Neighborhood environmental factors linked to hospitalizations of older people for viral lower respiratory tract infections in Spain: a case-crossover study | Environmental Health | Full Text
We conducted a bidirectional case-crossover study (all patients serve as their controls) in individuals aged 65 years or older who had a hospital admission due to viral LRTI in Spain during 2013–2015. This study was approved by the Research Ethics Committee (Comité de Ética de la Investigación; CEI PI 81_2021) of the Instituto de Salud Carlos III (Madrid, Spain). All the extracted information was completely anonymous and did not require the patients’ consent.
The link in space-time between environmental factors and MBDS data was established as follows: i) The environmental data [temperature (°C) and relative humidity (%)] and ambient air pollutants [SO2 (μg/m3), CO (μg/m3), NO2 (μg/m3), O3 (μg/m3), PM10 (μg/m3)] from the meteorological stations distributed throughout the territory were geolocated in space as a reference point (latitude-longitude). ii) Each patient in the study had their spatial location through the residential zip code (geographical area), from which the centroid was extracted, and geolocation in space as a reference point (latitude-longitude) was obtained. iii) Once both data sources were geolocated in space, each patient was linked to the meteorological station closest to their home. iv) The MBDS had the date of hospital admission and each meteorological station of the measurement date, so the link of dates was simple. The mean distance from each residential zip code to its nearest meteorological station was 8.99 km 95% CI (8.69, 9.28).”
This study indicates that low temperatures, high relative humidity, and high concentrations of NO2, O3, PM10, and CO are associated with increased hospital admissions due to viral LRTI in patients 65 or older. Our data support the monitoring of environmental factors to assess the risk of hospital admissions and advise minimizing exposure to air pollutants in older people.
This study was performed for all 12 months, instead of only the colder months (December–March), when there were more hospitalized patients than during the warmer months (April–November). It is so because we wanted to analyze if there were associations between outdoor environmental pollution and LRTI hospitalizations at any time of the year (cold and warm seasons). As we showed, the epidemiological wave of viral LRTI occurs during the cold months (December–March), but there were also LRTI viral infections in the other months of the year, including summer.
Changes in weather conditions affect the respiratory system enabling the spread of infection-causing pathogens [7, 11]. These changes can increase the risk of viral LRTI and cause pneumonia, bronchitis, and other respiratory tract pathologies in older adults [9, 10]. An increase in the number of inflammatory cells and fibrinogen concentration has been observed during cold exposure, damaging the respiratory system, and leading to urgent hospitalizations and possible death [7, 11]. Besides, lower temperatures increase pathogens’ stability, abundance, survival, and infectivity [7]. High humidity increases the infectivity of viruses because humidity stabilizes the droplets that carry the pathogen from person to person through the air [7]. Our study found a higher risk of hospitalization for viral LRTI among older adults ≥65 years exposed to low temperatures and high relative humidity before hospital admission. Low temperatures and high humidity are associated with a higher risk of viral LRTI [24,25,26]. Our data agree with previous data showing that ambient temperatures below the reference levels potentiate respiratory tract infections and increase hospital admissions in older adults [7, 9, 10]. However, some studies show discordant data on temperature concerning our research [27, 28], partially justified because not all regions of the world have the same seasonal pattern of LRTI, finding differences in the circulation of respiratory viruses according to geographic characteristics [29,30,31].
NO2 is an irritating pollutant related to the high traffic that penetrates deep into the lung, causing respiratory diseases, including viral LRTI [2, 6]. NO2 causes an imbalance in the Th1/Th2 differentiation (increased IL-4/IFN-γ ratio) and the activation of the JAK-STAT pathway, damaging the lung cell membrane and increasing airway inflammation [32]. Our study found an elevated risk of hospital admissions due to viral LRTI associated with short-term exposure to NO2 in older people. Our findings are consistent with other reports on short-term [33] and long-term [34] exposure to outdoor NO2 and COVID-19 in older people with respiratory failure. It may be due to NO2 inhalation oxidizing proteins and lipids and altering the immune system [35]. However, discordant studies did not show any association between outdoor NO2 and LRTI in older people [36], suggesting that outdoor NO2 may impact viral LRTI in combination with other environmental pollutants rather than NO2 itself [37, 38].
O3 is a potent and toxic oxidizing gas that arises in the stratosphere or the troposphere after various reactions from photochemical smog [2, 6]. Its absorption usually occurs by inhalation, which can penetrate deeply into the lungs due to its low solubility in water. O3 reacts with cells lining the airways, stimulating their receptors and nerve endings and leading to oxidative stress, inflammation, and decreased total lung capacity [39]. Our findings are consistent with previous reports that found significant associations between short-term exposure to ambient O3 and increased risk of pneumonia hospital admissions among older adults [40, 41]. However, discordant studies did not find a relationship between outdoor O3 and LRTI hospital admissions [42, 43].
