The impact of exposure to particulate air pollution from non-anthropogenic sources on hospital admissions due to pneumonia

To the Editor:

Community-acquired pneumonia is a significant cause of morbidity and mortality among older adults [1]. The role of air pollution as a risk factor for pneumonia hospitalisations and mortality has been investigated [2, 3] with most evidence coming from studies in North American and European cities, where anthropogenic sources are predominant in generating air pollution.

Particulate non-anthropogenic air pollution originating from dust is a common public health risk. Being located between the Sahara and the Arabian deserts (the world's largest dust-belt), the Negev region of Israel is exposed to extremely high levels of particulate matter originating from natural dust storms. During dust storms in this region, particulate matter levels can significantly exceed those defined as acceptable in terms of air quality and human health (50 µg·m⁻³) with hourly concentrations of 100–5000 µg·m⁻³ [4]. Dust particles reach the southeastern Mediterranean by two main trajectories: one from the west (North Africa–Sinai–Negev) and the second from east (Arabian Desert–Negev) [5] with dust particles having somewhat different mineralogical and chemical compositions. In the winter, dust storms from the western sources are most prevalent, whereas dust storms from both west and east directions are frequent in the spring depending on the synoptic system [6]. Most of the intense storms with higher dust concentrations are associated with the western sources [4].

The unique combination of a centralised modern medical system and urban population residing in this arid and hot region makes the Negev an ideal “environmental laboratory” for studying the health effect of global environmental change such as desertification and global warming.

Previously, we have shown an association between dust exposure and risk of hospitalisation due to chronic obstructive pulmonary disease (COPD) [7] and asthma [8]. The aims of the present analyses were 1) to investigate the association between hospitalisations due to pneumonia and exposure to particles with a 50% cut-off aerodynamic diameter of 2.5 µm (PM_2.5) and 10 µm (PM₁₀) and 2) to identify individual characteristics that might modify the potential health effect.

Data from 4257 patients with 5611 hospitalisations admitted to Soroka University Medical Center (SUMC; Beersheba, Israel), a 1000-bed tertiary hospital between 2003–2013 due to pneumonia (ICD-9 codes 487, 486, 481, 480.8, 514, 482.41 and 482.8) were included in this analysis. SUMC is the only hospital for a population of 730 000 in Southern Israel. The following patient level data were obtained using the centralised electronic medical records database: diabetes (ICD-9 code 250), cardiovascular disease (ICD-9 codes 390–429), hypertension (ICD-9 code 401), COPD (ICD-9 codes 490, 491, 492 and 496) and socio-demographic data, such as sex and age.

Exposure assessment was based on a hybrid satellite-based spatio-temporally resolved model incorporating daily satellite-remote sensing data at a spatial resolution of 1×1 km [9]. Briefly, we make use of a new algorithm developed by the US National Aeronautics and Space Administration (NASA): the Multi-Angle Implementation to Atmospheric Correction (MAIAC) [10], which provides aerosol optical depth (AOD) data at high resolutions. Using mixed-model frameworks, we regressed daily PM₁₀ and PM_2.5 mass concentration from the Ministry of Environmental Protection against AOD, temporal predictors (obtained from the Technion Center of Excellence in Exposure Science and Environmental Health air pollution monitoring database) and spatial predictors (obtained through the Israeli Central Bureau of Statistics and Survey Bureau mapping service). When AOD data were not available due to cloud coverage and non-retrieval days, we fitted a generalised additive model with a thin-plate spline term of latitude and longitude to interpolate the estimates. Good model performance was achieved, with out-of-sample cross validation R2 of 0.79 and 0.72 for PM₁₀ and PM_2.5, respectively. Both model predictions had little bias, with cross-validated slopes (predicted versus observed) of 0.99. More in depth details can be found in Kloog et al. [9]. The daily average concentrations of the pollutants, estimated throughout the studied region, were assigned for each patient based on proximity to his geocoded home address. Daily data on air temperature and relative humidity for the study period were obtained from a central monitoring site located in the centre of the largest city of the region. We obtained daily meteorological data including temperatures (in °C), and relative humidity from two monitoring stations located in the centre of the largest city (Be'er-Sheva) in the Negev area. PM₁₀ and PM_2.5 levels exceeded the World Health Organization-recommended daily guideline of 50 and 25 μg·m⁻³ [11] in 36% and 17.1% of the study period days, respectively. The climate in the study region is relatively hot and dry, reaching a maximal daily average temperature of 31.4°C during the study period.

The data were analysed using the case–crossover approach. This association was estimated by odds ratios with 95% confidence intervals using conditional logistic regression model. Analyses were performed using R statistical software, version 3.0 (R Foundation for Statistical Computing, Vienna, Austria).

Using the novel exposure assessment we were able to show that natural particulate matter exposure increases the risk for hospital admission for pneumonia. The association between particulate matter was observed 5 days after the exposure (OR 1.051, 95% CI 1.013–1.091, for increase in 10 units of PM_2.5). The most susceptible subgroups to the PM_2.5 and PM₁₀ exposure were male patients older than 65 years (OR 1.108, 95% CI 1.037–1.184, and OR 1.019, 95% CI 1.000–1.040, respectively) or elderly patients with cardiovascular disease (OR 1.072, 95% CI 1.023–1.123, and OR 1.016, 95% CI 1.002–1.030, respectively) (table 1).

Desertification and global warming trends pose significant global ecological and environmental problems. Proximity of the Negev area to the Sahara and the Arabian deserts as important sources of mineral dust, highlights the importance of our findings and enhance our understanding of respiratory morbidity and its association with non-anthropogenic air pollution. Our finding with respect to association between exposure to particulate matter and risk of hospitalisation due to pneumonia are consistent with the results from a study from Ontario, Canada [2], and a study from the USA [3, 12]. Thus, our study strengthens the suggestion that air pollution may act as an irritant and induce defensive responses in airways, such as increased mucus secretion and increased bronchial hyperreactivity [13]. In addition, particulate matter has been shown to produce free radicals and oxidative stress on lung cells. These reactions might lead to the tissue inflammation, resulting in exudative discharge to the alveoli. Radiologically, this will be evident as a consolidate, and the diagnosis of pneumonia can be established. Animal studies have shown an increased vulnerability to PM₁₀ in animals with cardiopulmonary disease [14] and exacerbation of ongoing pneumococcal infection after exposure to concentrated ambient PM_2.5 [15].

In conclusion, community-acquired pneumonia is a significant cause of morbidity and mortality among older adults. We found that short-term exposure to natural particulate matter increases the risk for hospital admission for pneumonia, particularly for older patients or patients with pre-existing cardiovascular disease.