Impact of seasonal biometeorological conditions and particulate matter on asthma and COPD hospital admissions

Abstract

Climate change and air pollution are pressing public health concerns, necessitating monitoring of their impact, particularly on respiratory diseases like obstructive lung diseases. This retrospective study analyzed medical records of patients hospitalized at the Warmia and Mazury Centre for Pulmonary Diseases in Olsztyn, Poland (2012–2021) for asthma and chronic obstructive pulmonary disease (COPD) exacerbations. Data included meteorological factors such as temperature, humidity, wind speed, precipitation, and levels of PM2.5 and PM10. The Humidex was utilized to assess thermal discomfort, considering various meteorological and thermal seasons. Findings indicated seasonal variability in asthma and COPD exacerbations. During winter, poorer air quality due to higher PM2.5 and PM10 levels correlated with increased exacerbations (r = 0.283, p < 0.05; r = 0.491, p < 0.001). In summer, discomfort from meteorological conditions led to more hospital admissions. Humidex values strongly correlated with admissions for obstructive diseases (R² = 0.956 for asthma; R² = 0.659 for COPD), with July and August showing statistically higher admission rates (p < 0.05). The study highlights the significant impact of air pollution and meteorological conditions on exacerbations of asthma and COPD, with Humidex serving as a valuable predictor during summer months.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-024-84739-9.

Introduction

Combating climate change is a crucial element of the global public health strategy. Such strategy should account for both the adaptation to the changing climatic conditions and measures to reduce emissions of harmful substances¹. Effective integration of these measures can significantly improve general population health, including reducing the incidence of diseases such as asthma and chronic obstructive pulmonary disease (COPD), which are particularly sensitive to meteorological conditions². In times of dynamic climate change, the development of strategies requires a systematic updating of input data, taking into account the geographical specificities of a given area. Climate change contributes to significant shifts in the occurrence of some diseases. Diseases that thus far have been characteristic of tropical areas become to appear in regions of a moderate climate, while diseases typical of particular regions may change their patterns of occurrence. These shifts have serious implications for public health and require the adaptation of disease prevention and control strategies in order to effectively manage new epidemiological threats^3,4.

The monitoring of these phenomena requires attention and reasonable interpretation, especially given that biometeorological relationships are specific to different geographical areas. Relationships between climate, air pollution, and public health are often complicated and multidimensional. Adaptational behaviours to climate change can be illustrated by a rapid increase in the demand for air conditioning systems that help in managing extreme temperatures, though these systems also generate heat and do not always mitigate the effects of air pollution^5–7. In cities, air pollution caused by household fuel combustion and/or car emissions is a great challenge.

It should be noted that climate change significantly influences meteorological dynamics, including the speed and directions of winds as well as intensity and spatial and temporal distribution of precipitation. These modifications may lead to the shifts in the distribution and levels of harmful substances in the atmosphere, e.g. via wind-related dispersal of pollutants, while the changes in the quantity and intensity of precipitation may impact on the washout of pollutants, which has significant implications for the quality of air as well^8–11.

According to the World Health Organisation (WHO), approx. 7 million people die due to air pollution annually¹². This is caused mainly by particulate matter (PM), including PM2.5 (PM with a aerodynamic diameter of less than 2.5 μm), which settles in the respiratory tract, leading to an abnormal immune inflammatory response and impaired epithelial cell function^13,14. Yao et al. demonstrated that PM2.5 may worsen lung function, thus contributing to the development of COPD¹⁵. An increase in PM2.5 levels by 2.4 µg/m³ leads to a decrease of forced expiratory volume (FEV1) by 101.7 ml¹⁶. Moreover, an increase in PM2.5 of 10 µg/m³ is associated with an increase in emergency admissions in ambulatory care (hospital admissions or emergency department admissions) by 1.4–2.5% ¹⁷. A short exposure to PM2.5 increases the number of inflammatory biomarkers, which may be a common denominator for adult asthma^18,19. Considering climate change and greater air pollution, the monitoring of the impact of these factors on the incidence of obstructive lung diseases – asthma and COPD – becomes particularly significant^20–22. As estimated by the WHO, approx. 235 million of people have asthma and approx. 251 million of people have COPD worldwide²³. It is known that the incidence of these diseases and exacerbation of their symptoms can be associated with the level of air pollution^24,25. This, as already mentioned, depends, among other factors, on a number of meteorological parameters, including ambient temperature, relative air humidity or wind speed²⁶.

