Serosurveillance Identifies an Endemic Hotspot of Lassa Fever in Faranah, Upper Guinea

Abstract

Background

Lassa fever, caused by Lassa virus (LASV), is a rodent-borne disease endemic to West Africa. Despite growing interest from the global health community, the overall disease burden remains poorly understood due to persistent underreporting and inadequate diagnostic capacities. This study aimed to assess LASV infection dynamics in rural Guinea, addressing gaps in understanding its epidemiology in endemic settings.

Methods

A cross-sectional serosurveillance study was conducted across 6 villages in Faranah, Guinea, involving 1306 participants. Serum samples were analyzed for LASV-specific antibodies using enzyme-linked immunosorbent assay, and a subset confirmed via immunofluorescence and neutralization assays. Statistical and mathematical models were used to estimate key epidemiological parameters, including the force of infection, antibody waning rates, and infection prevalence.

Results

Overall, IgG seroprevalence was 82.3%, while IgM prevalence stood at 1.8%. Seroprevalence of IgG exceeded 50% by age 5 and increased steadily with age, indicating lifelong exposure. Antibody waning rates suggested an average IgG persistence of 58 years. Mathematical analysis showed that halving LASV prevalence would require a 10-fold reduction in the force of spillover, highlighting the challenges of effective control measures.

Conclusions

The seroprevalence observed in this study was markedly higher than reported elsewhere in West Africa, identifying Faranah as a hotspot for LASV transmission. Despite the high seroprevalence and estimated force of infection, the number of reported acute cases is very low, suggesting that LASV may frequently cause pauci- or asymptomatic infections or be commonly misdiagnosed for other diseases. These findings underscore that LASV is more widespread than assumed in rural Guinea.

This study identifies Faranah, Guinea, as a Lassa fever hotspot with unusually high IgG seroprevalence (>80%), suggesting widespread exposure and frequent asymptomatic infections. It also highlights challenges for surveillance, diagnosis, and control in endemic rural settings.

Lassa fever, caused by Lassa mammarenavirus (LASV), is a significant public health concern in West Africa [1]. Infections primarily occur through direct or indirect contact with rodents or their excreta [2, 3]. Among the rodent species identified as potential LASV reservoirs [4], Mastomys species are considered the main source of human infections due to their frequent presence in and around human dwellings [5–7]. Human-to-human transmission can also occur in various settings, but superspreading events in healthcare facilities can particularly amplify outbreaks [8]. However, genomic studies suggest that rodent-to-human transmission remains the predominant pathway for infections, especially in rural areas [9]. Clinically, approximately 80% of infected individuals present with mild symptoms such as low-grade fever and general malaise, while severe infections involving hemorrhage and multiorgan failure occur in about 20% of cases [10]. Although the overall case-fatality risk in endemic regions is relatively low (2%), it can rise up to 50% among hospitalized patients during outbreaks [10]. Consequently, the World Health Organization and Coalition for Epidemic Preparedness Innovations classified Lassa fever among the diseases with pandemic potential that should be prioritized for research and development to enhance resilience for public health emergencies [1].

The lack of standardized epidemiological studies in LASV-endemic countries complicates efforts to assess the global disease burden and develop effective control measures [11]. Early estimates suggested that 200 000–300 000 LASV infections occur annually, resulting in 5000–10 000 deaths [12]. However, these figures may significantly underestimate the true incidence due to persistent underreporting, particularly from remote rural areas. More recent models estimate that >900 000 people in West Africa are infected annually [13], with projections indicating a potential increase driven by climate change and land-use shifts [14]. Cross-sectional population studies within endemic regions have reported LASV seroprevalence rates ranging from 5% to 50%, reflecting the highly heterogeneous epidemiological landscape of the disease [15, 16]. Despite these high seroprevalence rates, active human infection is rarely detected in rural communities, partly due to the overlap of LASV disease symptoms with more commonly diagnosed diseases like malaria [17, 18]. For example, previous seroprevalence studies in Guinea, such as the one conducted by Lukashevich et al, reported seroprevalence rates as high as 40% [19]. Yet, only anecdotal reports of acute Lassa fever cases have emerged from the country, underscoring a significant gap in case detection and clinical diagnosis [17, 18, 20, 21]. This discrepancy highlights the urgent need to strengthen surveillance and diagnostic capacity to better understand and manage the disease burden in endemic regions.

