Bijlage 2 Ingekomen brieven i.r.t. coalitieakkoord - deel 2
Tekst van het document
Bijlage 2 Ingezonden brieven i.r.t coalitieakkoord Sociale Leefomgeving 1. Augeo Foundation: Oproep steun voor kinderen die thuis geweld meemaken 2. Caring Doctors, De Groene Huisarts, PAN Nederland, Groen Zorg Alliantie: Voedseladvies Gemeenten 2026 3. EEN tegen EENZAAMHEID: Oproep om de lokale aanpak op eenzaamheid te versterken 4. Divosa, het Juridisch Loket, Sociaal Werk Nederland, de Nederlandse Vereniging van Rechtswinkels, Nederlandse Schuldhulproute (Geldfit), Landelijke Organisatie Sociaal Raadslieden, Vereniging Humanitas, Vereniging Schuldhulpmaatje Nederland, Leger des Heils en Stichting Alliantie Vrijwillige Schuldhulp: Oproep om te investeren in sociaaljuridische hulp 5. GGD: Oproep om gezondheid een centrale plek te geven in het coalitieakkoord 6. De heer Slingenberg: Vergeet mensen met een beperking niet! 7. Smartphonevrij Opgroeien Nederland: Oproep om de maatschappelijke norm te verschuiven (later smartphonegebruik, bewust omgaan met social media). 8. VeiligheidNL: Oproep om verkeersveiligheid en valpreventie op te nemen in het akkoord 9. 113 Zelfmoordpreventie: Oproep om suïcidepreventie op te nemen in het coalitieakkoord. Fysieke Leefomgeving 10. Beraad Vlieghinder Moet Minder: Aandachtspunten met betrekking tot hinder van vliegverkeer 11. Eindhoven Airport: Brief over groeiwensen van de luchthaven 12.
Federatie Ruimtelijke Kwaliteit: Oproep om lokale kwaliteitsadvisering m.b.t. fysieke leefomgeving te versterken en aandachtspuntenlijst voor gemeenten bij het borgen van omgevingskwaliteit 13. Fietsersbond: Oproep om fietsveiligheid te verbeteren 14. Gedeputeerde Stijn Smeulders: Oproep Samen maken we Brabant verkeersveilig 15. HomeFlex: Position Paper over Flexwonen 16. Knarrenhof Boxtel: Oproep tot het faciliteren van minimaal één Knarrenhof- project in elke kern 17. Minister van VRO: Oproep tot gezamenlijke aanpak woningmarktproblematiek met het belang van de woningzoekende voorop 18. Podium Boxtel: Verzoek om aan verkenning naar nieuwe locatie ook Het Rabobankpand mee te nemen 19. Regionale Samenwerking Wonen: Wensenlijst woningcorporaties 20. Seniorenvereniging Boxtel: Pleidooi voor een toegankelijke en veilige leefomgeving en langer zelfstandig thuis wonen 21. The lab of Thought: Oproep om meer ruimte te maken voor verhalen uit de gemeenschap over mobiliteit en de straat 22. Veilig Verkeer Nederland: 3 keuzes voor veiliger verkeer in uw gemeente Programma Duurzaam 23. Energieregio Noordoost-Brabant: Overdrachtsdocument voor de nieuwe bestuursperiode 24. De heer Slof: Aandacht gevraagd voor wetenschappelijk onderzoek naar de relatie tussen blootstelling aan bestrijdingsmiddelen en gezondheidsrisico’s 25.
Waterschap de Dommel: brief met onderwerpen op het gebied van water om op te nemen in het bestuursakkoord 26. De heer Van der Zee: oproep op een strategisch woningbouwbeleid met groen als onderlegger te ontwikkelen Programma Economie, Cultuur en Sport 27. ANWB en Wandelnet: oproep om te investeren in lopen en wandelen in de gemeente 28. Huis 73: Oproep om te blijven investeren in een sterke bibliotheekvoorziening 29. MDL-fonds: oproep tot het plaatsen van meer openbare toiletten 30. Sportsector: Oproep om sport en bewegen een stevige plek te geven in het coalitieakkoord 31. Stichting ERM: Oproep om gemeentelijk erfgoed beter te behouden 32. Stichting Kennisalliantie Bedrijventerreinen Nederland: Oproep ontwikkeling bedrijventerrein en andere werklocaties 33. VNONCW en MKB Nederland: 11-puntenplan Voor vitale stads- en dorpscentra Programma Bestuur en organisatie 34. Commissaris van de Koning in de provincie Noord-Brabant: brief om aandacht te vragen voor het thema weerbaarheid en veerkracht op lokaal niveau 35. De heer Jonkers: oproep om een keus te maken tussen bestuurlijk samengaan of zelfstandig blijven zonder ambtelijke fusie 36. Open State Foundation: Oproep invoering lobbyregister 37. Tienskip: Jongeren en de lokale democratie 38.
Voor Ons Nederland, Justice for Prosperity, Stichting Weerbare Democratie en Bits of Freedom: Oproep om te kiezen voor digitale autonomie Ons verhaal: Ruimte voor energie - 1 • Overdrachtsdocument voor de nieuwe bestuursperiode • Ons verhaal ‘Ruimte voor energie’ Energieregio Noordoost-Brabant Ons verhaal: Ruimte voor energie - 2 Energieregio Noordoost-Brabant (NOB) is een regionaal samenwerkingsverband van tien gemeenten, Provincie Noord-Brabant, waterschappen Aa en Maas en De Dommel, netbeheerder Enexis en maatschappelijke partners. Samen werken we aan de uitvoering van de Regionale Energiestrategie (RES 1.0) en aan de versnelling van de energietransitie. Via de Energietafel wordt bestuurlijk vormgegeven aan gezamenlijke afspraken en regionale prioriteiten. Onze samenwerking kenmerkt zich door een enthousiaste en oplossingsgerichte werkwijze, met ruimte voor lokale verschillen. Per 1 januari 2026 is RES Noordoost-Brabant verdergegaan onder de naam Energieregio Noordoost-Brabant, met een verbrede samenwerking rond warmte, het energiesysteem en voortzetting van de RES-opgave. Lees meer over de naamswijziging Ruimte voor energie Sinds 2020 werkt Energieregio NOB aan de energietransitie richting een klimaatneutraal Nederland in 2050.
In de verhaallijn Ruimte voor energie staat onze visie op het Energiesysteem van de Toekomst centraal, waarin lokale opwek, gebruik en opslag van energie en integrale afstemming van vraag, aanbod en infrastructuur nodig zijn om knelpunten zoals netcongestie te verminderen. Stand van zaken In 2021 hebben de gemeenten binnen onze regio afgesproken om gezamenlijk een duurzame opwekopgave te realiseren van 1,6 TWh (bod RES 1.0). Die doelstelling is nog niet gerealiseerd. Hoewel veel zon- en windprojecten in een volgende fase zijn gekomen, is realisatie vóór 2030 niet haalbaar zonder extra besluitvorming en oplossingen. Lees de voortgangsrapportage, juli ‘25 In 2025 hebben we een energiesysteemverkenning uitgevoerd om inzicht te krijgen in mogelijke ontwikkelingen van het regionale energiesysteem. Het rapport laat zien dat de energietransitie in onze regio vraagt om een fundamentele herinrichting van het energiesysteem met elektriciteit als belangrijkste drager én met grote gevolgen voor het elektriciteitsnet. De verkenning biedt inzicht in toekomstige keuzes en knelpunten en vormt de basis voor vervolgbesluiten over de ontwikkeling van het Energiesysteem van de Toekomst. Lees het rapport Energiesysteemverkenning Noordoost Brabant We werken ook samen aan energiebesparing en duurzame warmte. In RES 1.0 is het doel van 11 % energiebesparing in 2030 ten opzichte van 2017 vastgelegd.
