Multi-pathogen serosurveillance in Tonga examining demographic correlates and geographic clustering of 17 pathogens Authors: Harriet LS Lawford, Ofa Tukia, Joseph Takai, Sarah Sheridan, Selina Ward, Holly Jian, Beatris Mario Martin, Reynold ‘Ofanoa, Stefanie Vaccher, Fiona Angrisano, Leanne J Robinson, Colleen L Lau Published: 2026 ABSTRACT AbstractIntegrated serosurveillance enables concurrent assessment of population immunity to vaccine-preventable diseases (VPDs) and infectious pathogens. We aimed to estimate seroprevalence of antibodies (Abs) to 17 pathogens to provide insights to inform public health strategies to control and eliminate infectious diseases in Tonga. A targeted cross-sectional survey was conducted in May-July 2024 among participants aged ≥ 5 years from Tongatapu, Ha’apai, and Ongo Niuas. Dried blood spots were collected and assayed on a multiplex bead assay (MBA) containing 26 antigens from 17 pathogens. Seropositivity was defined using antigen-specific cut-offs. Associations with seropositivity were assessed, and geographic clustering was investigated. 1,786 of 1,787 (98.9%) participants (median age 19 years; 46% male) had valid MBA results. In our target population, LF Ab seropositivity increased with age, with high seroprevalence in Ha’apai (Wb123 Ab: 13.1%; Bm14 Ab: 8.8%). Dengue seroprevalence was > 90% among > 50-year-olds and were highest in Tongatapu. Seroprevalence for VPD was high; however, measles seroprevalence (86.1%) was below theoretical elimination thresholds despite historically high vaccination coverage. Geographic clustering was significant (p < 0.01) for most pathogens, with strongest clustering seen at village-level. Clear demographic and geographic correlates of seroprevalence in this target population highlight focal transmission and potential immunity gaps. MBA platforms offer a cost-effective tool to support targeted elimination strategies and vaccination campaigns in the Pacific.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1038/s41598-026-55309-y.Keywords: Integrated serosurveillance, Multiplex bead assays, Infectious diseases, Elimination, TongaSubject terms: Diseases, Immunology, Medical research, Microbiology Introduction Serological surveillance (or serosurveillance) provides critical insights into population-level immunity to vaccine-preventable diseases (VPDs) and other infectious pathogens by measuring the prevalence of antibody (Ab) biomarkers to pathogen-specific antigens (Ag). Ab seropositivity reflects past infection, vaccination, or ongoing/chronic infection. Integrated serosurveillance (ISS), as defined by the World Health Organization (WHO), is the implementation of population-based surveys to collect and analyse samples for simultaneous estimation of prevalence of Ab against multiple pathogens.1 Frequent ISS offers an efficient means to monitor multiple pathogens simultaneously. National and regional elimination programs can be supported by using ISS to identify waning immunity and immunity gaps for VPDs, as well as the prevalence of infectious diseases nearing elimination, or that are assumed to be eliminated.2 ISS can therefore help provide early warning signals of potential disease outbreaks by identifying susceptible populations or previously unrecognised pathogen exposure.2 Neglected tropical diseases (NTDs) are a group of 21 infectious diseases, mainly affecting populations in tropical and subtropical regions, that have historically received less attention and funding than other illnesses. In the Roadmap for Neglected Tropical Diseases 2021–2030,3 WHO outlined ambitious goals to eliminate NTDs, with targets set for elimination of at least one NTD in 100 countries by 2030, and complete global eradication of two NTDs. Similarly, WHO’s Immunization Agenda 2030 has disease-specific elimination goals, particularly global elimination of measles and rubella by 2030.4 However, despite the announcement by WHO that all 22 Pacific Island Countries and Territories (PICTs) have been verified as having eliminated endemic measles and rubella,5 the Western Pacific Region alone reported an increase in measles cases of 743% between 2022 and 2024, likely due to gaps in vaccination coverage and surveillance.6 Both the NTD Roadmap 2021–2030 and the Immunization Agenda 2030 have highlighted the importance of sustained surveillance, particularly integrated surveillance, to achieve and monitor elimination goals. The Healthy Islands Vision,7 endorsed by Pacific Health Ministers in 1995 and reaffirmed in 2015, provides an overarching regional framework that links human health, environmental wellbeing, and resilient health systems. Despite important progress in reducing several communicable diseases, including those targeted for elimination, PICTs remain vulnerable to disease outbreaks, re-emerging diseases, and persistent transmission driven by extreme weather events, challenges in reaching remote communities, and fragmented financing. These pressures underscore the importance of strong, resilient surveillance and early warning systems to safeguard elimination gains and support regional health priorities. Multiplex bead-based assays (MBA), developed using multianalyte profiling (xMAP®) technology from several companies including Luminex (http://www.luminexcorp.com/) and Bio-Rad (https://www.bio-rad.com/), use fluorogenic emissions to detect responses when specific Ag-Ab binding occurs.8 MBAs enable concurrent testing of Ab against up to approximately 100 Ag from a single dried blood spot (DBS),9 supporting a multi-pathogen surveillance approach. MBA for population-level multi-pathogen surveillance can be both cost-effective and less resource intensive compared to time-consuming single-analyte assays. However, interpretation of Ab responses requires caution, as serological markers may reflect cumulative exposure/infection and may not be able to distinguish between past and current infection. Further, seropositivity can be influenced by cross-reactivity and variation in assay performance; therefore, careful selection of Ag and validated seropositivity thresholds, together with consideration of assay specificity and epidemiological context, are essential to minimise misinterpretation arising from cross-reactivity or non-specific Ab responses. Successful and sustained disease elimination depends on accurate surveillance. MBAs have been used for surveillance of eliminated and near-eliminated diseases, such as lymphatic filariasis (LF), with Ab seromarkers showing high area under the Receiver-Operating Characteristic Curve (AUC) values to classify patients as Ag-positive and demonstrating their potential to strengthen LF surveillance.10 For VPDs, serosurveys are considered the gold standard for measuring population immunity; MBA can leverage serosurveys by estimating population immunity to multiple VPDs in one sample and thereby identify disparities in seroconversion among pathogens included in multivalent vaccines.2 Indeed, MBA have exhibited higher sensitivity for measles and rubella immunoglobulin G (IgG) than single-plex ELISAs.11 However, assay performance characteristics may vary across Ag targets, and interpretation of MBA results should consider potential differences in sensitivity and specificity between markers. MBA has