To investigate the true incidence of LASV infection in Guinea and identify risk factors such as the abundance of the reservoir rodent, we conducted a cross-sectional serosurveillance study in 6 villages in Upper Guinea [22, 23]. Here, we report human seroprevalence data from the Faranah region of Upper Guinea, which notably differs from findings of previous seroprevalence studies conducted in other parts of West Africa. To explore the epidemiological implications of our findings, we used mathematical susceptible-infectious-recovered (SIR) models to estimate key transmission parameters, including the force of infection and the duration of waning immunity. Additionally, we estimated the true prevalence of human LASV infection by accounting for undiagnosed cases likely missed due to underreporting and evaluated the potential impact of spillover reduction (eg, via rodent control) on the human LASV prevalence in such communities.

MATERIALS AND METHODS

Study Sites and Participant Selection

From January to March 2014, we conducted a cross-sectional serosurveillance study in rural villages of the Faranah prefecture in Upper Guinea by recruiting human subjects from all age groups. We selected the Faranah region due to its known LASV endemicity [19] and our long-term research presence in the area since 2003 [5, 6, 24]. Six study villages (Brissa, Dalafilani, Damania, Sokourala, Sonkonia, and Yarawalia) were chosen based on specific criteria: remoteness from paved roads, populations not exceeding 1000 inhabitants, accessibility within a 45-minute drive from Faranah, and the highest proportion of LASV-positive rodents (Supplementary Figure 1, Supplementary Table 1, Supplementary Excel File 1).

We calculated a target sample size of 1200 individuals (approximately 200 per village) to ensure statistical power for a planned follow-up study [19, 24]. However, this follow-up study was not carried out due to the West African Ebola outbreak (2014–2016). To randomly select households (6 individuals on average) within each village, we spun a bottle at the village center to determine a starting direction and sampled all households along that line until reaching the village edge. This procedure was repeated in the opposite and perpendicular directions. After obtaining informed consent, all household members aged 5 years and older were recruited. Anticipating potential dropouts, we oversampled and ultimately recruited 1306 individuals, distributed as follows: 234 from Brissa (32 households), 211 from Dalafilani (34 households), 203 from Damania (32 households), 213 from Sokourala (19 households), 235 from Sonkonia (36 households), and 210 from Yarawalia (35 households). Blood samples were collected via venipuncture from all study participants and sociodemographic data were recorded. Blood samples were centrifuged, and serum was aliquoted on-site each sampling day. The serum was subsequently frozen at −20°C in Faranah and transported to Hamburg (Germany) for further analysis.

Serology

All human serum samples were tested for immunoglobulin G (IgG) (indicative of past infection) using an enzyme-linked immunosorbent assay (ELISA), and for immunoglobulin M (IgM) (indicative of recent infection) using an immunofluorescent assay (IFA). In the ELISA procedure, 25 µL of diluted serum (1:20 in 1% Triton X-100 phosphate-buffered saline [PBS]) and 25 µL of diluted biotinylated nucleoprotein (1:800 in Triton-1% bovine serum albumin–PBS) were incubated overnight at 4°C in CD32-coated plates. The following day, plates were washed 3 times with 0.5% Tween-PBS, and detection was performed by adding streptavidin-peroxidase solution for 1 hour (see detailed protocol in Gabriel et al [25]). The cut-off value was calculated as 1.5 times the sum absorbance of the 3 negative calibrators (German donors), with sera above this value considered IgG positive. This ELISA protocol was previously validated with a specificity of 95.4% and a sensitivity of 90.1% using negative control samples from Ghana and Germany and positive controls from Nigeria [25].