De actuele cijfers laten inmiddels een besparing van ongeveer 17,2 % zien, grotendeels dankzij externe factoren zoals gestegen energieprijzen en (tijdelijk) gedragseffecten. Energiebesparing blijft echter een belangrijk aandachtspunt gezien de stijgende vraag naar elektriciteit en de opgave van de warmtetransitie. Om inwoners te ondersteunen bij woningverduurzaming, werken we samen in een Meerjarig Collectief Ontzorgingstraject (MCO). Lees meer over MCO De uitvoering van duurzame warmte in de regio is gestart, maar de transitie moet de komende jaren verder worden opgeschaald om de bijgestelde warmteambities richting 2030 te realiseren. Alle gemeenten werken aan een nieuw Warmteprogramma dat het college van B&W uiterlijk in 2027 vaststelt. Het programma schetst de warmteaanpak voor de komende jaren. Het onderscheidt gebieden met hun toekomstige (verwachte) warmtebron. Als uw gemeente in het Warmteprogramma gebruik maakt van de aanwijzingsbevoegdheid zal de gemeenteraad gevraagd worden dit in het Omgevingsplan vast te leggen. OVERDRACHTSDOCUMENT Overdrachtsdocument voor de nieuwe bestuursperiode - 2 Ons verhaal: Ruimte voor energie - 3 Actuele ontwikkelingen Zienswijze op Ontwerp-Nota Ruimte Energieregio NOB heeft samen met regionale partners een zienswijze ingediend op de Ontwerp-Nota Ruimte, waarin energie als essentieel onderdeel van ruimtelijke keuzes wordt benadrukt.
In de zienswijze zijn diverse zaken opgenomen met betrekking tot energie. Verbreding naar andere beleidsvelden In meerdere bijeenkomsten voor bestuurders en raadsleden hebben we benadrukt dat de energietransitie een bredere aanpak vraagt. Warmte, netcongestie en het Energiesysteem van de Toekomst maken dat energie nauw verbonden is met wonen, economie en ruimtelijke ontwikkeling. Lees verder Tijdens de eerste editie van de Energiedag zette Energieregio NOB een stap in die verbreding. Lees hier het verslag Energiesysteem van de Toekomst De Energietafel heeft bestuurlijk groen licht gegeven voor de verdere uitwerking van het Energiesysteem van de Toekomst. Hiermee start de volgende fase waarin de uitkomsten van de energiesysteemverkenning worden vertaald naar concrete scenario’s, keuzes en handelingsperspectieven voor de regio. Lees meer Op 15 april verzorgen wij een deelsessie over het Energiesysteem van de Toekomst tijdens de regionale raadsbijeenkomst in het kader van het inwerkprogramma voor de nieuwe gemeenteraden. Op uw verzoek verzorgen we een presentatie tijdens een beeldvormende raadsbijeenkomst. Aandachtspunten voor de komende periode Voor de nieuwe bestuursperiode is de energietransitie een urgente opgave.
De komende jaren zijn bepalend voor het realiseren van de regionale afspraken én voor het voorkomen dat energie een beperkende factor wordt voor wonen, werken en economische ontwikkeling. Graag geven wij u een aantal aandachtspunten mee: Tijdige realisatie van de opwekopgave uit RES 1.0, waarbij nieuwe bestuurlijke afspraken in aanloop naar 2030 worden meegenomen richting gemeenteraden en lokaal worden verankerd. Actieve samenwerking met netbeheerders voor verzwaring van het elektriciteitsnet en afstemming buurtvolgorde met wijkuitvoeringsplannen in Warmteprogramma’s. Borgen van voldoende capaciteit en middelen voor energie als ordenend principe voor plannen en gebiedsontwikkelingen. Zorg voor een integrale aanpak in de organisatie naar het Energiesysteem van de Toekomst. Betrek lokale initiatieven en energiecoöperaties actief bij uitvoering en draag- vlak. En verbindt de energietransitie met het sociaal domein, met aandacht voor betaalbaarheid en energiearmoede. Netcongestie vraagt om een andere aanpak en soms om onorthodoxe maatregelen. Als Energieregio Noordoost-Brabant willen we hierin een actieve rol vervullen door energie integraal te verbinden met ruimtelijke ontwikkeling, wonen en economie. En door vraag en aanbod lokaal beter op elkaar af te stemmen.
Daarbij zetten we in op energiegemeenschappen, lokale energieconcepten en oplossingen die bijdragen aan een betrouwbaar en betaalbaar energiesysteem. Overdrachtsdocument voor de nieuwe bestuursperiode - 3 Ons verhaal: Ruimte voor energie - 4 Sinds 2020 werken we als Energieregio Noordoost-Brabant aan de energietransitie op weg naar een klimaatneutraal Nederland in 2050. Waar de focus voorheen alleen lag op energiebesparing en het duurzaam opwekken van energie, zijn er inmiddels flinke stappen gezet in de warmtetransitie. Gemeenten werken volop aan warmteprogramma’s en wijkaanpakken. De grote druk op het elektriciteitsnet vanwege de groeiende vraag en productie van duurzame energie, zorgt ervoor dat onze aandacht steeds meer verschuift naar het energiesysteem als geheel. Oftewel naar de ontwikkeling van het Energiesysteem van de Toekomst. In ‘Ruimte voor energie’ beschrijven we de beweging die we in Noordoost-Brabant voor ons zien en samen willen maken. Ruimte voor energie De beschikbare netcapaciteit in Nederland, en dus ook in Noordoost-Brabant, staat onder grote druk. Dat merken we bijvoorbeeld aan woningbouwplannen die vertraging oplopen doordat stroomaansluitingen niet (direct) beschikbaar zijn. Ook staan steeds meer bedrijven in de wacht voor een stroomaansluiting en transportcapaciteit.
Ze willen uitbreiden en meer werkgelegenheid creëren, maar zijn pas na 2033 of later bij de netbeheerder aan de beurt voor een aansluiting. Het kabinet streeft naar een CO2 vrij elektriciteitssysteem in 2040. Een ambitieus doel, maar de cijfers zijn confronterend: de vraag naar elektriciteit wordt drie keer zo groot als nu en de productie ruim vier keer zo groot. Dit vraagt nu en in de toekomst veel van onze landelijke en regionale elektriciteit-infrastructuur. Netcongestie is dan ook geen tijdelijk probleem. Netcongestie is ‘here to stay’. En dat raakt ons allemaal. Hoe zorgen we ervoor dat we wél voldoende duurzame elektriciteit hebben? Zodat bedrijven kunnen ondernemen en we economische kansen benutten. Dat we woningen op het net kunnen aansluiten en onze woningbouwopgave kunnen realiseren. Dat we bestaande woonwijken kunnen transformeren in aardgasvrije wijken. En, waarbij we tegelijkertijd oog hebben voor een gezonde en prettige leefomgeving? Dat kan door andere aanpakken te hanteren dan we nu doen. En door anders te kijken naar ons energiesysteem op weg naar een klimaatneutraal Nederland in 2050. Een andere benadering: van centraal naar decentraal De oplossing ligt niet in eindeloos wachten op uitbreiding van het centrale elektriciteitsnet. Netcongestie dwingt ons tot een andere benadering van ons energiesysteem.