Asthma and COPD are two common respiratory system diseases that, despite some common clinical features, differ in their pathophysiology and development mechanisms²⁷. Both pathological conditions are characterised by a chronic inflammation of the airways, leading to the obstruction in the air flow – reversible in asthma, and irreversible in COPD²⁸. Asthma is dominated by CD4 + T lymphocytes, eosinophils, and mast cells, which cause variable, often reversible, bronchial obstruction induced by environmental factors (allergens, infections, tobacco smoke, air pollutants, occupational factors) and genetic factors (predisposition to atopy and airway hyperresponsiveness)^29–32. Symptoms of asthma include wheezing, shortness of breath, chest tightness, and coughing³³. In contrast, COPD is a progressive condition in which CD8 + T lymphocytes and macrophages predominate, leading to the irreversible airway obstruction as a result of chronic inflammation and damage to lung tissue caused by harmful agents such as tobacco smoke, air pollutants, and industrial chemicals³⁴. Symptoms of COPD include chronic cough, expectoration of sputum, shortness of breath and recurrent respiratory tract infections. Both diseases are characterised by the thickening of the airway walls, goblet cell proliferation, hypertrophy of the mucous glands, and obstruction of the airway lumen caused by inflammatory exudates and mucus. However, in COPD inflammatory processes involve also the lung parenchyma, leading to damage of the lung interstitium, and consequently resulting in the loss of alveolar adhesions and reduced elastic recoil of the lungs^35,36. Thus, despite similar symptoms initially, the course of those two diseases is different. Both, however, are characterised by the increase in spasticity during exacerbation, which may be induced by external factors.

The effects of various environmental factors on the human body can be assessed with the use of basic physical parameters describing atmospheric conditions such as ambient temperature, relative humidity, wind speed, and many others, or chemical parameters such as C02 or PM10 concentrations in the air, whose common feature is that they are easily obtained using direct measurement (simple parameters). A whole range of complex indexes can be also used for this purpose, which a priori take into account numerous components³⁷. Our interest has been drawn by one of these – the Humidex. This index describes the perceived human comfort as influenced by two basic factors: air temperature and humidity³⁸. In our belief, apart from the Universal Thermal Climate Index (UTCI), whose usefulness in asthma has already been confirmed, it is one of the more interesting tools to assess human physiological responses to heat stress^39,40.

The understanding of the impact of climate change and air pollution on obstructive diseases is crucial to designing preventive and therapeutic strategies that can improve patients’ quality of life and reduce the burden of these diseases on healthcare systems. This paper presents the results of the study investigating the relationship between meteorological parameters and the occurrence of asthma and COPD in the climate of Central Europe, highlighting the significance of their monitoring and its adaptation in the context of global environmental changes.

The aim of this study was to investigate the relationships between meteorological parameters, levels of PM2.5 and PM10, and the frequency of hospital admissions for COPD and asthma. By analyzing the lagged effects of air pollutants (PM2.5, PM10) and meteorological factors on hospital admissions and quantifying relative risks (RR) across lag periods (0–10 days), the study seeks to identify critical windows of vulnerability, explore the temporal dynamics of environmental exposures, and provide actionable insights for public health interventions to mitigate respiratory health risks. Additionally, the study aims to determine whether the epidemiology of COPD and asthma exhibits similarities or differences based on these factors. Furthermore, the potential use of the Humidex as a variable in the epidemiological analysis of exacerbations of obstructive lung diseases was also explored.

Materials and methods

Study population

Data analysed in this study were obtained from medical records of patients hospitalised in the Warmia and Mazury Centre for Pulmonary Diseases in Olsztyn, Poland, in the period from 1 January 2012 till 31 December 2021. We examined information regarding the hospitalisation of patients diagnosed with J44 (chronic obstructive pulmonary disease) and J45 (asthma) according to the ICD-10 classification. Planned admissions related to desensitisation, rehabilitation, and treatment according to the protocol of the drug programme were extracted from the collected data. Consequently, only hospitalisations related to the onset of asthma or COPD symptoms were singled out, eliminating planned admissions not directly related to exacerbations of these conditions.

To ensure that the analysis was based on patients residing in the area for which air quality data were collected, we included only those with confirmed residency within Olsztyn, as determined by hospital admission records linked to residential postal codes. Cases where residency was uncertain or could not be verified were excluded from the study.

Furthermore, only adult patients aged 18 years or older were included in the analysis. After the initial selection, from the original 886 COPD records and 1941 asthma records, a database was created including 862 COPD records and 1419 asthma records. Then the collected data were mapped on meteorological data.

Meteorological data

Meteorological data for the period of 2012–2021, obtained from the Institute of Meteorology and Water Management of the National Research Institute (Poland) were processed for the needs of our study. The data included values of daily mean, maximum and minimum air temperature, wind speed, relative humidity, and precipitation. The Humidex (short for humidity index) expressed in °C was used to determine the human-felt temperature, the perception of which is formed under the influence of ambient thermal and humidity conditions, and whose scores accounting for air temperature (^oC) and relative humidity (https://www.ccohs.ca/oshanswers/phys_agents/humidex.html) were calculated with the use of the BioKlima programme (https://www.igipz.pan.pl/BioKlima.html)^41,42. Humindex was employed to determine whether thermal stress of the organism is related to asthma and COPD. Data regarding the levels of PM2.5 and PM 10 were obtained from the portal (www.powietrze.gios.gov.pl) administered by the Chief Inspectorate of Environmental Protection in Poland⁴³.

