Dutch Cohort Study Will Link Home Pesticide Exposure to Primary Care Outcomes
Residential exposure to agricultural pesticides has moved from a theoretical concern to a testable public health question in the Netherlands. A new retrospective dynamic cohort study will use primary care data to examine whether living near treated agricultural plots is associated with acute symptoms, recurring health conditions, respiratory exacerbations, and greater healthcare use. The protocol is designed to separate broad anxiety about environmental exposure from measurable clinical outcomes in routine care.
The study will follow people registered in Dutch general practices from 2011 through 2025, drawing on the Nivel Primary Care Database and exposure estimates generated by the Institute for Risk Assessment Sciences at Utrecht University. The cohort is large, about 900,000 individuals living in moderately urban to non-urban areas, which gives the analysis enough breadth to address common outcomes and some less frequent events. By using a dynamic design, the investigators can account for changing residency and evolving exposure patterns over time rather than freezing participants at a single baseline moment.
One sentence captures the core tension in this protocol: the exposure is environmental, but the clinical readout comes from ordinary family practice.
Measuring Exposure at the Home Address
The exposure assessment is unusually granular. Two forms of individual-level pesticide exposure will be linked to health outcomes: the amount of active substance used within buffers of 50, 100, 250, and 500 meters around homes, and modeled concentrations at the individual residence. These metrics matter because distance alone is a blunt surrogate; proximity can imply very different levels of actual exposure depending on crop type, application intensity, weather, and local geography.
The active substances were selected in advance using epidemiological and toxicological evidence, societal and scientific concern, and reported use patterns in the Netherlands. That selection process is important clinically because it avoids treating all pesticide compounds as interchangeable. Different active substances have different biological behavior, and a risk signal from one class cannot automatically be generalized to another.
By linking spatial exposure estimates to longitudinal records, the study can ask whether there is a dose-pattern relationship rather than a simple exposed-versus-unexposed contrast. In environmental epidemiology, that distinction is often what turns a public worry into a plausible causal hypothesis.
Symptoms, Exacerbations, and Everyday Care Use
The outcome set is broad enough to reflect both acute and recurring morbidity. Primary analyses will examine incidence and prevalence of health endpoints using longitudinal multilevel Poisson regression models. Separate models will assess exacerbations in patients with asthma and chronic obstructive pulmonary disease, along with total healthcare utilisation, using zero-inflated Poisson and negative binomial approaches as appropriate.
That analytic choice fits the clinical question. Respiratory disease is one of the most plausible pathways for pesticide-related harm in residential settings, since airborne drift or secondary exposure could aggravate airway inflammation in susceptible patients. Asthma and chronic obstructive pulmonary disease are also disorders where relatively small environmental changes can translate into extra visits, medication adjustments, and symptom flares, even when hospital-level events remain uncommon.
The protocol also makes room for acute and recurring symptoms and conditions outside the respiratory tract, though the abstract does not list each endpoint individually. A primary care database is well suited to this work because many environmentally triggered complaints, from transient symptoms to repeated consultations, never reach specialty care or hospitalization.
Why Primary Care Data May Be the Right Lens
Using general practice records gives the study a practical advantage: it captures the front door of the health system. Primary care is where patients present with nonspecific symptoms, where chronic airway disease is managed over time, and where recurrent healthcare use can be tracked more faithfully than in a single hospital dataset. It also reduces the bias that comes from restricting analysis to severe disease only.
The design is retrospective, but the cohort is dynamic, which gives the investigators flexibility to follow new entrants and changing residence patterns over a 14-year window. That is especially relevant for agricultural exposure, where a family may move closer to or farther from treated plots and where seasonal application cycles may create time-varying risk. The statistical models will adjust for relevant covariates at both the individual and neighbourhood level, an essential step because residential exposure is tied to rurality, socioeconomic factors, local healthcare access, and other contextual influences that can distort crude associations.
Power calculations suggested feasibility for the selected outcomes. Even so, the protocol acknowledges a hard boundary familiar to environmental epidemiology: small effects for low-prevalence outcomes may require larger populations than are currently available. In other words, the database is strong enough to answer some questions more cleanly than others.
Ethics, Governance, and the Limits of Consent
The study does not fall within the scope of the Medical Research Involving Human Subjects Act, so formal ethical approval is not required. Governance approval from the relevant bodies overseeing Nivel-PCD is being sought, and informed consent is not needed because the records are pseudonymised. Individual patients can still opt out through their general practitioner, preserving a route for refusal within routine care.
For a topic that has generated public concern, the governance framework matters. People living near agricultural land often worry about invisible exposures, but policy discussions rarely benefit from anecdote alone. A registry-based cohort can test whether those concerns map onto actual increases in symptoms, respiratory flares, or healthcare utilisation, while also showing where no signal exists.
The protocol does not claim in advance that pesticide exposure will cause disease, and it does not assume the opposite. It lays out a structured way to measure whether home-level proximity to agricultural active substances corresponds to a heavier burden of primary care contact, especially among patients with asthma or chronic obstructive pulmonary disease. The chief limitation is built into the question itself: rare outcomes and modest risk increases may stay difficult to detect even in a large cohort, leaving the most consequential answers to depend on how well exposure can be estimated and how faithfully routine records capture clinical change.