Het dwingt ons tot een integrale gebiedsaanpak waarin ruimtelijke ontwikkelingen, wonen, economie, mobiliteit én energie hand in hand gaan. En het dwingt ons tot het creëren van een decentraal systeem. Daarin komen vraag en aanbod van energie geografisch en in tijd, dichter bij elkaar in de eigen regio, op bedrijventerreinen en woonwijken. Het vraagt om een energiesysteem waarin we de samenhangende onderdelen – energievraag, energieaanbod en energie-infrastructuur – anders organiseren. ONS VERHAAL Netcongestie dwingt ons tot een integrale gebiedsaanpak waarin ruimtelijke ontwikkelingen, wonen, economie, mobiliteit én energie hand in hand gaan. Ruimte voor energie Ons verhaal: Ruimte voor energie - 5 Minister Hermans benadrukt in haar Kamerbrief van juni 2025 de noodzaak om vraag en aanbod in balans te brengen. Een uitdaging die we in Noordoost-Brabant nog volledig voor ons hebben. Maar hierin ligt ook onze kans. Door nú de regie te pakken, kunnen we efficiënter gebruikmaken van het energiesysteem én het systeem anders inrichten. Integrale gebiedsaanpak In de transitie naar het Energiesysteem van de Toekomst, moeten ruimte, wonen en economie samenkomen. Bestuurders, ondernemers en belangenorganisaties kunnen de gewenste economische groei niet meer los zien van energievraag- stukken.
Ambtenaren ruimtelijke ordening en projectontwikkelaars moeten vroegtijdig energiebehoeften meenemen in plannen voor gebiedsont- wikkeling. En nieuwbouwplannen en aardgasvrije wijken vragen om energiegemeenschappen, groepscontracten en publiek-private samenwerking. Kortom, we moeten samenwerken over de grenzen van verschillende sectoren heen. Een bedrijventerrein wordt niet alleen een plek voor economische activiteit, maar ook een energielandschap waar ondernemers energie produceren, opslaan, delen en verbruiken. Nieuwe woonwijken zijn in de toekomst gebieden waar bewoners in energiegemeenschappen samenwerken aan het opwekken, opslaan en delen van energie. Het energiesysteem, inclusief de benodigde infrastructuur, wordt een integraal onderdeel van gebiedsontwikkeling. Laten we samen ruimte maken voor energie in Noordoost-Brabant, zodat we in de toekomst prettig en gezond blijven wonen, werken en leven in onze mooie regio. Ruimte voor energie We staan voor een fundamentele keuze. We kunnen blijven wachten op uitbreiding van het centrale elektriciteitsnet en accepteren dat economische ontwikkeling en woningbouw achterblijven. Óf we pakken zelf regie. We bouwen aan een decentraal energiesysteem waarin vraag en aanbod bij elkaar liggen zodat we de bestaande en nieuwe energie-infrastructuren beter benutten.
Regie pakken vraagt moed, samenwerking en het loslaten van oude denkpatronen, niet alleen bij (decentrale) overheden, maar ook bij onze ondernemers, inwoners en maatschappelijke organisaties. Het bouwen aan een decentraal energiesysteem is een ruimtelijk, economisch én maatschappelijk vraagstuk. Het gaat om het creëren van ruimte, letterlijk en figuurlijk. Letterlijk, door slimmer om te gaan met de schaarse ruimte die we hebben en energie op te wekken waar deze wordt gebruikt. En figuurlijk, door ruimte te maken voor nieuwe denkwijzen, voor samenwerking over grenzen heen en voor lokale oplossingen. Vernieuwende samenwerking in Noordoost-Brabant Hoe doen we dat in de praktijk? Als tien gemeenten, provincie Noord- Brabant, waterschappen Aa en Maas en De Dommel, netbeheerder Enexis en maatschappelijke partners werken we al geruime tijd samen aan de energietransitie. Binnen de Energietafel maken we verdere stappen naar vernieuwende samenwerkingsvormen. Hierin staan co-creatie en innovatie centraal en willen we collega’s vanuit andere beleidsvelden nauw betrekken. Hierbij hebben we commitment aan elkaar én aan ons gezamenlijke doel: een decentraal toekomstbestendig energiesysteem. Laten we samen ruimte maken voor energie in Noordoost-Brabant, zodat we in de toekomst prettig en gezond blijven wonen, werken en leven in onze mooie regio.
https://doi.org/10.1038/s44360-026-00087-0 Mapping pesticide mixtures to cancer risk at the country scale with spatial exposomics Received: 21 August 2025 | Jorge Honles | 1,2, Juan Pablo Cerapio | 2,3, Claudia Monge | 2,4,5, Accepted: 17 February 2026 | Agnès Marchio | 2,4,5, Eloy Ruiz 2,6, Ramiro Fernández | 2,6, Sandro Casavilca-Zambrano | 2,7,8, Juan Contreras-Mancilla | 2,7,8, Published online: xx xx xxxx | Tatiana Vidaurre | 2,9,10, Thomas Condom | 11, Swann Zerathe | 12, Olivier Dangles | 13,14, Éric Deharo | 2,15, Javier Herrera-Zuñiga | 2,16, Check for updates | Pascal Pineau | 2,4,5 & Stéphane Bertani | 1,2 Despite decades of concern over the carcinogenic potential of agricultural pesticides, toxicological studies relying on single endpoints have yet to establish a definitive link between environmental pesticide exposure and cancer in real-world contexts. Here we use an integrative spatial Bayesian framework that merges high-resolution environmental pesticide risk modelling with comprehensive cancer registry data to map pesticide-linked cancer clusters in Peru with unprecedented precision.
Our process-based model, encompassing 31 key pesticide active ingredients, together with an innovative stratification of cancer cases by developmental lineage, reveals a robust spatial association between environmental pesticide exposure risk and cancer incidence. In pesticide-associated cancer hotspots, exposomic profiling of liver tissue—a primary target of chemical carcinogens—uncovers a distinct transcriptomic signature of pesticide exposure, implicating a non- genotoxic mode of action that disrupts core regulatory circuitries sustaining cell identity. Collectively, these findings strongly support a mechanistic link between pesticide exposure and cancer, challenging assumptions of human non-carcinogenicity derived from reductionist experimental models. This study redefines the exposome as a lineage-conditioned, mechanistically tractable framework and shows how complex pesticide mixtures can contribute to carcinogenic trajectories, with profound and far-reaching implications for global health policy and socio-ecological equity. Pesticides pose inherent risks to human health through multifaceted | tend to oversimplify real-world dynamics6,7. Moreover, the tumour- mechanisms1.
These risks arise from the effects of complex mixtures of | inducing potential of primarily non-genotoxic agents differs markedly active ingredients (AIs) and their degradation products, compounded | between rodent models and humans8. Although studies of mixtures by interactive effects, non-linear dose–response relationships and | yield valuable insights into the health risks from multi-agent exposures, variable exposure during the lifespan2–4. The pervasive presence of | the interplay between individual life histories and socio-environmental pesticides in food, water and ecosystems renders the characterization | context further complicates risk assessment4,6,7. Consequently, the of environmental exposures—a key component of the exposome—a | real-world carcinogenicity of pesticides remains insufficiently under- daunting challenge4. Observational studies often fall short in capturing | stood, hindering rigorous risk assessments and delaying effective pub- the complexity of pesticide exposures5, whereas experimental models | lic health interventions. Meeting these challenges requires heuristic A full list of affiliations appears at the end of the paper. e-mail: stephane.bertani@ird.fr Nature Health models with enhanced detection and predictive capabilities to identify | resolution and predictive precision20 (Supplementary Fig. 1).