also been used to successfully monitor helminth infections, particularly the NTD strongyloidiasis, with a recent study demonstrating the utility of MBA as an alternative diagnostic tool for current Strongyloides stercoralis infection.12Taenia solium taeniasis/cysticercosis is a zoonotic NTD that uses humans as its definitive hosts and pigs/boars as intermediate/secondary hosts.13 Human cysticercosis occurs after humans ingest T. solium eggs, acting as accidental intermediate hosts; migration of cysticerci to the human brain results in neurocysticercosis (NCC) where cysts develop in the brain.13 NCC is the most frequent preventable cause of epilepsy worldwide and is estimated to cause 30% of all epilepsy cases in countries where the parasite is endemic. Taeniasis microscopy has low sensitivity, however serology for taeniasis Ags results have shown high sensitivity and specificity.14Giardia lamblia and Cryptosporidium parvum are enteric protozoan parasites commonly associated with diarrheal disease in humans; Ag for these pathogens have successfully been incorporated into MBA panels in serosurveys in several countries.15 Tonga was validated by WHO as having eliminated LF as a public health problem in 2017.16 For LF, elimination as a public health problem is defined by WHO as an antigenemia prevalence of less than 1% in 6-7-year-old children (in areas where W. bancrofti is endemic and Aedes is the primary vector17. In 2024, the Tonga Operational Research for Post-validation Surveillance for Elimination of Lymphatic Filariasis (TOPELF 2024) was conducted to ensure that there was no undetected LF resurgence in Tonga.16 Alongside Ag testing for LF, TOPELF 2024 collected DBS samples for MBAs, the results of which are presented in this current study. While the primary aim of TOPELF was to support post-validation surveillance of LF in Tonga following elimination as a public health problem, the MBA platform facilitated simultaneous measurement of Ab responses to multiple pathogens. This provides a valuable opportunity to generate integrated seroepidemiological insights across a broader range of diseases, including other NTDs, arboviruses, and VPDs. Detailed analyses of LF Ag16 and antifilarial Ab markers18 in relation to post-validation surveillance are presented elsewhere. The present study therefore focuses on the wider multi-pathogen serological profile of the study population and aimed to estimate seroprevalence of 17 pathogens to provide insights to inform public health strategies to prevent, control, and eliminate infectious diseases in Tonga. Our specific objectives were to estimate seroprevalence by (i) sex (males vs. females); (ii) age group (5–10 years, 11–19 years, 20–50 years, > 50 years); (iii) island group; (iv) demographic associations for seropositivity; and (v) geographical clustering of seropositivity. Methods Ethics approval The study was approved by the National Health Ethics and Research Committee of the Tonga Ministry of Health (Ref#20240129) on 22 February 2024 and ratified by The University of Queensland’s Human Research Ethics Committee (Project#2024/HE000493) on 12 March 2024 and Alfred Ethics Committee (Project#669/24) on 28th October 2024. All methods were performed in accordance with the approved ethical guidelines and regulations. Processes for obtaining consent have been described.19 Briefly, written consent was sought from participants (parent/guardian consent for children aged < 18 years), who were advised that they could revoke their consent at any time without prejudice. Written information, consent forms, and surveys were provided/conducted by local field teams in a culturally sensitive manner in Tongan or English. Study setting The Kingdom of Tonga is an archipelago of 176 islands, 36 of which are inhabited, spanning approximately 800 km from North to South. Administratively, it is divided into five divisions: Tongatapu, ‘Eua, Ha’apai, Vava’u, and the remote Ongo Niuas island group.20 There is a total population of approximately 100,179 people, around 70% reside on Tongatapu, the largest island and home to the capital, Nuku’alofa. Tonga has a tropical climate with mean annual temperatures ranging between 23°C and 28°C. Study design The study protocol was co-designed with the Tonga Ministry of Health (MOH) and has been described in detail elsewhere.16 In brief, the survey aimed to identify communities and sub-populations at increased risk of LF using historical data and local knowledge. Four survey components were implemented: (1) community surveys, (2) primary schools, (3) high schools and boarding schools, and (4) a diabetes outpatient clinic at Vaiola Hospital (Nuku’alofa), the largest hospital in Tonga (Fig. 1). As this was a targeted surveillance survey designed to prioritise areas and sub-populations with higher probability of LF infection rather than a probability-based, nationally representative sample, the findings should be interpreted as proportions seropositive within the sampled groups and not as nationally generalisable prevalence estimates for Tonga. Map of Vaiola Hospital and selected primary schools, high schools, and communities by island group (Tongatapu, Ha’apai, and Ongo Niuas). Tonga Operational Research for Post-validation Surveillance for Elimination of Lymphatic Filariasis 2024. Reprinted from The Lancet Regional Health – Western Pacific, Vol. 57, Lawford HLS, Tukia O, Takai J, et al., Persistent lymphatic filariasis transmission seven years after validation of elimination as a public health problem: a cross-sectional study in Tonga, Article 101,513, Copyright (2025), with permission from Elsevier. Timing of survey A targeted cross-sectional survey was conducted between May and July 2024 in two phases: 9 May-6 June 2024 (Tongatapu and Ha’apai) and 11–19 July 2024 (Ha’apai and Niuatoputapu [Ongo Niuas]). Sample size and household selection Participant selection procedures to reach the target sample size of 1,955 participants have been reported.16,19 As described, 15 households and five backup households per community were randomly chosen by the Tonga Statistics Department using a random number generator. Backup households were used if selected households were uninhabited or unavailable. Household members were defined as those who primarily resided in or had slept in the household the previous night. All participants aged ≥ 5 years from households, primary and high schools (including boarding schools), the outpatient clinic, and volunteer groups were eligible and invited to participate. Data and sample collection Data were collected electronically using EpiCollect5 (version 76.1.0; https://five.epicollect.net) on smartphones. Questionnaires were used to captured demographic information (age, sex, birthplace, village, occupation) and travel history (domestic and international) over the previous 12 months. GPS coordinates of households and schools were recorded where possible. Approximately 300–400 µL of blood was collected via finger prick into heparin-coated BD Microtainer® tubes for LF Ag testing using Alere™ Filariasis Test Strips (Abbott, Scarborough, ME). DBS were prepared on Cellabs Tropbio Filter Paper Blood Collection Disks™ for all participants. Multiplex bead assays The chemical coupling of Ags to MagPlex