg/m2 Kilogram per square meter Baseline sociodemographic characteristics stratified by trimester Abbreviations: ANC Antenatal care, BMI Body Mass Index, n Frequency, % Percentage, SD Standard Deviation, ≤ Less than or equal to, ≥ Greater than or equal to, < Less than, > Greater than, kg/m2 Kilogram per square meter Maternal health service utilisation and programmatic factors stratified by trimester Maternal healthcare utilisation patterns and related programmatic factors were also assessed (Table 1). Most of the pregnant women (41.8%) attended antenatal care (ANC) between 4 and 7 times, while 38.7% accessed radio or television almost daily, twice weekly. Perceived access to healthcare facilities was a challenge for many, with 71.4% reporting distance to health facilities as a major problem. Ownership and use of malaria prevention interventions varied. More than half (59.7%) of the women owned a long-lasting insecticidal net (LLINs). With respect to IPTp-SP uptake, 64.3% of participants had received ≤ 2 doses. Frequent uptake of iron-folate supplements was reported by 58.7% of the pregnant women. Environmental and household practices stratified by trimester Environmental exposure and household exposure factors were also examined (Table 1). Nearly half of the participants resided within 50 m of potential mosquito breeding sites, including drainage tunnels (34.2%), refuse dumping sites (34.0%), and overgrown vegetation (39.0%). Proximity to livestock shelters was reported by 28.1% of women. Household practices varied, with 47.8% of participants storing water in unsealed containers and 34.2% not engaging in frequent gutter draining as a larval control measure. Meanwhile, 44.2% of women reported frequent evening outdoor activity, and 75.3% frequently used mosquito repellents. Prevalence of asymptomatic malaria in pregnancy (AMiP), parasitaemia density, and Plasmodium species distribution stratified by trimester The overall prevalence of AMiP was 13.4%, with a significantly higher burden in the second trimester (χ2 = 10.96, p < 0.001) (Table 2). This pattern indicates that the risk of AMiP is not evenly distributed across gestation but peaks during the second trimester. With respect to parasitaemia density, 4.1% of women had high parasite counts (5,000–99,999 parasites/µL). A significant decline in parasitaemia density was observed from the second to the third trimester (χ2 = 11.65, p = 0.009). With respect to Plasmodium species distribution, Plasmodium falciparum was the most predominant (9.7%). Its prevalence was notably higher in the second trimester (9.1%) compared to the third trimester (0.6%). In addition, Plasmodium malariae accounted for 1.1% of infections, while 2.6% of women were co-infected with P. falciparum and P. malariae. Overall, these differences in species distribution across trimesters were statistically significant (χ2 = 11.73, p = 0.008) (Table 2). Table 2Prevalence of asymptomatic malaria and Plasmodium species distribution stratified by trimesterVariablen [%]2nd trimester (337 [72.9])3rd trimester (125 [27.1])χ2p-valueAsymptomatic malaria in pregnancy (AMiP) Positive62 [13.4]56 [12.1]6 [1.3]10.96< 0.001 Negative400 [86.6]281 [60.8]119 [25.8]Parasitaemia density High (5000-99,999 parasites/µL blood)19 [4.1]16 [3.5]3 [0.6]11.650.009 Moderate (1000–4999 parasites/µL blood)23 [5.0]22 [4.8]1 [0.2] Low (< 1000 parasites/µL blood)20 [4.3]18 [3.9]2 [0.4] Negative400 [86.6]281 [60.8]119 [25.8]Plasmodium species Plasmodium falciparium 45 [9.7]42 [9.1]3 [0.6]11.730.008 Plasmodium malariae 5 [1.1]4 [0.9]1 [0.2] Plasmodium falciparium and malariae 12 [2.6]10 [2.2]2 [0.4] Negative400 [86.6]281 [60.8]119 [25.8]Total462 [100]337 [72.9]125 [27.1]97.28< 0.001Abbreviations: df Degree of freedom, n Frequency, % Percentage, χ2 Chi-square, * p < 0.05, < Less than, > Greater than, µL Microlitre Prevalence of asymptomatic malaria and Plasmodium species distribution stratified by trimester Abbreviations: df Degree of freedom, n Frequency, % Percentage, χ2 Chi-square, * p < 0.05, < Less than, > Greater than, µL Microlitre Decomposing disparities in AMiP in the second and third trimesters stratified by individual covariates The Oaxaca-Blinder decomposition analysis revealed that the disparity in AMiP between the second and third trimesters was 0.118. 