Background-A study was undertaken to investigate the relationship betw
een daily hospital admissions for asthma and air pollution in London i
n 1987-92 and the possible confounding and modifying effects of airbor
ne pollen. Methods-For all ages together and the age groups 0-14, 15-6
4 and 65+ years, Poisson regression was used to estimate the relative
risk of daily asthma admissions associated with changes in ozone, sulp
hur dioxide, nitrogen dioxide and particles (black smoke), controlling
for time trends, seasonal factors, calendar effects, influenza epidem
ics, temperature, humidity, and autocorrelation. Independent effects o
f individual pollutants and interactions with aeroallergens were explo
red using two pollutant models and models including pollen counts (gra
ss, oak and birch). Results-In all-year analyses ozone was significant
ly associated with admissions in the 15-64 age group (10 ppb eight hou
r ozone, 3.93% increase), nitrogen dioxide in the 0-14 and 65+ age gro
ups (10 ppb 24 hour nitrogen dioxide, 1.25% and 2.96%, respectively),
sulphur dioxide in the 0-14 age group (10 mu g/m(3) 24 hour sulphur di
oxide, 1.64%), and black smoke in the 65+ age group (10 mu g/m(3) blac
k smoke, 5.60%). Significant seasonal differences were observed for oz
one in the 0-14 and 15-64 age groups, and in the 0-14 age group there
were negative associations with ozone in the cool season. In general,
cumulative lags of up to three days tended to show stronger and more s
ignificant effects than single day lags. In two-pollutant models these
associations were most robust for ozone and least for nitrogen dioxid
e. There was no evidence that the associations with air pollutants wer
e due to confounding by any of the pollens, and little evidence of an
interaction between pollens and pollution except for synergism of sulp
hur dioxide and grass pollen in children (p<0.01). Conclusions-Ozone,
sulphur dioxide, nitrogen dioxide, and particles were all found to hav
e significant associations with daily hospital admissions for asthma,
but there was a lack of consistency across the age groups in the speci
fic pollutant. These associations were not explained by confounding by
airborne pollens nor was there convincing evidence that the effects o
f air pollutants and airborne pollens interact in causing hospital adm
issions for asthma.