Natural Gas Development and Human Health in PA: Let's Get the Facts Straight

By Nick Deiuliis President and CEO, CNX Resources

Poor policy favors superficial optics and follows manufactured storylines. Sound policy aims for substantive improvement and values rational decisions based on objective data. ​

Today a poor policy path is being promoted to harm a life-sustaining industry, manipulating concerns for human health as a convenient tool. It’s happening in Pennsylvania with its natural gas industry and local communities. ​

A ground game is churning that fabricates a storyline of the natural gas industry hurting residents: causing asthma, causing childhood cancers, and adversely impacting newborns. Making and broadcasting disgusting and baseless accusations to vilify and take down a noble and societally crucial endeavor, and in the process hurting the very region and communities the opportunists claim to work in the interests of.

This past August, the University of Pittsburgh (Pitt) released the results of its studies on public health impacts from natural gas development in southwestern Pennsylvania. The studies left much to be desired; they suffered from fatal design flaws, many of them self-inflicted.

Yet the studies failed to find causation that would link natural gas development to health problems. The biggest ‘bombshell’ findings were a pair of very weak (and in one case, nonsensical) associations of natural gas development and two discrete health issues. No smoking guns identified.

But the rollout, presentation, and subsequent reporting of the findings (or lack thereof) were staged to garner maximum speculation, innuendo, and debasing of the natural gas industry. An all-too-common occurrence in media and academia today.

Mountain of Evidence Prior to the Pitt Studies

Over the years prior to the recent Pitt studies, substantial research has been conducted by numerous organizations on the impacts of natural gas development on public health. ​ The studies yielded disappointing results for those hoping to link shale development with human health risks: the expected risks have not materialized while efforts to find clear causation of natural gas development on health risks have come up short.

Consider the massive body of scientific work and measurement performed on point prior to this summer.

In the then-largest study of its kind, a 2015 Yale-led investigation found no evidence that trace contamination of organic compounds in drinking water wells near Marcellus shale development in northeastern Pennsylvania came from deep hydraulic fracturing shale horizons, underground storage tanks, well casing failures, or surface waste containment ponds.

The Yale study was followed by a Duke-led effort in 2017 to assess the impacts of natural gas development on groundwater in northern West Virginia. The Duke study concluded that there was a clear indication for the lack of groundwater contamination and subsurface impact from shale-drilling and hydraulic fracturing. And that trace metals associated with potential health impacts also showed no correlation with proximity to shale gas activities. ​

In 2018 the University of Cincinnati assessed the risk of methane making its way into groundwater in the Utica shale region of Ohio. The study found no relationship between methane concentration or source in groundwater and proximity to active gas well sites. ​ The study did, however, show a decrease in methane concentration in some regularly monitored wells during the study period. And that pH and conductivity did not change as shale gas drilling increased, nor with distance to the nearest shale gas well. Data did not indicate intrusion of frac fluids. ​ ​ ​

The Pennsylvania Department of Environmental Protection (DEP) and Department of Health (DOH) were also busy in 2018, conducting studies and issuing reports that concluded Marcellus shale emissions did not exceed National Ambient Air Quality Standards (NAAQS) limits and that the emissions are not expected to be harmful to healthy citizens. Estimated additional lifetime cancer risks were found to be very low from exposures to chemicals near natural gas activity (average levels of carcinogenic chemicals detected were generally similar to levels typically seen in ambient air in mixed urban, suburban, and rural areas across the US). And significantly fewer Air Quality Index (AQI) days worse than “Good” were measured at a monitoring site near natural gas facilities versus local comparison sites.

Penn State performed a 2018 study of groundwater in rural regions of Pennsylvania (Bradford County) where natural gas development is present. The study found only rare instances of possible gas contamination amid an overall trend of improving water quality despite heavy Marcellus shale development. The Penn State researchers saw possible contamination by natural gas in only 0.5% of the nearly 1,400 shale wells studied in heavily drilled Bradford County. The remaining water chemistry data highlighted that groundwater either improved or remained level from samples taken prior to the 1990s.

"The most interesting thing we discovered was the groundwater chemistry in one of the areas most heavily developed for shale gas – an area with 1400 new gas wells – does not appear to be getting worse with time, and may even be getting better," said the director of Penn State's Earth and Environmental Systems Institute.

2023 Pitt Studies Findings

The recent Pitt studies, spanning years and millions of dollars in expenditures, showed no ‘causation’ and a limited, highly questionable pair of ‘associations’ between natural gas development and two specific health issues. ​

In statistics, it is important to differentiate between ‘association’ and ‘causation.’ Two variables may be associated without a causal relationship. For example, there is a statistical association between the number of people who drowned by falling into a pool and the number of films Nicholas Cage appeared in for a given year (they indeed do show an association when tracked over a specific period of time). However, there is obviously no causal relationship.