73.7% of the disparity in AMiP during the second and third trimesters was explained by differences in endowments (characteristics) between groups [0.087 (95% CI: 0.042, 0.092; p = 0.001)]. This indicates that most of the disparity arises from differences in the distribution of these characteristics between the two groups of women. Thus, it is not mainly driven by differences in the strength or direction of their associations with AMiP. The remaining 26.3% of the disparity was due to differences in how risk factors influenced AMiP across trimesters (0.031 [95% CI: 0.014, 0.052], p < 0.001). Differences due to characteristics (Explained component [E]) Several measurable sociodemographic characteristics (Table 3) contributed positively to widening disparities in AMiP between the second and third trimesters. Belonging to the poorest and middle wealth groups contributed 9.2% (0.008 [0.006, 0.013], p < 0.001) and 3.4% (0.003 [0.002, 0.007], p < 0.001) of the total gap, respectively. Similarly, residing in rural areas accounted for 6.9% (0.006 [0.004, 0.009], p < 0.001) of the disparity, reflecting the elevated malaria exposure typical of rural environments. Housing type also played a substantial role, with mud-thatch and mud-iron sheet dwellings explaining 9.2% (0.008 [0.006, 0.014], p = 0.004) and 5.8% (0.005 [0.003, 0.016], p < 0.001) of the trimester disparity, respectively. Furthermore, lack of formal education (4.6%; 0.004 [0.002, 0.006], p = 0.003) also influenced the disparity positively, hence widening the gap. In contrast, being unmarried (-6.9%; -0.006 [-0.010, -0.003], p < 0.001) and unemployment (-8.1%; -0.007 [-0.011, -0.003], p = 0.021) also influenced the disparity by narrowing the direction of the gap. Table 3Oaxaca-Blinder decomposition of disparities in asymptomatic malaria in pregnancy (AMiP) stratified by sociodemographic factorsExplanatory variableDifferences due to Characteristics (E)% of total gap95% CI (E)p-value (E)Differences due to Coefficients (C)% of total gap95% CI (C)p-value (C)Age Mean ± SD [29.30 ± 3.61 (95% CI: 28.97–29.63)] 18–25-0.006-6.9[-0.005, 0.003]0.1440.001-3.2[-0.100, 0.102]0.182 26–300.0033.5[-0.011, 0.018]0.1540.008-25.8[-0.034, 0.050]0.217 31–40RefRefMarital status Not married-0.006-6.9[-0.010, -0.003]< 0.001*-0.013-41.9[-0.081, -0.013]0.002* MarriedRefRefEducation No formal education0.0044.6[0.002, -0.006]0.003*-0.014-45.2[-0.054, -0.034]< 0.001* Formal educationRefRefLabour force Unemployed-0.007-8.1[-0.011, -0.003]0.021*0.00929[0.002, 0.071]0.023* EmployedRefRefPlace of residence Rural0.0066.9[0.004, 0.009]< 0.001*0.0039.7[ 0.001, 0.005]0.012* UrbanRefRefHousehold category Mud with thatch0.0089.2[0.006, 0.014]0.004*0.01445.2[0.012, 0.023]0.001* Mud with iron sheets0.0055.8[0.003, 0.016]< 0.001*-0.007-22.6[-0.012, -0.002]< 0.001* Blocks with iron sheetsRefRefWealth index Poor0.0089.2[0.006, 0.013]< 0.001*-0.001-3.2[-0.005, -0.002]0.009* Middle0.0033.4[0.002, 0.007]< 0.001*-0.017-54.8[-0.045, -0.012]< 0.001* WealthyRefRefMedia exposure (Listening to the radio and watching television) Less than twice a week0.0022.3[-0.002, 0.004]0.1620.00722.6[-0.023, 0.038]0.131 At least twice a week-0.007-8.1[-0.003, 0.002]0.1580.00619.4[-0.020, 0.032]0.246 Almost every dayRefRefAbbreviations: CI Confidence Interval, n frequency, % percentage, SD Standard Deviation, < Less than Oaxaca-Blinder decomposition of disparities in asymptomatic malaria in pregnancy (AMiP) stratified by sociodemographic factors Abbreviations: CI Confidence Interval, n frequency, % percentage, SD Standard Deviation, < Less than Within clinical and programmatic factors (Table 4), several malaria-preventive and healthcare utilisation indicators contributed meaningfully to the explained gap. Specifically, severe (9.2%; 0.008 [0.005, 0.012], p = 0.012), moderate (6.9%; 0.006 [0.003, 0.008], p = 0.016), and mild (4.6%; 0.004 [0.002, 0.005], p = 0.003) anaemia all contributed significantly to widening the disparities in AMiP between the trimesters. In addition, suboptimal uptake of IPTp-SP (≤ 2 doses) accounted for 71.3% (0.062 [0.030, 0.082], p = 0.001), while infrequent iron-folate supplementation contributed 18.4% (0.016 [0.012, 0.019], p = 0.034) of the explained disparities in AMiP. Similarly, ≤2 malaria diagnostic tests during ANC explained 24.1% (0.021 [0.014, 0.043], p = 0.014), and infrequent use of mosquito repellents accounted for 50.6% (0.044 [0.021, 0.061], p = 0.005) of the trimester gap. Finally, reporting illness confusion also widened the explained gap by 6.9% (0.006 [0.003, 0.009], p = 0.001). Table 4Oaxaca-Blinder decomposition of disparities in asymptomatic malaria in pregnancy (AMiP) stratified by clinical and programmatic factorsExplanatory variableDifferences due to Characteristics (E)% of total gap95% CI (E)p-value (E)Differences due to Coefficients (C)% of total gap95% CI (C)p-value (C)Gravidity Multigravida-0.0066.9[-0.079, 0.006]0.185-0.011-35.5[-0.046, 0.025]0.167 Secundigravida0.0022.3[-0.015, 0.018]0.248-0.017-54.8[-0.061, 0.024]0.108 PrimigravidaRefRefParity Multiparous0.0078.1[-0.027, 0.040]0.1990.0039.7[-0.002, 0.148]0.055 NulliparousRefRefBody Mass Index (BMI) Obese-0.0044.6[-0.005, 0.004]0.2820.00619.4[-0.052, 0.125]0.123 Overweight-0.009-10.3[-0.008, 0.008]0.1840.01651.6[-0.035, 0.069]0.134 Underweight-0.001-1.2[-0.008, 0.005]0.147-0.007-22.6[-0.029, 0.028]0.159 Normal weightRefRefAnaemia statusMean haemoglobin ± SD [9.78 ± 1.89 (95% CI: 9.60–9.94)] Severe Anaemia0.0089.2[0.005, 0.012]0.012*-0.015-48.4[-0.022, -0.008]0.004* Moderate Anaemia0.0066.9[0.003, 0.008]0.016*-0.004-12.9[-0.007, -0.001]0.001* Mild Anaemia0.0044.6[0.002, 0.005]0.003*0.01651.6[0.011, 0.036]< 0.001* Non-anaemicRefRefFrequency of