The study, led by Stefan Modzelewski and colleagues, enrolled 104 participants: 85 patients with schizophrenia and 19 healthy controls, recruited from three clinical centers within a 15-kilometer radius of a standardized weather station in the Podlaskie Voivodeship, a region of northeastern Poland known for pronounced seasonal variation. Each participant underwent a single session combining clinical assessment with blood sampling, and serum levels of four immune messengers — TNF-α, IFN-γ, IL-6, and IL-4 — were measured using a multiplex cytokine assay in a subsample of 84 subjects.

What makes the study methodologically distinctive is its dual approach to quantifying weather exposure. The first, a local approach, decomposed wind direction into north–south (meridional) and east–west (zonal) vector components using trigonometric transformation, weighted by wind speed and a stability index reflecting how consistently the wind blew from one direction over a three-day window. The second, a synoptic approach, was far more ambitious: the researchers used the NOAA HYSPLIT trajectory model to trace the seven-day backward path of air masses arriving at 1,500 meters altitude over the study site, classifying each as Arctic, Polar Continental, Polar Maritime, or Tropical based on its geographic origin and characteristic properties. Two researchers classified the trajectories independently while blinded to participants’ clinical status, achieving strong inter-rater agreement (Cohen’s κ = 0.819), with disagreements resolved by a third assessor.

The headline finding emerged from the synoptic classification. Among patients with schizophrenia, Tropical air masses — those advecting from the south and typically associated with warmer conditions in the Polish climate — were linked to significantly lower concentrations of both IFN-γ and TNF-α than any other air mass type, with Kruskal–Wallis tests yielding p-values of 0.016 and 0.001 respectively, and all adjusted post-hoc comparisons below 0.05. No such differences appeared for IL-6 or IL-4, suggesting the effect was not a generalized suppression of immunity but a selective shift in Th1-associated, pro-inflammatory signaling.

More striking still was the interaction with health status. When the researchers modeled the relationship between air mass type and log-transformed IFN-γ, they found that Tropical air was associated with lower IFN-γ in patients but higher IFN-γ in healthy controls compared with Polar Maritime and Arctic air — interaction contrasts of approximately 0.93 and 1.15, both with p < 0.001. In other words, the same environmental condition that dampened a pro-inflammatory signal in patients amplified it in healthy people. Under Polar Maritime and Arctic air masses, patients showed markedly higher IFN-γ than controls; under Tropical air, that between-group gap vanished entirely. The full interaction model explained roughly 52 percent of the variance in log IFN-γ, although the authors caution that this figure reflects the joint contribution of health status, air mass type, and their interaction rather than air mass type alone.

The researchers interpret these patterns through the lens of

The researchers interpret these patterns through the lens of immunoseasonality, a framework holding that weather-related factors modulate immune activity in ways that can destabilize the clinical course of psychiatric illness. The directionally opposite IFN-γ responses observed here are consistent with a health-status-dependent form of this modulation: the same synoptic condition appears to push immune signaling in opposite directions depending on whether the immune system is already dysregulated. In schizophrenia, a body of literature has documented alterations in peripheral cytokines, including elevations of pro-inflammatory mediators that may reflect low-grade neuroinflammation. IFN-γ is a signature cytokine of Th1-type adaptive immunity, and TNF-α is a central pro-inflammatory mediator produced by macrophages and other immune cells. A selective shift in these two signals — without parallel changes in IL-6 or IL-4 — suggests that air mass exposure may preferentially engage a specific arm of the immune response rather than producing a blanket inflammatory or anti-inflammatory effect.

The biological plausibility of such an effect rests on the fact that an air mass is not merely wind. By the time an air mass arrives over a region, it carries an integrated history of the surfaces it crossed: its temperature, humidity, pressure profile, particulate and pollen burden, and even its pollutant content reflect days of transit. Arctic and Polar Continental air masses reaching northeastern Poland in the colder months tend to be cold, dry, and dense, while Tropical air is warmer and typically more humid. Physiological responses to these conditions — vasoconstriction, changes in mucosal barrier function, shifts in autonomic tone, altered vitamin D-related pathways, and stress hormone dynamics — could plausibly converge on immune signaling. The authors argue that this integrative quality is precisely why macro-scale air mass classification may outperform single-variable or local wind-direction measures as an environmental proxy: a wind vector decomposition captures only the local direction of airflow, whereas an air mass category summarizes a coherent meteorological regime.