According to the updated Köppen-Geiger climate classification, the research area, that is north-eastern (NE) Poland, is characterised by a warm-summer humid continental climate (Dfb)⁴⁴. The annual pattern shows an evident seasonality of weather conditions, with the occurrence of four seasons. During the study period (2012–2021), the average annual air temperature was 8.6 °C, January was the coldest month with the average temperature of -1.9 °C, and July was the hottest month with the average temperature of 18.8 °C. The average annual precipitation was 664 mm (SD = 133 mm), and relative humidity was 78% (SD = 2%). The annual records of air temperature and precipitation in the study period in relation to air pollution with PM2.5 and PM10 are presented in Fig. 1.

Fig. 1.

Meteorological conditions (air temperature and precipitation) against air pollution with PM2.5 and PM10 in the period of 2012–2021 in Olsztyn (NE Poland).

The analysis of weather conditions was carried out for each month and accounting for thermal seasons according to the following criteria: warm period (thermal summer) with average daily temperature > 15 °C; cold period (thermal winter, pre-winter, pre-spring) with average daily temperature < 5 °C⁴⁵.

Meteorological conditions and air quality in the study period are presented in Table 1.

Table 1.

Average monthly values of selected atmospheric environment parameters in the period of 2012–2021 in relation to asthma- and COPD-related hospital admissions.

Parameter	Months												Average (sum)
	Jan	Feb	Mar	Apr	May	Jun	Jul	Aug	Sept	Oct	Nov	Dec
Sum of asthma admissions	151	145	140	112	115	94	95	103	119	129	110	106	118 (1419)
Sum of COPD admissions	107	89	99	64	71	58	71	53	58	78	58	56	72 (862)
Average temperature, oC	-1.9	-0.7	2.8	8	13.2	17.3	18.8	18.3	13.8	8.7	4.5	0.9	8.6
Minimum temperature, oC	-4	-3.3	-0.9	2.9	7.7	11.9	14	13.3	9.6	5.4	2.5	-1.1	4.8
Maximum temperature, oC	0.3	2.3	7	13.5	18.9	22.9	24.2	23.8	18.9	12.6	6.6	2.9	12.8
Wind speed, m/s	3.3	3.1	3.2	3.2	3	2.8	2.7	2.3	2.6	2.9	3	3.4	3
Relative humidity, %	88	83	74	67	69	70	74	73	79	84	90	89	78
Humidex	n/a	n/a	n/a	n/a	17.3	23.1	25.6	25	19	n/a	n/a	n/a	22.0
Precipitation, mm	44.2	29.8	37.5	35.2	59.7	69.0	99.0	66.9	66.8	65.8	44.3	46	664.2
PM2.5, µg/m3	24.8	27.1	20.6	14	10.6	9.2	10	9.8	10.4	17	19.5	20.4	16.1
Total number of days with PM2.5 > 25 µg/m3*	120	125	94	26	2	0	2	2	2	55	78	82	588
No days with asthma at PM2.5 > 25 µg/m3	56	73	47	6	1	0	0	0	1	19	29	26	258
% of asthma admissions (PM2.5 > 25 µg/m3)	37.1	50.3	33.6	5.4	0.9	0.0	0.0	0.0	0.8	14.7	26.4	24.5	18.2
No days with COPD at PM2.5 > 25 µg/m3	44	34	34	10	2	0	0	0	0	21	13	12	170
% of days with COPD at PM2.5 > 25 µg/m3	41.1	38.2	34.3	15.6	2.8	0.0	0.0	0.0	0.0	26.9	22.4	21.4	19.7
PM10, µg/m3	29.7	31.6	26.6	21.8	17.3	15.5	15.8	16.3	17.9	24.9	26	23.8	22.3
Total number of days with PM10 > 50 µg/m3**	37	39	31	7	1	0	0	0	0	17	13	9	154
No days with asthma at PM10 > 50 µg/m3	18	24	19	1	0	0	0	0	0	7	2	3	74
% of asthma admissions (PM10 > 50 µg/m3)	11.9	16.6	13.6	0.9	0.0	0.0	0.0	0.0	0.0	5.4	1.8	2.8	5.2
No days with COPD at PM10 > 50 µg/m3	17	20	17	3	0	0	0	0	0	5	3	0	65
% COPD admissions (PM10 > 50 µg/m3)	15.9	22.5	17.2	4.7	0.0	0.0	0.0	0.0	0.0	6.4	5.2	0.0	7.5

Significant values are given in bold.

*Thresholds set for Interim Target 4: *PM2.5 24-hour mean is 25 µg/m³ and **PM10 24-hour mean is 50 µg/m³; n/a – not applicable.

Statistical analyses

Values of selected meteorological, biometeorological, and air pollution parameters were used to determine the relationship between the conditions of the atmospheric environment and the occurrence of respiratory diseases. The data used for analytical purposes were obtained from a comprehensive dataset containing daily records of asthma and COPD patient counts along with various explanatory factors from three main categories: (1) Meteorology represented by average temperature (T_avr), humidity, and total precipitation; (2) Biometeorology represented by Humidex values; (3) Air pollution represented by concentrations of particulate matter (PM2.5 and PM10).

