Durability and effectiveness of insecticide-treated nets in Mali: a longitudinal gSG6-P1 biomarker-based assessment of children’s exposure to Anopheles bites Authors: Ibrahim Traore, Moussa B. M. Cisse, Dipomin François Traoré, Yacouba Dansoko, Tidiani Sinayoko, Amadou Diakité, Lazeni Konaté, Alou Yacouba Sangaré, Jean-Marie Sanou, Alice Dembelé, Mariam S. Sangaré, Aldiana Kadidia Maiga, Mamadou Sow, Mahamadou Magassa, Franck Remoué, Ousmane Koita Published: 2025 ABSTRACT BackgroundInsecticide-treated nets (ITNs) remain a key intervention in malaria prevention. However, their protective effectiveness may decline with physical deterioration, even when usage remains high. This study assessed the impact of ITN physical integrity on children’s exposure to Anopheles mosquito bites over a three-year period in Mali, using gSG6-P1 biomarker as an innovative immuno-epidemiological indicator of human exposure.MethodsA three-year prospective cohort study was conducted from 2018, 2019, and 2020 in two rural health districts of Mali: Kéniéba using Yorkool® ITNs and Kita using PermaNet® 2.0 ITNs. A total of 586 children under five years old were enrolled and followed annually across 30 villages randomly selected into the two districts. Household surveys captured ITN ownership, usage patterns, and net condition. Net physical integrity was evaluated using proportional hole index (pHI). Blood samples were collected each year and analysed for anti-gSG6-P1 IgG levels, expressed as ΔOD. Net condition and specific IgG levels were analysed across time points and stratified by site and ITNs type.ResultsITN usage remained high (> 75%) across all survey years, but the proportion of serviceable nets declined significantly, particularly for Yorkool® (49% at 36 months versus 78% for PermaNet® 2.0). Median anti-gSG6-P1 IgG levels increased concurrently, indicating rising exposure to Anopheles bites as net integrity deteriorated.Children sleeping under Yorkool® nets showed higher specific IgG levels than those using PermaNet® 2.0, suggesting reduced protective performance of Yorkool® over time.ConclusionThis study demonstrates that ITN effectiveness decreases as physical deterioration advances, even when usage is maintained. Monitoring net integrity through the gSG6-P1 biomarker provides an innovative, field-adapted approach to anticipate ITN protection failures and to support evidence-based decision-making in malaria control programmes. Background Malaria remains a leading cause of morbidity and mortality in sub-Saharan Africa, particularly among children under five years of age. According to the World Health Organization, malaria caused an estimated 263 million cases and 597,000 deaths globally in 2023, with approximately 76% of deaths occurring among children under five, underscoring the urgency of effective vector control strategies [1]. In Mali, despite the implementation of multiple malaria control strategies, including seasonal chemoprevention and universal coverage campaigns for insecticide-treated nets (ITNs), malaria remains endemic and disproportionately affects young children. ITNs, the primary vector control tool deployed at scale, have significantly contributed to reducing malaria transmission, morbidity, and mortality across the continent [2, 3]. These nets provide protection by physically blocking mosquitoes and chemically repelling or killing them through insecticide treatment, thereby offering personal protection while reducing vector density and lifespan. Nevertheless, recent studies have raised concerns regarding the durability and sustained effectiveness of ITNs under real-life, operational conditions. Although ITNs are designed to remain effective for at least three years, their physical integrity and bioefficacy often deteriorates more rapidly due to wear and tear, washing practices, and local environmental conditions [4]. Moreover, vector resistance to pyrethroids (the insecticide class historically used in most conventional ITNs until the introduction of ‘next-generation nets’) has become increasingly widespread in West Africa, further compromising their efficacy [5–7]. These challenges highlight the need to monitor not only ITN coverage and use but also their functional condition and protective value over time. Current evaluations often rely on aggregated indicators such as net ownership and usage rates. However, these metrics fail to capture individual-level variations in exposure to Anopheles bites [8–10]. Furthermore, traditional entomological tools, such as human landing catches or indoor residual collections, are labor-intensive, context-dependent, and often insufficient to assess the residual malaria transmission risk. This limitation is particularly important in areas where low mosquito densities, outdoor biting behaviour, or partial net use may still result in significant exposure and infection risk. There is a pressing need for sensitive and scalable