Norris MR, Bielory L. Allergies and Asthma in a Warming World: A Threat to Public Health. Harvard Public Health Review 2021; 34.
The prevalence of allergic diseases has risen in industrialized countries over the past 50 years.(Arbes, Gergen, Elliott, & Zeldin, 2005; Meng, Nagarajan, Son, Koutsoupias, & Bielory, 2016) Likewise, pollen counts have risen over the same period of time and are predicted to continue rising.(Zhang, Bielory, Cai, Mi, & Georgopoulos, 2015; Zhang, Bielory, Mi, et al., 2015) This is not only problematic for the millions of people who suffer from seasonal allergies affecting the eyes and upper airways (nose, sinuses, and eustachian tubes) but also for those who suffer from lower inflammatory respiratory disorders (e.g. asthma). A significant factor contributing to this observation has been climate change. Sensitivity to seasonal aeroallergens (i.e. pollen) has frequently been used to study the relationship between pollution, climate change, and allergic inflammatory respiratory disorders (e.g. rhinitis and asthma).(D’Amato et al., 2015)
Over a century of research has gone into attempts to elucidate the relationship between pollen grains and seasonal allergies. The traditional view is that people with genetic susceptibility need to reach a “threshold level” of allergen exposure before symptoms will develop (such as eye itching or nasal congestion). In this model, intact pollen grain counts are traditionally used as a surrogate for the level of allergen exposure a person will have. The Pollen Count paradigm using direct pollen counts for symptom prediction is based on multiple premises, namely (1) symptom severity has a linear correlation with atmospheric pollen grain counts and (2) pollen allergenicity (the allergen content of an individual pollen grain) is constant.
For over 50 years, we have known an individual’s intrinsic threshold level will change throughout a pollen season. This phenomenon has historically been called “priming”.(Connell, 1969) Essentially, the theory suggests receptors that evoke allergic symptoms will respond to lower concentrations of pollen allergen stimulation over time as the receptors themselves become primed for future response with each activation. This is important as it directly relates to medication utilization, thus to overall increases in healthcare costs throughout an allergy season.(Durham et al., 2014) It is important to understand that the allergic inflammatory response and resulting allergic symptoms (e.g. eye itching and redness) are biphasic, meaning they have an early-phase and late-phase response. The early-phase response typically starts within seconds to minutes and lasts for 20 to 30 minutes.(Abelson, Chambers, & Smith, 1990; Bacon et al., 2000) Late-phase responses occur between 4 and 24 hours in 33% to 100% of patients at high levels of allergen exposure.(Bonini et al., 1990; Choi & Bielory, 2008) Additionally, the allergenicity of pollen grains vary by grain, throughout a season, and by year – even if released from the same plant.(Buters et al., 2012) This stems from how a plant’s growth and production of pollen is affected by atmospheric carbon dioxide and ozone levels, seasonal and annual differences in temperature, and regional geography. As plants are exposed to warmer temperatures and higher levels of carbon dioxide we can expect to observe more vigorous plant growth, more pollen production, and more allergen per pollen grain than otherwise expected.(Zhang, Bielory, Mi, et al., 2015; Ziska et al., 2011) As our understanding of pollen’s impact on the human body and the environment’s impact on pollen production improves, we’re beginning to discern non-linear patterns linking allergy symptom severity to pollen grain exposure.
Most of what we treat are symptoms stemming from the early-phase response, which is evidenced by pollen provocation studies performed by pharmaceutical companies. The inadequate treatment of late-phase responses is what leads to a chronic disease state. Several recent studies correlating grass pollen counts and symptom severity scores show clinical symptoms appear to have a direct correlation with rising pollen grain counts up until a point where symptoms continue to worsen but in a non-linear fashion.(Caillaud et al., 2012; Durham et al., 2014; Kiotseridis et al., 2013)[Table 1] The emergence/persistence of late-phase responses may contribute to our understanding of why symptom severity scores continue to rise despite the likely saturation of receptors responsible for the early-phase response. These late-phase responses are less likely to respond to traditional seasonal allergy medications. Thus, with increasing pollen counts we can expect medication utilization to increase while also expecting the efficacy of said medications to decrease.(Prince, Norris, & Bielory, 2018) This translates into greater healthcare costs and greater morbidity scores from seasonal allergy sufferers.
Another issue with using pollen counts to predict symptoms is highlighted by allergic asthma. Intact pollen grains are too large to penetrate lower airways; however, pollen exposure is a known trigger for asthma. This led to the discovery of allergenic airborne particles, called paucimicronic particles or orbicules, that are much smaller than the intact pollen grain and can reach the lower respiratory tract. Allergic asthma is further exacerbated by air pollution. In the absence of pollen, airborne pollutants have the ability to cause lower airway inflammation and irritation.(D’Amato, Liccardi, D’Amato, & Holgate, 2005) And by attaching to the surface of pollen grains and orbicules, pollutants modify their potential to induce allergic reactions.(D’Amato et al., 2005) Grass pollen has been found to be a strong independent non-linear predictor of asthma hospital admission when analyzed with a multi-pollutant model, with an increasing effect on asthma admission up to a threshold of 30 pollen grains/m3 (at which point the effect remained stable).(Erbas et al., 2007).