Causation, on the other hand, means that the exposure produces the effect.

The Pitt studies found no causation from unconventional shale development to any of the health risks studied. Rather, researchers stretched to find two associations using skewed measurements, atypical definitions, and not attempting to account for key environmental and other factors that have proven demonstrable impacts on health.

The studies relied on a very limited proximity metric which doesn’t identify any exposure pathways, assumes constant emissions, and ignores critical factors like weather, work, air dispersion, lifestyle choices and known existing hazards. In addition, the studies assumed all natural gas wells “are created equal.”

Despite these flaws and limitations, the researchers acknowledged, “No evidence was found to support an association between exposures to [natural gas] activities and other environmental factors and the risk of leukemia, [central nervous system] tumors, and malignant bone tumors, including [Ewing’s Family of Tumors].” An extremely low lymphoma association correlation was found, underscoring the limited methodologies employed.

Asthma exacerbations were not linked with proximity to wells in pad preparation, drilling or hydraulic fracturing phases, regardless of how close wells were to homes or the number of wells nearby. Curiously, the only association claimed was to the natural gas production phase, when little to no activity occurs on the pad and emissions are minimal.

Ironically, Pitt’s own data on asthma in western Pennsylvania show a 50% decline of severe asthma cases between 2014 and 2020, even as natural gas production in the study region increased by over 200%. And across the study period, air quality in the Western Pennsylvania region has generally improved, with PM2.5 particulate matter, an asthma trigger, declining in the eight-county study area to well below NAAQS (EPA). Reviewing the Asthma and Allergy Foundation of America’s “Most Challenging Places to Live with Asthma” informs that no region (including Pittsburgh) with natural gas development falls in the top 20, but Philadelphia, Allentown, and Harrisburg each do.

The Pitt study found birthweights for mothers living close to natural gas facilities remain in the normal healthy range, and no association to other adverse birth outcomes. The average birthweight was within the national average of 2400-4000 grams, and the greatest reduction in birthweight associated with natural gas well exposure was only 0.8% below the average, still well within healthy ranges. The researchers pointed out this “poses little health risks.” Interestingly, the odds of preterm birth were higher for those living with no natural gas activity near the mother’s residence during pregnancy.

Fatal Flaws of the Pitt Studies

The Pitt studies suffer from fundamental limitations in design and methods that, coupled with how the findings were presented to the public, raised anxiety unwarranted by actual data. ​

Researchers never visited shale gas sites, refused opportunities to do so, didn’t take air or water samples, or generate any new, original data or measurements. Statistical speculation trumped actual measurement.

If the researchers had spent time in the field, they would have seen how natural gas development is safe, well-regulated, and produced here better than anywhere else in the world.

In addition, the researchers relied solely upon statistical models and static locational information. They ignored key influential factors like actual emissions, wind, air dispersion, weather patterns, not to mention other potential environmental sources of exposure or outdoor hazards.

In a Question-and-Answer document regarding the research studies, the Pennsylvania DOH cautioned: “Establishing cause-effect relationship in environmental studies is very difficult. In many cases, it is also not generally feasible to be able to gather information on or understand all the possible factors that may impact health such as genetics, other exposures over a lifetime and lifestyle factors that may impact the health outcome in question.”

Consider the finding of association of natural gas development to lymphoma (0.006%-0.0084% association between diagnosis and well location) and no association with other cancers, including Ewing Sarcoma. The researchers primarily relied on the relation of the disease and how close sufferers lived to a fracking facility. Because data was limited to information found on birth certificates, the studies wrongfully assumed people lived at the same address for up to 29 years, while ignoring daily travel to locations like schools and workplaces. Such an approach would be laughable if it were not for the serious issues at hand.

The researchers admit the cancer study did not adequately account for variables the American Cancer Society lists as common lymphoma causes. These include genetic predisposition, infection, and exposure to radiation (such as the Canonsburg, Pennsylvania uranium waste facility where government monitoring has shown higher radiation levels).

The asthma assessment suffered from similar methodology flaws, failing to account for known asthma triggers including indoor and outdoor air pollutants, and refusing to present this data to the public. Nor did Pitt researchers explain why they labeled all asthma exacerbation cases as “severe” when the standard in medical studies is to categorize asthma exacerbation cases as mild, moderate, and severe. Without explanation, researchers broke these medical research norms. ​

The asthma data were limited to information found in medical records. Meaning that while smoking status was accounted for, a child’s exposure to secondhand smoke was not. Other known asthma triggers that were ignored, as identified by the CDC, include indoor and outdoor air pollutants, dust mites, mold, and pests. This leaves a huge gap in the potential other external factors that are known to trigger and exacerbate asthma.