metrics that directly measure human exposure to mosquito bites, beyond traditional entomological methods to evaluate the effectiveness of ITNs under routine conditions and to inform timely net replacement policies [11, 12]. In this context, immuno-epidemiological biomarkers, such as the IgG response to the Anopheles gambiae salivary peptide gSG6-P1 offer a promising approach to directly measure recent human exposure to mosquito bites [11]. This biomarker reflects both fine-scale and large-scale variations in human-vector contact, temporally and spatially, and has been successfully used to evaluate the impact of vector control interventions such as ITNs or indoor residual spraying (IRS) [13–15]. Unlike entomological indicators, anti-gSG6-P1 antibody levels provide direct evidence of exposure at the individual and populational levels, independent of self-reported net use, and can help detect early failures in vector control coverage or effectiveness. Building upon this innovative approach, the present study conducted a longitudinal assessment of children’s exposure to Anopheles mosquito bites in Mali, using anti-gSG6-P1 IgG levels as a serological biomarker. Specifically, the study evaluated the protective efficacy of two types of ITNs brands (Yorkool® and PermaNet® 2.0) distributed during the 2017 national campaign and followed their performance over a three-year period under operational conditions. It was hypothesized that children sleeping under physically damaged nets would exhibit higher IgG levels against gSG6-P1, reflecting increased exposure to Anopheles bites. The findings are intended to provide actionable evidence to guide malaria control programmes in optimizing ITNs replacement strategies and improving vector control outcomes in vulnerable populations. Methods Study sites description The study was conducted in two rural health districts in southwestern Mali: Kéniéba and Kita, located in the Kayes region. Both sites belong to the Sudanian climatic zone, characterized by a long dry season (November–May) and a rainy season (June–October), during which malaria transmission is most intense. Malaria is endemic with moderate transmission, with Plasmodium falciparum as the predominant parasite species and Anopheles gambiae s.l. as the primary vector [16–18]. Mean annual rainfall ranges between 800 and 1100 mm, peaking between July and September. The region is predominantly rural, with extensive agriculture, traditional gold mining activities, and temporary ponds that serve as vector breeding sites. The transmission season typically starts in July, peaks between September and October, and declines by December [16–18]. A map of the study area is presented in Fig. 1.Fig. 1: Location of surveyed villages in Kéniéba (n = 15) and Kita (n = 15) districts, Mali, for ITN durability and biomarker assessments (2018–2020) [17]. The map shows the geographical distribution of study villages in Kéniéba (green) and Kita (orange) districts, where ITN durability monitoring and anti-gSG6-P1 biomarker assessments were conducted over three years (2018–2020) : Location of surveyed villages in Kéniéba (n = 15) and Kita (n = 15) districts, Mali, for ITN durability and biomarker assessments (2018–2020) [17]. The map shows the geographical distribution of study villages in Kéniéba (green) and Kita (orange) districts, where ITN durability monitoring and anti-gSG6-P1 biomarker assessments were conducted over three years (2018–2020) Study period and design This study was performed post 2017 ITN universal distribution campaign, with evaluations performed at three time points (12, 24, and 36 months post distribution) corresponding to 2018, 2019, and 2020 respectively. The evaluation focused on two ITNs types: Yorkool® nets, distributed in Kéniéba, and PermaNet® 2.0 nets, distributed in Kita. PermaNet® 2.0 (polyester 100 denier, deltamethrin 55 mg/m2, surface-coated) and Yorkool® (polyester, 75 denier, deltamethrin 55 mg/m2, polymer-incorporated) differ in polymer composition and insecticide binding technology. A summary of physical and chemical characteristics is provided in Table 1. No IRS or additional vector control campaigns occurred in either district during the study period. Housing characteristics, including wall type, roof material, and the presence of open eaves, were documented by direct observation using a standardized checklist, as these features influence mosquito entry. Seasonal rainfall data for 2018–2020 were extracted from published climatic datasets [16–18] to contextualize temporal variation in malaria transmission and mosquito exposure. At each annual survey (2018–2020), data collection was performed during the dry season (November–December), following the end of the rains, to capture the residual effects of malaria transmission and net protection. ITN integrity assessments and blood sampling were conducted concurrently. with ach survey round lasting approximately 