To examine the lagged effects of meteorological variables and particulate matter (PM) on daily asthma admissions, we employed distributed lag models (DLMs) using generalized linear models (GLMs) with a Poisson distribution and log-link function. DLMs are well-suited for time-series health studies, as they capture both immediate and delayed effects of exposures on health outcomes⁴⁷. Predictors included meteorological parameters (air temperature, total precipitation, relative humidity, wind speed) and air pollution (PM2.5 and PM10) with lag times ranging from 0 to 10 days. Additionally, the biometeorological factor expressed by Humidex was considered in DLM analysis for warm period. Lagged variables for each predictor were generated by shifting the daily values up to 10 days prior to each outcome observation.

The outcome variables were the daily counts of asthma and COPD, modeled as a function of lagged predictors, fitted separately for each lag and predictor in cold and warm periods. The relative risk (RR) and 95% confidence intervals (CIs) were computed for each lagged variable, with , where represents the model-estimated coefficient. Significant associations (p < 0.05) were identified to explore immediate and delayed health impacts. Analyses were performed using Python’s statsmodels library⁴⁸. All models controlled for baseline admissions via an intercept term. Results were interpreted with consideration of potential collinearity, seasonal patterns, and other temporal confounders⁴⁹.

To evaluate temporal differences among asthma- and COPD-related admissions, distinguished by meteorological, and biometeorological factors as well as air pollution, analysis of variance (one-way ANOVA), followed by a post-hoc Duncan’s Multiple Range Test (MRT) (p ≤ 0.05), was performed. Pearson’s correlation coefficients were calculated for asthma- and COPD-related admissions and environmental factors. Prior to the analyses, the dataset aggregated by month was checked for normality using the Shapiro-Wilk test at p < 0.05.

In this study, we employed a Sankey diagram to visually represent the hierarchical influences of different explanatory factors on the number of patients with asthma and COPD⁵⁰. Correlation analysis was conducted to determine the strength and direction of the relationships between each explanatory factor and the patient counts. The factors were then ranked (in percentages) based on the absolute values of their correlation coefficients to identify the most influential factors. A Sankey diagram consists of two panels: panel 1 presents the associations between asthma and COPD patients with individual explanatory factors, while panel 2 presents explanatory factors grouped to provide more general insights. The diagram was constructed using the Plotly library in Python⁵¹.

Results

In the study, we examined 1419 records of patients with asthma (550 males and 869 females) and 862 records of patients with COPD (405 males and 457 females). The detailed age distribution of the patient population is provided in Table S4. In the asthma group, the average age of females was 59.51 years (SD: 14.99 years), and of males it was 58.39 years (SD: 14.62 years). The average length of hospital stay was 4.25 days (SD: 3.77 days) for females and 5.23 days (SD: 5.40 days) for males. The most frequent comorbidities in this patient group were thyroid disorders (290 cases), arterial hypertension (653 cases), and cardiac insufficiency (75 cases). In the COPD group, the average age of females was 69.38 years (SD: 7.86 years), and 69.68 years (SD: 9.52 years) for males. The average length of hospital stay was 13.96 days (SD: 6.91 days) for females and 12.92 days (SD: 7.56 days) for males. The most frequent comorbidities included thyroid disorders (150 cases), arterial hypertension (467 cases), and cardiac insufficiency (137 cases). Both asthma and COPD patients were most often hospitalised during winter months.

The number of hospital admissions of patients with exacerbation of asthma and COPD was largely dependent on atmospheric conditions and air pollution represented by the levels of PM2.5 and PM10. This is demonstrated with a Sankey diagram (Fig. 2). The diagram effectively illustrates these relationships, providing a clear hierarchical visualization of how specific factors contribute to the patient outcomes. Each factor is quantitatively linked to the diseases, with the width of the connections representing the strength of the contribution. The contribution of air pollution to the incidence of asthma and COPD is highest for overall air pollution (85.9%), with PM10 slightly prevailing over PM2.5 (45.6% and 40.3%, respectively). As regards meteorological factors (81%), air temperature contributed in 33.2%, relative humidity in 10.1%, while precipitation in 16.4%. Humidex as a biometeorological indicator contributed equally to both respiratory diseases, comprising the total values of 33.1%.

Fig. 2.

Sankey plot for the relationships between respiratory diseases and environmental factors (21 flows between 12 nodes). Numbers at each node denote % contribution of each factor. Total inputs = Total outputs = 200.00%.

While the diagram highlights the significant role of both air pollution and meteorological factors in affecting respiratory health, single explanatory factors show either positive or negative relationships with asthma and COPD incidents (Table S1). Positive correlations were revealed between respiratory diseases and PM2.5 (r = 0.205 for asthma and r = 0.366 for COPD, p < 0.05), PM10 (r = 0.217 for asthma and r = 0.435 for COPD, p < 0.05). Negative correlations were found between respiratory diseases and air temperature (r = − 0.207 for asthma and r = − 0.252 for COPD, p < 0.05) (Table S1).

