Used ventilation air filters have been shown to reduce indoor environmental quality and worker performance and increase symptoms, with effects stronger after reaction of filters with
ozone. We analyzed data from the US EPA Building Assessment Survey and Evaluation (BASE) study to determine if
ozone and specific filter media have interactive effects on building-related symptoms (BRS). We analyzed a subset of 34 buildings from the BASE study of 100 US office buildings to determine the separate and joint associations of filter medium [
polyester/synthetic (PS) or fiberglass (FG)] and outdoor
ozone concentration (above/below the median, 67.6 microg/m(3)) with BRS. Using logistic regression models and general estimating equations, we estimated odds ratios (
ORs) and 95% confidence intervals for the association of filter medium,
ozone, and filter medium x
ozone with BRS. Relative to FG + low
ozone, PS alone or high
ozone alone, were each significantly (P < 0.05) associated only with
fatigue/difficulty concentrating (
ORs = 1.93 and 1.54, respectively). However, joint exposure to both PS + high
ozone, relative to FG + low
ozone, had significant associations with lower and upper respiratory,
cough,
eye, fatigue, and
headache BRS (
ORs ranged from 2.26 to 5.90). Joint
ORs for PS + high
ozone for lower and upper respiratory and
headache BRS were much greater than multiplicative, with interaction P-values <0.10. Attributable risk proportion (ARP) estimates indicate that removing both risk factors might, given certain assumptions, reduce BRS by 26-62%. These findings suggest possible adverse health consequences from chemical interactions between outdoor
ozone and PS filters in buildings. Results need confirmation before recommending changes in building operation. However, if additional research confirms causal relationships, ARP estimates indicate that appropriate filter selection may substantially reduce BRS in buildings, especially in high-
ozone areas.
PRACTICAL IMPLICATIONS: The results indicate that a better understanding of how filters interact with their environment is needed. While the mechanism is unknown and these findings need to be replicated, they indicate that the joint risk of BRS from
polyester/synthetic filters and outdoor
ozone above 67.6 microg/m(3) is much greater than the risk from each alone. These findings suggest potential reductions in BRS from appropriate selection of ventilation filter media or implementing strategies to reduce
ozone entrained in building ventilation systems. If the relationships were found to be causal, filter replacement and
ozone abatement should be undertaken.