Off- cancers attributable to pesticide exposure7. | site contamination driven by long-range transport extended up to To address this critical gap, we developed an advanced geospatial | 30–50 km beyond cultivated land (≈1,600 km2). The highest envi- modelling approach that offers a transferable blueprint for global | ronmental pesticide exposure risks were concentrated in the Andean environmental health assessments. By integrating high-resolution, | highlands and slopes, especially along the western flank and south- process-based environmental models with Bayesian inference that cap- | ern coastal areas, where limited precipitation exacerbates pesticide tures population-level heterogeneity in pesticide–cancer risk relation- | accumulation. Conversely, the Amazon basin and northern coast ships, our approach maps exposure risk at a fine spatial scale and links | consistently exhibited lower apparent risk levels (Fig. 1 and Extended statistical associations to mechanistic pathways of carcinogenesis. | Data Fig. 2).
Peru, a cradle of plant domestication9, combines a rich agricultural | To validate real-world model performance, we analysed hair sam- heritage with acute socio-ecological pressures that push against plane- | ples from 50 individuals residing in environmentally distinct pesti- tary boundaries10,11, rendering it a compelling setting for modelling pes- | cide risk zones19. Biomonitored levels of contaminating AIs and their ticide-related cancer risk. Its diverse geography—from Pacific coastal | degradation products exhibited significant spatial autocorrelation deserts to Andean highlands and Amazonian rainforests—shapes het- | (global bivariate Moran’s I = 0.42; z-score = 1.49; P = 0.044), closely erogeneous agricultural practices and pesticide use across a rapidly | aligning with modelled exposure risk estimates (Extended Data Fig. 3 modernizing sector. Environmental monitoring has revealed alarming | and Supplementary Table 1). Model sensitivity to climate variability levels of contamination12, while the country’s epidemiological transi- | was assessed by comparing two contrasting El Niño–Southern Oscil- tion has brought cancer to the forefront of public health priorities13,14 . | lation (ENSO) phases.
Predicted risk significantly increased during Stark spatial variability in cancer incidence, reflecting socio-ethnic | the 2015 El Niño episode compared with neutral conditions in 2019 disparities intertwined with territorial dynamics, raises the possibility | (P = 2.2 × 10−16), particularly in the Andes and northern coast (Extended of environmental factors as potential drivers13,14. Moreover, distinctive | Data Fig. 4). This preliminary finding suggests that climate events may molecular cancer patterns in Andean–Amazonian regions15–18 , where | heighten local exposure risk by altering pesticide use patterns21 and, Indigenous and peasant communities bear disproportionate pesticide | in turn, the transport and partitioning dynamics that govern their exposure burdens19, further underscore the imperative to elucidate the | environmental distribution. carcinogenic consequences of pesticide hazards.
Spatial cancer distribution by cell lineage ontogeny Results | Building on the pesticide risk model, we mapped the spatial distribu- Process-based environmental pesticide risk mapping | tion of cancer risk across Peru using data from the Peruvian National We developed and validated a high-resolution environmental model to | Cancer Institute (INEN) registry—the country’s most comprehensive assess pesticide risk across Peru and to forecast regions with elevated | source of cancer records13,14—for the years 2007 to 2020. Diagnoses exposure. Based on pesticide transport and degradation principles, | were validated by expert pathologists and classified according to the the model computes—on a national scale—the environmental fate | International Classification of Diseases, 10th revision (ICD-10). Key vari- of the 31 most commonly used pesticides in the country (Fig. 1 | and | ables—including birth date, sex, residential address at diagnosis, diag- Extended Data Table 1). To mirror real-world applications, data on AIs | nosis date and ICD-10 code—were curated to exclude recurrent cases, from regulatory sources were cross-validated against field surveys of | duplicates and inconsistencies. Residential histories were verified 650 agrochemical retailers (Extended Data Fig. 1).
Notably, none of the | against the 2007 and 2017 national censuses to ensure a minimum of pesticides included in the model are classified as group 1 carcinogens | 5-year residency before diagnosis, yielding a dataset of 158,072 primary by the International Agency for Research on Cancer (IARC). | cancer cases (C00–C96). Cases were then geocoded using a self-hosted To map environmental pesticide exposure risk, we integrated spa- | application programming interface (API) and assigned to districts to tial data on soil properties and monthly resolved hydrometeorological | match the spatial resolution of the environmental pesticide risk model. variables (2014–2019) to approximate pesticide transport, degrada- | Recognizing that cancer development is shaped by developmental tion and dispersion from application sites to downstream deposition | lineage programmes22, | 23, we stratified the dataset using a molecular zones (Fig. 1 | ). A hierarchical framework aggregated simulation outputs | framework based on histogenesis and germ layer origin, as proposed across all 31 AIs onto a grid with 100 m × 100 m resolution, generating | by Berman24 (Fig. 2a). This strategy moves beyond organ-based classi | - a normalized risk score for each grid cell (scale 0–100).
District-level | fications by grouping cancers according to cell ontogeny, enabling the calibration, fine-tuned using 2018 cultivated land cover data, enhanced | identification of hallmarks shared among tumours arising from organs predictive accuracy by aligning model outputs with local agricultural | with a common developmental lineage. Although often overlooked activity (Extended Data Fig. 2). The model thus captures cumulative, | in cancer classification, lineage dependency has profound implica- long-term risk from pesticide mixtures by concurrently estimating the | tions for environmental epidemiology. At the molecular level, lineage environmental behaviour of all 31 AIs, thereby characterizing tempo- | dependency mechanisms are well characterized, with lineage-specific rally stable exposure risk surfaces that reflect persistent contamination | master transcription factors (MTFs) orchestrating the autoregulatory regimes rather than short-term variability. | core regulatory circuitries (CRCs) that define cell identity in normal By integrating a process-based framework with empirical data, | tissues25 while sustaining cancer cell fitness in a lineage-dependent our model reconstructs plausible pesticide exposure scenarios at the | manner26–28.
Modulated by extrinsic cues and nuclear hormone recep- district scale (median area = 207.4 km2), capturing the spatial footprint | tors28, these lineage-specific MTFs constitute systemic vulnerabilities of routine human–environment interactions. To our knowledge, no | that can drive cancer cell transformation in response to environmental other system combines national coverage, high spatial resolution | stressors29,30. This rationale underpinned our adoption of lineage- and multi-year temporal depth to model chronic exposure risk to a | based stratification in cancer cluster analysis, offering a biologically comprehensive panel of key pesticides, making it uniquely suited for | grounded lens through which to interrogate environmental drivers spatial epidemiology in Peru. | of tumorigenesis24. Nationwide, the model mapped risk levels for 1,793 of 1,874 dis- | Within this framework, cancers were stratified into six main cat- tricts (95.7%; 1,245,451.8 km2), demonstrating its broad geographic | egories (Fig. 2a). Endoderm- and ectoderm-derived tumours were clas | - coverage despite minor data gaps (Fig. 1 | and Extended Data Fig. 2). sified as either surface (n | = 57,890; 36.6%) or parenchymal (n | = 32,781; Zones of moderate and high risk encompassed more than one-third | 20.7%).