microspheres (beads) (Luminex Corp., Austin, TX) were provided by the US Center for Disease Control (US CDC) and Burnet Institute. Briefly, DBS were eluted into 96-well plates, diluted to a final concentration of 1:400 and assayed for Ab on an MBA containing 26 Ag-coupled beads from 17 pathogens grouped into five disease categories: NTDs (LF [Bm14 and Wb123], trachoma [Pgp3], yaws [Rp17 and TmpA], taeniasis [Es33], cysticercosis [T24H], and strongyloidiasis [SsIR]), arboviruses (dengue virus [DENV] 1–4 [Ns1-1, Ns1-2, Ns1-3, Ns1-4], chikungunya [Chik E1], and Zika [Zika NS1]), VPDs (tetanus [Tet], diphtheria [Dip], measles [MeV], rubella [RuV] pertussis [Pt], SARS CoV-2 [Spike, N, RBD591), and enteric protozoa (cryptosporidiosis [Cp17 or Cp23] and giardiasis [Vsp3 or Vsp5]). Sample plates were read on a Bio-Plex 200 instrument (Bio-Rad, Hercules, CA). For each Ag, median fluorescence intensity minus background (MFI-bg) signal was used to determine seropositivity. Thresholds for seropositivity were determined using one of three approaches depending on the availability of reference samples: (1) the mean plus three standard deviations (SD) of the MFI of known negative sera; (2) receiver operating characteristic (ROC)-derived MFI cut-offs based on panels of known positive and negative samples, where positives were defined by PCR or reference serological assays (e.g., TPPA/RPR) and negatives included sera from individuals with no evidence of infection, such as non-travellers from non-endemic settings; or (3) standard curves generated using WHO International reference serum standards. Using these standards, concentrations were calculated using linear regression fit to log-transformed values from dilution series with seroprotective thresholds IU values. Thresholds corresponding to partial protection (> 0.01-<0.1 IU/mL), full protection (≥ 0.1-<1.0 IU/mL), and maximum (≥ 1 IU/mL) antibody levels were calculated for diphtheria and tetanus. Thresholds for measles were > 153 mIU/mL, rubella > 9.36 IU/mL, and pertussis 5 IU/mL. Definitions and interpretation of seropositivity Pathogens were grouped into four broad disease groups: NTDs (LF, trachoma, yaws, taeniasis, cysticercosis, strongyloidiasis), arboviruses (dengue virus [DENV] 1–4, chikungunya, Zika), VPDs (measles, rubella, diphtheria, tetanus, pertussis, SARS CoV-2), and enteric protozoa (cryptosporidiosis and giardiasis) acknowledging that some pathogens, such as dengue virus, can fall into more than one category. Definitions and interpretation of seropositivity are summarised below and in Table 1. Table 1Definition and interpretation of pathogen-specific antibody markers. adapted from Carcelen et al.24 with permission from the publisher.Disease groupingPathogenAntigen abbreviationInterpretationMedian Fluorescence Intensity cut-offCut-off determinationSource of AntigenCoupling concentration (ug/Ag/1.25e7 beads)Coupling pHNeglected tropical diseasesLymphatic filariasis (Wuchereria bancrofti) Wb123 Wb123 is a larval antigen. Antibodies are expressed by the L3 larval stage transmitted by mosquitoes. Seropositivity suggests exposure and potential ongoing transmission.403Mean + 3SD (CDC controls)Recombinant CDC307.2 Bm14 Bm14 seropositivity indicates past or current exposure but not necessarily recent transmission. Seropositivity to either antigen was considered indicative of past/current LF infection. Antibodies are long lasting.1467Mean + 3SD (CDC controls)Recombinant CDC1207.2Trachoma(Chlamydia trachomatis) Pgp3 Reflects historical and cumulative exposure to C. trachomatis; early indicator of transmission but cannot distinguish ocular from urogenital or perinatal exposure. Estimates restricted to < 14 years.714ROCRecombinant CDC1207.2Yaws(Treponema pallidum pertenue) Rp17 Current infection or previous exposure to T. pallidum subsp pallidum (syphilis) or T. pallidum subsp pertenue (yaws). Marker of historical infection.108ROC (Maximum Youden’s J)Recombinant CDC157.2 TmpA Current infection to T. pallidum subsp pallidum (syphilis) or T. pallidum subsp pertenue (yaws). Marker of recent infection.704ROC (Maximum Youden’s J)Recombinant CDC307.2Taeniasis(Taenia solium) Es33 Reflects past/current infection with adult pork tapeworm (taeniasis).534ROCRecombinant CDC105Cysticercosis(Taenia solium) T24H Reflects past/current infection with cysts (cysticercosis).304ROCRecombinant CDC1205Strongyloidiasis(Strongyloides stercoralis) SsIR Reflects past/current S. stercoralis infection. High titres can be indicative of current chronic infection.2505ROCRecombinant CDC207.2 + 2 M UreaArbovirusesDengue virus(Seroptypes 1–4)NS1-1Reflects past/current infection with respective DENV serotype.5410ROCMeridian Life Sciences127.2NS1-210,068Meridian Life Sciences127.2NS1-310,092Meridian Life Sciences127.2NS1-4807Meridian Life Sciences127.2Chikungunya virus Chik E1 Reflects past/current infection chikungunya virus infection.5914ROCMeridian Life Sciences7.57.2Zika virus Zika NS1 Reflects past/current infection Zika virus infection.18,013ROCMeridian Life Sciences307.2Vaccine-preventable diseasesDiphtheria(Corynebacterium diphtheriae)Diphtheria toxoidVaccine acquired immunity. Detectable Ab above minimal threshold (≥ 0.01 IU/mL) suggestive of vaccination or past infection, though levels may not guarantee protective immunity.159Standard curve (0.01 IU/mL)WHO Int TE-3605Tetanus(Clostridium tetani)Tetanus toxoid505Standard curve (0.01 IU/mL)WHO Int 10/26212.55Pertussis(Bordetella pertussis)Pertussis toxinDetectable antibodies above minimal threshold (5 IU/mL for pertussis)163Standard curve(5IU/mL)WHO Int 06/142207.2Measles(Measles virus)Measles virusDetectable antibodies above minimal threshold (0.153 IU/mL)1363Standard curve (0.153 IU/mL)RUBI-1-941507.2Rubella(Rubella virus)Rubella virusDetectable antibodies above minimal threshold (9.363 IU/mL)1210Standard curve (9.363 IU/mL)RUBI-1-94307.2COVID-19(SARS-CoV-2)Nucleocapsid protein (N)Usually natural infection except in places where nucleocapsid is in vaccines.2578ROCGenscript3.47.2 RBD519 Vaccine-induced immunity.2365ROCRecombinant CDC157.2Spike glycoprotein S1-receptor binding domainVaccine acquired or natural infection.584ROCRecombinant CDC65Enteric protozoaCryptosporidiosis Cp17 Reflects past/current Cryptosporidiosis infection.244ROCRecombinant CDC6.85 Cp23 1377ROCRecombinant CDC12.55Giardiasis Vsp3 Reflects past/current Giardia infection.312ROCRecombinant CDC605 Vsp5 445ROCRecombinant CDC605 Definition and interpretation of pathogen-specific antibody markers. adapted from Carcelen et al.24 with permission from the publisher. Trachoma (Chlamydia trachomatis) Yaws (Treponema pallidum pertenue) Taeniasis (Taenia solium) Cysticercosis (Taenia solium) Strongyloidiasis (Strongyloides stercoralis) Dengue virus (Seroptypes 1–4) Diphtheria (Corynebacterium diphtheriae) Tetanus (Clostridium tetani) Pertussis (Bordetella pertussis) Standard curve (5IU/mL) Measles (Measles virus) Rubella (Rubella virus) COVID-19 (SARS-CoV-2) Among the LF Abs, Wb123 is larval Ag, suggesting that Wb123 Ab-positive results are more likely to be associated with exposure to infected mosquitoes (and therefore an indicator of ongoing transmission) than Bm14 Ab-positive, which indicates past or current LF exposure but not necessarily recent transmission.10 