Air pollution exposure induces a decrease in type II interferon response: A paired cohort study

Abstract

Background: While air pollution is a major issue due to its harmful effects on human health, few studies focus on its impact on the immune system and vulnerability to viral infections. The lockdown declared following the COVID-19 pandemic represents a unique opportunity to study the large-scale impact of variations in air pollutants in real life. We hypothesized that variations in air pollutants modify Th1 response represented by interferon (IFN) γ production.

Methods: We conducted a single center paired pilot cohort study of 58 participants, and a confirmation cohort of 320 participants in Nice (France), with for each cohort two samplings at six months intervals. We correlated the variations in the production of IFNγ after non-specific stimulation of participants' immune cells with variations in key regulated pollutants: NO₂, O₃, PM_2.5, and PM₁₀ and climate variables. Using linear regression, we studied the effects of variations of each pollutant on the immune response.

Findings: In the pilot cohort, IFNγ production significantly decreased by 25.7% post-lockdown compared to during lockdown, while NO₂ increased significantly by 46.0%. After the adjustment for climate variations during the study period (sunshine and temperature), we observed a significant effect of NO₂ variation on IFNγ production (P=0.03). In the confirmation cohort IFNγ decreased significantly by 47.8% and after adjustment for environmental factors and intrinsic characteristics we observed a significant effect of environmental factors: NO₂, PM₁₀, O₃, climatic conditions (sunshine exposure, relative humidity) on variation in IFNγ production (P=0.005, P<0.001, P=0.001, P=0.002 and P<0.001 respectively) but not independently from the BMI at inclusion and the workplace P=0.007 and P<0.001 respectively).

Interpretation: We show a weakening of the antiviral cellular response in correlation with an increase of pollutants exposition.

Funding: Agence Nationale de la Recherche, Conseil Départemental des Alpes-Maritimes and Region Sud.

Keywords: Air pollution; Immune response; Interferon gamma; Lockdown; Nitrogen dioxide.

Figure 1

Flowchart of participants: for the pilot cohort (a) and the confirmation cohort (b). Note: CADAM is Alpes-Maritimes Departmental Administrative Center.

Figure 2

Evolution of the concentrations of the air pollutants at Nice East station (a) and West station (b): pilot cohort. The main and secondary vertical axis show the concentrations of pollutants in µg/m³. The horizontal axis show the average values between February 2020 and February 2021 for each pollutant, with colour areas corresponding to the sampling periods (red for the lockdown = T1, from 27/03/2020 to 26/06/2020 and green for the resumption of activities = T2, from 23/09/2020 to 14/01/2021). A vertical red dotted line represents the end of lockdown (30/06/2020). Horizontal blue, orange and yellow dotted lines are the annual average regulatory values (2021 guidelines) for NO₂ (10 µg/m³), PM10 (15 µg/m³), and PM2.5 (5 µg/m³).

Figure 3

Evolution of the concentrations of the air pollutants at Nice East station (a) and West station (b) and Airport station (c) confirmation cohort. The main and secondary vertical axis show the concentrations of pollutants in µg/m3. The horizontal axis show the average values between July 2020 and July 2021 for each pollutant. Horizontal blue, orange and yellow dotted lines are the annual average regulatory values (2021 guidelines) for NO₂ (10 µg/m³), PM10 (15 µg/m³), and PM2.5 (5 µg/m³).

Figure 4

Exposure mapping of nitrogen dioxide over the study period for the confirmation cohort. The background map shows the exposure concentration in µg/m³ (between 28 July 2020 and 21 July 2021) for nitrogen dioxide (average of daily maximums). The 3 dotted ellipses represent the 3 exposure zones (or workplaces) in the analysis, according to the location of the participants and the sensor considered. The 4 inclusion centers are marked with a red square. CADAM is Alpes-Maritimes Departmental Administrative Center. The white arrow points north.

Figure 5

Comparison of medians of immunity, pollution and climate variables between T1 and T2 for the pilot (a on left) and confirmation (b on right) cohorts. Each boxplot represents the median of variation (median of differences) and its 95% confidence interval for each variable between T1 and T2. Test de Wilcoxon pour données appariées. The horizontal axis represents values in units (IU/mL for IFNγ, µg/m³ for pollutants, °C for temperature, % for relative humidity, km/h for wind speed and minutes for sunshine). For the pilot cohort N=58 and for the confirmation cohort N=320. *P≤0.05, ^⁎⁎P≤0.01, ^⁎⁎⁎P≤0.001.

Figure 6

Comparison of the quartiles of IFNγ variation as a function of the nitrogen dioxide variation for the pilot cohort (a on left) and confirmation cohort (b on right). Each panel represents a heatmap for ΔNO₂. Each column represents a quartile (Q1 to Q4): quartiles of IFNγ change in % (T2-T1/T1). The colours represent the intensity of exposure variations for each participant in each quartile. For the pilot cohort N=58 and for the confirmation cohort n=320.*P≤0.05, ^⁎⁎P≤0.01, ^⁎⁎⁎P≤0.001.

Figure 7

Multiple linear regression curve between IFNγ variation and pollutants variation. The curves represent the estimated marginal means after fitting the multivariate model for the variation of IFNγ as a function of workplace and the variation in NO₂ (a), O₃ (b) and PM₁₀ (c)