Mesoderm-derived cancers were subdivided into non-mes- of the national territory, surpassing reference models in both spatial | enchymal (n | = 31,881; 20.2%) and mesenchymal (n | = 21,518; 13.6%) Nature Health Nature Health Article https://doi.org/10.1038/s44360-026-00087-0 2014 L%ro = (Q/P) × f × exp(-3×ln2/DT50Soil) × 100/(1 + Kd) Environmental pesticide risk score 100 0 DT50Soil Kd Koc 300 km N Cultivated land cover Soil organic carbon (SOC) Plant interception Pesticide physicochemical properties Terrain slope Precipitation Surface run-off Pesticide application Process-based modelling incorporating 31 pesticides Nationwide 100 m × 100 m resolution grid Topography-driven routing of cumulative overland pesticide flow between grid cells 2019 2018 Fig. 1 | High-resolution environmental model of pesticide risk across Peru, designed to support spatial cancer epidemiology. This process-based model estimates cumulative pesticide risk across spatial domains and over a 6-year period, both of which are pertinent to chronic human exposure. It integrates spatial data on soil properties and hydrometeorological variables—updated monthly from 2014 to 2019—into a transport framework grounded in surface pesticide mobility and degradation processes, formally represented by an equation described in Methods (see also Supplementary Methods).
This framework estimates the environmental fate and dispersal of the 31 most commonly applied pesticides in Peru (Extended Data Table 1). Overland pesticide flow is simulated by routing surface run-off from cell to cell along topography- driven pathways, modelling transport from handheld spraying sites typical of Peruvian agriculture to downstream deposition zones. Outputs across all 31 AIs are aggregated onto a 100 m × 100 m grid to compute a normalized cumulative risk score per cell (scale 0–100), accounting for pesticide mixtures. The map shows the geographical distribution of this score, colour-scaled in red from light (lower risk) to dark (higher risk), highlighting zones of potentially elevated environmental pesticide exposure risk across Peru. Model accuracy is enhanced through calibration using district-level cultivated land cover data (as in Extended Data Fig. 2).
a | b Nervous system (C47, C70–C72) | Respiratory tract (C30–C34) | 300 km Cutaneous melanoma (C43) | Oropharyngeal tract | N (C00–C06, C09–C14) Salivary glands | Thyroid (C73) (C07–C08) | Lung (C34) Thymus (C37) | Gastrointestinal tract (C15–C21, C26) Breast (C50) | Biliary tract (C23–C24) Hepatopancreatic | Non-melanoma of skin system (C22, C25) | (C44) Kidney (C64) Prostate (C61) Bladder (C67) Haematological cancers | Uterus (C53–C55) (C81–C96) | Reproductive system (C51–C52, C60) Musculoskeletal system | Developmental lineage: (C40–C41, C49) | Endoderm/ectoderm parenchymalEndoderm/ectoderm surface Mesenchyme Non-mesenchymal mesoderm Neuroectoderm, neural plate Primitive cells d | Global Moran’s I = 0.37; z-score= 7.53; P = 2.5 × 10 | –14 20 km RR = 1.85; CI: 1.39−2.42 5 RR = 2.30; CI: 1.83–2.84 c | P = 0.165 RR = 2.44; CI: 2.02–2.93 | Geographical location code0 | 0 8 4 70 | Pesticide risk score 0 0 | 1 | 2 | 3 | 4 | 5 | n = 52 | n = 156 | n = 64 | n = 107 | n = 40 | n = 17 Fig. 2 | Stratification by cell lineage ontogeny reveals spatial cancer clustering | risk as a function of estimated pesticide exposure within the same area. associated with pesticide exposure risk.
| a, Developmental mapping of | Overlaid is a kernel density estimate heatmap of liver cancer cases, illustrating major cancer types across anatomical systems24, with ICD-10 codes indicating | the spatial alignment of the RR maximum with a geostatistical cluster—confirmed corresponding tumour sites. b | , Map of Peru showing pesticide-associated cancer | by Moran’s I | statistic (global Moran’s I | = 0.37; z-score = 7.53; P = 2.5 × 10−14). hotspots (RR > 1), stratified by developmental lineage. RRs were estimated | Locations are slightly jittered to preserve privacy. Colour code (a | –d) and number using a Bayesian geostatistical model fitted with INLA (Supplementary Fig. 2). | of pesticide-associated cancer hotspots (b | and c | ): endoderm/ectoderm-derived The inset highlights a hotspot in the Chanchamayo Province ( Junín region). | parenchyma, red (n | = 52); endoderm/ectoderm-derived surface, violet (n | = 156); c, Violin plots showing the distribution of RR estimates across lineages. P | value | mesenchyme, green (n | = 64); non-mesenchymal mesoderm, yellow (n | = 107); is from a Kruskal–Wallis test. d | , Left: a spatial overlay identifying a multi-lineage | neuroectoderm (neural plate), blue (n | = 40); primitive cells, grey (n | = 17). hotspot of pesticide-associated cancers in the Chanchamayo Province.
RRs are | Illustration in a | created in BioRender; Bertani, S. https://biorender.com/jo8oap5 reported with 95% CIs. Right: a three-dimensional Bayesian posterior mean RR | (2026). surface for liver cancer (C22.0), depicting spatial variation in modelled cancer types. Neuroectodermal tumours, originating from the neural plate, | power—extraembryonic trophoblastic tumours (C58; n | = 318; 0.2%) formed a distinct group (n = 7,852; 5%). Primitive cell tumours—includ- | and odontogenic tumours of endodermal/ectodermal origin (C75.2; ing those of germ cell origin (n = 2,076; 1.3%) and those derived from | n = 4; 0.002%)—were excluded from analysis. totipotent or multi-potent precursors (n = 194; 0.1%)—were grouped | To identify cancer clusters based on developmental lineage, we under a ‘primitive cells’ category. Cancers of undetermined origin | then computed standardized incidence ratios (SIRs) at the district (C76–C80; n = 3,558; 2.3%) and tumours with insufficient geostatistical | level. Districts with SIR > 1, indicating a higher-than-expected cancer Nature Health incidence relative to the general population, were identified as cancer | ethnic groups, including the Asháninka, Kakinte, Nomatsigenga, clusters. This approach enabled the generation of high-resolution maps | Quechua and Yanesha38 (Extended Data Fig. 8).
Spatial autocorrelation delineating lineage-specific cancer clusters across Peru (Extended | analysis based on residential locations uncovered a distinct cluster Data Fig. 5). | of liver cancer (C22.0) within this hotspot (Fig. 2d), characterized by strong and statistically significant spatial dependence (global univari- Lineage-defined cancer clusters associate with pesticide risk | ate Moran’s I = 0.37; | z-score = 7.53; | P = 2.5 × 10 | −14). We applied Bayesian inference using the integrated nested Laplace approximation (INLA)—a robust framework for estimating complex | Liver cancer aligns with pesticide exposure hotspots spatial models31—to forecast regions where a substantial share of the | The endoderm-derived liver is a primary target site for carcinogens in cancer burden is probably attributable to environmental pesticide | mammals39–41, making it an ideal organ for validating pesticide-related exposure. Specifically, the method optimizes the geolocation of hot- | carcinogenic risks.