ANC visits 1–30.0022.3[-0.004, 0.008]0.1870.00516.1[-0.030, 0.039]0.285 4–7-0.001-1.3[-0.003, 0.002]0.162-0.006-19.4[-0.021, 0.008]0.141 ≥ 8RefRefIllness confusion Yes0.0066.9[0.003, 0.009]0.001*-0.019-61.3[-0.035, -0.003]0.005* NoRefRefDistance from household to health facility Major problem0.0011.3[-0.005, 0.006]0.263-0.026-83.9[-0.055, 0.003]0.083 Not a problemRefRefOwnership of long-lasting insecticidal nets (LLINs) Yes-0.012-13.8[-0.026, -0.002]0.011*-0.013-41.9[-0.032, -0.007]0.023* NoRefRefNumber of IPTp-SP taken ≤ 2 (Sub-optimal)0.06271.3[0.030, 0.082]0.001*-0.011-35.5[-0.034, -0.006]< 0.001* ≥ 3 (Optimal)RefRefFrequent uptake of iron folate Not frequent0.01618.4[0.012, 0.019]0.034*-0.002-6.5[-0.005, -0.003]0.025* FrequentRefRefMalaria diagnostics at ANC ≤ 20.02124.1[0.014, 0.043]0.014*-0.005-16.1[-0.030, -0.002]0.283 ≥ 3RefRefFrequent use of mosquito repellents Not frequent0.04450.6[0.021, 0.061]0.005*-0.012-38.7[-0.019, -0.004]< 0.001* FrequentRefRefAbbreviations: ANC Antenatal care, BMI Body Mass Index, CI Confidence Interval, n frequency, % percentage, SD Standard Deviation, ≤ Less than or equal to, ≥ Greater than or equal to, < Less than, kg/m2 Kilogram per square meter Oaxaca-Blinder decomposition of disparities in asymptomatic malaria in pregnancy (AMiP) stratified by clinical and programmatic factors Abbreviations: ANC Antenatal care, BMI Body Mass Index, CI Confidence Interval, n frequency, % percentage, SD Standard Deviation, ≤ Less than or equal to, ≥ Greater than or equal to, < Less than, kg/m2 Kilogram per square meter Furthermore, some environmental and household practices (Table 5) also made strong positive contributions. In particular, frequent evening outdoor activity accounted for 58.6% (0.051 [0.022, 0.073], p < 0.001), while proximity to overgrown vegetation (≤ 50 m) explained 42.5% (0.037 [0.023, 0.081], p = 0.004) of the trimester disparity. Table 5Oaxaca–Blinder decomposition of disparities in asymptomatic malaria in pregnancy (AMiP) stratified by environmental exposure and household practicesExplanatory variableDifferences due to Characteristics (E)% of total gap95% CI (E)p-value (E)Differences due to Coefficients (C)% of total gap95% CI (C)p-value (C)Frequent evening outdoor activity Yes0.05158.6[0.022, 0.073]< 0.001*0.00619.4[0.003, -0.009]< 0.001* NoRefRefProximity of the household to the drainage Tunnel Close (≤ 50 m)0.0078.1[-0.055, 0.071]0.115-0.001-3.2[ -0.013, 0.010]0.115 Far (> 50 m)RefRefProximity of the household to the refuse dumping site Close (≤ 50 m)-0.001-1.3[-0.067, 0.066]0.2840.0013.2[-0.014, 0.014]0.284 Far (> 50 m)RefRefProximity of the household to overgrown vegetation Close (≤ 50 m)0.03742.5[0.023, 0.081]0.004*-0.008-25.8[-0.015, -0.005]0.009* Far (> 50 m)RefRefProximity of the household to livestock shelters Close (≤ 50 m)-0.016-18.4[-0.076, 0.043]0.585-0.004-12.9[-0.003, 0.002]0.539 Far (> 50 m)RefRefFrequent draining of household gutters as larval control practices Yes0.01517.2[-0.010, 0.042]0.2450.02167.7[-0.014, 0.049]0.234 NoRefRefWater storage practices Storage without a seal-0.001-1.3[-0.043, 0.040]0.9360.0013.2[ -0.004, 0.014]0.936 Storage with a sealRefRefAbbreviations: CI Confidence Interval, n frequency % percentage, ≤ Less than or equal to, ≥ Greater than or equal to, < Less than, > Greater than Oaxaca–Blinder decomposition of disparities in asymptomatic malaria in pregnancy (AMiP) stratified by environmental exposure and household practices Abbreviations: CI Confidence Interval, n frequency % percentage, ≤ Less than or equal to, ≥ Greater than or equal to, < Less than, > Greater than Differences due to coefficients [Unexplained component (C)] Among the unexplained components, several covariates (as presented in Tables 3 and 4, and 5) reinforced disparities through their differential effects. Mild anaemia, unemployment, mud-thatch housing, frequent evening outdoor activity, and rural residence accounted for 51.6% (0.016 [0.011, 0.036], p < 0.001), 29% (0.009 [0.002, 0.071], p = 0.023), 45.2% (0.014 [0.012, 0.023], p = 0.001), (19.4%, 0.006 [0.003, 0.009], p < 0.001), and 9.7% (0.003 [0.001, 0.005], p = 0.012) of the unexplained gap in AMiP between the second and third trimester, respectively. Conversely, several factors exhibited protective effects that narrowed the unexplained gap in AMiP between the second and third trimesters. Specifically, not being married reduced the unexplained disparity by -41.9% (-0.013 [-0.081, -0.013], p = 0.002), while having no formal education accounted for a reduction of -45.2% (-0.014 [-0.054, -0.034], p < 0.001). Likewise, residing in mud-iron sheet houses narrowed the unexplained gap by -22.6% (-0.007 [-0.012, -0.002], p < 0.001). Belonging to the poor and middle wealth groups contributed reductions of -3.2% (-0.001 [-0.005, -0.002], p = 0.009) and − 54.8% (-0.017 [-0.045, -0.012], p < 0.001), respectively. Comparable narrowing effects were also observed among clinical factors. Severe anaemia reduced the unexplained disparity by -48.4% (-0.015 [-0.022, -0.008], p = 0.004), while moderate anaemia accounted for − 12.9% (-0.004 [-0.007, -0.001], p = 0.001). Similarly, several programmatic factors demonstrated compensatory coefficient effects. Suboptimal IPTp-SP uptake (≤ 2 doses) narrowed the disparity by -35.5% (-0.011 [-0.034, -0.006], p < 0.001), infrequent iron-folate uptake by -6.5% (-0.002 [-0.005, -0.003], p = 0.025), ownership of LLINs by -41.9% (-0.013 [-0.032, -0.007], p = 0.023), and infrequent use of mosquito repellents by -38.7% (-0.012 [-0.019, -0.004], p < 0.001). Furthermore, illness confusion contributed a substantial narrowing effect of -61.3% (-0.019 [-0.035, -0.003], p = 0.005), while proximity to overgrown vegetation similarly reduced the unexplained component by -25.8% (-0.008 [-0.015, -0.005], p = 0.009). Discussion Prevalence of AMiP, parasitaemia density, and Plasmodium species This study assessed the prevalence of asymptomatic malaria in pregnancy (AMiP), parasitaemia density, and Plasmodium species distribution among pregnant women in their second and third trimesters. It was revealed that a notable proportion of pregnant women carried malaria parasites without showing symptoms. This estimate is broadly consistent with findings from other sub-Saharan African countries, where prevalence has typically ranged between 20% to over 40% [24, 25]. Similarly, studies conducted in Eastern and Central regions of Ghana have reported higher prevalence ranging between 43.4% and 44.4% [26, 27]. Regionally, similar patterns have been observed in Ethiopia [24] and Burkina Faso [28], where asymptomatic carriage among pregnant women remains common despite scaled-up interventions. The comparatively lower prevalence observed in the present study may reflect improved uptake of antenatal services and malaria prevention strategies. However, it still aligns with the broader epidemiological profile of AMiP across the sub-region [29]. Contextually, the difference in rates may indicate differences in endemicity, local transmission intensity, seasonal patterns, coverage of malaria control interventions, improved access to healthcare, and greater community sensitisation about preventive practices [30]. Rather than simply reflecting reduced transmission, this observed prevalence of AMiP may indicate a shift towards more heterogeneous malaria epidemiology. The study also revealed a significantly higher prevalence of AMiP in the second trimester compared to the third trimester. One possible explanation is the temporary decline in maternal immunity that occurs during mid-pregnancy, which increases susceptibility to infection [31]. As pregnancy progresses, partial immune recovery may occur, leading to lower infection rates in the third trimester [32]. More importantly, differences in malaria parasitaemia observed between the comparison groups are consistent with the established protective effect of intermittent preventive treatment with sulfadoxine-pyrimethamine (IPTp-SP) during pregnancy [33, 34]. As a key malaria control strategy, IPTp-SP is designed to reduce parasite burden and prevent new infections. Pregnant women in their second trimester may not have completed the full course of IPTp-SP, leaving them more exposed to infection. Therefore, lower parasitaemia levels among exposed groups are expected within this clinical context [33, 34]. This observation supports earlier reports indicating that the timing and completeness of preventive interventions influence malaria risk across different stages of pregnancy [33, 34]. Concerning the distribution of Plasmodium species, P. falciparum accounted for nearly all infections. Alongside, a smaller proportion were co-infected with P. falciparum and P. malariae. This aligns with the well-established pattern in sub-Saharan Africa where P. falciparum remains the principal species due to its high transmission efficiency and ability to evade host immunity [35, 36]. Consequently, this biological advantage enables it to outcompete other Plasmodium species in areas where multiple species coexist [37]. In contrast, the relatively low proportion of mixed infections observed in this study contrasts with reports from certain regions of West and Central Africa, where co-infections with P. malariae and P. ovale are more common [38, 39]. Such discrepancies may stem from ecological variations in vector species distribution, differing malaria transmission intensities, or the diagnostic sensitivity of tools used [40]. These findings suggest that, although progress has been made in reducing the burden of malaria among pregnant women, asymptomatic infections remain a significant and under-recognised reservoir [31]. The observed pattern highlights the need for enhanced screening strategies during ANC and for strengthening the implementation of existing preventive interventions. Decomposing disparities in AMiP in the second and third trimesters stratified by individual covariates The Oaxaca-Blinder decomposition analysis indicated that a substantial