That interpretation is supported by the study’s internal contrast between its two exposure models. The continuous meridional and zonal wind components, averaged over three days, showed no association with cytokine concentrations, while the categorical synoptic classification produced the significant and interaction effects described above. If weather-related immune modulation were driven simply by the direction from which air flows into a locality, the vector model should have detected it. Its failure to do so, alongside the success of the trajectory-based classification, points toward the bundled properties of the air mass — rather than wind direction per se — as the operative exposure. This distinction matters for future research design, because it implies that studies reducing weather to temperature, humidity, or wind speed in isolation may miss the composite signal that synoptic categories preserve.

Notably, neither exposure model showed any association with PANSS scores, the standard measure of positive, negative, and general psychopathology used in the study. This absence is important for calibrating expectations. The cross-sectional design captured a single moment in each participant’s clinical course, and symptom severity at one time point may be a relatively insensitive outcome for an environmental exposure whose effects, if real, might accumulate over days to weeks or manifest more clearly as changes in relapse risk, hospitalization timing, or symptom fluctuation. The cytokine findings should therefore be read as evidence of a measurable biological response to synoptic conditions, not yet as evidence that air mass type moves symptom scores. Bridging that gap — from immune signal to clinical trajectory — remains the central unanswered question.

Several limitations temper the findings and are acknowledged by the authors. The sample is modest, particularly the 19 healthy controls whose IFN-γ values drive the interaction effect, and the cytokine analysis was restricted to the 84 participants with complete laboratory data. The cross-sectional design cannot establish temporality: blood was drawn once, and the air mass classification reflects the seven days preceding that draw, but unmeasured confounders — infections, medication changes, sleep, diet, physical activity, and smoking — could each interact with both weather and immunity. Antipsychotic and other psychotropic medications themselves influence cytokine profiles, and the study population, drawn from a single region of northeastern Poland, may not generalize to climates where the local air mass climatology differs. The authors also note that the large variance explained by the interaction model should not be attributed to air mass type alone, since health status and its interaction with exposure jointly contribute to that figure.

Within those constraints, the study opens several concrete avenues. Longitudinal designs with repeated cytokine sampling in the same patients across changing synoptic regimes would allow each individual to serve as their own control, substantially increasing statistical power and clarifying within-person dynamics. Linking air mass exposure to admission records and relapse timing across larger cohorts and multiple geographic regions would test whether the immunoseasonal hypothesis translates into clinically meaningful prediction. Mechanistic work could examine intermediate pathways — such as barometric pressure sensitivity, thermoregulatory strain, or airborne allergen and pollutant loads carried by different air masses — that might mediate the immune shifts observed here. Replication of the directionally opposite IFN-γ pattern in independent samples is the essential first step, given that the finding is exploratory and derived from a pilot cohort.

The broader significance of the work lies in its methodological proposal. By treating the advected air mass, traced backward over seven days with a standard atmospheric trajectory model, as the unit of environmental exposure, the study offers psychiatry a framework borrowed from synoptic climatology — one that has been used in epidemiology for heat, cold, and air pollution health effects but rarely applied to severe mental illness. If the preliminary signal holds, weather-sensitive immune modulation in schizophrenia could eventually be monitored using routinely available meteorological trajectory data, potentially identifying periods of elevated biological risk before clinical deterioration becomes apparent. For now, the study stands as a carefully constructed proof of concept: the atmosphere above a patient, summarized by the origin of the air they breathe, may leave a detectable and health-status-dependent imprint on the immune molecules circulating in their blood.

Subject of Research: Air mass advection and immune dysregulation in schizophrenia: an exploratory pilot study

Article Title: Air mass advection and immune dysregulation in schizophrenia: an exploratory pilot study

Article References: Modzelewski, S., Oracz, A. J., Suprunowicz, M., Śladewska, A., Froń, O., Kosidło, J. W., Matowicka-Karna, J., & Waszkiewicz, N. (2026). Air mass advection and immune dysregulation in schizophrenia: an exploratory pilot study. Frontiers in Psychiatry, 17, Article 1950134. https://doi.org/10.3389/fpsyt.2026.1950134

Image Credits: AI Generated

DOI: 10.3389/fpsyt.2026.1950134

Keywords: schizophrenia, air masses, cytokines, IFN-gamma, immunoseasonality, weather, biometeorology, TNF-alpha, psychiatry, inflammation, HYSPLIT, pilot study