Prevalence and determinants of urinary schistosomiasis among school-aged children: a comparative risk analysis of demographic, environmental and water-use factors, in Anambra State

Abstract

Background

Urinary schistosomiasis is a neglected tropical disease that remains endemic in much of sub-Saharan Africa, with Nigeria bearing a considerable burden. Despite ongoing mass drug administration, infection continues to affect school-aged children in endemic communities. This study assessed the prevalence of urinary schistosomiasis among school-aged children in parts of Anambra State, Nigeria, and examined demographic, environmental, and water-use behaviours associated with infection.

Methods

A cross-sectional study was conducted between August and November 2025 in rural, semi-urban, and urban communities selected for freshwater use in domestic, agricultural, and recreational activities. Urine samples from 388 children aged 5–14 years were analysed using dipstick and filtration methods, and structured questionnaires were used to assess knowledge, attitudes, and practices.

Results

The overall prevalence was 13.1%, with higher rates in rural communities. The infections were mostly light to moderate, with a few heavy cases, and haematuria was uncommon. Knowledge of schistosomiasis was generally poor; however, participants reported relatively positive attitudes and acceptable hygienic practices. Multivariate analysis identified haematuria and mean attitude score as associated factor of infection, with higher attitude scores associated with lower odds of infection. However, this result should be interpreted with caution due to the low internal consistency of the attitude scale. The higher burden in rural communities highlights existing health inequities and the need for geographically targeted, equity-focused control strategies. This underscores the importance of multisectoral collaboration for sustainable control of schistosomiasis, in line with global neglected tropical disease elimination targets and the Sustainable Development Goals.

Conclusion

The findings highlight persistent transmission driven by behavioural and environmental factors and support the need for combined control strategies that integrate behavioural interventions, improved diagnostics, and environmental management.

Background

Urogenital schistosomiasis is a neglected tropical disease (NTD) endemic in Sub-Saharan Africa, with Nigeria accounting for a considerable burden [1, 2]. Previous research works have reported prevalence estimates ranging from 9.5% to 15% [3]. Schistosoma haematobium, Schistosoma mansoni, and Schistosoma japonicum are implicated in the infection. The World Health Organisation (WHO) categorizes schistosomiasis as a neglected tropical disease of global concern and a significant contributor to morbidity in endemic regions.

The disease remains one of the most persistent water-related parasitic infections globally, with over 100 million people requiring preventive treatment annually, the majority of whom reside in sub-Saharan Africa [4]. Epidemiological modelling studies have highlighted focal environmental conditions and human water-contact patterns as key determinants of transmission intensity.

Nigeria has recorded schistosomiasis cases across the country’s thirty-six states [3, 5], with over 100 million people at risk [1]. In 2015, the federal Ministry of Health reported a nationwide prevalence of 9.5% [6], although prevalence varies by region, with some areas reporting higher rates [7–10] than others [11–13]. This variation can be attributed to human factors such as age [14], gender [11, 15], as well as environmental factors, including temperature and altitude, which influence host survival, reproduction, level and frequency of rainfall, which provide temporary habitats for the host [16]. Studies across Nigeria have reported varying prevalence of urinary schistosomiasis, ranging from low to hyperendemic levels depending on geography and environmental conditions. A national review by Ezeh et al. (2019) [3] reported a high prevalence in Ogun State, while Adamawa State showed comparatively low levels, with several southern and northern states falling within moderate-to-hyperendemic zones. Similarly, a cross-sectional study [17] in a South-Western Nigerian community (Ogun state) reported a high prevalence of 44.1% among school-aged children, while the study conducted in Anambra State [11] reported a moderate prevalence of 14.5% and highlighted proximity to water-bodies as well as water-contact activities as a significant association to infection.

Additionally, knowledge gaps, misconceptions, and poor healthcare-seeking behaviour contribute to the spread of the infection [5, 18]. Schistosomiasis prevalence often exhibits strong spatial clustering, with communities near freshwater bodies such as rivers, dams, and irrigation canals consistently showing significantly higher infection rates than inland communities with improved water infrastructure.