proportion of the disparity (73.7%) was explained by differences in sociodemographic, clinical (obstetric), programmatic, and environmental factors. In contrast, 26.3% was attributable to differences in the effects of these factors across trimesters. This pattern suggests that inequalities in exposure and access to preventive measures play a more dominant role than differential biological susceptibility alone [41]. Similar evidence from Ghana and other West African settings shows that variations in malaria risk during pregnancy are strongly shaped by disparities in ANC utilisation, uptake of intermittent preventive treatment, and living conditions [15, 42, 43]. These differences are not explained by trimester-specific physiological changes alone. The contribution of the unexplained component highlights the potential role of context-specific behavioural practices [44]. It also points to the timing and quality of intervention delivery [43]. In addition, it may reflect unmeasured factors such as immunity and parasite dynamics [44, 45]. Taken together, these findings suggest that the higher burden of asymptomatic malaria in later pregnancy is not solely a function of gestational progression [30]. It also reflects underlying structural and programmatic gaps that persist across the continuum of ANC. Differences attributable to characteristics (Explained component [E]) The predominance of the explained component implies that compositional differences between trimester groups play a critical role in shaping the risk of malaria [10]. Regarding sociodemographic factors, economic disadvantage and residency and its characteristics emerged as a central drive of the observed disparities. Specifically, belonging to poorer wealth categories, living in mud-thatch or mud-iron sheet houses, and residing in rural areas collectively account for over 34% of the trimester gap in AMiP. These findings align with studies from Nigeria and Ghana that identified poor housing quality, rural residency, and low socioeconomic status as consistent predictors of malaria infection during pregnancy [10, 30, 46]. From a macro-structural perspective, such disparities echo the unequal distribution of environmental and infrastructural protection. For instance, poorer households are more likely to reside in mosquito-prone areas with inadequate drainage, limited screening, and less access to preventive interventions [30, 47]. Also, rural residency, which explained nearly 7% of the gap, follows the same pattern. Thus, rural environments often combine ecological risk with lower coverage of ANC and preventive interventions [30, 48]. Again, housing type further compounded exposure differences, with mud-thatch and mud-iron sheet dwellings contributing substantially to the gap. This converges with ecological and epidemiologic evidence that housing quality is a major determinant of vector entry and indoor biting [10, 49]. Mechanistically, poorly constructed walls and roofs increase indoor mosquito density and contact time [10], thereby raising asymptomatic carriage. In addition, educational attainment showed a smaller but meaningful contribution. Lack of formal education is a known correlate of both low malaria literacy and reduced ANC attendance [15, 21]. This supports a social gradient interpretation, where differential knowledge, income, and empowerment translate into unequal adherence to preventive guidelines or interventions [15]. Interestingly, marital status and employment exerted countervailing effects, narrowing the disparity. This may indicate contextual shifts in household dependency and ANC decision-making among unmarried or unemployed women [10, 30]. This highlights that social protection factors, such as community outreach or partner-independent ANC engagement, might mitigate disadvantage in later trimesters [50]. The most pronounced structural gaps were observed in clinical and programmatic factors. Anaemia, especially across the various severity levels, contributed notably to explaining the disparity. This observation aligns with the established concept of gestational immunomodulation, where physiological changes during pregnancy temporarily alter immune responses [51, 52]. The observed contribution likely mirrors the physiological decline in maternal iron stores in the second trimester []. In addition, delayed initiation or inconsistent use of iron-folate supplementation may further heighten susceptibility to parasitaemia during this period [53]. This is because iron deficiency weakens maternal immune function and lowers resistance to infection [53]. In contrast, adequate and regular iron-folate intake supports red blood cell production and maintains optimal haemoglobin levels, both of which are essential for oxygen transport and effective immune response [51]. When supplementation is delayed or irregular, anaemia and micronutrient deficiencies tend to persist, thereby reducing the body’s capacity to suppress Plasmodium multiplication [51, 52]. This effect becomes more