15 days per district. Study population description The study was conducted in 30 villages (15 in Kéniéba and 15 in Kita) representing a mix of rural and semi-urban communities. According to 2017 administrative data from National Malaria Control Programme (NMCP) [19], the total population of these districts was approximately 294,273 inhabitants in Kéniéba and 516,751 in Kita. The population is composed predominantly of Bambara, Malinké, Khassonké and Fulani ethnic groups [18]. The study targeted children aged 6–60 months, living in households that received ITN during the 2017 national distribution campaign. Study nets and the populations cohort’s constitution In each district, a total of 15 villages were selected using a random-number generator in Microsoft Excel (Version 2016), based on official ITN distribution lists provided by the NMCP. Within each selected village, 10 households were randomly chosen for follow-up. All selected households had received ITNs during the 2017 national campaign. Sample size calculations were designed to detect a 20% difference in the proportion of serviceable nets between ITN types at 36 months (α = 0.05, power = 0.80). With 15 villages per district and 10 households per village (n = 300 households, ≈ 580 children), the study achieved adequate statistical power and operational feasibility. Children were eligible if aged 6–60 months, resided permanently in the household, and regularly slept under an ITN. Exclusion criteria included temporary visitors, children with chronic illnesses or febrile conditions at the time of sampling, and those whose caregivers did not provide consent. The same children were followed over the three annual survey rounds whenever possible. New eligible children present in the household at a given survey were also enrolled, without replacing those who had aged out, moved away, or were lost to follow-up. This approach ensured that the cohort represented the actual under-five population residing in the study households at each survey round [11]. Analyses at each time point, therefore, represent repeated cross-sectional assessments of all eligible children available during each survey year. Monitoring the physical durability of ITN Standardized questionnaires were administered electronically using tablets equipped with the Open Data Kit (ODK) platform. Data on household characteristics, net type, ITN ownership, hanging status, washing practices, and usage were collected and securely stored on the Laboratory of Applied Molecular Biology (LBMA) server in Bamako. The physical integrity of each net was assessed using the WHO-recommended Proportional Hole Index (pHI) with the four standard hole-size categories (0.5–2 cm, 2–10 cm, 10–25 cm, and > 25 cm) [20]. Nets were classified as being in good condition (pHI < 64), serviceable (pHI ≤ 642), or too torn (pHI > 642). These assessments were performed at 12-, 24-, and 36-month intervals post-distribution. The same trained field team performed all assessments to minimize inter-observer variability. Evaluation of human exposure to Anopheles bites using the gSG6-P1 Biomarker Capillary blood was collected by finger prick onto filter paper (Whatman® 903) to create dried blood spots (DBS), which were air-dried and stored at − 20 °C with desiccants until analysis. Specific IgG antibody responses to the Anopheles gambiae salivary peptide gSG6-P1 were quantified using an indirect ELISA protocol as described by Remoue et al. [21]. The gSG6-P1 peptide was synthesized in lyophilized form by Genepep® (France). ELISA assays were performed on NUNC® round (U) ELISA plates. Each sample was tested in duplicate, and a third well served as a blank control containing only antigen. A biotinylated anti-human IgG antibody conjugated to streptavidin-peroxidase was used as the secondary antibody. Each plate included pooled high, medium, and low positive controls to monitor inter-plate variability. No universal cut-off was defined; instead, ΔOD was calculated as the mean optical density of duplicate antigen wells minus the background from antigen-free wells. Intra-assay variability remained below 10%, confirming analytical reliability, as previously described by Traore et al. (2020) [10]. Optical density (OD) readings were expressed as:\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta {\text{OD}}\, = \,({\text{OD}}_{{1}} + \,{\text{OD}}_{{2}} )/{2}\, - \,{\text{OD}}_{{{\text{blank}}}} .