In Peru, hepatocellular carcinoma (HCC)—the most spots, defined as regions where the lower bound of the 95% credible | prevalent liver cancer subtype—disproportionately affects young, interval (CI) for the relative risk (RR) exceeds 1, indicating a geostatis- | non-cirrhotic individuals with Indigenous ancestry18,42, particularly in tically significant association between modelled pesticide risk and | central regions such as Junín15,43. Tumours in these patients consistently observed cancer clustering. | exhibit a molecular profile that diverges from conventional classifica- Stratifying cancers by developmental lineage improved the INLA | tions15,18. Although hepatitis B virus (HBV) is widely considered the prin- model’s predictive performance relative to conventional organ-based | cipal risk factor for liver cancer in this population, the frequently occult classifications32, broadening its applicability, particularly for less com- | nature of HBV infection—characterized by low viral DNA levels44—along mon cancers. Lineage-informed modelling identified 436 hotspots | with the distinctive clinical and molecular features of the disease15,18,42, across Peru, underpinned by a robust spatial structure and a mod- | raises questions about whether HBV alone accounts for these cases.
elled association between pesticide exposure and cancer incidence | Geostatistical analyses pinpointed a liver cancer cluster in Junín (Fig. 2b,c, Extended Data Fig. 6 and Supplementary Fig. 2). RRs within | (Fig. 2d), implicating environmental pesticide exposure as a potential these hotspots ranged from 1.14 to 9.38 (mean ± s.d. = 2.52 ± 1.42), indi- | driver. To investigate this link, we performed transcriptomic pro- cating a spatially consistent and epidemiologically relevant increase | filing on paired HCC and non-tumour liver (NTL) samples from 36 in cancer risk associated with environmental pesticide exposure. | non-cirrhotic patients residing in pesticide-associated cancer hot- Although RR estimates did not significantly differ across cell lineage | spots18. Pesticide-induced hepatotoxicity is primarily mediated by groups (P | = 0.165; Fig. 2c | ), the distribution and extent of hotspots non-genotoxic mechanisms8,45, although cumulative effects from AI varied by lineage, suggesting potential lineage-specific susceptibility | mixtures and their degradation products may elicit modest geno- to local pesticide exposure patterns (Fig. 2b and Extended Data Fig. 6). | toxic effects46.
Using the sample enrichment score (SES), an optimized The most extensive at-risk zones were associated with endodermal | algorithm for single-sample gene set enrichment47, we detected a and ectodermal epithelial cancers—primarily affecting the gastro- | gene expression signature indicative of exposure to non-genotoxic intestinal tract (C15–C21), lungs (C34) and skin (C44)—followed by | carcinogens48 (Extended Data Table 2). Predominantly present in NTL non-mesenchymal, mesoderm-derived malignancies such as those of | samples (P | = 5.1 × 10−4), this signature points to early exposure preced- the female genital organs (C51–C58) and kidney (C64) (Fig. 2a–c and | ing malignant transformation (Fig. 3a). We next scored NTL samples Extended Data Fig. 6). | for gene expression signatures linked to established risk factors for Geospatial mapping of pesticide-associated cancer risk revealed | liver cancer49, including HBV infection, high alcohol consumption, an intricate interplay among population disparities, land use and Peru’s | steatosis and foodborne genotoxins such as aflatoxin B1 (Extended Data diverse geography. Risk was predominantly concentrated in rural areas | Table 2). Statistical analyses confirmed preferential tumour induction experiencing intense anthropogenic pressure.
Along the semi-arid | by non-genotoxic carcinogens (Friedman test; Q = 112.6, | P = 2 × 10 | −23), Pacific coast, prominent hotspots coincided with zones of modern | a pattern visually apparent in Fig. 3b (see also Supplementary Table 2 agriculture on reclaimed and fertilized land33, notably in Ancash and | and Supplementary Fig. 3). To determine whether this non-genotoxic Piura (north of Lima) and in Ica (south of Lima; Fig. 2b | and Extended | signature was specific to the pesticide-associated cancer hotspots in Data Fig. 1). In the Andes, smaller hotspots emerged in inter-Andean | Peru or reflected broader environmental exposure patterns, we com- valleys, where steep terrain accelerates pesticide surface run-off34, | pared the Peruvian dataset with transcriptome datasets from France, probably intensifying local exposure and fostering cancer cluster | Taiwan and Turkey. The results confirmed that the signature was unique formation. This pattern was particularly pronounced in the high- | to Peruvian patients, with no analogous signal detected in any other lands of northern Peru (Amazonas and Cajamarca) and southern Peru | cohort (P = 6.8 × 10 | −15; Fig. 3a,b | ). (Apurímac, Ayacucho and Cusco; Fig. 2b and Extended Data Fig. 1).
| Prompted by geospatial cancer cluster analyses stratified by devel- Further east, across the Amazon basin, widespread hotspots pertaining | opmental lineage, we examined the expression of nine lineage-specific to non-mesenchymal, mesoderm-derived cancers spanned the mon- | MTFs that orchestrate CRCs and drive endoderm-derived hepato- tane and lowland rainforests in the central regions (Huánuco, Junín and | biliary carcinogenesis26 (Extended Data Table 2). Remarkably, NTL Pasco; Fig. 2b and Extended Data Figs. 1 and 6), aligning with deforesta | -samples from Peruvian patients exhibited significantly higher SES for tion fronts driven by agricultural expansion35–37 (Extended Data Fig. 7). | these MTFs than those from France, Taiwan and Turkey (P = 5.7 × 10 | −16; Including deforestation as a fixed covariate significantly improved the | Fig. 3c and Supplementary Fig. 4), indicative of CRC disruption. These INLA model fit, as evidenced by reductions in the deviance informa- | findings suggest early systemic disruption of the MTF-driven autoregu- tion criterion (DIC = 2,804.24; ΔDIC = 5.73) and the widely applicable | latory loop sustaining CRC-dependent hepatocyte identity50, occur- information criterion (WAIC = 7,221.95; ΔWAIC = 3.74), reinforcing | ring before malignant transformation.
This preneoplastic alteration the spatial association between agricultural pressure and pesticide- | is consistent with a field-effect-like phenomenon in histologically associated cancer risk. | normal liver tissue42, whereby some hepatocytes, despite retain- Notably, pesticide-associated cancer hotspots in central Peru— | ing functionality, probably enter an unstable steady state within a amid regions of agricultural deforestation—warrant careful scrutiny, | saddle-node epigenetic landscape51 (Fig. 3d). While this epigenetic as they overlap with Andean–Amazonian Indigenous and peasant | landscape is shaped by multiple layers of regulation, our data indicate territories, where socio-economic disparities and land-use pressures | that aberrant DNA methylation18—triggered by pesticide-induced CRC compound exposure risks35–37 . Among these, a notable multi-lineage | disruption52 and maintained in a lineage-specific manner by the poly- hotspot emerged in Junín (Fig.
2b,d), a region inhabited by Indigenous | comb repressive complex 2 (PRC2)53,54 , a key effector in Peruvian liver Nature Health a | Non-genotoxic exposure signature | MTF gene setbc P = 6.8 × 10 | –15 | P 1= 5.7 × 10 | –16 60 | P = 5.1 × 10 | –4 | P = 2.5 × 10 | –9 40 20 0 PE-NTL | FR-NTL | TW-NTL | TR-NTL n = 36 | n = 10 | n = 37 | n = 20 PE-HCC | FR-HCC | TW-HCC | TR-HCC | 0 | 1 n = 36 | n = 81 | n = 81 | n = 17 | Non-genotoxic d | Metastable Exposure to pesticide mixtures | functional hepatocyte Nuclear receptor? HBV-induced disruption Core transcriptional regulatory circuitry SE | MTF | A | Cancer cell SE | MTF | B | Epigenetic landscape with saddle-node topology Me3 Me3 Me3 Me3 SE | MTF | X PRC2-maintained unstable steady state Clonal expansion Fig. 3 | Transcriptomic evidence of non-genotoxic carcinogenesis and early | plotted individually as points. Peruvian samples are matched pairwise. lineage-specific regulatory disruption in pesticide-associated liver cancer. | P values are from Kruskal–Wallis tests followed by Dunn’s multiple-comparison a, Raincloud plots of SES for a curated non-genotoxic exposure signature in | test with correction and from two-sided paired Wilcoxon tests. d | , Schematic HCC and NTL samples48 (Extended Data Table 2).