Studies have reported an increasing prevalence of the parasite in communities with frequent water-contact behaviour, such as washing, bathing in streams, fishing, and recreational activities. Aquatic snails are well known as intermediate hosts of schistosomiasis and are essential for the parasite’s development, multiplication, and subsequent human infection, especially in children and young adults, who are believed to have more contact with water bodies [19]. Schistosomiasis is categorized into three main stages: cercarial dermatitis, which arises from the penetration of the skin by cercariae and causes skin inflammation; acute schistosomiasis, occurring between 3 and 8-weeks post-infection and presenting with weight loss, nausea, vomiting, diarrhoea, fever, and weakness; and chronic schistosomiasis, arising due to granulomatous inflammation in various tissues [5, 20].

Despite periodic mass drug administration of praziquantel, school-based deworming programmes, and health education campaigns aimed at reducing transmission, urinary schistosomiasis continues to affect school-aged children in many endemic communities. Although the prevalence of infection has been extensively reported, studies have examined the environmental and behavioural factors that place these children at risk, as well as how their daily actions and inactions contribute to its spread. There is a need to revisit this by comparing previous data with current research to better inform whether educational strategies, in particular, have been tremendously helpful in disease control. In addition, in Anambra State, where the present study was conducted, there is limited information on how environmental and behavioural factors influence transmission and, consequently, the prevalence of the disease. This study aimed to determine the prevalence of urinary schistosomiasis among school-aged children and to examine the demographic, environmental, and water-use behaviours associated with the infection, to provide evidence to support more targeted and effective control measures.

Methods

Study area and design

The study, involving multi-stage sampling for selecting study participants, purposive community selection, parasitological assessment, and socio-demographic data collection, was conducted between September and November 2025 in selected communities of Anambra State. With an estimated population of over 5 million people, Anambra State is located in southeastern Nigeria and bounded by Delta State to the west, Imo and Rivers States to the south, Enugu State to the east, and Kogi State to the north. The inhabitants are predominantly Igbo, with a population engaged in trade, commerce, small- and medium-scale industries, and agriculture (notably yams, cassava, and rice farming). The mixed urban and rural lifestyle provides varying levels of access to water, sanitation, and healthcare. These socio-demographic and environmental factors are critical in shaping the spatial epidemiology of diseases in the state. The study locations (Fig. 1) were purposively selected based on water bodies, such as lakes, and slow-moving streams, used for domestic, agricultural, and recreational purposes to ensure inclusion of high-risk populations relevant to the study objective. Communities included Amansea and Aguleri, Awka and Onitsha, and Nimo and Agulu, ensuring diverse ecological and socio-demographic settings, particularly regarding water-use practices.

Fig. 1.

Map showing the study area and selected communities in Anambra State, Nigeria

Amansea (Awka North LGA) and Aguleri (Anambra East LGA) are rural communities situated along the Ezu and Anambra Rivers, respectively, while the urban sites were Awka and Onitsha. Awka (Awka South LGA), though urbanised, contains small streams, including one that runs through Nnamdi Azikiwe University and flows into Amansea, thereby facilitating moderate human contact with surface water. Onitsha, a densely populated metropolitan city along the river Niger, is characterised by frequent human–water interactions. Semi-urban communities comprised Nimo (Njikoka LGA) and Agulu (Anaocha LGA), where water bodies such as Agulu Lake and the Mkpukpa and Ogbujilekwe Rivers are used for agricultural and domestic activities.

Sample size, sampling technique

A multi-stage sampling technique was employed in the study. The study area was first stratified into three settlement types: rural, urban, and semi-urban. From each, two communities were purposively selected to account for ecological variation in water use. Lastly, children aged five to fourteen in the selected communities were randomly selected to participate in the study. Preference was given to communities closest to the water bodies. A proportional allocation method was used to determine the number of children selected from each community, ensuring equal representation across the sampled communities. Random selection of participants within each settlement was done to minimise selection bias.

The sample size was calculated using the single-population proportion formula.

where: n = required sample size

z = 1.96 (standard normal deviate at 95% confidence)

P = expected prevalence of urinary schistosomiasis

d = desired precision (0.05).

The 14.5% prevalence reported by Aribodor et al. [5] was used, yielding a sample size of 190 participants. Because participants were clustered within communities, a design effect (DEFF) of 2 was applied to account for within-cluster similarity, by inflating the sample size, preventing underestimation of the sample size. The adjusted sample size was calculated as:

with n_SRS = 190 and DEFF = 2 (using a design of 2, which is common in developing countries), thus giving a value of 380 after computation. To further account for possible non-response, a 10% non-response rate was added, bringing the final sample size to 418 participants.