For this study, seropositivity to either Ab was considered indicative of past/current LF infection. For trachoma, Pgp3 Ab seropositivity reflects cumulative exposure to Chlamydia trachomatis and is considered an early indicator of transmission, but cannot distinguish ocular infection from urogenital exposure or perinatal transmission21 thus seroprevalence estimates for these pathogens were restricted to < 14-year-olds only. For yaws, Rp17 Ab seropositivity is considered indicative of past infection, whilst TmpA Ab are indicative of recent infection. Dual positivity suggests current or very recent yaws, and for the purpose of this paper, will be considered the main measure of yaws infection. Similarly, MBAs cannot differentiate yaws infection from syphilis, thus seropositivity estimates for yaws were also limited to ≤ 14-year-olds only. For T solium taeniasis [Es33]/cysticercosis [T24H], seropositivity to Es33 Ab only was considered indicative of past/current taeniasis (adult pork tapeworm), T24H Ab seropositivity only indicative of past/current cysticerci (cysticercosis), and dual-positivity indicative of past/current infection to both stages. Lastly, SsIR Ab seropositivity was considered indicative of past/current strongyloidiasis infection. For the arborviruses, seropositivity to Ns1-1 Ab, Ns1-2 Ab, Ns1-3 Ab, Ns1-4 Ab, Chik E1 Ab, and Zika NS1 Ab was considered indicative of past/current DENV1-4 infection, chikungunya virus, and Zika virus, respectively. For diphtheria, tetanus, pertussis, measles, and rubella, seropositivity was defined using assay-specific minimal Ab thresholds indicating detectable immune responses consistent with prior vaccination or infection, noting that this may not be sufficient for full protection against infection or disease.22 These thresholds were distinguished from established correlates of protection associated with individual-level immunity, and from population-level herd immunity thresholds used to guide vaccination targets and elimination programmes. For SARS CoV-2, Ab to nucleocapsid were interpreted as markers of prior infection, while antibodies to spike and receptor-binding domain (RBD) may arise from either vaccination or infection. COVID-19 vaccines used in Tonga during the study period were primarily spike-based; therefore nucleocapsid (N) seropositivity is likely to reflect natural infection. Lastly, for enteric protozoa, Cp17 Ab seropositivity was considered indicative of acute, recent cryptosporidiosis infection, Cp23 Ab seropositivity for past infection, and dual-positivity for recent infection with an established immune response.23 Seropositivity to either Vsp3 or Vsp5 was considered indicative of past/current giardiasis infection. Statistical analysis Data were analysed using Stata statistical software (StataCorp, Version 18.0, College Station, TX). Summary statistics with 95% confidence intervals (95% CI) for Ab seroprevalence were calculated for age groups (5–10 years, 11–19 years, 20–50 years, and > 50 years), sex (male or female), and island group (Tongatapu, Ha’apai, or Niuatoputapu [Ongo Niuas]). Logistic regression was used to assess associations between demographics and seropositivity for each Ab. Variables with p < 0.2 on univariable analyses were tested using multivariable logistic regression; variables were sequentially removed from the multivariable models to arrive at the most parsimonious models, in which variables with p < 0.05 were retained. As analyses were conducted separately for each Ab outcome, final model structures differed between pathogens. Estimated seroprevalence (presented as percentages) was calculated using the ‘margins’ command in Stata following multivariable logistic regression models. Marginal predicted probabilities and 95% CIs of seropositivity for each covariate category. Regression models were developed separately for each Ab outcome, therefore predicted probabilities were derived from pathogen-specific models. Mixed-effects logistic regression models (Stata command ‘melogit’) were used to examine associations between age, sex, and seropositivity while accounting for geographic clustering of participants. Random intercepts were included for village and island group to reflect the hierarchical structure of the data, with participants nested within villages and villages within island groups. Participants recruited through schools or the diabetes outpatient clinic were asked to report their home village, and clustering was therefore defined according to village of residence rather than recruitment site. All villages were included in the analysis regardless of sample size. Intraclass correlation coefficients (ICCs) were estimated following each model (Stata command ‘estat icc’) to quantify the proportion of variance attributable to village- and island-level clustering. ICC values range from zero to one; higher values indicate more homogeneity, suggesting a high degree of clustering at that level. Results Study population demographics A total of 1,787 participants were enrolled, with a mean age of 29.5 years (range 5–89); 46% were male. Of these, 1,768 (98.9%) had valid MBA results. Most participants (64%) were recruited from Tongatapu, 22% from Niuatoputapu, and 14% from Ha’apai, reflecting the targeted survey design. Overall, 20% (n = 358) of participants were aged 5–10 years old, 30% (n = 531) 11–19 years old, 28% (n = 496) 20–50 years old, and 22% (n = 381) > 50 years old. Seroprevalence estimates among the target population varied substantially across pathogens and are summarised for all Abs in Supplementary Table 1 (stratified by age and sex) and Supplementary Table 2 (stratified by island group). Logistic regression was used to identify associations for seropositivity (Supplementary Tables 3a-c). After adjusting for age group, sex, and island group, estimated seroprevalence showed strong age-related increases for most pathogens (Supplementary Tables 4a-f). Results are summarised by disease group below. Neglected tropical diseases Overall Ab seroprevalence in the target population ranged from 0.2% (95% CI: 0.0-1.2) for yaws to 6.9% (95% CI: 5.1–9.2) for trachoma. Compared to females, significantly higher Ab seroprevalence was seen among men for LF (Bm14 [males = 6.5% [5.0-8.4] and females = 3.7% [2.7–5.1]; p = 0.002] and Wb123 [males = 6.5% [5.0-8.4] and females = 3.7% [2.7–5.1]; p = 0.009]) and strongyloidiasis (males = 3.1% [2.0-4.6] and females = 1.6% [0.9–2.6]; p = 0.03). Females had significantly higher Ab seroprevalence for cysticercosis (males = 3.1% [2.2–4.4] and females = 1.5% [0.9–2.6]; p = 0.03) (Fig. 2, Panel A). Seroprevalence significantly increased with age for the LF Abs (Bm14 [p < 0.001] and Wb123 [p < 0.001]) and strongyloidiasis (p < 0.001). Ab seroprevalence estimates for trachoma were limited to participants ≤ 14 years (n = 545) and was higher among 5-10-year-olds (8.4% [5.7–11.7]) vs. 11-14-year-olds (1.9 [0.9–3.4]; p < 0.001) (Fig. 2, Panel B). Significantly higher Ab seroprevalence against LF Ag were seen in Ha’apai (Wb123 = 13.1% [9.5–17.9] and Bm14 = 8.8% [5.8–13.0]) compared to Tongatapu (Wb123 = 1.7% [1.1–2.6] and Bm14 = 0.5% [0.2–1.2]) or Niuatoputapu (Wb123 = 9.4% [6.8–12.7] and Bm14 = 6.5% [4.4–9.4]) (Fig. 2, Panel C). Seroprevalence of neglected