b | , Scatter plots of min–max | summarizing the proposed mechanism by which pesticide mixtures induce scaled SES values in NTL samples, comparing the non-genotoxic signature with | early systemic disruption of the CRC formed by super-enhancers (SEs) and transcriptomic signatures linked to major HCC risk factors. A dashed identity | MTFs—an autoregulatory loop essential for sustaining hepatocyte identity50. line (x | = y | ) highlights deviations between signatures. c | , Raincloud plots of SES | This disruption propels cells into an unstable, preneoplastic steady state situated for a curated gene set of lineage-specific MTFs implicated in hepatobiliary | within a saddle-node epigenetic landscape51—a departure from Waddington’s carcinogenesis26. In a | –c, Peruvian samples (PE; red; n | = 36 NTL and n | = 36 HCC) model. Characterized by aberrant DNA methylation and maintained by PRC218, are compared with cohorts from France (FR; blue; n | = 10 NTL and n | = 81 HCC), this metastable state constitutes a transformation-prone, epigenetically Taiwan (TW; orange–brown; n | = 37 NTL and n | = 81 HCC) and Turkey (TR; green; entrapped configuration that destabilizes cell-fate commitment. Dysregulation n = 20 NTL and n | = 17 HCC); lighter and darker shades denote NTL and HCC, | of PRC2, frequent in HBV infection55, may trigger the collapse of this precarious respectively.
Raincloud plots in a | and c | combine box plots with kernel density equilibrium, culminating in malignant transformation. Schematic in d | adapted estimates. Box plots indicate the median (centre line), interquartile range | with permission from ref. 51 | , Elsevier; icons in d | created in BioRender; Bertani, S. (IQR, box) and 1.5 × IQR bounds (whiskers), with outliers beyond the whiskers | https://biorender.com/3v25ddg (2026). cancer18—compromises cellular stability, thereby undermining proper | Our advance stems from integrating a state-of-the-art, process- cell-fate commitment and predisposing the tissue to malignant trans- | based environmental risk model with Bayesian spatial epidemiology formation. Subsequent dysregulation of PRC2 through its SUZ12 subu- | and transcriptomic validation. This strategy enabled high-resolution nit during HBV infection55 may breach this unstable epigenetic milieu, | mapping of cumulative environmental exposure risk to pesticide mix- acting as a tipping point that precipitates malignant transformation. | tures—unprecedented in scope—with each AI individually deemed non-carcinogenic56. The resulting risk surfaces were linked to spatial Discussion | patterns of cancer incidence.
By stratifying comprehensive cancer This study links pesticide exposure to cancer risk in a nationwide | registry data according to developmental lineage, we redefined can- cohort, revealing mechanistic pathways of environmentally driven | cer clustering within a coherent ontogenetic framework, uncover- carcinogenesis, with important implications for global health policy. | ing hallmark patterns often obscured by conventional organ-based Nature Health classifications24. This biologically anchored approach affords a | reductionist toxicology7, our approach offers a scalable and transfer- nuanced assessment of pesticide-associated cancer risk, exemplified | able framework for global environmental health assessment. Crucially, by the identification of an atypical molecular subtype of liver cancer | our findings underscore the imperative of embedding socio-ecological in Andean–Amazonian Indigenous patients18, whose aetiology had | equity within regulatory policy—an essential step towards mitigating remained elusive. | ecological harm and protecting vulnerable populations from environ- Our findings indicate that pesticide exposure may contribute | mentally driven cancers. to the risk of liver cancer in this population.
At the molecular level, chronic exposure to non-genotoxic pesticide mixtures—even at or | Methods below individual experimental no-observed-adverse-effect levels3— | Ethics and data protection compliance disrupts lineage-dependent transcriptional programmes and fosters | The study complies with Peru’s Personal Data Protection Law 29733 cancer development. Transcriptomic analyses show that such expo- | (Regulation PDPL 016-2024-JUS). The INEN Cancer Registry is author- sures perturb the expression of lineage-specific MTFs in normal tissue, | ized by the Peruvian National Authority for the Protection of Personal destabilizing the regulatory networks that sustain cell identity25,50 and | Data under registration number RNPDP-EP 4794. Ethical approvals undermining the epigenetic safeguards that enforce cell-fate commit- | were granted by the INEN Institutional Review Board under reference ment52. This aberrant transcriptional state probably increases cellular | numbers 113-2014-CIE/INEN, 407-2016-CIE/INEN and 049-2024-DICON/ vulnerability to extrinsic cues—whether environmental, metabolic, | INEN. All procedures adhered to the ethical principles of the WMA infectious or inflammatory30,49—that may act as tipping points for | Declaration of Taipei on research involving health databases, big data malignant transformation, most notably through mild or even occult | and biobanks.
HBV infection in Peruvian patients with liver cancer44. These find- ings raise a critical question: do disruptions in lineage-specific MTF | Survey of agrochemical retailers expression accumulate over time, or are they imprinted during early | A survey, described in detail elsewhere60, was conducted in 2020 to development57, in line with the DOHaD (developmental origins of | identify the pesticides most commonly used in Peru and to detect any health and disease) hypothesis58? While our data implicate the liver—a | continued use of obsolete or banned products61, as mandated by the primary target of chemical carcinogens39–41—as a sentinel site of lineage- | Peruvian National Agricultural Health Service (SENASA). A total of specific disruption of MTF activity, it remains undetermined whether | 650 agrochemical retailers—both formal (as listed in the Integrated analogous mechanisms occur in other exposed tissues, which were not | Agricultural Input Management System (SIGIA); https://servicios. assessed here.
Further investigation of additional organ sites identified | senasa.gob.pe/SIGIAWeb/sigia_consulta_empresa.html; accessed 24 through geospatial clustering is warranted to determine whether this | October 2020) and informal (including itinerant and market-based lineage-dependent susceptibility reflects a broader systemic response | vendors)—were randomly sampled across five agriculturally diverse to pesticide mixtures. | regions: Ayacucho, Huancavelica, Huánuco, Junín and Pasco (Extended Beyond its molecular insights, our study reveals pressing socio- | Data Fig. 1). environmental challenges. In regions where intensive agriculture, | A pretested, semistructured questionnaire62 was used to doc- unsustainable land management and limited healthcare coalesce10,13 , | ument commercial pesticide names and application frequencies. the dispersal of pesticides not only undermines ecological resilience59 | Interviews were conducted face to face by a native Spanish-speaking but also exacerbates enduring health inequalities. Geospatial model- | agronomist, with responses recorded electronically using KoboTool- ling reveals that high-risk zones for pesticide-associated cancer are | box (v.2.020.25). The response rate reached 98.9%, ensuring repre- disproportionately concentrated in rural areas experiencing intense | sentative coverage of local pesticide practices.