Study population

The study population comprised school-aged children resident in selected communities. These children represent the high-risk population for urinary schistosomiasis and were thus selected to assess the burden of infection and associated factors. A list of eligible children willing to participate in the study was obtained from community schools and household listings. From the above list, participants were selected using simple random sampling and assigned a unique identification code.

Inclusion and exclusion criteria

Participants were included if they were aged 5–14 years, had lived in the selected community for at least six months, had parental consent, and had completed assent forms. Female children who were menstruating at the time of sample collection were excluded to reduce the risk of false-positive haematuria results.

Data collection

The questionnaire was developed based on previously published studies [21–23] on schistosomiasis epidemiology, with an emphasis on validated instruments assessing knowledge, attitudes, and practices. Items were adapted to suit the local context of the study area. The information obtained was used to assess exposure patterns and infection status. Sample collection was carried out between September to November 2025.

Socio-demographic and risk factor assessment

A structured questionnaire was designed and administered to study participants during sample collection. The questionnaire was designed to focus on four main sections: demographic characteristics, knowledge, attitudes, and practices and behaviours related to Schistosoma haematobium infection. It was designed to assess participants’ awareness of the disease, their perception of risk factors, and their behavioural practices that may increase the risk of infection with the parasite.

For the study, a data collection tool was designed using closed-ended, easy-to-understand statements that are age-appropriate for the study group. Additionally, to reduce information bias, trained interviewers administered the questionnaires using age-appropriate language, ensuring consistency in data collection across study sites.

Validity and reliability of questionnaires

According to Heale and Twycross [4], questionnaire validation is the process of ensuring the instrument measures what it is intended to measure. The questionnaire underwent initial validation to ensure adequate representation of the constructs of interest. Subsequently, it underwent expert review by two domain experts, and their recommendations were incorporated into the revised instrument. A pilot study was then conducted to pre-test the questionnaire for clarity, consistency, and contextual relevance. Data obtained from the pilot phase were further analyzed to determine the internal consistency of the instrument using Cronbach’s alpha and the Kuder-Richardson Formula 20 (KR-20).

where:

• k = number of items

• pi = proportion of participants who got the item / correct (or “yes”)

• qi = 1 – pi

• σ² total = variance of the total test score (sum of all items).  

This assesses the internal consistency of dichotomous survey items. Thus, it was used to complement Cronbach’s alpha in assessing participants’ knowledge. Additionally, Cronbach’s alpha was used to assess the internal consistency of the knowledge, attitude, and practice survey items. All analyses were conducted in R version 4.5.1 using the Psych package.

Urine sample collection

Participants were each given a well-labelled sterile sample collection tube for fresh urine. Sample collection took place between 10:00 am and 2:00 pm, the peak period for schistosome egg extraction to enhance diagnostic yield. All samples were under strict monitoring to ensure proper collection, preserved on ice packs, and transported to the laboratory for microscopic examination. Urinary samples were analysed using filtration and the urine test strip (dipstick) method.

Dipstick and visual analysis of urine samples

Dipstick analysis was used to detect haematuria, a documented marker of S. haematobium infection. For this purpose, each sample was tested by immersing the strip in urine for approximately 10 s, and the result was compared with the manufacturer’s reference standard. In addition, all samples were visually assessed for clarity and the presence or absence of blood.

Parasitological examination

Parasitological assessment was performed using filtration to detect and quantify schistosome eggs in each sample. The samples were first homogenised, and 10 ml of each was extracted with a sterile syringe and filtered through a membrane filter paper to collect any schistosome eggs, if present. The membrane filter was gently removed after filtration, placed on a microscope slide, and mounted for observation at 10x and 40x magnification to identify trapped eggs by their distinctive terminal spine. For quality control, the samples were examined in duplicate by independent microscopists blinded to participants’ questionnaire responses to reduce observer bias during parasitological assessment. When the two microscopists disagreed, a third experienced microscopist independently reviewed the slides and resolved discrepancies, providing the final reading.