To further assess the predictive performance of the IgG nucleoprotein ELISA, we tested all samples from Yarawalia (n = 210) using IFA, as described by Wulff and Lange [26] and in more detail in the Supplementary Material. To determine the diagnostic accuracy of the ELISA relative to the IFA—considered the reference standard for LASV IgG detection—we generated a receiver operating characteristic (ROC) curve. In addition, we performed a neutralization assay using a replication-competent vesicular stomatitis virus (VSV) expressing LASV Glycoprotein (GP) from lineage IV strain Josiah (VSVΔG/LASVGP), following the protocol described by Müller et al [27]. ELISA results were converted to signal-to-noise ratios and correlated with log-transformed neutralization titers using Pearson correlation. Neutralization titers ≤8 were classified as negative and excluded from this correlation.

Statistical Analysis

To explore the relationships between serostatus and participant characteristics, we employed generalized linear mixed models (GLMMs). These models used IgG or IgM serostatus as response variables, with sex and age as explanatory variables and village as a random effect. Furthermore, to account for the nonrandom sampling of participants within villages, we included household as a nested random effect within villages to appropriately model the inherent clustering in our data. We assumed a binomial distribution with a logit-link function to account for the binary nature of the serostatus data. To contextualize our findings, we conducted a literature review focusing on age-specific IgG seroprevalence data. We extracted relevant data from the 6 most recent or frequently cited LASV serosurveillance studies conducted on healthy individuals using cross-sectional designs [12, 18, 19, 24, 28, 29]. Generalized additive models (GAMs) were used to model the seroprevalence, as 1 study showed a nonlinear (first increase followed by decrease) correlation with age [30]. These models included age as a smoothed fixed effect. This combined approach provided a broader epidemiological perspective and helped interpret our study's results within the context of existing LASV research. We conducted our analyses using R version 4.2.2 software (R Development Core Team, 2017). For all GAMs, the optimal level of smoothing for the age variables was determined through cross-validation using the built-in function in the R package MGCV.

Mathematical Modeling

We developed a mathematical modeling approach to estimate key epidemiological parameters of LASV transmission in 6 communities in Faranah, Guinea. Using age and IgG serostatus from human serum samples, we inferred the force of infection, antibody waning rates, and LASV prevalence. We assumed constant, community-specific exposure from rodent spillover, rapid recovery with antibody formation, and uniform antibody decay. An SIR model allowed us to estimate the likelihood of observed seroprevalence patterns using maximum likelihood estimation. We validated the model with simulated data and used it to explore how changes in spillover rates could affect active LASV infection prevalence. We refer to the Supplementary Material for a full explanation of the modeling approach and the R files.

Ethics

This project received ethics approval from the Ethics Committee of Charité–Universitätsmedizin Berlin (reference number EA2/177/11) and the Guinean National Ethics Committee for Health Research (reference number 12/CNERS/12) and was further endorsed by a formal letter of approval from the Minister of Health, which was presented to key health officials in the Faranah region, including the Director of Health for the region and the prefectural health service. During the prospection stage, the project was introduced to the inhabitants of each participating village through a detailed presentation explaining the study's objectives, stages, and components. Following this, the consent previously granted by village chiefs and the broader community was reaffirmed. Informed consent was primarily obtained in writing. When written consent was not feasible, oral consent was collected in the presence of a witness who provided written confirmation. Participation in the study was strictly voluntary. For minors, consent was required from at least 1 parent or guardian. Participants were offered sandwiches, juice, and medical consultations as a token of appreciation. Additionally, any village resident experiencing illness received a free medical consultation, regardless of their participation in the study, ensuring that the benefits extended to the entire community.

RESULTS

The serological analysis revealed an overall LASV-specific IgG prevalence of 82.3% (95% confidence interval [CI], 80.2%–84.3%; Supplementary Table 2), indicating that most participants had experienced LASV exposure at some point. Consistent with this high cumulative exposure, we found an IgM prevalence of 1.8% (95% CI, 1.2%–2.6%), suggesting that recent infections occurred frequently in these populations.