$$\end{document}ΔOD=(OD1+OD2)/2-ODblank. No threshold for positivity was predefined. Instead, antibody levels were treated as continuous variables to assess relative exposure across time points, net types, and conditions. Statistical analysis All statistical analyses were conducted using Stata version 14 and GraphPad Prism version 10. Socio-demographic characteristics (sex distribution, age groups) and ITNs usage patterns (use the previous night, weekly frequency) were analysed as categorical variables and compared across sites and survey rounds using the chi-squared test. Missing demographic variables (e.g., age or sex) were recorded as ‘NA’ and retained in analyses when other relevant data were available, to minimize bias from casewise deletion. For the ITNs cohort, the primary variable analysed was the proportion of nets in serviceable condition (pHI ≤ 642). Mean proportions were calculated annually per site, and year-to-year differences were tested using chi-squared comparisons. For the children cohort, the main outcome was the median optical density (ΔOD) reflecting anti-gSG6-P1 IgG levels. Because ΔOD values were non-normally distributed, non-parametric tests were used: the Wilcoxon-Mann–Whitney test for two-group comparisons and the Kruskal–Wallis test for multiple-group comparisons. Children exposure (ΔOD) was not classified into fixed ITN ‘torn’ or ‘intact’ net categories. Instead, ΔOD medians were compared across survey years, sites, and ITN physical integrity under routine use. All tests were two-sided, and a p-value < 0.05 was considered statistically significant. Ethical considerations The study protocol was approved by the Ethics Committee of the National Institute of Research in Public Health (Institut National de Recherche en Santé Publique INRSP), Mali (Approval No. 02/2018/CE-INRSP). Written informed consent was obtained from ITNs owners and parents or legal guardians of all participating children. All individual-level data were anonymized to ensure confidentiality. Results Characteristics of the study population by village A total of 586 children under five years old were enrolled across 30 villages in two districts: Kéniéba (n = 225) and Kita (n = 361). The study design ensured comparability between the two ITNs types, Yorkool® (Kéniéba) and PermaNet® 2.0 (Kita), under similar environmental settings. Table 2 summarizes socio-demographic characteristics and ITNs ownership and usage patterns over the study period. The sex ratio remained balanced across all survey years in both sites, with no significant differences between males and females (p = 0.797 in Kéniéba; p = 0.874 in Kita). Table 1Technical specifications of ITN brand monitoredDistribution SiteBrandMaterialDenierShapeInsecticide contentBinding technologyKéniébaYorkool®Polyester75RectangularDeltamethrin 55 mg/m2Polymer-incorporatedKitaPermaNet® 2.0Polyester100RectangularDeltamethrin 55 mg/m2Surface-coated Table 1Technical specifications of ITN brand monitoredDistribution SiteBrandMaterialDenierShapeInsecticide contentBinding technologyKéniébaYorkool®Polyester75RectangularDeltamethrin 55 mg/m2Polymer-incorporatedKitaPermaNet® 2.0Polyester100RectangularDeltamethrin 55 mg/m2Surface-coated Technical specifications of ITN brand monitored However, the age distribution varied significantly (Chi-squared = 27.82, p = 0.001 in Kéniéba; Chi-squared = 24.22, p = 0.002 in Kita), with fewer children under 12 months and a growing proportion in the 36–60 months over time. This shift may have influenced exposure trends, as younger children can differ in sleeping arrangements and ITN adherence. ITNs ownership was 100% across all years in both sites. ITNs usage increased significantly over time (Kéniéba: Chi-squared = 9.36, p = 0.009; Kita: Chi-squared = 6.67, p = 0.035), with a higher proportion of children sleeping under an ITN every night. At 36 months, last-night use was very high in both study sites (Kéniéba: 76.3% to 95.6%; Kita: 95.7% to 100%; Table 2). In addition, the proportion of children using ITNs every night within a week also increased (Kéniéba: from 72.3% to 95.6%; Kita: from 96.6% to 99.2%) (Table 2). Despite these behavioural improvements, physical integrity declined steadily. In Kéniéba (Yorkool®), the proportion of nets in serviceable condition (pHI ≤ 642) dropped from 84.9% at 12 months to 48.8% at 36 months (at the end of the study). In Kita PermaNet® 2.0 , it decreased from 95.4% to 77.7% (Table 2). IgG levels to gSG6-P1 salivary peptide in children across study sites and years IgG antibodies levels against the gSG6-P1 peptide were measured to assess exposure to Anopheles mosquito bites across survey years (Figs. 2A, B). Significant temporal variation in ΔOD values was observed over time, and levels were consistently higher in Kéniéba than in Kita (Kruskal–Wallis: H = 154, p < 0.0001).Fig. 2Specific IgG (anti-gSG6-P1) levels in children by study site and survey year (ΔOD medians with point distributions)Table 2Socio-demographic characteristics of the study population by study siteVariableSite 1 KéniébaChi-squared (p-value)Site 2 KitaChi-squared (p-value)Yorkool®PermaNet® 2.012 months (N = 119)24 months (N = 61)36 months (N = 45)12 months (N = 117)24 months (N = 123)36 months (N = 121)SexFemale50 (42%)27 (44%)17 (38%)0.453 (0.797)61 (52%)62 (50%)65 (54%)0.268 (0.874)Male69 (58%)34 (56%)28 (62%)56 (48%)61 (50%)56 (46%)Age (in months) < 1220 (16.8%)7 (11.5%)2 (4.4%)27.82 (0.001)11 (9.4%)6 (4.9%)8 (6.6%)24.22 (0.002)12–2329 (24.4%)4 (6.6%)6 (13.6%)12 (10.3%)36 (29.2%)15 (12.3%)24–3519 (15.9%)17 (27.9%)7 (15.5%)25 (21.3%)23 (18.7%)30 (24.7%)36–4728 (23.5%)13 (21.3%)21 (46.6%)28 (23.9%)33 (26.8%)28 (23.1%)48–6020 (16.8%)20 (32.8%)8 (17.7%)40 (34.1%)23 (18.7%)38 (31.4%)NA3 (2.5%)01 (2.22%)1 (0.8%)2 (1.6%)2 (1.6%)ITN assesmentN = 245N = 144N = 45N = 305N = 147N = 99% in Serviceable condition (pHI ≤ 642)84.9%70.3%48.8%22.32 (< 0.0001)95.4%91.8%77.7%19.31 (< 0.0001)% in Good condition (pHI < 64)65.7%47.2%33.3%15.15 (0.0005)92.7%71.4%63.6%26.97 (< 0.0001)% Too torn (pHI > 642)15.1%29.1%51.1%22.32 (< 0.0001)4.5%8.1%22.2%19.31 (< 0.0001)ITN ownership (Parent) No0%0%0%0 (1)0%0%0%0 (1) Yes100%100%100%100%100%100% ITN use Last night No28 (23.5%)8 (13.1%)2 (4.4%)9.36 (0.009)5 (4.3%)7 (5.6%)0 (0%)6.67 (0.035) Yes90 (75.6%)52 (85.2%)43 (95.6%)112 (95.7%)114 (92.6%)120 (99.1%) NA1 (0.8%)1 (1.6%)002 (1.6%)1 (0.9%)# time in a week 0 (Never)25 (21%)6 (9.8%)2 (4.4%)12.89 (0.045)3 (2.6%)2 (1.6%)1 (0.8%)10.16 (0.118) 1–6 (Often)8 (6.7%)4 (6.5%)0 (0%)1 (0.8%)6 (4.9%)0 (0%) 7 (Every day)86 (72.3%)50 (81.9%)43 (95.6%)113 (96.6%)115 (93.5%)120 (99.2%) NA01 (1.6%)0000 Specific IgG (anti-gSG6-P1) levels in children by study site and survey year (ΔOD medians with point distributions) Table 2Socio-demographic characteristics of the study population by study siteVariableSite 1 KéniébaChi-squared (p-value)Site 2 KitaChi-squared (p-value)Yorkool®PermaNet® 2.012 months (N = 119)24 months (N = 61)36 months (N = 45)12 months (N = 117)24 months (N = 123)36 months (N = 121)SexFemale50 (42%)27 (44%)17 (38%)0.453 (0.797)61 (52%)62 (50%)65 (54%)0.268 (0.874)Male69 (58%)34 (56%)28 (62%)56 (48%)61 (50%)56 (46%)Age (in months) < 1220 (16.8%)7 (11.5%)2 (4.4%)27.82 (0.001)11 (9.4%)6 (4.9%)8 (6.6%)24.22 (0.002)12–2329 (24.4%)4 (6.6%)6 (13.6%)12 (10.3%)36 (29.2%)15 (12.3%)24–3519 (15.9%)17 (27.9%)7 (15.5%)25 (21.3%)23 (18.7%)30 (24.7%)36–4728 (23.5%)13 (21.3%)21 (46.6%)28 (23.9%)33 (26.8%)28 (23.1%)48–6020 (16.8%)20 (32.8%)8 (17.7%)40 (34.1%)23 (18.7%)38 (31.4%)NA3 (2.5%)01 (2.22%)1 (0.8%)2 (1.6%)2 (1.6%)ITN assesmentN = 245N = 144N = 45N = 305N = 147N = 99% in Serviceable condition (pHI ≤ 642)84.9%70.3%48.8%22.32 (< 0.0001)95.4%91.8%77.7%19.31 (< 0.0001)% in Good condition (pHI < 64)65.7%47.2%33.3%15.15 (0.0005)92.7%71.4%63.6%26.97 (< 0.0001)% Too torn (pHI > 642)15.1%29.1%51.1%22.32 (< 0.0001)4.5%8.1%22.2%19.31 (< 0.0001)ITN ownership (Parent) No0%0%0%0 (1)0%0%0%0 (1) Yes100%100%100%100%100%100% ITN use Last night No28 (23.5%)8 (13.1%)2 (4.4%)9.36 (0.009)5 (4.3%)7 (5.6%)0 (0%)6.67 (0.035) Yes90 (75.6%)52 (85.2%)43 (95.6%)112 (95.7%)114 (92.6%)120 (99.1%) NA1 (0.8%)1 (1.6%)002 (1.6%)1 (0.9%)# time in a week 0 (Never)25 (21%)6 (9.8%)2 (4.4%)12.89 (0.045)3 (2.6%)2 (1.6%)1 (0.8%)10.16 (0.118) 1–6 (Often)8 (6.7%)4 (6.5%)0 (0%)1 (0.8%)6 (4.9%)0 (0%) 7 (Every day)86 (72.3%)50 (81.9%)43 (95.6%)113 (96.6%)115 (93.5%)120 (99.2%) NA01 (1.6%)0000 Socio-demographic characteristics of the study population by study site In Kéniéba, median ΔOD increased from 0.097 at 12 months to 0.249 at 24 months, reaching 0.349 at 36 Month. In Kita, it increased from 0.112 at 12 months to 0.246 at 24 months, and then declined to 0.164 at 36 months. The increase between 12 months and 36 months was less steep in both sites, suggesting a possible behavioural or ecological buffering. Site comparisons revealed no significant difference in IgG levels between Kéniéba and Kita at 12 and 24 months (p > 0.05). However, at 36 months, the difference became statistically significant (Mann–Whitney U = 4927, p < 0.0001), with Kéniéba showing markedly higher antibody levels, indicating a possible faster decline in ITNs effectiveness (Fig. 2B). This suggests that ITNs effectiveness in reducing mosquito exposure may have declined more rapidly in Kéniéba, potentially due to differences in net physical condition (hole index). Association between anti-gSG6-P1 IgG levels and ITNs physical condition Figure 3 intentionally integrates these immunological data with ITN physical integrity indicators to illustrate the inverse relationship expected under operational conditions. Each panel shows median ΔOD values (scatter-plots, left axis) alongside the percentage of ITNs in serviceable condition (pHI ≤ 642, bars, right axis) at 12, 24, and 36-month.Fig. 3Integrated relationship between ITN physical integrity (pHI) and exposure to Anopheles bites (IgG anti-gSG6-P1), by study site and years. (A) Kéniéba (Yorkool®); (B) Kita (PermaNet® 2.0). Panels show ΔOD medians (left axis) and % serviceable ITNs (pHI ≤ 642; right axis) at 12, 24, and 36 months Integrated relationship between ITN physical integrity (pHI) and exposure to Anopheles bites (IgG anti-gSG6-P1), by study site and years. (A) Kéniéba (Yorkool®); (B) Kita (PermaNet® 2.0). Panels show ΔOD medians (left axis) and % serviceable ITNs (pHI ≤ 642; right axis) at 12, 