Data analysis

The following R packages were utilised during data management, statistical analysis, machine learning, psychometric evaluation, spatial analysis, and graphical presentation of results: R packages including car, psych, dplyr, vip, pROC, caret, randomForest, likert, stringr, ggplot2, tidyr, sf, naijR, readxl, RColorBrewer, viridis, and patchwork. These packages supported data import, cleaning, w and manipulation; descriptive and inferential statistical analyses; psychometric assessment; predictive modelling; receiver operating characteristic (ROC) analysis; spatial data processing; and the generation of publication-quality figures and visualisations.

Both logistic regression and Random Forest models were employed to provide complementary explanatory and predictive insights into infection risk. Logistic regression was used to estimate the direction, magnitude, and statistical significance of associations between predictor variables and infection status through odds ratios and confidence intervals. In contrast, the Random Forest model was applied to identify important predictor variables, capture potential non-linear relationships and interactions among variables, and assess predictive performance. The combined use of both approaches enhanced the robustness of the analysis by enabling comparison of findings across traditional statistical and machine learning frameworks.

Infection intensity was graded as light (1–24 eggs/10 ml urine), moderate (25–49 eggs/10 ml), and heavy (> 50 eggs/10 ml) [24, 25]. Although the WHO classifies infections as light or heavy, a three-tier classification was used for finer-grained stratification. Descriptive analysis of demographic variables was performed using BaseR and dplyr. Survey data were cleaned, geocoded by Local Government Area, and visualised using dplyr, sf, naijR, ggplot2, RColorBrewer, and BaseR packages.

Knowledge was assessed as the proportion of correct responses, while attitude and practice were measured on three-point Likert scales and analysed using BaseR and the Likert package. KAP responses were coded such that safe practices had higher scores, and correlations with age were assessed using BaseR. Prevalence was calculated from laboratory-confirmed cases. Chi-square or Fisher’s exact tests were used to assess associations among categorical variables, while the Mann–Whitney U test was applied for continuous variables, respectively. Variables with p < 0.05 were included in multivariate analysis. Multivariate analysis used a random forest to assess variable importance, followed by logistic regression with stepwise selection. ANOVA assessed effect size; ROC curves evaluated AUC; ggeffects visualised predictions; and the performance and caret packages assessed model fit and accuracy.

Results

Descriptives

A total of 388 participants were enrolled in the study (Table 1). The sociodemographic characteristics of the respondents were analysed to describe the study population and provide context for interpreting the distribution of infection and associated risk factors. Key variables, including age, sex, parental occupation, and settlement type, were assessed to capture differences in participant background across rural, semi-urban, and urban settings. Sampling was balanced across the three study areas; urban, rural and suburban However, unfortunately for the urban communities, a relatively high non-response rate was observed, due to incomplete or inappropriate questionnaire responses and the failure of some participants to return the urine samples for parasitological analysis, leaving the total number of participants as urban (n = 98), suburban (n = 137), and rural (n = 153) areas. This pattern may be attributed to several contextual factors. Urban participants often demonstrate lower levels of sustained engagement in community-based health surveys due to time constraints and a reduced sense of community participation compared to rural settings. In addition, there may also be increased reluctance among urban participants to provide biological samples, especially in school and household-based surveys. Collectively, these factors likely contributed to reduced data completeness in the urban clusters. A mean age of 9.3 years was recorded, with the histogram (Fig. 2) indicating a fairly normal distribution.

Table 1.

Socio-demographic characteristics of participants (N = 388)

Variable	Category	Frequency (n)	Percentage (%)
Sex	Female	139	35.8
	Male	151	38.9
	Missing	98	25.3
Location	Rural	153	39.4
	Semi-Urban	137	35.3
	Urban	98	25.3
Parents’ Occupation	Artisan	46	11.9
	Civil servant	137	35.3
	Farmer	91	23.5
	Trader	113	29.2
	Mixed/Other	2	0.52
Haematuria Status	Cloudy	8	2.06
	Gross	8	2.06
	Microscopic	34	8.76
	No	338	87.1
Age (years)	Mean age	-	9.27

Fig. 2.

Histogram of the age distribution of participants

Reliability analysis

Reliability analysis showed very low internal consistency for the knowledge scale (Cronbach’s alpha = 0.065; KR-20 = 0.078), indicating that the items may not represent a single underlying construct. The attitude scale also demonstrated low reliability (Cronbach’s alpha = 0.381); however, removal of one item (“urinating in rivers and streams is wrong”) increased the coefficient to 0.47, suggesting only marginal improvement. In contrast, the practice scale exhibited excellent internal consistency (Cronbach’s alpha = 0.94).