tropical diseases by sex (A: males represented in dark purple bars and females in light purple bars), age (B: 5–10 years [white bar], 11–19 years [light purple bar], 20–50 years [medium purple bar], and > 50 years [dark purple bar]), and island group (C: Tongatapu represented by the white bar, Ha’apai by the light purple bar, and Niuatoputapu by the dark purple bar). Tonga Operational Research for Post-validation Surveillance for Elimination of Lymphatic Filariasis 2024. Trachoma and yaws restricted to participants aged ≤ 14 years (total = 595 participants). Significance tested using univariable logistic regression. ***p < 0.001; **p < 0.01; *p < 0.05; n.s = “non-significant”. Older participants had significantly higher adjusted odds ratios (aORs) for seropositivity for both LF Ab (Bm14 [Wald χ²(3) = 31.28; p < 0.001] and Wb123 [Wald χ²(3) = 32.46; p < 0.001]) and strongyloidiasis (Wald χ²(3) = 13.69; p = 0.003), whilst those aged 11–19 years had significantly lower aORs for seropositivity for cysticercosis vs. those aged 5–10 years (aOR = 0.35 [0.13–0.95]; p = 0.04) with no association seen for older age groups. Compared to females, males had significantly higher aORs for seropositivity for both LF Ab (Bm14 [aOR: 2.95 (1.61–5.40); p = < 0.001] and Wb123 [aOR: 2.12 (1.35–3.35); p = 0.001]) and strongyloidiasis (aOR: 2.22 (1.16–4.30); p = 0.02). Males had significantly lower aOR for Ab for cysticercosis compared to females (aOR: 0.50 (0.25–0.99); p = 0.048). Compared with outer-island participants, those from Tongatapu had significantly lower aORs for seropositivity for both LF Ab (Bm14 [Wald χ²(3) = 40.19; p < 0.001] and Wb123 [Wald χ²(3) = 56.55; p < 0.001]). aORs for seropositivity for trachoma was significantly higher in Ha’apai compared to Tongatapu only (aOR: 2.16 [1.01–4.64; p = 0.048]), but higher aORs were found for Ab seropositivity for cysticercosis in Tongatapu vs. Niuatoputapu only (aOR: 0.32 [0.11–0.92]; p = 0.03) (Supplementary Table 3a). For both LF Ab, estimated seroprevalence was low among participants aged < 20 years (< 1%) and increased with age, from 5% in 20-50-year-olds to 13% in > 50-year-olds among females and 14% in 20-50-year-olds to 30% in > 50-year-olds among males in Ha’apai. In Niuatoputapu, estimated seroprevalence increased from 4% in 20-50-year-olds to 11% in > 50-year-olds among females and 12% in 20-50-year-olds to 27% in > 50-year-olds among males. Cysticercosis seropositivity was highest in Tongatapu, particularly among females and adults aged 20–50 years, while predicted probabilities in Ha’apai and Niuatoputapu were low and mostly non-significant across all sex and age groups. Estimated seroprevalence of strongyloidiasis Ab increased with age and more common in males than females. Estimated seroprevalence of trachoma Abs were highest in Ha’apai (12%) (Supplementary Table 4). Arboviruses Among the target population, females had significantly higher seroprevalence for Ab for Zika virus (males = 18.5% [15.9–21.4] and females = 23.8% [21.1–26.6]; p = 0.007), chikungunya virus (males = 30.0% [26.8–33.3] and females = 37.1% [34.0-40.2]; p = 0.002), and all dengue serotypes (p < 0.001) compared to males (Fig. 3, Panel A). Ab seroprevalence estimates increased significantly with age for Zika virus (p < 0.001), chikungunya virus (p < 0.001), all dengue serotypes (p < 0.001) (Fig. 3, Panel B). By island group, significantly higher seroprevalence for all arboviruses was seen in Tongatapu (DENV-1 = 63.3% [60.4–66.0]; DENV-2 = 63.2% [60.3–65.9]; DENV-3 = 64.1% [61.3–66.9]; DENV-4 = 70.5 [67.8–73.1]; Zika virus = 24.9 [22.5–27.5]; chikungunya virus = 39.9% [37.1–42.8]) compared to Ha’apai (DENV-1 = 57.4% [51.1–63.4]; DENV-2 = 49.0% [42.8–55.2]; DENV-3 = 50.6% [44.4–56.8]; DENV-4 = 64.9 [58.8–70.6]; Zika virus = 18.3 [14.0-23.6]; chikungunya virus = 28.7% [23.4–34.6]) or Niuatoputapu (DENV-1 = 52.7% [47.7–57.7], p = 0.001; DENV-2 = 47.3% [42.3–52.3], p < 0.001; DENV-3 = 47.8% [42.8–52.8], p < 0.001; DENV-4 = 62.3% [57.4–67.1], p = 0.007; Zika virus = 13.0% [10.0-16.7], p < 0.001; chikungunya virus = 19.5% [15.8–23.8], p < 0.001) (Fig. 3, Panel C). Seroprevalence of arboviruses by sex (A: males represented in light blue bars and females in dark blue bars), age (B: 5–10 years [white bar], 11–19 years [light blue bar], 20–50 years [medium blue bar], and > 50 years [dark blue bar]), and island group (C: Tongatapu represented by the white bar, Ha’apai by the light blue bar, and Niuatoputapu by the dark blue bar). Tonga Operational Research for Post-validation Surveillance for Elimination of Lymphatic Filariasis 2024. Significance tested using univariable logistic regression. ***p < 0.001; **p < 0.01; *p < 0.05; n.s = “non-significant”. Older participants had significantly higher aORs for Ab seropositivity for all arboviruses. Compared to males, females had higher aORs for Ab seropositivity for all arbovirus except DENV-1. Participants from Tongatapu had higher aOR for Ab for all arboviruses compared with outer-island participants (Supplementary Table 3b). Arboviral Ab seropositivity showed marked age increases, particularly in Tongatapu. The predicted Ab seroprevalence against DENV-1, DENV-2, and DENV-3 in Tongatapu exceeded 90% among adults aged > 50 years (Supplementary Table 4). Vaccine-preventable diseases Compared to males, females had significantly higher Ab seroprevalence for SARS CoV-2 (RBD591 Ag [p = 0.003] and Spike protein [p = 0.02]), measles (p < 0.001), rubella (p < 0.001), and tetanus (p = 0.001) (Fig. 4, Panel A). Seroprevalence estimates were higher among older age groups for SARS CoV-2 (RBD591 Ag [p = 0. 001], Spike protein [p = 0. 001], N [p < 0.001]), measles (p = 0.001), and rubella (p = 0.001). Seroprevalence estimates were higher among younger age groups for diphtheria (p < 0.001), tetanus (p < 0.001), and pertussis (p = 0.01) (Fig. 4, Panel B). Significantly higher seroprevalence was seen in Tongatapu vs. other island-groups for VPD Abs (measles [p < 0.001], rubella [p = 0.01], diphtheria [p = 0.01], tetanus [p < 0.001], pertussis [p < 0.001]) (Fig. 4, Panel C). Seroprevalence of vaccine-preventable diseases by sex (A: males represented in light green bars and females in dark green bars), age (B: 5–10 years [white bar], 11–19 years [light green bar], 20–50 years [medium green bar], and > 50 years [dark green bar]), and island group (C: Tongatapu represented by the white bar, Ha’apai by the light green bar, and Niuatoputapu by the dark green bar). Results show minimal threshold of protection (≥ 0.01 IU/mL [0.153 IU/mL for measles; 9.36 IU/ml for rubella; 5 IU/mL for pertussis]) Tonga Operational Research for Post-validation Surveillance for Elimination of Lymphatic Filariasis 2024. Significance tested using univariable logistic regression. ***p < 0.001; **p < 0.01; *p < 0.05; n.s = “non-significant”. Older participants had significantly higher aORs for all Abs for SARS CoV-2 (RBD591, Spike, N) and measles than those aged 5–10 years. Older participants had significantly lower aORs for diphtheria, tetanus, and pertussis. Participants aged 10–19 years had lower aORs for measles seroprevalence compared to those aged 5–10 years. Females had significantly higher aOR for Ab for tetanus, measles, and rubella compared to males. Participants