pronounced during the second trimester, when physiological demands rise, and exposure to malaria parasites is typically higher [54]. Again, suboptimal IPTp-SP uptake explained 71% of the total disparity, while infrequent iron-folate supplementation and ≤2 malaria diagnostic tests during ANC accounted for 18% and 24%, respectively. These gaps are typical of early-to-mid pregnancy, when many women delay ANC initiation [55]. Similar trimester-related coverage gaps have been documented in Tanzania and Ghana, where first or second ANC visits often occur too late for timely IPTp-SP administration [56, 57]. This suggests that programmatic bottlenecks, rather than behavioural neglect, perpetuate trimester differences in infection risk [46]. Likewise, the low use of mosquito repellents and inconsistent symptom recognition further demonstrate how knowledge and fragmented service delivery interact to widen disparities. In many rural and low-resource settings, health education on malaria prevention during pregnancy is often limited to bed net use, with little emphasis on supplementary protective measures such as repellents or avoidance of evening outdoor exposure [15]. This narrow focus, coupled with irregular health promotion activities, leads to limited awareness and low uptake of alternative preventive options [15]. Similarly, the inconsistent recognition of malaria-related symptoms can be traced to two main factors. It mirrors inadequate counselling during antenatal visits, where pregnant women may not receive sufficient guidance on subtle or atypical signs of malaria [30]. Secondly, the asymptomatic nature of infection often masks disease perception, leading many women to underestimate their susceptibility [58]. As a result, reduced awareness and perceived risk are likely to lower care-seeking motivation and delay timely preventive action [15]. When viewed as a whole, these gaps indicate that preventive knowledge alone is insufficient when not reinforced through continuous, integrated ANC counselling and community-based education. Furthermore, some behavioural and ecological exposures were also key contributors. In particular, frequent evening outdoor activity and proximity to overgrown vegetation contributed notably to the explained gap in trimester prevalence. These findings are consistent with ecological exposure theory, which posits that behavioural patterns and environmental modification mediate disease distribution, even in the presence of preventive efforts [59]. In malaria-endemic zones, such micro-environmental factors often mirror broader deficits in urban planning and public sanitation. In this context, the observed patterns have been reported in rural sub-Saharan Africa, where environmental modification around households strongly determines vector abundance [10, 30]. This confirms the value of environmental management as part of ANC-linked malaria control efforts. Differences attributable to coefficients (Unexplained component [C]) The unexplained component mirrors differences in the effects of covariates on AMiP risk between trimesters, rather than differences in their distribution. In this study, both reinforcing (positive) and compensatory (negative) coefficient effects were identified, indicating complex interactions between trimester-specific immune dynamics and contextual factors. Several contextual factors exerted a stronger influence in the second trimester, thereby widening the unexplained disparity in AMiP. The greater contribution of unemployment and poor housing quality indicates that socioeconomic disadvantages are likely to interact with the timing of care access and preventive outreach. In many low-resource settings, early ANC visits are less consistent, and women in lower-income households may experience delayed coverage of IPTp or health education, magnifying trimester differences [60]. Similarly, the amplified effects of rural residence and inadequate housing point to persistent structural vulnerabilities, limited vector control infrastructure, weak surveillance, and environmental conditions that favour transmission [61]. These patterns correspond with regional evidence that malaria risk is not uniformly distributed but shaped by spatial inequities in service provision and environmental management [30]. In this context, the stronger coefficients in the second trimester may indicate that systemic support mechanisms, such as outreach or household-level interventions, begin to take effect only later in pregnancy. Conversely, several variables narrowed the unexplained gap, implying that their structural impact weakened or reversed as pregnancy progressed. The reduced influence of wealth gradients and low educational attainment, for instance, suggests that ANC attendance and programmatic interventions become more uniformly applied in later trimesters. This trend aligns with observations from antenatal programme evaluations in sub-Saharan Africa, where repeated facility contacts help offset socioeconomic barriers to malaria