These findings indicate that the knowledge and attitude components should be interpreted with caution, while the practice scale is likely to provide more reliable estimates. The low reliability of the knowledge scale may reflect the heterogeneous and multidimensional nature of the items, which were designed to capture distinct aspects of knowledge. As the instrument was intended primarily for descriptive and item-level analyses rather than as a unidimensional scale, the knowledge items were retained. Thus, the reliability results informed a cautious interpretation rather than the exclusion of the instrument.

KAP analysis

More than 90% of participants were unaware of schistosomiasis and of how it spreads, reflecting persistent gaps in health literacy, this may be attributed to limited integration of neglected tropical diseases such as schistosomiasis into routine school health education programmes. Knowledge of schistosomiasis was generally low among participants (Fig. 3), indicating limited knowledge of the disease’s symptoms and awareness that it spreads through urine; only just over half were aware of this. However, a significant proportion correctly recognised the importance of proper hygiene, such as wearing footwear and avoiding urination in rivers, indicating good hygienic practices among participants. On the other hand, participants generally demonstrated positive attitudes toward preventive behaviours (Fig. 4). This suggests an underlying positive trend among participants; however, the mean attitude Likert scale score was 2.3 on a 3-point scale, indicating that the sampled population was, on average, “not sure” about matters regarding the spread of the disease.

Fig. 3.

Knowledge of participants towards urogenital schistosomiasis

Fig. 4.

Participants’ attitude analysis

In the practice section, participants exhibited a positive disposition toward safe practices, as indicated by the Likert plots in Fig. 5. Participants reported an average score of 2.58 on a 3-point scale, indicating a strong tendency to agree with safer practices. Given that a large proportion of participants had not heard of the disease or were not knowledgeable about its symptoms, our data suggest the existence of good hygienic practices that can be built on in subsequent education campaigns. Although Rural participants had significantly higher knowledge, they exhibited the lowest attitude and practice scores. Taken together with the Random Forest findings, this suggests a gap between reported practices and their effective implementation, likely reflecting behavioural or structural barriers that limit impact. Stratified analysis (Fig. 6) showed that, despite higher levels of knowledge in rural communities, infection risk remained highest, indicating that knowledge alone is insufficient for protection and that contextual and behavioural factors drive transmission. This observed gap between knowledge and protective practices is consistent with the knowledge-behaviour gap framework, in which awareness does not automatically translate into behavioural change due to structural constraints such as limited access to safe water sources, habitual water-contact practices, environmental exposures, and prevailing social norms within endemic communities. Furthermore, the stronger influence of practice-related variables in rural areas highlights the need for targeted, context-specific interventions that improve not just awareness, but the effectiveness and consistency of preventive behaviours.

Fig. 5.

Participants’ practice analysis

Fig. 6.

Mean attitude score by stratum

Prevalence of schistosomiasis

Among the 388 participants in our study, 51 tested positive for the parasite, yielding a prevalence of 13.1%. Disaggregating the prevalence by study site, rural communities had the highest prevalence at 19.6%, followed by the semi-urban communities at 10.2%, and the lowest being the urban communities at 7.1%. Of these, 7.7% (n = 30) were located in rural areas, whereas only 1.8% (n = 7) of all Schistosome-positive cases were in urban settlements. Additionally, only 1.8% (n = 7) of all participants who tested positive for schistosome had haematuria. Finally, egg counts ranged from 1 to 50 eggs per 10 ml of urine, and Fig. 7 shows the participants’ haematuria status.

Fig. 7.

Prevalence of haematuria

An analysis of infection intensity based on egg count revealed that the majority of participants across all settlement types had no detectable infection. Overall, 86.6% of individuals had no infection, while 13.0% had light infections and only 0.5% had moderate infections; no heavy infections were recorded. When stratified by settlement type, rural communities had a higher proportion of infected individuals, with 19.4% presenting light infections and 1.3% moderate infections. In contrast, semi-urban and urban communities had predominantly uninfected participants, with only 9.7% and 7.6% having light infections, respectively; no moderate or heavy infections were observed. These findings indicate that infection intensity was generally low, with light infections predominating, though the burden was relatively higher in rural areas.

