![UOGD exposure assessment in an example ZIP code and month (Washington, Pennsylvania 15301, August 2015)
a, The locations of active UOGD wells, the 1 × 1 km grid population density and the prevailing monthly wind direction. b, The calculation of PE and DE metrics for an example grid in the ZIP code, which is outlined in bold in a. The proximity-based UOGD exposure (PE=IDWall=∑i=1N1di)\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$({\mathrm{PE}} = {\mathrm{IDW}}_{{\mathrm{all}}} = \mathop{\sum} \limits_{i=1}^{N} \frac {1} {d_i})$$\end{document} was calculated as the inverse-distance-weighted (IDW) of wells in all directions within a circular buffer with a radius of 5 km (set for illustration purposes) and was used in Model I of Analysis Set I. The UOGD exposure contributed by upwind wells (IDWup=∑j=1N1dj)\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$({\mathrm{IDW}}_{{\mathrm{up}}} = \mathop{\sum} \limits_{j=1}^{N} \frac {1} {d_j})$$\end{document} was calculated using the IDW of all wells that fall within the windward circular quadrant (pink-shaded area in b). The ratio between IDWup and IDWall is defined as the downwind-based exposure (DE = IDWup/IDWall) and was used in Model II of Analysis Set I. N, number of wells in circular buffer; n, number of wells in circular sectional buffer; d, distance between the well and the grid centre.](https://www.researchgate.net/publication/358152697/figure/fig3/AS:1127576001101826@1645846476355/UOGD-exposure-assessment-in-an-example-ZIP-code-and-month-Washington-Pennsylvania_Q320.jpg)
- A preview of this full-text is provided by Springer Nature.
- Learn more
Preview content only
Content available from Nature Energy
This content is subject to copyright. Terms and conditions apply.
Articles
https://doi.org/10.1038/s41560-021-00970-y
1Department of Environmental Health, Harvard T. H. Chan School of Public Health, Boston, MA, USA. 2Department of Biostatistics, Harvard T. H. Chan
School of Public Health, Boston, MA, USA. 3Department of Epidemiology, Harvard T. H. Chan School of Public Health, Boston, MA, USA. 4Present address:
Division of Occupational and Environmental Medicine, Lund University, Lund, Sweden. ✉e-mail: lol087@mail.harvard.edu
Oil and natural gas development from low-permeability geo-
logical formations (known as unconventional oil and gas
development (UOGD)) has rapidly expanded over the past
decade. As of 2015, more than 100,000 onshore UOGD wells have
been drilled using directional drilling combined with multi-stage
high-volume hydraulic fracturing (fracking)1. A total of 17.6 million
US residents currently live within one kilometre of at least one active
well2. The annual percentage of newly completed oil and gas wells
that target unconventional formations increased from 2.3% in 2001
to 47% in 2015, and then to 71% in 2019. Compared with conven-
tional oil and gas development (COGD), UOGD generally involves
longer construction periods, larger well pads and requires larger
volumes of water, proppants and chemicals during the multi-stage
hydraulic fracturing process3. Owing to the rate of expansion and
larger theoretical environmental impacts, it is critical to study the
health effects and exposure pathway(s) of UOGD.
UOGD activities—including pad construction, well drilling,
hydraulic fracturing and production—have been associated with
increased human exposure to harmful agents4–7. UOGD-related
primary air contaminants include volatile organic compounds
(or VOCs)8, nitrogen oxides9 and naturally occurring radioac-
tive materials10,11. UOGD operations have also been associated
with elevated concentrations of organic compounds12, chloride
and total suspended solids in drinking water13. Higher levels of
UOGD-associated non-chemical exposures, such as noise14 and
night light15, have also been reported in nearby neighbourhoods.
Previous health-effects studies have found significant associations
between proximity-based exposure to UOGD and adverse prena-
tal16–19, respiratory20, cardiovascular21 and carcinogenic outcomes22.
The association between exposure to UOGD and all-cause
mortality among the elderly has not been quantified. In addition,
previous studies were conducted in specific geographical locations
and thus did not evaluate the heterogeneity in exposures and out-
comes across large geographical regions. Previous studies also did
not investigate the exposure pathway(s) through which UOGD
activities could lead to adverse health effects, primarily due to the
lack of any large-scale measurement of UOGD-sourced pollutants
in some intensively drilled regions. To address these gaps in the data
and characterize the spatiotemporal gradients of the UOGD-sourced
agents, investigators have designed proximity-based exposure (PE)
metrics of varying complexity23. Most of these PE metrics assumed
a uniform distance decay in the concentrations of UOGD-related
agents in all directions. Although this assumption largely holds for
noise and light pollution, which travel similarly in all directions, it
does not account for the directional dispersion of UOGD-sourced
airborne or waterborne pollutants in nearby environments. PE met-
rics could be improved by incorporating the transport mechanisms
of UOGD-sourced agents, such as wind direction and underground
water flow24. Accounting for the directional dispersion of agents
would also enable the investigation of potential exposure pathways.
Following the process shown in Fig. 1, we built an open cohort
of 15,198,496 Medicare beneficiaries (136,215,059 person-years)
residing in our study area (Fig. 2), which includes all major US
UOGD regions (Supplementary Note 1) from 2001 to 2015. We
also gathered location, construction and production records for
more than 2.5 million oil and gas wells. Rather than solely relying
on PE metrics, we calculated downwind-based exposure (DE) met-
rics, which incorporate the wind direction in the exposure assess-
ment (Fig. 3). On the basis of these two exposure metrics (PE and
DE), we conducted two sets of analyses (Analysis Set I and II) to
investigate whether or not living in proximity to and downwind of
UOGD wells is associated with higher mortality risks in Medicare
Exposure to unconventional oil and gas
development and all-cause mortality in
Medicare beneficiaries
Longxiang Li 1 ✉ , Francesca Dominici 2, Annelise J. Blomberg 1,4, Falco J. Bargagli-Stoffi 2,
Joel D. Schwartz1,3, Brent A. Coull1,2, John D. Spengler1, Yaguang Wei 1, Joy Lawrence1 and
Petros Koutrakis1
Little is known about whether exposure to unconventional oil and gas development is associated with higher mortality risks in
the elderly and whether related air pollutants are exposure pathways. We studied a cohort of 15,198,496 Medicare beneficia-
ries (136,215,059 person-years) in all major US unconventional exploration regions from 2001 to 2015. We gathered data from
records of more than 2.5 million oil and gas wells. For each beneficiary’s ZIP code of residence and year in the cohort, we cal-
culated a proximity-based and a downwind-based pollutant exposure. We analysed the data using two methods: a Cox propor-
tional hazards model and a difference-in-differences design. We found evidence of a statistically significant higher mortality
risk associated with living in proximity to and downwind of unconventional oil and gas wells. Our results suggest that primary
air pollutants sourced from unconventional oil and gas exploration can be a major exposure pathway with adverse health effects
in the elderly.
NATURE ENERGY | VOL 7 | FEBRUARY 2022 | 177–185 | www.nature.com/natureenergy 177
Content courtesy of Springer Nature, terms of use apply. Rights reserved