Data regarding daily average PM concentrations clearly show the coexistence of low air temperatures and high PM levels.

High levels of PM2.5 and PM10 are strongly correlated with more frequent asthma- and COPD-related hospital admissions. The average concentration of PM2.5 during the study period was 16.1 µg/m³. The average monthly concentrations exceeded daily and annual norms in winter months. The greatest % ratio for PM2.5 above the normal values, i.e. 27.1 µg/m³ (WHO), occurred in February, whilst the annual norm for PM10, i.e. 10 µg/m³ (WHO), was mostly exceeded from January till March (Table 1)⁴⁶.

As regards obstructive lung diseases, the correlation between PM2.5 and PM10 with the number of hospital admissions demonstrates evident differences in the winter half-year and the summer half-year. The linear regression analysis for total monthly asthma-related hospital and average concentrations of PM (Fig. S2) shows a weaker correlation as compared to COPD, for which the values are high in wintertime: R² for PM2.5 = 0.212, and for PM10 = 0.352. In summertime, no significant relationship was revealed between PM levels and the occurrence of obstructive lung diseases. Moreover, in the months of the winter half-year, the values of the WHO norm for PM2.5 was exceeded 4-fold, and for PM10 over 2.5-fold (Fig. 3, Fig. S2).

Fig. 3.

Monthly variations in asthma and COPD admissions in relation to PM2.5 and PM10 levels.

As presented in Figs. 2 and 4, meteorological conditions are significant in increasing exacerbations of asthma and COPD, particularly in wintertime.

Fig. 4.

Relationships between the average temperature and asthma and COPD hospital admissions.

When analysing relationships between meteorological parameters and the occurrence of various diseases frequently complex biometeorological indexes are employed³⁹. One of these is the Humidex that measures human thermal-humidity discomfort⁵². In summertime when Polish climate is dominated by high temperatures and high humidity, asthma- and COPD-related hospital admissions were recorded when the following Humidex classes: slight, moderate and strong discomfort were observed (Fig. 5). No Humidex values representing very strong discomfort due to thermal-humidity conditions were revealed.

Fig. 5.

Distribution of asthma and COPD admissions against Humidex ratings.

(source: https://www.ccohs.ca/oshanswers/phys_agents/humidex.html). The range of Humidex smaller than 29 denotes no discomfort, 30–34 some possible slight discomfort, 35–39 moderate discomfort, 40–45 possible strong discomfort, more than 46 very strong discomfort.

Aggregate data regarding the distribution of the number of hospital admissions of patients with asthma and COPD exacerbation in relation to Humidex classes for the summer period (Tmax > 22 ^oC) are presented in Fig. 5. Worthy of notice is the nearly identical distribution of patients with asthma and COPD against the four Humidex groups. During the study period, out of all hospitalised patients with asthma (1419) and COPD (862), 17% and 18%, respectively, were admitted during the summer when Tmax > 22 °C. The highest patient numbers, approximately 30%, were recorded when thermal-humidity discomfort was minimal or unnoticeable (Humidex < 29). However, a considerable proportion of ‘summer’ patients, 32% for asthma and 29% for COPD, presented during periods of slight to moderate discomfort, with Humidex values in class II-IV (30–45). Notably, during periods of strong discomfort (Humidex class IV, 40–45), only four patients were admitted to hospital.

The average number of hospital admissions in July and August (the highest Humidex values) is statistically greater than in the remaining months of the warm period (p < 0.05). An analogous and very similar relationship was noted for PM concentrations in the cold period (Fig. 6).

Fig. 6.

Fluctuations in the number of asthma and COPD hospital admissions and Humidex values in the warm period (in particular months), when the average daily temperature exceeded 15 °C. B. Fluctuations in the number of asthma and COPD hospital admissions in the cold period (in particular months), when the average daily temperature was below 5 °C. The same letter symbols denote homogeneous groups, not statistically significantly different in Duncan’s MRT at p < 0,05.

This index is also correlated with the number of hospitalisations due to obstructive lung diseases: r² = 0.956 for asthma and r² = 0.659 for COPD (Fig. S1).

The lagged risk analysis for asthma and COPD admissions, considering both meteorological factors and air quality predictors (PM10 and PM2.5) across specific lag times (0–10 days), revealed several significant associations (Fig. 7, Table S3). For asthma admissions during the cold period (days with temperatures < 5 °C), PM10 showed a significant increasing association. The risk of admission increased by 2% on the day of admission (RR: 1.02, 95% CI 1.00–1.04, p = 0.049) and by 4% at lag + 7 days (RR: 1.04, 95% CI 1.01–1.06, p = 0.001). Conversely PM2.5 demonstrated a slightly protective effect at lag + 7 days, reducing asthma admissions by 3% (RR: 0.97, 95% CI 0.94–0.99, p = 0.005). During warm periods, significant RR predictors of asthma admissions included the Humidex at lag + 1 day, PM10 at lags + 7 and + 10 days, relative humidity (RH) at lag + 0 and + 1 day, average daily air temperature (Tavr) at lag + 0 day and minimum daily air temperature at lag + 7 days. Notably, the Humidex was associated with a 13% short-term increase in asthma admissions at lag + 1 day (RR: 1.13, 95% CI 0.97–1.31, p = 0.012). PM10 and Tmin contributed to increases in asthma admissions by 7% and 17%, respectively, at lag + 7 days. RH had a protective effect, reducing asthma admissions by 3% over short-term lags (0–1 days).