A univariate correlation analysis of the average knowledge, attitude, and practice variables, as well as age, showed weak correlations, except for a moderate correlation (r = 0.6) between average attitude and practice. Testing the categorical variables (sex, parents’ occupation, haematuria, clarity of urine, among others) using Chi-square analysis, “eating snail meat”, haematuria, and clarity of urine showed significant associations with the infection status of the participants. On the other hand, the Wilcoxon test identified the mean attitude score as the only numerical variable significantly associated with participants’ infection status. Although the result shows this association, it may overstate the strength of the attitude as a predictor of infection due to low Cronbach’s alpha values. Based on the univariate analysis, variables with significant associations with the infection were included in the multivariate analysis. The variables were subjected to a random forest analysis of predictor importance, as shown in Fig. 8. This revealed haematuria as having the most importance, while sex had the least importance as a predictor. Based on this, a logistic binomial regression model was designed using a forward stepwise approach.

Fig. 8.

Variable importance for the random forest model

The regression analysis identified haematuria as a predictor of infection, while the mean attitude score was an associated factor of our model. The model showed that a unit increase in mean attitude score is associated with a 2.22-fold reduction in the odds of infection. Participants with microscopic haematuria were 2.5 times as likely to be infected as those classified as cloudy haematuria. The logistic regression model performed well, with an area under the curve (AUC) of 0.842 (95% CI: 0.773–0.911) and a Tjur’s coefficient of 0.438, indicating robust discrimination between infected and non-infected individuals. During development, haematuria was found to confound the association with urine clarity and semi-urban location, altering effect estimates by 50–65% and reducing statistical significance; hence, these predictors were not used. A performance check using the performance package showed some residuals outside the estimated probabilities, but these had no influence on the model. Positive predictive checks, collinearity [6], and quantile residuals (Fig. 9) all indicated acceptable model fit.

Fig. 9.

Fitness of model using the positive predictive test, Collinearity, and Quantile residuals

The classification model demonstrated high overall performance, with an accuracy of 92.76% (95% CI: 89.71–95.14%), significantly above the No Information Rate of 86.82% (p = 0.0001436). The Kappa statistic of 0.6207 indicates substantial agreement between predicted and actual classifications. McNemar’s test (p = 0.0003298) suggested some asymmetry in misclassifications. Sensitivity was 52.94%, and specificity was 98.81%, reflecting the low prevalence of positives (13.18%) and the model’s conservative positive-prediction bias. The PPV was 87.10%, the NPV 93.26%, and the balanced accuracy 75.88%. Predicted probabilities showed 100% infection for gross haematuria, 71% for microscopic haematuria, and 2–38% across attitude scores, increasing to 4–54% after adjusting for haematuria, highlighting its dominant influence on infection risk.

Rural, semi-urban, urban likelihood ratio test using generlized linear model, respectively

A forward multiple regression analysis was conducted by location type (rural, sub-urban, and urban) to better capture variability than by individual towns, which showed minimal and non-significant effects. Haematuria was a significant predictor across all strata, with the strongest effect in rural areas (LR = 85.7, p < 0.001) and the weakest in urban areas (LR = 6.58, p = 0.04). The mean attitude score was only significant in the suburban stratum (Fig. 10). Although likelihood ratio tests indicated overall model relevance within each stratum, no individual predictors were statistically significant in the regression model, suggesting possible non-linear relationships or interactions, which were further explored using Random Forest modelling.

Fig. 10.

Mean knowledge score (%) across rural, semi-urban, and urban strata

Random Forest classification confirmed (Fig. 11) that haematuria is the strongest predictor of transmission across all strata. Practice-related variables were more influential in rural areas, consistent with the significantly higher prevalence observed there (19.6%, n = 153). In line with this, a logistic regression model from our data indicates that those in Sub-Urban and Urban areas had 53% and 68%, respectively, low odd of having the infection, which was statistically significant in our model.

Fig. 11.

Variable importance of predictors of schistosomiasis across stratum

Discussion

This study provides evidence on the prevalence, behavioural correlates, and public health significance of urinary schistosomiasis among school-aged children in Anambra State, Nigeria. Integrating parasitological findings with behavioural data (knowledge, attitudes, and practices) offers insight into transmission dynamics across rural, semi-urban, and urban settings.