When we stratified results by age, we found that IgG prevalence exceeded 50% by 5 years of age and continued to rise through adolescence and adulthood (GLMM estimate slope = 0.036 ± standard error [SE] = 0.005, Aikaike Information Criterion [ΔAIC] = 56.9; P < .0001). By age 10–15 years, >70% of participants had already tested positive for IgG (Figure 1). Correspondingly, IgM prevalence remained around 2% in all age groups (estimate = −0.0037 ± SE = 0.013, ΔAIC = −1.9; P = .777), suggesting a consistent pattern of LASV exposure throughout life. We found no significant differences in LASV IgM (estimate = 0.218 ± SE = 0.049, ΔAIC = −1.81; P = .658) or IgG (estimate = 0.019 ± SE = 0.157, ΔAIC = −1.9; P = .9012) prevalence between men and women. Compared to other cross-sectional studies of healthy populations, our IgG prevalence was considerably higher across all ages (Figure 2). Although we initially aimed to correlate human seroprevalence (IgG and IgM) with rodent abundance and LASV prevalence in rodents per village (Supplementary Table 1), the uniformly high human seroprevalence across villages precluded meaningful variation for such analyses. Therefore, no further information on these aspects is provided.

Figure 1.

Age-specific immunoglobulin G (IgG) prevalence for Lassa virus is presented across 6 villages in Faranah province, as analyzed using generalized linear mixed models. The left panel shows the predicted IgG prevalence with colored envelope lines indicating adjusted standard error intervals, while the right panel displays immunoglobulin M (IgM) prevalence. For IgM, standard error envelopes were estimated using a fixed-effects model due to convergence issues with the random cluster effects.

Figure 2.

Age-specific Lassa virus immunoglobulin G (IgG) prevalence in humans from different cross-sectional studies found in the literature that also investigated apparently healthy populations in endemic settings. Solid lines represent seroprevalence estimates derived from generalized additive models, accompanied by colored envelopes that indicate standard errors. Only data from Faranah were used from Lukashevich et al [19].

When applying an age-structured SIR model and maximum likelihood estimation (Supplementary Material, equation 5), we obtained values for the force of infection (λ) in each village ranging from 0.10 to 0.23 new infections per susceptible individual per year (Table 1). These high rates align with the elevated IgG prevalence, indicating high LASV exposures. This model further estimated the duration of antibody persistence to be around 58.8 years (waning rate ρ = 0.017), suggesting that once infected, individuals maintain LASV-specific antibodies for most of their lives. Substituting estimated values of model parameters into equation 7 (Supplementary Material) predicts a steady-state prevalence of active LASV infection at 8–10 × 10⁻⁴, consistent with our IgM findings that also suggest active infection in all sampled villages. Using the same equation, we also simulated the effect of reducing spillover (ie, lowering λ) on the prevalence of active infections (I). The mathematical analyses show that λ must decrease substantially to reduce LASV prevalence meaningfully. For example, halving the LASV prevalence requires approximately a 10-fold reduction in λ (Figure 3).

Table 1.

Susceptible-Infectious-Recovered Model Estimates for the Force of Infection and the Predicted Steady-State Prevalence of Lassa Virus Infection for 6 Villages in the Faranah Region, Upper Guinea, Based on Age-Specific Immunoglobulin G Prevalence

Village	FOI (λ)	Prevalence (I)
Brissa	0.10	8.2 × 10−4
Dalafilani	0.13	8.8 × 10−4
Damania	0.22	9.5 × 10−4
Sokourala	0.22	9.6 × 10−4
Sonkonia	0.23	9.7 × 10−4
Yarawalia	0.11	8.3 × 10−4

Abbreviation: FOI, force of infection.

Figure 3.

Predicted human prevalence of current infection as a function of the force of infection (λ) attributable to spillover from infected rodent populations. Each colored dot represents 1 of the study villages in Faranah.