In our study, O3 was the most critical environmental factor because it was strongly associated with viral LRTI hospital admissions, increasing with longer delay times. Interestingly, the epidemiological wave of viral LRTI occurred during the cold months (December–March), when O3 levels were lower compared to the warmer and hotter months (May–September) when older people spend much more time outdoors. The impact of O3 on the LRTI severity depends on several factors, such as viral epidemiological characteristics and O3 exposure (outdoor activities, O3 concentrations, exposure time, and susceptibility to air pollutants). The O3 sources in winter are practically the same as in summer, mainly for chemical reactions between O3 precursors in the atmosphere, such as NOX and volatile organic compounds from combustion associated with cars, planes, trains, power plants, oil refineries, factories, or evaporation of organic compounds from standard consumer products (paints, cleaning products, and solvents) [44]. O3 levels increase when their precursor emissions react in the presence of sunlight, warm temperatures, and light winds (warm seasons). When winter arrives, the temperature and solar radiation decrease, and most of the warm air rise, displacing O3 to the upper layers of the atmosphere [44]. However, it should also be noted that Spain has a Mediterranean climate characterized by hot summers, low winds, and intense solar radiation; and cool winters that are slightly cloudy and rainy. It affects the physical-chemical processes of O3 formation, which is why O3 continues to be generated in the cold months, with production peaks on specific days when the temperature and solar radiation are higher [45].
PM10 can be inhaled through small liquid or solid droplets that invade the lungs and cause long-term severe respiratory problems. PM10 has a long half-life, allowing it to spread to distant destinations, where people become exposed [2, 6]. PM10 causes lung damage by increasing inflammation and airspace epithelial permeability [46]. Several studies have demonstrated an association between particulate matter up to 2.5 μm in size (PM2.5) and emergency visits for severe viral respiratory diseases in older patients [34, 47, 48]. Unlike our study, other studies reported no increase in LRTI hospitalizations related to PM10 [8], likely due to varying ambient PM10, weather conditions, and co-pollutants in different geographic areas.
CO is generated mainly during incomplete hydrocarbon combustion from internal combustion engines, waste incinerators, coal power plants, and the oil industry. CO diffuses quickly across the pulmonary membrane triggering proinflammatory responses in the airways [49]. CO is a “silent killer” that binds to hemoglobin in the blood, forming carboxyhemoglobin that displaces oxygen, reduces oxygen-carrying capacity, and decreases the release of oxygen to tissues, increasing the risk of asphyxia-related deaths [50]. Inhalation of CO can be toxic to the respiratory system, causing asthma exacerbation [51] and increased hospital admission for chronic obstructive pulmonary disease [52]. Our data concord with other studies that found an association between outdoor CO levels and hospital admissions for viral LRTI [53,54,55] and pneumonia [56]. Nevertheless, another report has not shown significant associations between CO and respiratory and LRTI hospital admissions [57,58,59]. These controversial results can be due to densely populated areas, urban congestion, and heavy traffic load, where the predominant air pollutants are NO2 and particulate matter. Therefore, the effects of the CO’s co-emission with these airborne pollutants may confound the contribution of CO in air pollution on health [60].
Strengths and limitations of the study
Our study also has several strengths that must be considered: (i) this is a nationwide study with a very high number of older adults over 65 years of age with a viral LRTI hospital admission, something challenging to reach with any other database; (ii) we use a bidirectional case-crossover design that minimizes the impact of the absence of fundamental variables in the regression analysis [21].
The most important limitations are the following: (i) The retrospective design may introduce biases; (ii) the lack of relevant clinical information for the correct interpretation of the data since medical history data (comorbidities and treatments) may affect hospital admission and a stratified analysis would have provided exciting information in this regard; (iii) the diagnostic bias because in the MBDS there was no specific code for the diagnosis of LRTI, and we used ICD-9-CM codes previously used in high impact factor publications [47, 61], but we do not really know the accuracy of the MBDS for LRTI diagnoses; (iv) the MBDS is anonymous and makes it difficult to control whether some older people over 65 have been hospitalized several times; (v) we did not analyze other emerging outdoor air pollutants, such as volatile organic compounds, including benzene; and (vi) lack of indoor air pollution data may have a significant impact on viral LRTIs because most people, especially the older population, spent more time indoors [62], facilitating the transmission of viral LRTIs among everyone.