The frequency of storms has increased in temperate and subtropical areas.(D’Amato, Cecchi, Annesi-Maesano, & D’Amato, 2018) Increases in severe weather over the past few decades have often been attributed to climate change. Thunderstorms in particular are linked to increased rates of asthma exacerbations. One hypothesis is that water causes paucimicronic particles to be released from pollen grains that get aerosolized during storms, which in turn results in increased allergen exposure and increases in respiratory symptoms in pollen-allergic individuals.(D’Amato, Annesi-Maesano, Vaghi, Cecchi, & D’Amato, 2018) One of largest recorded asthma epidemics documented in medical literature occurred in Melborne, Australia and was attributed to thunderstorm asthma. Grass pollen concentrations were >100 grains/m3 when a gust front crossed the city that resulted in a 10˚C temperature drop, 70% humidity increase, and increased concentration of aerosolized particulate matter.(Lindstrom et al., 2017) Within 30 hours there were 3,365 excess respiratory-related presentations to emergency departments (672% increase) with a 476 excess asthma-related admissions to hospitals (992% increase) that disproportionately affected individuals of Indian, Sri Lankan, or southeast Asian birth (compared to prior 3 years).(Hew et al., 2019; Thien, 2018) Although only 28% of these people had doctor-diagnosed asthma, 87% reported symptoms of seasonal allergic rhinitis (runny nose, sneezing, itchy nose, watery itchy eyes) in the springtime.(Hew et al., 2019) Although this is a unique case where multiple environmental factors came together to create a public health crisis, similar events have been recorded in Italy and the United Kingdom.(Davidson, Emberlin, Cook, & Venables, 1996; Murray et al., 1994) Events such as these highlight the need to develop validated models for forecasting seasonal allergy symptoms, educate pollen-sensitive individuals of the potential dangers thunderstorms present during a pollen season, and develop plans for coordinating a health care response to such crises.
Although sensitivity to aeroallergens do not have a significant effect on overall mortality, they do affect morbidity, decrease work/school productivity, and lead to higher healthcare costs. As overall pollen grain counts and individual pollen grain allergenicity increase, we expect to see more people sensitized to pollen while simultaneously expecting the efficacy of traditional seasonal allergy medications to decrease. Warmer annual temperatures are already resulting in longer, more severe pollen allergy seasons. Models for forecasting the impact of climate change on aeroallergens have been developed,(Zhang, Bielory, Cai, et al., 2015) however our ability to utilize and access raw pollen count data is presently limited as aeroallergen monitoring is often performed by private, self-funded collectors. This is an area of climate research that would benefit from additional funds to assist with data collection efforts and the creation of a centralized public database.
There are many factors that impact an individual’s health that are indirectly related to measurable changes in climate. Critical to physicians, patients, and public health stakeholders is elucidating the relationship between atmospheric pollen grain concentrations and symptomatic presentation. Individual variability within a population and regional differences in allergenicity of individual pollen grains have confounded attempts to define critical atmospheric pollen levels to date. By pairing environmental tracking with personalized healthcare data, individual and regional patterns may emerge that lead to the development of evidence-based guidance that could better assist health care providers and stakeholders in their missions to reduce burden of disease stemming from allergic disease.
Table 1: Population studies correlating grass pollen counts to symptom severity scores
|Eye Itching and Redness||Linear Relationship||Non-Linear Relationship|
|Kotseridis 2013 (Sweden)||0 grains/m3 to 70 grains/m3||>70 grains/m3|
|Caillaud 2012 (France, Switzerland)||0 grains/m3 to ~90 grains/m3||>90 grains/m3|
|Durham 2014 (Multinational)||0 grains/m3 to ~90 grains/m3||>90 grains/m3|
|Nasal Symptoms||Linear Relationship||Non-Linear Relationship|
|Kotseridis 2013 (Sweden)||0 grains/m3 to 80 grains/m3||>80 grains/m3|
|Caillaud 2012 (France, Switzerland)||0 grains/m3 to ~80 grains/m3||>80 grains/m3|
|Durham 2014 (Multinational)||0 grains/m3 to ~80 grains/m3||>80 grains/m3|
|Lower Respiratory Symptoms||Linear Relationship||Non-Linear Relationship|
|Kotseridis 2013 (Sweden)||0 grains/m3 to 70 grains/m3||>70 grains/m3|
|Caillaud 2012 (France, Switzerland)||Linear Relationship||Linear Relationship|
|Durham 2014 (Multinational)||Not Discussed||Not Discussed|
Table 1: At a population level, nasal and ocular symptom severity have a linear relationship with pollen grain counts up to a point of approximately 80 grains/m3 and 70-90 grains/m3, respectively.