To assess the diagnostic reliability of the IgG ELISA, we additionally screened 210 serum samples from Yarawalia using both IFA and neutralization assays. When compared to IFA (Supplementary Figure 2), the ELISA showed 4 false positives (specificity = 98%) and 1 false negative (sensitivity = 99%) (area under the ROC curve = 0.996; Supplementary Table 3). In contrast, the seroprevalence detected by the neutralization assay (55.7% [95% CI, 48.7%–62.5%]) was significantly lower than that measured by ELISA (73.3% [95% CI, 66.8%–79.2%]) and IFA (71.9% [95% CI, 65.3%–77.8%]). We also observed a weak correlation between ELISA signal-to-noise values and neutralization titers (Pearson r = 0.31), suggesting that while ELISA effectively detects past LASV exposure, it may not fully capture the presence or magnitude of functional neutralizing antibodies (Figure 4).

Figure 4.

Comparison between the Lassa virus (LASV) nucleoprotein enzyme-linked immunosorbent assay (ELISA) (Gabriel et al [25]) and a vesicular stomatitis virus ΔG/LASV glycoprotein–based neutralization assay [27]. Left: Boxplots of immunoglobulin G (IgG) ELISA signal-to-noise ratios categorized by neutralization outcome (positive or negative). Middle: Pearson correlation between ELISA signal-to-noise values and the logarithm of neutralization titers. Right: Age-specific LASV seroprevalence based on neutralization titers (titers ≤8 considered negative), with fitted values and 95% confidence intervals from a binomial generalized additive model.

DISCUSSION

Our study revealed an exceptionally high LASV IgG seroprevalence (83% overall) compared to other serosurveys among healthy individuals in West Africa [11] (Figure 2). Notably, more than half of the children had detectable LASV IgG antibodies by the age of 5 years, and seroprevalence continued to increase with age, suggesting constant exposure to the virus throughout life (Figure 1). This finding aligns with the consistent IgM prevalence (1%–2%) across all age groups. These patterns highlight the pervasive presence of LASV in the daily life of these communities in Faranah, likely driven by frequent interactions with rodent reservoirs, their excretions, or contaminated food or materials in their households. Interestingly, Klempa et al [24] reported a similar age-related increase in LASV IgG prevalence in 2 villages near our study area, likely sharing comparable ecological and social conditions. However, Bausch et al [18], Kernéis et al [28], and Lukashevich et al [19] reported LASV IgG levels remaining constant across age classes, and in some cases, even decreasing with age (eg, Grant et al [29]). In addition, McCormick et al [12] described an initial increase in LASV IgG prevalence with age followed by a decline after age 30 years. These findings may suggest that in other communities, LASV infection predominantly occurs in younger individuals, possibly due to specific exposure activities like hunting, after which IgG titers decline over time as immunity wanes.

Differences in serological methodologies likely contribute to the variability in LASV seroprevalence reported across studies. In our investigation, we used a novel ELISA to detect LASV nucleoprotein-specific IgG antibodies, which likely offers enhanced accuracy compared to older assay formats. This ELISA was thoroughly validated by Gabriel et al [25] and has since been employed in multiple seroepidemiological studies [31–33]. Given the unexpectedly high IgG seroprevalence observed in our study population, we further validated our findings by testing a subset of samples from the village of Yarawalia using 2 independent methods: an IFA and a neutralization test [27]. We observed strong concordance between the ELISA and IFA results, both of which detect binding IgG antibodies. Although the neutralization assay yielded a lower seroprevalence estimate (55%) compared to the ELISA (73%), this discrepancy is consistent with previous reports showing that neutralizing antibodies are detected in only about half of Lassa fever survivors [34], and that nucleoprotein-binding antibodies do not necessarily confer neutralizing capacity [35]. To our knowledge, this is the first population-based study to report such high levels of LASV-neutralizing antibodies. Collectively, these findings strengthen the conclusion that the communities surrounding Faranah represent a true hotspot of LASV transmission, and that the observed high seroprevalence is unlikely to be an artifact of the serological methods employed.