Fig. 7.

Effects of daily PM2.5, PM10 and humidex on relative risk (RR) of asthma and CODP admissions over lag times of 0–10 days. The statistical significance of lagged RRs of asthma and COPD hospitalization estimates for significant predictors is presented in Table S3.

For COPD admissions, significant associations with winter air quality emerged at longer lags compared to asthma. PM10 increased COPD admissions immediately, with a 5% increase observed at lag + 0 day (RR: 1.05, 95% CI 1.02–1.08, p < 0.001) and a 6% increase at lag + 1 day. The effect of PM10 ranged from a 5% increase at lag + 2 days to a 3% increase at lag + 10 days. PM2.5, however, exhibited a slight protective effect, reducing COPD admissions by 3–7%. Tmin demonstrated a protective effect on COPD admissions at lags + 1 and + 2 days but increased the risk by 10% at lag + 6 days (RR: 1.10, 95% CI 1.00–1.21, p = 0.05).

During warm periods, short-term lags (0–1 days) of the Humidex were associated with increased risks of COPD admissions, with relative risks of 18% and 31%, respectively. Significant predictors also included Tmax at lag + 3 days (RR: 1.32, 95% CI 1.05–1.65, p = 0.035) and wind speed (Vw) at lag of 4 days since exposure (RR: 1.50, 95% CI 1.08–2.09, p = 0.035). Similarly to asthma, RH had a protective effect reducing COPD admissions by 4% and 5% at lags + 1 and + 2 days since exposure, respectively.

Discussion

The results of our study revealed a marked seasonal variability as regards emergency hospital admissions due to asthma and COPD. This relationship can be described with a simple observation: ‘fewer in summertime, more in wintertime’ (Figs. 3 and 4). Yet, in order to fully understand why this is so and which of the analysed parameters is of the greatest significance, different meteorological parameters require a closer scrutiny.

Authors of a number of studies describe relationships only between single meteorological parameters and asthma and COPD, primarily indicating a positive correlation as regards high and low temperatures^53–55. They highlight that nonoptimal ambient temperatures can impact lung function negatively⁵⁶. For instance, studies conducted in Spain by Hervása et al. revealed relative humidity and maximal wind speed as risk factors for asthma in children⁵⁷. These authors found no significant relationships between asthma and atmospheric pressure and rainfall.

Undoubtedly, both cold and heat waves are associated with a higher risk of exacerbations of obstructive lung diseases⁵⁸. Our study results confirm a seasonal increase in the number of hospital admissions due to these diseases, especially in wintertime when ambient temperature is lower (Fig. 4).

Shi et al., based on the data collected in Ganzhou (China), reported a positive correlation between COPD and extremely low ambient temperature (in their study that is temperature < 3.3 °C) and extremely low humidity (< 47.8%)⁵⁹. The comparison of our results with those obtained in Spain and in Ganzhou, China, requires accounting for climatic differences. In Poland, cold waves are more intense, and wintertime average temperatures are considerably lower than in these two other regions⁶⁰. Ganzhou is located in the humid subtropical climate (Köppen – Cfa), and Olsztyn (Poland) in the warm-summer humid continental climate (Köppen – Dfb). During our study period, the maximum lowest recorded temperature was − 27^oC, while the lowest relative humidity was 33%. The comparison of these data is therefore inherently subject to a reference-point error. Analogously, Fishe et al., based on data obtained in Florida, USA, characterised by the humid subtropical and tropical savanna climates (wg Köppen – Cfa and Aw), reported a relationship between exacerbations of respiratory diseases and heat waves in summertime and a reversal relationship in wintertime⁶¹. But they discuss the area with average temperatures about 30^oC in the summer and 18 °C in winter. In Poland, and especially in its north-eastern region, the former temperatures, despite climate change, are regarded as extreme (Fig. 1). The Polish climate shares more similarities with that in Finland (Köppen De – continental subarctic or boreal climates), where temperatures drop below 0 °C in wintertime, and a similar relationship with the cold as in NE Poland can be observed⁵⁵. In our study, we did not find any statistically significant relationships for precipitation and relative humidity (Table S1).

The comparison of the impact of air pollution with PM, especially with PM2.5, on exacerbations of obstructive lung disease is equally problematic. In the previously referred to study from Florida, USA, seasonal variability of PM2.5 levels (higher in wintertime) was reported, but their average values are often several times smaller as compared to central and eastern Europe (Table 1)⁶¹. It is hardly surprising, then, that in our study PM2.5 and PM10 are the strongest factors impacting on asthma- and COPD-related hospital admissions (Table S1; Figs. 2 and 6).