School-aged children aged 5–14 constituted the study population, with the majority of participants in the middle childhood age groups. This age range has consistently been identified as the most vulnerable to urinary schistosomiasis due to increased engagement in recreational and domestic water-contact activities such as swimming, bathing, washing, and fetching water from natural freshwater sources. The observed age distribution in the study is therefore consistent with previous reports from endemic regions in Nigeria and sub-Saharan Africa [3, 24], where infection burden is typically concentrated among children of primary school age. Similarly, this has been reported in recent studies from Cameroon and Tanzania, where school-aged children were identified as the most vulnerable population because of frequent domestic and recreational contact with contaminated freshwater sources [14, 21]. This is likely attributable to frequent contact with water bodies through activities such as swimming, washing, and water collection, which increase exposure to cercariae. Although sex has been identified as a predictor of infection in some studies, this was not evident in the present study, supporting the view that gender-related differences in infection are more reflective of culturally defined exposure patterns than intrinsic biological susceptibility [19].

An overall prevalence of 13.1% was recorded, consistent with reports from similar endemic settings in Nigeria, where prevalence ranges from 9.5% to 15.0% [1, 10]. This finding is also comparable to the report by Aribodor et al. [17] in Anambra State, Nigeria, which documented persistent transmission among adolescents despite ongoing control programmes. Similarly, Onyekwere et al. [6] reported continued endemicity among Nigerian schoolchildren in communities where water-contact activities are frequent. However, this estimate is lower than the findings reported by Okoye et al. [26] in Anambra State and other rural communities, where prevalence ranges from 30% to 50% [27, 28]. A higher prevalence was observed in rural communities (7.7%) than in urban communities (1.8%), reflecting greater reliance on natural freshwater sources for domestic, occupational, and recreational activities [25]. A similar rural–urban pattern has been reported in previous studies [5]. The lower prevalence observed in urban areas further supports the role of improved access to water and sanitation in reducing transmission. Improved Water, Sanitation and Hygiene (WASH) infrastructure has been associated with reduced exposure to cercaria-infested water bodies and lower schistosomiasis transmission risk in South-Western Nigeria and other West African endemic settings [28]. Stratified analysis of prevalence across communities showed variation; this gradient highlights the strong influence of environmental exposures and behavioural differences. In addition to prevalence, the analysis of infection intensity showed that most infections were light, with very few moderate cases and no heavy infections recorded. This pattern suggests low-intensity transmission across the study area, although the relatively higher proportion of infections in rural communities indicates sustained exposure and ongoing transmission in these settings. The predominance of light infections, which are often asymptomatic, remains epidemiologically important, as they may contribute to ongoing parasite transmission within the population.

Haematuria remains a useful clinical indicator in the diagnosis of Schistosoma haematobium infection; however, its sensitivity declines in low-intensity infections [29]. The egg counts observed in this study (1–50 eggs/10 ml urine) suggest predominantly light infections, which may be asymptomatic yet epidemiologically important in sustaining transmission. The low prevalence of haematuria among infected children highlights the limitations of relying solely on haematuria for diagnosis, particularly in low-transmission settings, and underscores the need for more sensitive diagnostic approaches.

Behavioural analysis showed weak associations between knowledge, attitude, and practice, with a comparatively stronger association between attitude and practice [27, 29]. However, the reliability of the attitude scale was low (Cronbach’s alpha = 0.381), which limits confidence in its consistency as a measure of a single construct. As such, associations involving attitude should be interpreted with caution. The association between snail meat consumption and infection status, although statistically significant, does not indicate a direct transmission route and may instead reflect broader environmental or cultural exposure patterns.

The analysis identified attitude score as significantly associated with infection status; however, given the cross-sectional design, this finding should be interpreted strictly as an association rather than a causal relationship. Similar observations have been reported in Kano State, Nigeria, where knowledge of schistosomiasis did not consistently translate into protective practices because of persistent environmental exposure and reliance on unsafe water sources [28]. Although the estimated odds ratio suggests a relationship, temporality cannot be established, and the scale’s low internal consistency further limits interpretability.

The use of multivariate modelling, including logistic regression and random forests, strengthens the study’s analytical framework. The exclusion of urine clarity due to confounding with haematuria reflects appropriate epidemiological judgement. The model demonstrated good discriminative ability (AUC = 0.842), exceeding commonly accepted thresholds for epidemiological risk models, which ranged from 0.70 to 0.80.