The exceptionally high LASV seroprevalence observed in this study can be attributed to several interrelated factors. Selection bias at the village level may have skewed our findings, as our study specifically targeted endemic villages where a high LASV prevalence was previously observed in rodents (Supplementary Table 1). In contrast, the studies of McCormick et al [12], Grant et al [29], and Kernéis et al [28] used 2-stage cluster sampling designs, providing more representative community data, while Bausch et al [18] and Lukashevich et al [19] recruited participants through hospitals or public calls, potentially enriching their cohorts with symptomatic individuals. Furthermore, ecological differences between villages likely played a significant role. For instance, in the Faranah region, black rats (Rattus rattus)—an invasive species not known to transmit LASV but able to outcompete Mastomys—are scarce. This lack of competition may have allowed Mastomys, the primary LASV reservoir, to thrive in households, thereby increasing the risk of human exposure [6, 36]. Alternatively, other studies (eg, Grant et al [29] and Kernéis et al [28]) also examined periurban communities where competition between Rattus and Mastomys may have interfered with LASV transmission. Geographic coverage differences further complicate comparisons. For example, whereas our study and that of Klempa et al [24] focused on rural villages in Faranah, studies by Grant et al [29] and Kernéis et al [28] sampled more extensively across Sierra Leone and Guinea, potentially diluting site-specific patterns. Changes in human behavior, climate, and land use over the last decades may have also genuinely increased LASV transmission in the Faranah region [14], as indicated by the higher seroprevalence found in our 2014 sampling cohort compared to earlier studies by Lukashevich et al [19] and Klempa et al [24] performed in the same region. Another explanation could be a shift in bushmeat consumption from big (becoming rare) to smaller wildlife animals, including Mastomys [37]. These factors underscore the need for more standardized study protocols to improve our understanding of LASV transmission dynamics across West Africa.

A striking feature of our analyses is the contrast between the high LASV seroprevalence and the low number of reported acute human cases in the Faranah region (and generally in Guinea) [18]. One plausible explanation is that Lassa fever primarily acts as a pauci- or asymptomatic childhood disease in these communities, similar to other pathogens that exhibit age-related severity patterns, such as severe acute respiratory syndrome coronavirus 2 [38]. Some evidence indeed supports the idea that children may experience milder LASV disease, which could explain its apparent clinical invisibility in younger populations [39, 40]. Furthermore, early exposure to LASV during childhood may reduce the severity of Lassa fever disease upon reinfections later in life. Another explanation is that many febrile cases currently diagnosed as malaria or other febrile illnesses may represent undetected LASV infections within these communities [17, 41]. Correspondingly, our modeling effort estimated that, at any given time, 1 individual (prevalence is 8–10 per 10 000 people) is actively infected with LASV in a typical village (assuming approximately 1000 inhabitants). These findings align with recent model-based studies suggesting that LASV infections in West Africa may exceed 1 million annually [13, 42], clearly surpassing historical estimates [12]. In either case, both explanations highlight the urgent need for longitudinal studies to better understand the age-dependent severity, reinfection dynamics, and clinical presentation of LASV in these highly endemic communities [43].

Furthermore, our study estimates a substantially slower LASV IgG antibody waning compared to other studies [12], highlighting critical uncertainties in understanding the long-term dynamics of LASV immunity and their implications for epidemiology. Unfortunately, empirical evidence on LASV IgG seroreversion is limited and primarily focused on periods <2 years after acute infection [34]. Based on our field data and the maximum likelihood function (Supplementary Material, equation 5), we estimate a seroreversion rate corresponding to an average duration of 58 years (ρ = 1/58), indicating that most individuals retain detectable IgG levels for most of their lives. Consistent with these findings, Bond et al reported anecdotal evidence of long-lasting LASV IgG antibodies persisting for at least 40 years in 2 individuals [44]. In contrast, McCormick et al conducted a prospective cohort study that estimated an average seroreversion duration of 15.6 years [12]. This variability presents a significant challenge to accurately predicting LASV incidence. If IgG seroreversion occurs more frequently than estimated by our model, the incidence rates are likely to be underestimated. For instance, applying McCormick and colleagues' antibody waning estimates (Supplementary Material, equation 7) to the Faranah communities would result in a 2-fold increase in endemic LASV prevalence—from 1 infection per 1000 individuals in our study to 2 per 1000 based on McCormick and colleagues' model [12].