From the perspective of public health, of particular importance is the relationship between the number of exacerbations of obstructive lung diseases when air quality is poor because of PM concentrations far exceeding the established norm (even 4-fold) in wintertime (Fig. 3)⁶².

Considering meteorological and biometeorological parameters as well as air pollution with PM, we concluded that PM (PM2.5 and PM10) (Tables S1, S2) contributes the most to the increase of hospital admissions due to exacerbations of asthma and COPD (Fig. 2). This is consistent with other studies that report that fine particulate matter can penetrate deep into the lungs and settle in the airways and lung tissue, thus leading to an abnormal immune inflammatory response and impaired epithelial cell function. Consequently, this results in, i.a., inflammation and exacerbation of obstructive diseases^63–65. It is important to remember that concentrations of PM2.5 and PM10 show an evident association with atmospheric conditions (Fig. 1, Table S1) and depend on low ambient temperatures, wind speed, and precipitation^66–71.

Our study highlights the complex and time-dependent relationships between environmental exposures and asthma and COPD exacerbations. The strong association between PM10 and increased asthma admissions at lag + 7 aligns with previous research indicating that coarse particulate matter in the air contributes to respiratory inflammation and hospitalizations, particularly for asthma⁷². The association between PM10 and increased COPD admissions shows immediate and short-term delayed impacts on respiratory health what is consistent with the outcomes of the same authors who demonstrated the adverse effects of coarse particulate matter on respiratory health through mechanisms of airway inflammation and oxidative stress⁷². In contrast, the protective effects of PM2.5, though counterintuitive, may reflect collinearity with PM10 or interactions with other unmeasured confounders. This underscores the need for caution when interpreting isolated pollutant effects in multivariable contexts⁴⁹. The delayed impacts of temperature and precipitation suggest potential indirect pathways, such as changes in air quality, allergen concentrations, or viral activity, which merit further exploration⁷³.

Future studies should integrate stratified analyses by season and explore non-linear relationships using distributed lag non-linear models (DLNMs). Moreover, our results need to be referred to the parameter that we did not examine, that is a seasonal variability of upper respiratory tract infections. Exacerbations of COPD are more frequent in wintertime (Fig. 6, Fig. S2)⁷⁴. In Poland, season-related infections occur from October till April, with the onset of infections generally 10 days following the first autumn break⁷⁵. Studies on animal models carried out by Lowen et al. revealed that low temperature and lower relative humidity increase the risk of the influenza virus transmission⁷⁶. The stability of aerosolized virus is highest at relative humidity of 20–40%, lowest at 50%, and acceptable at 60–80%. The crucial role in the virus transmission is played by the so-called ‘droplet nuclei’ whose behaviour changes depending on humidity⁷⁷. The virus infectivity is highest at low relative humidity and decreases as humidity increases. An increase in the number of upper respiratory tract infections may contribute to the larger number of exacerbations of asthma and COPD^78,79.

In our study, PM2.5 and PM10 levels are strongly correlated with each other (0.944, p < 0.001) (Table S1), but they are also strongly negatively correlated with the average ambient temperature (− 0.799 and − 0.693, p < 0.01; respectively). Moreover, normal levels of PM are exceeded only in the cold period (Table 1; Fig. 3). The most interesting result is, however, the relationship between PM and the number of hospital admissions due to obstructive lung diseases calculated for the winter season (meteorological season, not the calendar one). Correlations are strong and amount to r = 0.283 for asthma and PM2.5, r = 0.312 for asthma and PM10, r = 0.491 for COPD and PM2.5, and r = 0.574 for COPD and PM10, with p-value always being < 0.05 (Table S2).

There is a lack of publicly available cohort data at the European level that assesses and compares the impact of biometeorological factors on the frequency of exacerbations in obstructive respiratory diseases. Nevertheless, findings from local studies provide significant insights into this topic. Hoffmann et al., in a study published in Respiratory Medicine in 2022, based on research conducted in Berlin, demonstrated that wind speed significantly affects the frequency of COPD exacerbations⁸⁰. Similarly, an analysis of a Hungarian subpopulation revealed that a high diurnal temperature range and daily variability in dew point contribute to a 4.5% increase in the risk of COPD exacerbations, translating into a higher number of emergency department visits related to this condition⁸¹. In another study from 2022, the same authors additionally identified low temperatures, low dew points, and high atmospheric pressure as risk factors for a greater number of emergency department visits due to COPD⁸². Considering these findings, it appears justified to integrate various biometeorological parameters into more comprehensive indices, such as the Universal Thermal Climate Index (UTCI) or Humindex. This approach could enable more precise risk forecasting in the context of obstructive respiratory diseases.