from Tongatapu had higher aOR for Ab for all VPDs except Abs to SARS CoV-2 compared with outer-island participants (Supplementary Table 3c). Predicted measles and rubella Ab seroprevalence was uniformly high (> 85%) and highest among those aged > 50 years. Contrastingly, predicted Ab seroprevalence of diphtheria and pertussis was highest among 5-10-year-olds. SARS-CoV-2 estimated Ab seroprevalence was high across all age groups, exceeding 90% among ≥ 11-year-olds for Spike Ab and RBD591 Ab (Supplementary Table 4). Enteric protozoa Significantly higher Ab seroprevalence was seen for Ab for cryptosporidium among females compared to males (p = 0.005) (Fig. 6, Panel A) and among older compared to younger age groups (p < 0.001) (Fig. 5, Panel B). Significantly higher seroprevalence was seen in Tongatapu vs. other island-groups for cryptosporidium (p = 0.04) and in Ha’apai for giardia (p = 0.001). Intraclass correlation coefficients for clustering at the island (light shading) and village levels (dark shading) for neglected tropical diseases (purple bars), arboviruses (blue bars), vaccine preventable diseases (green bars) and enteric protozoa (orange bars). Tonga Operational Research for Post-validation Surveillance for Elimination of Lymphatic Filariasis 2024. Error bars denote 95% Confidence Intervals. Seroprevalence of enteric pathogens by sex (A: males represented in light orange bars and females in dark orange bars), age (B: 5–10 years [white bar], 11–19 years [light orange bar], 20–50 years [medium orange bar], and > 50 years [dark orange bar]), and island group (C: Tongatapu represented by the white bar, Ha’apai by the light orange bar, and Niuatoputapu by the dark orange bar). Tonga Operational Research for Post-validation Surveillance for Elimination of Lymphatic Filariasis 2024. Significance tested using univariable logistic regression. ***p < 0.001; **p < 0.01; *p < 0.05; n.s = “non-significant”. Older participants had significantly higher aORs for cryptosporidium Abs. Participants from Tongatapu had higher aOR for Ab for cryptosporidium and giardia compared with outer-island participants (Supplementary Table 3d). Estimated seroprevalence of giardia Abs were highest in Ha’apai (11%), whilst cryptosporidium seropositivity increased with age and was ≥ 90% among > 50-year-olds (Supplementary Table 4). Geographic clustering Geographic clustering was significant (p < 0.01) for nearly all Ab except SARS CoV-2, diphtheria, taeniasis, cysticercosis, strongyloidiasis, and cryptosporidium. Clustering was strongest at the village level, with LF showing the highest ICC (0.56 [95% CI: 0.27–0.81]) (Fig. 6 and Supplementary Table 5). Discussion This study presents population seroprevalence estimates of selected NTDs, arboviruses, VPDs, and enteric protozoa from a targeted cross-sectional study conducted in Tonga in 2024. Although the survey was not designed to provide nationally representative estimates, it offers a snapshot of cumulative exposure/infection across multiple pathogens within the sampled communities. Our findings demonstrate the utility of integrated multi-pathogen serosurveillance for characterising transmission dynamics and informing targeted surveillance and intervention strategies in Tonga. High seroprevalence of VPDs was found which, beyond natural exposure, likely reflects high national vaccination coverage. Reassuringly, and in contrast to similar serosurveys conducted in the region, 2moderate to high VPD seroprevalence was observed in younger age groups. The commonly cited ≥ 95% population immunity threshold required to sustain measles elimination was used as a reference for interpretation. However, this threshold was not achieved in this targeted study population. High seropositivity to cryptosporidium was observed among older adults residing in Tongatapu, which may warrant further investigation into assessments of, and possible interventions to, improve water quality and sanitation in this setting. Similarly, older adults had moderately high seroprevalence for dengue, reflecting historical and ongoing dengue transmission in the Pacific region. Interpretation of serological markers in MBAs requires careful consideration of the temporal dynamics of Ab responses. For many pathogens included in this study, seropositivity reflects cumulative exposure/infection rather than recent infection. Ab responses to pathogens such as dengue virus and cryptosporidium may persist for extended periods, and therefore cannot be used to infer recent transmission without additional epidemiological or molecular evidence. Similarly, some markers included in may have limited pathogen specificity or may reflect historical exposure/infection (e.g., Pgp3 Ab for trachoma and treponemal Abs that may arise from yaws or syphilis). Consequently, the serological findings reported here should be interpreted primarily as indicators of prior exposure/infection rather than direct evidence of recent or ongoing transmission. The use of serosurveillance to monitor pathogens that are not currently part of routine surveillance, including many NTDs, has been recognised as a valuable approach to improve disease burden estimates, particularly for low prevalence diseases that are approaching elimination, or that are considered eliminated.25 This targeted cross-sectional study was designed as a post-validation surveillance activity to identify any persistent LF transmission following Tonga’s validation of elimination of LF as a public health problem.16,19 LF Ab seroprevalence estimates were consistent with LF Ag findings that have been published;16 these findings will be discussed in detail elsewhere. Consistent with other serosurveys in the region, 2 we found Wb123 Ab seroprevalence was higher than that of Bm14 Ab, with Bm14 Ab seroprevalence approximating Ag prevalence.10 Wb123 Ab are generally considered markerse of recent infection with Wuchereria bancrofti compared to Bm14 Ab; however, neither marker can reliably distinguish recent from cumulative exposure, and seropositivity alone does not confirm active infection or ongoing transmission. Differences in Ag-specific Ab kinetics, assay performance, and programmatic history, including repeated MDA rounds, may contribute to these patterns. Interpretation of antifilarial Ab markers in post-elimination settings therefore remains an area of ongoing research and should be considered alongside our Ag-based surveillance findings from TOPELF. In this study, the estimated seroprevalence of trachoma among ≤ 14-year-olds was 6.9%. Similar to LF, trachoma has been targeted for elimination as a public health problem by WHO,3 with Fiji the most recent Pacific Island country to have achieved this. There is limited literature regarding trachoma in Tonga; a 1991 study of 4056 participants identified three individuals bilaterally blind from trachoma,26 and no subsequent cases have been reported.27 As seroprevalence to Pgp3 in young children is thought to reflect current and recent infection with Chlamydia trachomatis associated with trachoma transmission,28 additional data, particularly in children aged 1–5 years,29 would allow estimation of seroconversion rates and strengthen inferences about recent transmission in Tonga. While our study reported seroprevalence in ≥5-year-olds only, including younger age groups in future surveys would provide a more sensitive measure of potential transmission. Notably, significantly higher seroprevalence estimates were observed in Ha’apai, suggesting that targeted surveillance in this island group could provide a useful starting point for additional investigation. Encouragingly, only one possible case of yaws (dual positive for TmpA Ab and Rp17 Ab), an NTD that has been targeted by WHO for global eradication by 2030, was found. Further discussion with WHO and Tonga MOH is needed to determine next steps, but repeated serosurveys in this population would be valuable for detecting any changes in transmission status over time. We found unexpectedly high cryptosporidium seroprevalence (64.9% overall) in older age groups, indicating substantial exposure in our study population. In contrast, a longitudinal cohort study among children in Bangladesh found high Cp17 Ab and Cp23 Ab responses in the first year of life with a decline as age increased.30 Cryptosporidium seropositivity has been associated with environmental and socioeconomic factors including sanitation, income, and sex, as well as rurality.15,31 In our study, seroprevalence was higher in Tongatapu, which may reflect differences in population density, sanitation, or water quality compared with other island groups. However, seropositivity to enteric protozoa such as cryptosporidium could reflect either historic infection or ongoing low-level transmission, and interpretation should consider Ab persistence and life-course exposure patterns. While water, sanitation, and hygiene (WASH) conditions may contribute to transmission risk, infection may also occur through alternative pathways, including recreational freshwater contact or localised environmental contamination. In island settings where freshwater swimming or bathing is common, such exposures may not be directly attributable to deficiencies in water supply or sanitation infrastructure. Further investigation would therefore be required to better understand transmission pathways before attributing serological findings to specific environmental risk factors. We found cysticercosis and taeniasis seroprevalence to be low overall (2.4% and 0.8%, respectively). Cysticercosis seropositivity was more common among females in our study, as seen elsewhere.32 Variation between island groups was also seen, likely reflecting differences in livestock husbandry practices and sanitation. Given the risk of NCC, further investigation into clinical reports of NCC is warranted, along with discussion regarding targeted interventions in high-risk areas. Strongyloidiasis demonstrated significant associations with age and sex, while cysticercosis was associated with both geographical location and sex. Similar to our study, previous studies have also reported higher strongyloidiasis seroprevalence among older individuals and males.33,34. The geographical differences in seropositivity seen in this study reflect disparities in vaccination access and pathogen exposure, highlighting both the diversity of Tonga’s island groups and the challenges of delivering health services to remote populations. Distinct geographic trends were seen for seropositivity to VPDs and arboviruses, with lower seroprevalence (suggesting lower vaccination uptake and lower arbovirus infection) observed in the outer islands compared to Tongatapu. These findings likely reflect the combined challenges of limited access and population movement to the more remote island groups. In Tonga, reported coverage of the first and second dose of measles-containing vaccine has been high, ranging from 96 to > 99% from 2003 to 2024.35 For our 2024 study population, this corresponds to participants aged approximately 5–22 years, among whom measles seroprevalence was lower at 80%. Administrative vaccination coverage estimates may not correspond directly to measured seroprevalence due to factors such as inaccuracies in reporting or population denominators, incomplete two-dose uptake, waning immunity over time, or variability in vaccination delivery. Further, as measles vaccine potency is highly temperature-sensitive, maintaining a cold-chain is paramount for ensuring vaccine efficacy. In Tonga, travel to the remote island groups often incurs substantial costs and logistical challenges, which may affect vaccine accessibility, delay timely delivery, and could compromise cold-chain maintenance. These factors may contribute to geographic differences in seroprevalence, including the lower measles seroprevalence observed in Ha’apai and Niuatoputapu, although this cannot be confirmed in the present study. As all 22 PICTs have been validated as having eliminated measles and rubella,5 continued surveillance of population immunity is essential to prevent future outbreaks. It should be noted that the Ab thresholds used in this study represent assay-defined levels indicating detectable immune responses consistent with prior vaccination or infection rather than a direct measurement of established correlates of protection. Consequently, seroprevalence estimates may not correspond precisely to the proportion of individuals protected against infection. In addition, Ab waning and assay performance characteristics may influence estimates, particularly across age groups. These factors should therefore be considered when interpreting seroprevalence relative to population-level elimination targets. Dengue transmission is driven by mosquitoes feeding on viraemic humans, and is strongly associated with population density, movement, and urbanisation,36,37 thus the remote and sparsely populated settings of Ha’apai and Niuatoputapu would be considered less conducive to sustained outbreaks. Indeed, a major dengue outbreak occurred in Tonga in 2025, resulting in 845 cases and three deaths. The majority of confirmed cases were from the island groups of Tongatapu, ‘Eua, and Vava’u, with only four cases reported from Ha’apai and none from Niuatoputapu. Subsequent serosurveys are likely to show increased dengue seroprevalence reflecting this outbreak in Tongatapu. To date, two dengue vaccines (CYD-TDV [Dengvaxia®, manufactured by Sanofi Pasteur] and TAK-003 [Qdenga®, manufactured by Takeda Pharmaceutical Company]) have received clinical approval. Dengvaxia® was found to increase the risk of hospitalised and severe dengue among dengue-naïve children aged 2–8 years, with this excess risk emerging in year 3 of follow-up; no consistent increase in risk was observed in vaccinated individuals aged 9–16 years.37 Whilst understanding baseline seroprevalence to dengue serotypes will be invaluable for guiding vaccine policy and assessing program impact, dengue serotypes are highly cross-reactive in MBAs, and further work is needed to determine seropositivity to individual serotypes. Further, as our dengue serotype-specific results are based on NS1 Ab responses, which should be interpreted as indicative of exposure patterns rather than definitive serotype attribution given known cross-reactivity among flaviviruses and between dengue serotypes. A key limitation of our study is that we used a target sampling approach of higher-risk communities and selected sub-populations at greatest risk of LF. As a result, the