24, and 36 months In both sites, increases in anti-gSG6-P1 levels were strongly associated with declines in the proportion of serviceable ITNs, confirming the link between ITNs durability and human-vector contact. This association was more pronounced in Kéniéba, where faster net deterioration corresponded with greater exposure. In Kéniéba (Yorkool®), when 85% of ITNs were serviceable at 12 months, IgG levels remained low (ΔOD = 0.097). As serviceability declined to 70.3% at 24 months, ΔOD increased significantly to 0.249 (p < 0.001). At 36 months, when only 48.8% of nets were still serviceable, IgG levels increased further to 0.349, although the increase between 24 and 34 months was not statistically significant (p = 0.08). In Kita, IgG levels followed a similar pattern. From ΔOD = 0.112 at 12 months (with 95.4% of nets serviceable), levels increased to 0.246 at 24 months (92% nets serviceable, p < 0.001), and to 0.164 by 36 months (77.7% nets serviceable, p = 0.0022). The slower decline in net condition in Kita aligns with the lower exposure levels observed in children. Figure 3 confirms that declining net integrity is closely associated with increased exposure to Anopheles bites, even when ITNs usage remains high. This underscores the critical role of net durability in sustaining malaria prevention impact. The opposing trends demonstrate the progressive increase in exposure to mosquito bites as ITN integrity declines. This figure intentionally integrates exposure dynamics (see Fig. 2) with ITN durability indicators to illustrate their inverse relationship under operational conditions. Discussion This longitudinal study assessed the protective effectiveness and physical durability of two types of ITNs under operational conditions in Mali, using anti-gSG6-P1 IgG responses as a biomarker of children’s exposure to Anopheles mosquito bites. The salivary peptide gSG6-P1 has become a valuable biomarker for evaluating human exposure to Anopheles bites, especially in the context of vector control strategies. Its operational relevance has been demonstrated across diverse settings and interventions. In Burkina Faso, Senegal, Angola and Vanuatu, it was used to assess ITNs and mixed vector control strategies [8, 11, 13, 22, 23], while in Tanzania, it revealed residual exposure in low mosquito density setting [24]. In Ghana, it contributed to evaluating malaria transmission dynamics [25]. In urban and peri-urban settings such as Cameroon, Senegal, Kenya, and Côte d’Ivoire, gSG6-P1 highlighted spatial variations in exposure linked to housing quality and vector behaviour [8, 10, 25–27]. These studies validate its sensitivity and suitability for integration in vector-control impact assessments. Despite high and increasing ITNs usage over three years, we observed a significant rise in children’s exposure to mosquito bites, strongly associated with declining net integrity. This trend was evident as both study sites, Kéniéba (Yorkool®) and Kita (PermaNet® 2.0), with a more rapid loss of protection in Kéniéba. Several studies have shown continuous exposure to mosquito bites despite high ITNs use [28–32]. In this study, ITNs performance varied by product type. Yorkool® nets in Kéniéba showed faster physical deterioration and were associated with higher IgG levels in children compared with PermaNet® 2.0 in Kita. Although both nets share polyester construction with deltamethrin at equivalent target doses, prior evaluations suggest PermaNet® 2.0 is more resistant to wear-and-tear under field conditions [33–35], whereas independent data on the long-term durability of Yorkool® remain limited and context-dependent. Previous durability-monitoring studies in several African countries also reported faster deterioration of certain polyethylene-based ITNs compared to polyester design [33–35]. These findings support ongoing comparative evaluations of ITNs brands in real-life settings [36]. At 36 months, only 49% of Yorkool® nets in Kéniéba were still in serviceable condition. In contrast, in Kita, where PermaNet® 2.0 was distributed, 78% serviceable. These findings align with reports showing that PermaNet® 2.0 demonstrates superior durability under operational conditions compared with some other ITNs brands, including Yorkool® [33, 34]. The poorer durability observed in Yorkool® compared to PermaNet® 2.0 likely reflects differences in net fabric mechanical resilience (polymer strength and weave density) rather than insecticidal formulation, as both products contain deltamethrin at equivalent target doses (55 mg/m2). Similar patterns have been reported in several studies [33, 34]. These findings emphasize that net fabric composition and mechanical resilience are critical determinants of ITN performance under operational conditions. This reinforces the importance of field-based product performance monitoring. The use of anti-gSG6-P1 IgG response as a salivary biomarker provided a