Our study revealed that while respiratory diseases are determined by air pollution in wintertime, in the summer the increase in asthma- and COPD-related hospital admissions is caused by weather conditions. We arrived at this conclusion using the Humidex for the warm period, that accounts for both ambient temperature and relative humidity⁵². The analysis between Humidex values and asthma and COPD yielded significant relationships (Figs. 5 and 6). Regression analysis between monthly Humidex values, reflecting warmer and more humid atmospheric conditions in summertime months (May-October), and the number of hospital admissions both due to asthma and COPD demonstrates a strong cause-and-effect relationship, confirmed by a high coefficient of determination R² of, respectively, 0.659 and 0.956 (Fig. S1). Additionally, analysis of variance (one-way ANOVA, Dunacan post-hoc test, p < 0,05) confirmed that the average number of hospital admissions in the warmest months, i.e. July and August (the highest Humidex values), was statistically greater than in the remaining months of the warm period (Fig. 6).

To the best of our knowledge, this is the first study to use the Humidex (a biometeorological index) to assess the frequency of exacerbations of asthma and COPD. We believe that this index is of high utility potential as a predictor of exacerbations of obstructive lung diseases during the summer. Nevertheless, considering the geographical variability of climatic conditions and the adaptation of inhabitants to specific conditions, it requires additional site-specific verification.

Limitation

Our study, despite providing valuable information as regards the impact of meteorological conditions and air pollution on the number of hospital admissions due to COPD and asthma in Olsztyn, has some serious limitations that need to be accounted for when interpreting the results.

First, data were collected exclusively in Olsztyn, which may limit the generalisation of the results to other regions characterised by different climatic conditions, air pollution levels, or demographic structure. Differences in infrastructure, heating methods, or industrial activities may affect air pollution levels and their impact on public health.

Second, analyses were based on air pollution levels and meteorological conditions measured in specific locations, which does not account for individual differences in exposure to pollution. Differences in lifestyle, time spent outdoors, workplace conditions and ventilation systems may considerably impact on the actual risk of individual patients.

Additionally, data regarding air pollution and meteorological conditions were available only as regards particular time intervals. The lack of continuous measurements may have resulted in under- or overestimation of the impact of short-term pollution peaks on the health of patients with COPD.

Furthermore, the study focuses solely on hospitalized cases, which represent severe exacerbations. This approach excludes milder cases treated in outpatient settings, potentially limiting the comprehensiveness of the findings. Future studies should incorporate outpatient data to capture the full spectrum of disease severity and environmental impacts.

We also attempted to limit the influence of potential confounding factors by analyzing the data by season and considering annual trends in hospitalizations. However, residual confounding related to unmeasured factors, such as individual socioeconomic status, smoking habits, or pre-existing comorbidities, may still be present and could have influenced the results.Finally, the lack of data on patients’ adaptive behaviours, such as using air purifiers, wearing protective masks, and changing their lifestyle on high pollution days, may have affected the accuracy of health risk estimates.

Despite these limitations, our study provides meaningful insights into the relationship between air pollution, meteorological conditions, and respiratory health, offering a foundation for further research in this field.

Conclusions

1. Air pollution is the major risk factor for exacerbations of asthma and COPD. We confirmed a strong correlation between high levels of particular matter (PM2.5 and PM10) and an increase in hospital admissions due to exacerbations of asthma and COPD, especially in wintertime.

2. Meteorological conditions and respiratory diseases: such parameters as temperature, humidity, and wind speed, significantly impact on the occurrence of asthma and COPD. In wintertime, this impact is most likely secondary and related to their direct effect on particulate matter concentrations. The assessment of the impact of meteorological factors on the occurrence of obstructive lung diseases should take into account seasonality. The understanding of these relationships is crucial in the context of progressive climate change. Our findings indicate that the health impacts of meteorological variables and particulate matter in the air are not only immediate but may also manifest with delays of several days.

3. A biometeorological index – Humidex – that integrates the impact of temperature and humidity on thermal comfort may be a useful tool in assessing the influence of meteorological conditions on human health. Higher Humidex values in summertime are associated with a two-fold increase in the number of hospital admissions due to obstructive lung diseases. Further studies are necessary in order to confirm these results and the use of Humidex in preventive strategies.

4. The results of our study highlight the need for further research covering different geographical regions and longer observation periods. A comprehensive approach, taking into account both air pollution factors and meteorological conditions, will enable the development of effective preventive and therapeutic strategies that will improve the quality of life of patients with asthma and COPD and reduce the burden on healthcare systems.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Author contributions

All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by A.RW., E.D., K.G.L. The first draft of the manuscript was written by A. R.W. and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Funding

The authors declare that no funds, grants, or other support were received during the preparation of this manuscript.

Data availability

The datasets analysed during the current study are not publicly available due to the sensitive nature of medical data and patient confidentiality regulations but are available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.

Ethical statement

All procedures were performed in compliance with relevant laws and institutional guidelines and have been approved by bioethics committee at the Warmia-Masuria Medical Chamber in Olsztyn, Poland on June 17, 2024 (L.Dz. WMIL-KB/98/2024). As a retrospective research informed consent was waived by bioethics committee at the Warmia-Masuria Medical Chamber in Olsztyn, Poland.