Non-linear short-term effects of airborne pollen levels with allergenic capacity on asthma emergency room admissions in Madrid, Spain


Dr Aurelio Tobías, Departamento de Estadística y Econometría, Universidad Carlos III de Madrid, C/ Madrid 126, E-28903 Getafe, Spain.


Background Several time-series studies have shown positive associations between pollen levels and asthma morbidity. However, few studies have included air pollution as a potential confounder when examining this relationship, and they have only done so on a linear basis.

Objective To investigate the potential non-linear short-term effects of the types of pollen with allergenic capacity across the whole range of exposure on the daily number of asthma-related hospital emergencies in Madrid for the period 1995–1998.

Methods Data were collected on the daily number of asthma emergency room admissions and daily average levels of major types of pollen with allergenic capacity: Olea, Plantago, Poaceae and Urticaceae. We used Poisson regression with generalized additive models, controlling for trend and seasonality, meteorological variables, acute respiratory infections and air pollutants. To study the non-linear effect of pollen levels on asthma emergency room admissions, pollens were categorized into five groups defined on the basis of their respective distributions.

Results The strongest associations were registered in respect of a lag of 1 day for Urticaceae, a lag of 2 days for Plantago, and a day lag of 3 days for Poaceae. Adjustment for the different types of pollen failed to result in major changes. Non-linearities were evident for pollen levels. The greatest increase in risk vs. the reference category (minimum value – 50th percentile) was observed for the categories between the 95th and 99th percentiles, with increases of 32.0% for Plantago, 32.2% for Poaceae and 24.6% for Urticaceae, and between the 99th percentile and the maximum value, with increases of 31.3%, 78.7% and 49.8%, respectively. Olea was not related with asthma emergency room admissions.

Conclusions Pollens with allergenic capacity in Madrid are positively associated with asthma-related hospital emergencies. These associations remain stable when simultaneous adjustment is made for the four types of pollen and for air pollutants and meteorological variables. In terms of non-linear effects, Poaceae register the strongest association with asthma emergency room admissions.