sensitive, scalable, and individual-level indicator for recent mosquito exposure. This aligns with findings from Angola [11, 22], Senegal [9, 37], Kenya [25], Côte d’Ivoire [10], and Burkina Faso [38], which validated gSG6-P1 as a reliable indicator for evaluating vector control effectiveness where classical entomological tools are limited. In the present study, specific IgG levels clearly reflected site-specific differences in exposure. In Kéniéba, ΔOD increased from 12 to 36 months post-distribution, tracking the physical deterioration of Yorkool® nets. Similar patterns were reported in Benin, where rapid net degradation was associated with increased exposure and higher malaria incidence [12]. In Kita, children’s median ΔOD values were consistently lower compared to Kéniéba. Seasonal transmission may also have shaped specific IgG trends. In Mali, malaria peaks after the rainy season (July–October), and any mismatch between peak exposure and sampling could influence IgG levels. Variability in rainfall or larval habitat availability may have amplified or reduced vector contact across years. Similar trends linking declining ITN integrity with increased human exposure to mosquito bites have been reported in Benin [12]. These findings reaffirm that ITN effectiveness cannot be captured by ownership and usage indicators alone, as emphasized in multiple operational studies [4, 36, 39]. Importantly, the analysis did not compare specific IgG levels between predefined ‘torn’ versus ‘intact’ net users. Instead, it evaluated how progressive community-level decline in ITN condition influenced children’s exposure over time. This design reflects real-world programmatic monitoring, where the physical state of nets changes dynamically within households and survey rounds rather than remaining fixed at the individual level [11, 33, 40]. While self-reported ITNs usage remained high, exposure to Anopheles bites increased supporting growing evidence that physical deterioration is a primary driver of residual malaria transmission [6, 41]. The decline in anti-gSG6-P1 IgG levels observed in Kita in 2020, despite measurable deterioration in ITNs condition (serviceable: from 95 to 78%), was unexpected and highlights the complexity of exposure dynamics under operational conditions. This non-linear pattern diverged from the consistent increase observed in Kéniéba and challenges the assumption of a direct correlation between net degradation and mosquito exposure. Multiple factors may explain this pattern [2]. One plausible explanation is ecological variation, for example a reduction in Anopheles density following lower rainfall or environmental changes that disrupted larval habitats. Additionally, behavioural adaptations, such as improved ITNs use, net repair, housing improvement, indoor modifications, shifts in sleeping behaviour or complementary personal protection strategies may have contributed to reduced actual exposure, even in the presence of torn nets [42, 43]. This pattern reinforces the importance of interpreting gSG6-P1 data in accordance with seasonal transmission patterns, age-related behaviour, and micro-ecological variation, rather than assuming linear relationships between age and antibody level. Complementary entomological or meteorological data would be valuable for further exploration in future studies [11, 44]. This study findings support the integration of net condition assessments and immuno-epidemiological biomarkers into national malaria surveillance frameworks. While ITN ownership and usage remain essential indicators, they do not fully capture protection effectiveness at the population level. Unlike human landing catches (HLC) or CDC light trap collections, which provide point-in-time estimates and require intensive entomological expertise, the gSG6-P1 biomarker offers a temporally integrated, individual-level measure of human-vector contact. Anti-gSG6-P1 IgG responses typically rise within 10–14 days of exposure and decline rapidly when contact ceases, providing a sensitive marker of recent exposure [11, 44]. Its low-cost ELISA format and compatibility with dried blood spots make it suitable for large-scale surveillance and routine programmatic monitoring. Integrating serology can provide a cost-efficient, logistically lighter complement to standard ITN durability assessments. Samples can be processed in batches in national reference laboratories, whereas household integrity surveys require repeated field visits and trained personnel to classify each ITN according to WHO guidelines [20, 40]. Although serology does not replace direct physical assessments, combining both approaches can identify hidden protection gaps and guide timely replacement [13, 44]. Additionally, the observed disparities in product performance reinforce the need for evidence-based procurement. The relatively higher durability of PermaNet® 2.0 observed here is consistent with findings from