As a result, reinfections may be more frequent than currently anticipated, which could have implications for vaccination strategies. If primary infections in immunologically naive individuals are more severe than reinfections, early vaccination could be crucial in preventing severe disease. However, this assumption should be interpreted with caution, as disease severity can vary between primary and reinfections. Some viral infections, such as dengue, exhibit increased severity upon secondary exposure due to antibody-dependent enhancement, while others show no significant difference in severity between primary and secondary cases. To determine the optimal vaccination approach, it is essential to accurately distinguish between primary and potential reinfections—further highlighting the need for longitudinal studies.

Finally, we examined the effect of reducing spillover (eg, lowering λ by rodent control or behavioral changes) on the prevalence of active human infections (Figure 3). We observed that λ must decrease considerably to achieve a meaningful reduction in LASV prevalence. For instance, a reduction in prevalence by half necessitates an approximately 10-fold decrease in λ (Supplementary Material, equation 7). Achieving this level of reduction poses considerable logistical and ecological challenges, raising doubts about the feasibility of such interventions [22]. Moreover, the high force of infection observed in the region (λ = 0.1–0.23) may drive an epidemiological pattern in which infection predominantly occurs during early life, conferring long-term immunity. Under such conditions, Nuismer et al [45] demonstrated that spillover reduction may lead to unintended public health impacts. These outcomes may be driven by interventions (eg, rodent control or behavioral changes) that reduce LASV exposure during early life, potentially resulting in delayed primary infection during adulthood, when clinical manifestations may be more severe. In such settings, vaccinating children may be considered a safer and more practical alternative, as early induction of immunity could prevent the adverse consequences associated with a delayed age of LASV infection.

Supplementary Data

Supplementary materials are available at The Journal of Infectious Diseases online (http://jid.oxfordjournals.org/). Supplementary materials consist of data provided by the author that are published to benefit the reader. The posted materials are not copyedited. The contents of all supplementary data are the sole responsibility of the authors. Questions or messages regarding errors should be addressed to the author.

Notes

Acknowledgments. We would like to thank Inga Ziemer and Gédéon Bongo, who provided technical support with IgM testing on human samples and IgG testing on rodent samples. We are grateful to Fanta Berete and Mohamed Touré for their dedicated efforts in both rodent trapping and human sampling, and to Morlaye Sylla for his assistance in rodent trapping. Amara Camara played a key role in managing field operations for human and rodent sampling. Finally, we extend our gratitude to Dr Pema Grovogui for his medical support and assistance to rural communities during the first 2 years of the project. We also thank Petra Neubauer-Rädel for expert technical assistance related to the neutralization assay.

Author contributions. Conceptualization and writing–editing: J. M., S. L. N., and E. F.-C. Data collection: B. S., A. M. S., and E. F.-C. ELISA techniques and serology: S. G., B. B.-Z., T. S., H. M.-K., M. G., and E. F.-C. Human sampling design: M. B. Supervision of field work: N. M., E. F.-C., and A. M. S. Funding acquisition: E. F.-C., M. B., and N. M. All authors have either drafted the manuscript or reviewed it critically for important intellectual content, have approved the final version to be published, and agreed to be accountable for all aspects of the work.

Data availability. The study's de-identified dataset is available from the corresponding author upon reasonable request.

Declaration of generative AI. During the preparation of this work, the authors used ChatGPT in order to improve the grammar of the text. After using this tool, the authors reviewed and edited the content as needed and take full responsibility for the content of the publication.

Financial support. This work was supported by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) through the LAROCS (LAssa fever in Guinea and Sierra Leone: ROdent Control and Seasonality of human exposure to rodents) project 2013–2019 (reference numbers BO 3790/1&2-1 and FI 1781/1&2-1 to E. F.-C., M. B., and N. M.). This work was also supported by funding from the Horizon Europe program under grant agreement number 101191666 for the project “Identification of Novel Viral Entry Factors and Development of Antiviral Approaches” (DEFENDER) and the DFG (Projektnummer 197785619/SFB1021 to T. S.). J. M. is financially supported by the Department of Economy, Science, and Innovation of the Flemish government.