PFAS Health Study Component four: Data linkage study of health outcomes associated with living in PFAS exposure areas December 2021
Date
2021
Authors
Law, Hsei Di
Armstrong, Bruce
D'Este, Catherine
Randall, Deborah
Hosking, Rose
Lazarevic, Nina
Trevenar, Susan
Smurthwaite, Kayla
Lal, Aparna
Lucas, Robyn
Journal Title
Journal ISSN
Volume Title
Publisher
National Centre for Epidemiology and Population Health, The Australian National University
Abstract
Per- and polyfluoroalkyl substances (PFAS) are man-made chemicals that may have adverse
effects on the environment and human health. The primary aim of the PFAS Data Linkage Study
was to examine whether adverse health outcomes were more common in people who had lived in
towns with known PFAS contamination—Katherine in Northern Territory (NT), Oakey in
Queensland (Qld) and Williamtown in New South Wales (NSW) (the ‘exposure towns’). To do this,
we compared rates of selected health outcomes in these towns, to rates in other separate but
similar areas in Australia not known to have PFAS contamination (the ‘comparison areas’).
We conducted three separate studies that investigated four groups of health outcomes. Study 1
investigated maternal and infant (perinatal) health (15 outcomes); Study 2 examined childhood
development (6); and Study 3 investigated cancer (23) and deaths due to specific causes (4)—a
total of 48 outcomes. All studies used multiple data sources with records collected over many
years, which were linked to create richer datasets for analyses. All data used were originally
collected for administrative purposes. We only used data that did not identify individual people and
no direct contact was made with anyone whose data were included in the studies.
Over the three separate studies, for most of the health outcomes studied we did not conclude that
rates were higher in the towns than the comparison areas. For several health outcomes studied, we
observed higher rates in one but not the other two towns. These were: in Katherine, prostate cancer;
in Oakey, stillbirth, developmental vulnerability in two domains (physical health and wellbeing, and
communication skills and general knowledge) and laryngeal cancer; and in Williamtown,
postpartum haemorrhage (heavy blood loss following pregnancy), pregnancy-induced
hypertension (high blood pressure), kidney cancer and lung cancer. Rates of death from coronary
heart disease were higher in both Oakey and Williamtown.
For most of these health outcomes, we estimated the differences between the towns and
comparison areas to be relatively small. For others, the differences were of modest size, but our
estimates were imprecise, meaning the likely size of each difference could be anywhere between
quite small to quite large. Even though our studies included almost everyone who had ever lived in
the towns in the years we had available data (in some cases dating back to 1983), some of the
conditions studied are uncommon and we observed only a few cases. For these outcomes, we could
not precisely estimate the differences between the towns and comparison areas, and there is very
little we can say about whether a difference really exists.
Due to the nature of our studies, there were certain design limitations. We were unable to fully
account for certain risk factors (e.g. smoking) that could have led to observed differences in rates
(or lack of them) between the towns and comparison areas (‘confounding’). In particular, we were
not able to account for socioeconomic factors as well as we would have liked. This is important, as
socioeconomic conditions are strongly linked to health. In addition, some findings could have arisen
just by chance alone and not because an association truly exists. In light of the above, while there were higher rates of some adverse outcomes in individual towns,
the evidence suggesting that this was due to living in these areas was limited. We did not have
direct measurements of PFAS exposure and we cannot rule out that the higher rates were due to
chance or confounding. Further, there was low consistency in our observations across the three
towns (something we would not expect if PFAS caused an outcome), and there is limited evidence
from other studies observing similar results or explaining how potential biological processes can
result in PFAS causing these effects in humans. Overall, our findings are consistent with previous
studies, which have not conclusively identified causative links between PFAS and these health
outcomes.
Description
Keywords
Citation
Source
Type
Report (Commissioned)
Book Title
Entity type
Access Statement
Open Access