Commercial names were anthropogenic pressure. This pattern underscores the intricate inter- | standardized using OpenRefine (v.3.4.1), and unmatched entries were play between environmental degradation, land-use dynamics and | excluded. AIs and their Chemical Abstracts Service (CAS) registry num- socio-economic marginalization, placing underserved communities | bers were retrieved from the PubChem database (https://pubchem. at greatest risk and reinforcing the urgent need for targeted policy | ncbi.nlm.nih.gov/) using OpenRefine’s reconciliation function and interventions35. Moreover, preliminary evidence suggests that cli- | cross-validated against national regulatory records to confirm that mate phenomena such as ENSO shape land-use patterns and influ- | the final list of AIs accurately reflects pesticide mixtures present in ence pesticide dispersion21, thereby amplifying local exposure risk. | environmental contexts. As global warming advances, these dynamics are likely to intensify, further aggravating the burden of environmentally driven cancers | Pesticide inventory for environmental modelling among exposed populations. | A total of 31 AIs—comprising 19 insecticides, 7 fungicides and 5 herbi- Despite the robustness of our integrative approach, certain lim- | cides (Extended Data Table 1)—were retained for environmental mod- itations merit consideration.
First, individual pesticide exposures | elling. All, with the exception of monocrotophos (banned by SENASA were not directly measured but inferred from population-level spatial | in 2004), were authorized for use in Peru during the 6-year modelled proxies, introducing uncertainty regarding their timing, intensity | period (2014–2019). None was classified as carcinogenic to humans and chemical composition, as well as precluding direct attribution of | (group 1) in the IARC Monographs (https://monographs.iarc.who.int/ cancer onset at the individual level. Second, residual confounding from | list-of-classifications; accessed 28 May 2025), nor deemed extremely environmental, socio-economic or lifestyle factors cannot be entirely | hazardous (Ia) under the World Health Organization (WHO) classifica- excluded; nevertheless, the convergence of geospatial, aggregated epi- | tion of pesticides by hazard63. demiological and molecular evidence strongly supports the biological plausibility of our findings.
Future studies incorporating large-scale | Process-based environmental pesticide risk modelling biomonitoring, individual-level exposure assessment, multi-tissue | The environmental pesticide risk model was built on a grid of analyses and longitudinal molecular profiling will be essential to | 124,545,177 cells (100 m × 100 m; 0.01 km2 each), covering 96.7% of strengthen causal inference, evaluate generalizability and establish | Peru’s 1,287,072.5-km2 national territory. Environmental variables direct links between pesticide exposure and cancer incidence, building | integrated into the model included monthly precipitation and sur- on the framework established herein. | face run-off (millimetres per month; 2014–2019; PISCO_HyM_GR2M Taken together, this work establishes a unified probabilistic model | v.1.1), terrain slope (degrees) derived from elevation data (metres to quantify cancer risk across diverse tumour types and real-world | above sea level; 2014; EarthEnv-DEM90 v.1) and soil organic carbon pesticide exposure scenarios. By transcending the limitations of | (SOC, %) calculated from the Global Soil Organic Carbon Map v.1.5 Nature Health (2018; GSOCmap-FAO) and topsoil bulk density from the Harmonized | Hair samples were wrapped in aluminium foil, labelled at the root ends World Soil Database v.1.2 (2014; T_REF_BULK v.1.0).
All spatial layers | and stored dry at 4 °C for 2 weeks before analysis. A hair growth rate of were standardized to Universal Transverse Mercator (UTM) zone 18S | 1 cm per month was assumed67; the proximal 0.5 cm (embedded in the (WGS84, EPSG:32718), converted to raster format and resampled to | scalp) was excluded from analysis, as it is inaccessible for sampling. 100 m × 100 m resolution (≈3.2 arcsec at the equator) using bilinear | Levels of 67 AIs and degradation products were quantified in hair interpolation to ensure cross-dataset compatibility. | (pg mg−1)19, spanning 19 pesticides incorporated into the environmen- Physicochemical properties of AIs were sourced from the Chemi- | tal exposure risk model, as detailed in Supplementary Methods and cal Entities of Biological Interest (ChEBI) database (https://www.ebi. | Supplementary Table 1. ac.uk/chebi/), the Pesticide Properties Database (PPDB; https://sitem. herts.ac.uk/aeru/ppdb/en/) and PubChem. Environmental pesticide | Stratification of cancer registry data by developmental lineage risk for each AI was modelled using a lumped-parameter expression | The developmental lineage classification of neoplasms, originally developed by the Organization for Economic Co-operation and Devel- | developed by Berman24,68,69, is a validated taxonomy that organizes opment (OECD) expert group64.
This equation approximates key trans- | neoplasms by embryonic origin within a structured hierarchy of 39 port and dissipation processes—run-off mobilization, sorption and | classifiers across seven levels. To balance biological coherence with degradation—as follows: | statistical robustness, we stratified data from the INEN Cancer Registry Q | 100 | (2007–2020) to the fifth hierarchical level. The resulting schema was L% ro = ( | )×f×exp(−3×ln2/DT | 50Soil) × | as follows: P | (1+Kd) neoplasms where L% | ro is the estimated percentage of applied pesticide available | embryonic for run-off in a dissolved form; Q is the run-off to precipitation ratio, | primitive P calculated monthly for each grid cell; f | is a correction factor accounting | primitive_differentiating for terrain slope, plant interception and buffer zones (Supplementary | germ_cell Methods); DT50Soil is the soil dissipation half-life; and K | d is the soil–non_primitive water distribution coefficient, derived from the organic carbon distri- | endoderm_or_ectoderm bution coefficient (K | oc) and local SOC content (%OC) (as in Extended endoderm_or_ectoderm_surface Data Table 1).
| endoderm_or_ectoderm_parenchymal A total of 2,232 raster layers, representing monthly outputs for | odontogenic_epithelium 31 pesticides over 72 months (2014–2019), were averaged to produce | mesoderm the final environmental risk surface. Risk scores were normalized to a | mesenchyme 0–100 scale to enable both spatial and intercompound comparisons. | non_mesenchymal_mesoderm Model calibration was performed using the 2018 cultivated land cover | neuroectoderm_neural_plate (km2), derived from satellite imagery by the Peruvian Ministry of Agri- | extra_embryonic culture and Irrigation (MINAGRI)65. Zonal statistics were computed | trophoblastic using the 2021 district boundaries from the Spatial Data Infrastructure of Peru (IDEP). Normalized raster layers were overlaid with district | The full developmental lineage classification with taxonomy, polygons, and summary metrics (mean, median, minimum and maxi- | encoded in eXtensible Markup Language (XML) format and compris- mum) were calculated across all raster cells within each district. All | ing approximately 55,000 hierarchically organized neoplasia terms, computations were performed using QGIS v.3.28 LTR and R v.4.4.1 | was sourced from Berman’s seminal article24. A custom Python parser (packages: sf v.1.0-10, sp v.1.5-1 and terra v.1.8-70).
Database access | extracted the lineage mapping and exported it as a CSV file to stream- details are provided in Supplementary Table 3. Hard data underlying | line processing. Neoplasia terms were linked to ICD-10 Clinical Modifi- the environmental pesticide exposure risk map are publicly available | cation (ICD-10-CM) codes using a predefined crosswalk to concepts in at https://doi.org/10.6084/m9.figshare.29728463. | the US National Cancer Institute Thesaurus (NCIt_Maps_To_ICD10CM v.25.03e; https://evsexplore.semantics.cancer.gov/evsexplore/ Climatic phase determination | welcome). Discrepancies were systematically logged and curated by Climatic phases were characterized based on ENSO variability, defined | experts, with constraint-based mapping enforcing strict one-to-one by the 3-month running average Oceanic Niño Index (ONI) derived from | and hierarchical consistency. The resulting ICD-10-mapped develop- sea surface temperature anomalies in the Niño 1 + 2 region (0–10° S, | mental lineage classification was imported into a PostgreSQL relational 90–80° W). Monthly sea surface temperature data were obtained from | database, with SQL constraints applied to enforce referential integrity.