proportion of the study population who were seropositive to specific pathogens may differ from national distributions, either upward or downward depending on the epidemiology of each infection. Furthermore, differential sampling intensity between villages may affect clustering estimates, including potential inflation of village-level ICCs. For some pathogens, such as SARS-CoV-2, the absence of clustering may also reflect widespread transmission and population-level vaccination, which can reduce spatial heterogeneity in exposure and infection; in addition, the ability to detect clustering may vary depending on overall prevalence. These findings should therefore be interpreted as indicative of patterns within the sampled populations rather than as nationally-representative estimates. A further limitation is that seroprevalence estimates depend on assay-specific cut-off definitions; although validated approaches were used, alternative thresholds could yield different absolute estimates, particularly for pathogens with values near the threshold for seropositivity. Participants were recruited through multiple sources (communities, schools, and a single diabetes outpatient clinic). These recruitment sites represent different age strata (younger school-based populations and predominantly older clinic attendees) and may introduce heterogeneity. However, age was accounted for in analyses, and findings should be interpreted within the context of this structured recruitment strategy rather than as arising from unmeasured recruitment-based confounding. Finally, regression analyses were exploratory and should be interpreted with caution. Given the number of pathogens evaluated and the use of variable selection procedures, some associations may reflect chance findings arising from multiple testing rather than true epidemiological relationships. Overall, our findings demonstrate the utility of multi-pathogen serosurveillance using MBA platforms as an efficient, cost-effective means to generate high-resolution surveillance data, which can be used to support more informed and integrated public health strategies. Notably, the strongest clustering of seropositivity for most pathogens occurred at the village level, suggesting that interventions targeted at this scale could achieve the greatest public health impact. This finding could also suggest that participation in public health programs (e.g. vaccination or mass drug administration) is influenced at a community level, for example by village and church leaders, or prevailing community sentiment. Alternatively, clustering may reflect differences in programme reach or historical outbreak patterns between villages. While more research is needed to strengthen our interpretation of Ab responses, including standardisation of assays around international units, understanding inter-laboratory variability, refining methods for determining seropositivity cut-offs, and addressing cross-reactivity between related pathogens, our findings provide robust evidence supporting the incorporation of MBA into national and regional surveillance systems. If multiple serosurveys are conducted over time, results can be used to track the impact of public health interventions, or the impact of climate/environmental change on disease transmission. Importantly, if deployed through existing survey platforms and data pipelines, MBA can strengthen early detection of under-vaccination, emerging infections, and persistent transmission, thereby enhancing the global capacity to achieve and sustain elimination of NTDs and VPDs.25. Conclusion Overall, this study revealed distinct sero-epidemiological patterns of multiple infectious diseases in Tonga. Multi-pathogen serosurveillance identified clear demographic and geographic correlations of infection and immunity, with geographic clustering, particularly for LF, highlighting the focal nature of transmission and the importance of targeted, community-level interventions. MBA provides a scalable, cost-effective platform for integrated multi-pathogen surveillance across diverse pathogens, generating high-resolution data to guide evidence-based programmatic decisions, strengthen elimination strategies, and support coordinated public health responses in the Pacific region. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Acknowledgements We would like to thank all the staff at the Tonga Ministry of Health for their support and without whom this study would not have been possible. We sincerely thank Dr Pateli Patelisio, Dr Aline Teumohenga, Tinalasa Vunipola, Nancy Tokelau, and all members of the field teams for their enthusiasm, support, and dedication to the study. We thank the doctors, nurses, and health facility staff at Vaiola Hospital, Likamonu Hospital, Nomuka Hospital, and Uiha Hospital for accommodating and supporting the field teams. We thank Dr Patrick Lammie and Dr Katherine Gass for their assistance in interpreting the results. Lastly, we thank the US CDC for chemical coupling of beads for MBA. We acknowledge the PacELF library (https://pacelf.github.io/#/) for supporting the initial stages of survey design, including access to several relevant closed documents. Author contributions Conceptualisation and funding acquisition: OT, CL, SS. Methodology: OT, JT, RO, CL, SS, HL. Data collection: HL, SW, HJ, BMM, SS. Supervision: OT, CL. Formal analysis: HL. Antibody assays: SV, LR, FA. Data interpretation: CL, HL. Writing (original): HL.HL, HJ, and CL verified the data in the study and take full responsibility for the integrity of the data. All authors had full access to all the data in this study. All authors reviewed and approved the final manuscript. Funding This work received financial support from the Coalition for Operational Research on Neglected Tropical Diseases (COR-NTD), which is funded at The Task Force for Global Health primarily by the Bill & Melinda Gates Foundation (OPP1190754) and by the United States Agency for International Development through its Neglected Tropical Diseases Program. Under the grant conditions of the Foundation, a Creative Commons Attribution 4.0 Generic License has already been assigned to the Author Accepted Manuscript version that might arise from this submission. CLL and LJR were funded by Australian National Health and Medical Research Council (NHMRC) Investigator Grants (1193826; 2017630). Data availability All relevant data are within the paper. We are unable to provide individual-level antigen prevalence data and demographic data because of the potential for breaching participant confidentiality. The communities in Tonga are very small, and individual-level data such as age, sex, and village of residence could potentially be used to identify specific persons. Declarations 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. References Associated Data Supplementary Materials Supplementary Material 1 Data Availability Statement All relevant data are within the paper. We are unable to provide individual-level antigen prevalence data and demographic data because of the potential for breaching participant confidentiality. The communities in Tonga are very small, and individual-level data such as age, sex, and village of residence could potentially be used to identify specific persons.