The findings reinforce the importance of behavioural and environmental factors in schistosomiasis transmission. However, the observed disconnect between knowledge and practice suggests that knowledge alone may not translate into protective behaviours. Public health interventions should therefore extend beyond information dissemination to address behavioural and contextual drivers of exposure. Improvements in water, sanitation, and hygiene (WASH) infrastructure, alongside targeted behavioural interventions, remain critical, particularly in rural communities.

From an epidemiological perspective, the findings support moderate endemicity (13.1%) of urinary schistosomiasis in Anambra State, with transmission concentrated in rural and freshwater-dependent communities. The persistence of infection in peri-urban settings suggests the presence of localized transmission sites, such as seasonal streams, school environments, and informal water collection points. In addition, the age-related pattern observed suggests that shared recreational water contact may contribute to transmission dynamics among school-aged children. The relatively low sensitivity of haematuria further indicates that subclinical infections may sustain transmission and remain undetected during routine screening programmes.

Overall, while this study provides important insights into the epidemiology of urinary schistosomiasis in the study area, findings related to behavioural constructs should be interpreted with caution due to limitations in measurement reliability and study design. Future research should incorporate validated behavioural instruments and longitudinal approaches to better elucidate causal pathways.

Limitations of the study

The integration of clinical, behavioural, and attitudinal data, together with KAP surveys, urinalysis, and multivariate models, provided robust insights into how individual and environmental factors determine infection rates. However, the study had some limitations: the cross-sectional design may have limited the ability to establish a cause-and-effect relationship, and reliance on participants’ own accounts of behaviour could introduce response bias. Low haematuria and light infection intensities may have reduced the sensitivity of clinical indicators. The absence of age stratification in administering the KAP questionnaire may have introduced some age-related response variation, although interviewer assistance was used to minimise misunderstanding among younger participants. Measurement bias may have occurred due to the limited reliability of the knowledge assessment scale, which may have affected the precision of participants’ responses. Additionally, excluding menstruating girls from the study may have slightly affected the representativeness of that study population. Nevertheless, the study offers insights that can guide targeted interventions and the development of health education programmes in communities.

Conclusion

Urinary schistosomiasis in the study area is influenced by a combination of clinical and environmental factors, whereas behavioural variables should be interpreted with caution. Although low disease awareness and moderate attitude/practice scores were observed, these findings are limited by poor internal consistency of the knowledge (α = 0.065) and attitude (α = 0.381) scales. Consequently, knowledge and attitude were treated as exploratory indicators rather than robust inferential determinants. In contrast, urinalysis findings, particularly haematuria, provided more reliable epidemiological evidence of infection and clearer associations with disease occurrence.

Overall, the results point to persistent behavioural and environmental exposure risks, especially contact with potentially contaminated freshwater sources. The study therefore supports integrated control strategies that combine Water, Sanitation and Hygiene (WASH) improvements with targeted behavioural interventions to reduce rural freshwater exposure (such as limiting swimming, washing, and domestic water contact in unsafe streams). Strengthening community health education alongside improved access to safe water sources is essential for reducing transmission risk. However, conclusions regarding knowledge and attitude remain limited by measurement reliability and should be interpreted with caution.

Authors’ contributions

CEK: conceptualisation, study design and funding acquisition; ECI: data collection and analysis; CBI: study design and review of manuscript; UCO: statistical analysis and drafting of manuscript; ECA: Data collection and analysis, drafting of manuscript; WSN: study design, review of manuscript; UMO: Data collection and analysis, funding acquisition. All authors reviewed and approved the final draft of the manuscript.

Funding

This study was funded by the Tertiary Education Trust Fund (TETFUND) through the Institutional-Based Research grant.

Data availability

The datasets generated and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

Ethical approval was obtained from the Ethics Committee of Chukwuemeka Odumegwu Ojukwu Teaching Hospital, Awka, under reference number COOUTH/HREC/ETH.C/VOL.1/FN:05/046. Written or oral informed consent from parents/guardians was obtained following detailed information on the aim and objectives of the research. Prior to participation, all participants were provided with information on the research focus and willingness to participate in the study was confirmed by obtaining an assent at all stages of the study protocol. The study was conducted in accordance with the ethical standards of the institutional research committee and the principles of the Declaration of Helsinki. All participants whose laboratory investigations turned out positive were treated under the supervision of trained medical personnel. All authors have reviewed the manuscript and given approval for publication.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.