prevention [30]. Likewise, health system factors also appear to play a stabilising role. The diminishing influence of IPTp-SP uptake, LLIN ownership, and iron-folate supplementation on the disparity likely reflects improved programme consistency across gestational stages. As ANC visits increase, access to preventive guidelines or interventions becomes less stratified by income or residence, thereby equalising protection [62]. LLINs serve as a physical barrier that prevents mosquito bites during peak nocturnal biting hours [10, 15]. They also have an insecticidal effect that reduces vector survival and density within households [10]. Furthermore, the protective role of LLINs mirrors both individual-level behaviour and programme-level success [63]. Pregnant women who consistently use LLINs may also be more engaged with ANC services and health education, thereby benefiting from a broader package of preventive interventions [13]. This dual mechanism is particularly important in malaria-endemic settings. In such contexts, consistent exposure to infectious bites sustains transmission. Recommendations for policy and practice This study demonstrates that AMiP remains prevalent, with the highest burden occurring during the second trimester. Suboptimal IPTp-SP uptake and inconsistent preventive practices contribute substantially to this pattern. These findings point to the need for more targeted and programmatically grounded interventions within ANC services. In light of the elevated second-trimester burden, ANC services should move beyond symptom-based approaches by integrating routine asymptomatic malaria screening at defined gestational stages. Institutionalising scheduled parasite testing, particularly during mid-pregnancy, would enable earlier detection and treatment of infections that would otherwise remain unnoticed. This requires sustained availability of diagnostic tools, including rapid diagnostic tests and microscopy, alongside supportive supervision to maintain quality. Closely related to this is the observed gap in IPTp-SP uptake, which appears to drive much of the trimester disparity. Strengthening adherence to IPTp-SP protocols within ANC is therefore critical. Ensuring uninterrupted SP supply, reinforcing directly observed therapy (DOT), and systematically tracking completion of at least three doses as a core performance indicator would improve coverage and effectiveness. At the same time, the persistence of infection despite existing interventions highlights gaps in preventive behaviours. Greater emphasis should be placed on consistent use of long-lasting insecticide-treated nets (LLINs). This calls for a shift from distribution-focused strategies to behaviour-oriented approaches. Such approaches should include structured counselling and follow-up during ANC visits. They are particularly important for women in rural and peri-urban settings, where exposure risk is higher. Beyond facility-based care, the role of environmental and behavioural exposure cannot be overlooked. Integrating community-level malaria prevention strategies would complement biomedical interventions and help reduce transmission risk. These strategies include drainage maintenance and the elimination of stagnant water. They also involve targeted behaviour change communication to reduce unprotected evening outdoor activities. Furthermore, the interplay between malaria risk and maternal vulnerability underscores the importance of integrated ANC services. Strengthening routine haemoglobin monitoring, ensuring adequate iron-folate supplementation, and promoting appropriate nutrition can help reduce susceptibility to infection and mitigate adverse outcomes. To support sustained programme effectiveness, stronger monitoring systems are also required. Incorporating trimester-specific AMiP indicators into the national health information system would enable routine tracking of screening coverage, IPTp-SP uptake, and infection patterns across pregnancy stages, thereby improving data-driven decision-making. Study’s strengths and limitations This study offers several methodological and analytical strengths. It is among the few to apply the Oaxaca-Blinder decomposition to disentangle the sources of trimester disparities in asymptomatic malaria among pregnant women. This approach allowed for both quantitative partitioning of structural inequalities, thereby reinforcing understanding of the determinants of AMiP. Secondly, the study’s trimester-specific analytical design captures within pregnancy variations often masked in aggregated analyses, offering valuable insights for timing-specific interventions. Likewise, the inclusion of multidimensional variables, including sociodemographic, clinical, programmatic, and environmental domains, enhanced explanatory depth and ecological validity. However, some limitations should be acknowledged. The cross-sectional design limits causal inference, as temporal relationships between exposure and infection could not be definitively established. A