Senegal, Benin, Zambia and Kenya [33, 34]. As countries transition toward dual-active ITNs, product-specific evaluations should guide procurement and replacement strategies to maximize both biological efficacy and physical lifespan. Finally, incorporating biological, physical, and behavioural indicators into ITNs evaluations can enable more timely and targeted interventions, reducing the risk of silent transmission resurgence due to unnoticed net failure. This study’s strengths include its three-year longitudinal design, the use of a validated immuno-epidemiological biomarker, and real-world operational implementation. Unlike traditional entomological methods, IgG responses to mosquito salivary peptide provide a direct, sensitive measure of human-vector contact at the individual level [44]. Comparing two distinct ITN types under similar ecological conditions revealed product-specific differences in durability and associated protection, supporting the integration of durability monitoring and serological evaluation into national ITN monitoring systems. In settings with limited entomological capacity, the biomarker may serve as an early-warning tool for protection failure, especially when self-reported use remains high but physical condition deteriorates. A key limitation is the absence of entomological data in the present analysis, restricting the ability to directly correlate biomarker levels with mosquito densities or biting behaviour. Inclusion of entomological and parasitological data (malaria infection prevalence in children) and insecticidal efficacy data (human biting rate, cone bioassays, chemical residue testing) would offer a more comprehensive understanding of the mechanisms underlying protection loss. These complementary datasets have been collected and are currently being analysed for inclusion in a companion manuscript, which will further contextualize the immunological findings presented here. Additionally, the effects of outdoor biting, changing vector behaviour, and housing structure on exposure were not directly measured. As other studies have shown, traditional ITNs offer limited protection against early evening or outdoor feeding vectors [32], and exposure is often higher in poorly constructed housing [45, 46]. Seasonal variation in vector density is another confounder in serological monitoring. Because sampling occurred primarily during the dry season, IgG levels may underestimate exposure during high-transmission periods. Minimal loss to follow-up occurred between survey rounds, though this may introduce limited bias. Although the same children were followed when possible, unbalanced follow-up and household mobility limited the feasibility of individual-level mixed-effects analyses. Consequently, the analysis used repeated cross-sectional comparisons at each time point, which may underestimate within-subject correlation. Future analyses incorporating the full entomological dataset and multilevel models will allow more robust adjustment for repeated measures and ecological covariates [33, 40]. Lastly, while the present biomarker quantifies exposure, it does not differentiate indoor from outdoor biting. In summary, this study highlights the value of the IgG response to gSG6-P1 as a sensitive, field-adapted tool for monitoring ITNs effectiveness and vector exposure over time. It highlights how ITNs type, physical durability, transmission seasonality, and child age interact to shape individual exposure risk. These findings provide operationally relevant evidence to inform net replacement strategies, ITNs product selection, and integrated malaria vector surveillance in West African settings. Conclusion This study provides evidence that ITNs effectiveness in protecting children under five from Anopheles mosquito bites declines progressively over time, primarily due to physical deterioration, despite high usage rates. Using the immuno-epidemiological biomarker gSG6-P1, the study demonstrated that children’s exposure to mosquito bites increased as ITNs aged and became damaged, with notable differences in performance between net types. These findings highlight the importance of monitoring not only ownership and use but also physical durability and biological protection under real-world operational conditions. As malaria programmes evolve toward next-generation vector control tools, incorporating immuno-epidemiological and operational indicators will be essential to sustain current gains and close residual transmission gaps. They also reinforce the need for evidence-based ITN procurement and timely replacement strategies that ensure both chemical and physical performance are maintained throughout the product lifespan. These results support the integration of biomarkers of human exposure such as anti-gSG6-P1 IgG, together with systematic net-condition assessments into NMCP monitoring frameworks to guide targeted interventions and resource allocation.