key limitation is the comparison of IPTp-SP uptake across gestational trimesters. Women in the second trimester may not have had enough time to receive the recommended three doses. As a result, differences between second- and third-trimester women may be due to unequal exposure time rather than true differences in uptake or access to care. In addition, the reliance on microscopy for malaria diagnosis may have missed low-density or subpatent infections. These infections are often detectable only by more sensitive molecular methods such as polymerase chain reaction (PCR). This limitation may have led to outcome misclassification and an underestimation of prevalence. Furthermore, some environmental exposure variables were self-reported and may be subject to recall bias or misclassification. Lastly, the facility-based sampling approach, restricted to pregnant women attending ANC, may limit generalisability, as it excludes women who do not access ANC services and who may have different risk profiles. Conclusion The study reveals that trimester disparities in asymptomatic malaria in pregnancy (AMiP) are largely driven by modifiable programmatic, behavioural, and environmental factors rather than biological differences. The most significant contributors to the observed differences include suboptimal uptake of intermittent preventive treatment with sulfadoxine-pyrimethamine (IPTp-SP), inadequate use of mosquito repellents, and frequent evening outdoor exposure. Environmental factors, such as rural residence, poor housing quality, and proximity to mosquito breeding sites, further exacerbate these disparities. Conversely, improved coverage of long-lasting insecticidal nets (LLINs) and consistent iron–folate supplementation have a protective effect, helping to narrow the disparities. Addressing these inequities requires a focused effort on enhancing antenatal care (ANC)-based malaria screening, improving IPTp-SP uptake, and promoting preventive education regarding LLIN use and repellent application. Furthermore, integrating environmental management, strengthening social protection programs, and ensuring better access to health services will help mitigate the broader structural factors contributing to AMiP disparities. Closing this gap requires shifting from uniform approaches to trimester-responsive strategies that prioritise early ANC engagement, optimise IPTp-SP delivery, and embed community-level vector control within routine maternal health services. Supplementary Information Supplementary Material 1. Acknowledgements The author sincerely acknowledges Dr Fred Adomako-Boateng, Bono East Regional Director of Health, for granting permission to conduct this study across the selected health facilities within the region. Special appreciation is extended to Dr Nkrumah and the staff of the Antenatal Care (ANC) Unit of the Kintampo South District Hospital for their cooperation, assistance, and dedicated support during data collection. The author is also deeply grateful to all the pregnant women who voluntarily participated in the study, whose time and willingness made this research possible. Their contributions are highly valued and greatly appreciated. Abbreviations Antenatal Care Confidence Interval Long-lasting Insecticidal Nets Haemoglobin Intermittent Preventive Treatment with Sulfadoxine-Pyrimethamine World Health Organization Authors' contributions DB conceived and designed the study. DB, GA, LAA, SN, DH, and RBB collected all relevant data. DB curated the dataset. DB, LAA, SN, GA, DH, and RBB validated the data. DB performed the formal analysis, produced visuals, and interpreted the data. DB, SN, and LAA contributed materials and wrote the first draft of the manuscript. DB, LAA, DH, SN, GA, and RBB reviewed the first draft of the manuscript. All authors reviewed and approved the final manuscript. Funding No funding was received from any external source. Data availability The data supporting the findings of this study are not openly available due to sensitivity reasons. However, they are available from the author upon reasonable request. Declarations Ethics approval and consent to participate The study and all associated protocols were reviewed and approved by the Committee on Human Research, Publication, and Ethics (CHRPE), Kwame Nkrumah University of Science and Technology, School of Medical Sciences, under the Declaration of Helsinki (assigned approval number: CHRPE/AP/1081/23). Participants signed or thumbprinted a written consent form after receiving a thorough explanation before participating in the study. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Contributor Information Dennis Bardoe, Email: izyanjr@gmail.com. George Ahiaka, Email: ahiakagale@gmail.com. References Associated Data Supplementary Materials Supplementary Material 1. Data Availability Statement The data supporting the findings of this study are not openly available due to sensitivity reasons. However, they are available from the author upon reasonable request.