Conflicts and Bias

Fundamental to the credibility of research is that researchers are, in fact, independent and unbiased.

One of the lead researchers for these studies publicly advocated for increasing mandatory setbacks from shale gas activities during a 2021 public forum. Unfortunately, the researcher drew conclusions well before finalizing these studies and chose to advocate for a public policy that is being pushed by anti-energy activists seeking to ban natural gas development.

Consider the money trail to assess who benefits from the baseless speculation that ensues. Little if any new, empirical research was conducted in the studies. Of the $2.584 million in taxpayer funds spent on these studies, $1.5 million went to pay salaries and benefits of the researchers, while another $932,000 went for various, unspecified administrative fees. ​

A concerning lack of transparency stokes more concern about how the study was performed. The researchers failed to adhere to the provisions of the taxpayer-funded contract, which required it to conduct a public forum on at least an annual basis to advise on the status of the ongoing studies and, by extension, gain valuable input from all stakeholders affected by the studies (“On at least an annual basis, the Principal Investigator(s) and study team shall present study progress to date in a public forum in conjunction with the Department.” Section E.2.B (Work Plan) of Attachment 1; Contract number 4400018535). Despite claiming that it would “welcome open and collaborative conversations with the board (External Advisory Board) when we have data to share,” it appears that no open or collaborative conversations were held with the External Advisory Board or interested public stakeholders.

The studies considered proximity to only one potential exposure, unconventional natural gas wells, rather than taking a holistic approach of examining whether other activities or exposures may have contributed to adverse health impacts. For example, the study region hosts two uranium disposal sites which have been demonstrated to have uranium concentrations in groundwater that are 22x higher than the EPA’s maximum concentration limits. Additionally, pesticide exposure has been linked to the development of some cancers, leading credible researchers to examine the presence of treated croplands in proximity to incidents of cancer. Yet, the cancer study did not adequately factor in these or other potential exposures, as well as personal genetic or lifestyle factors that may influence health outcomes.

With respect to asthma, the Pitt researchers excluded residents of the city of Pittsburgh from their study population even though Pittsburgh is within the study region and many of its residents live in proximity to shale gas wells. No rational reason is offered for excluding Pittsburgh residents. But well-established research has shown that the rate of asthma in urban areas is meaningfully higher than the rate in rural settings. By removing Pittsburgh’s population from the study group, including a control population that would be expected to have statistically higher asthma rates, the study skews the findings by failing to draw from the control population in a consistent manner across the study region.

How the Process Works

One must understand the process used in the Pitt studies (and many others like it from across the country) to judge the findings objectively. The reactions of various parties upon the release of the studies provide a window into how this type of research has been co-opted to fulfill predetermined views of the natural gas industry by those opposed to it. ​ The intent is to build a process that functions as a positive feedback loop, with each subsequent link reinforcing the pre-desired outcome or view.

The studies used two primary datasets. First, the researchers access medical information of patients with maladies of concern, in this case childhood cancers, asthma, and birth outcomes. In these medical datasets, a home address is associated with the patient.

Second, researchers use locations of natural gas wells, and dates when they were drilled, hydraulically fractured, and producing. ​

The researchers then determine the strength of the statistical relationship between the home address of the patient and the distance to a well pad (or other natural gas infrastructure of interest such as a compression station).

It's a simple approach bordering on useless. It will fail to deliver any meaningful insight or solid conclusions beyond innuendo and rank speculation.

That’s why this methodology creates the need to use squishy descriptors such as ‘links’, ‘ties’, and ‘associations,’ but can’t be used to determine basic cause and effect. ​

The study authors say they welcome additional data and research. Yet regulatory agencies conduct extensive studies on actual air quality, radiation, and many other compounds of potential concern from industrial operations, including natural gas development. Environmental, health, and safety professionals at energy companies collect and analyze data regularly to protect their workforce, contractors, host communities and the environment. These extensive datasets are consciously ignored by those predisposed to creating worst-case speculative scenarios.

By design, the public sees only the alarmist headlines.

The reactions to these much-anticipated studies were telling. Some reactions were theatrical and prepared ahead of time (mostly journalists and environmental group bureaucrats with self-serving agendas) while others were raw with emotion emanating from genuine anguish (parents experiencing personal loss and tragedy). All the reactions were carefully commandeered by opportunists to nurture that positive feedback loop.

Constrained by the flaws of the study design, it was impossible to answer the questions of concerned residents, who came away empty-handed without the being told what is causing illnesses. ​

The researchers offered that “this was only the beginning” and “the first step,” both of which are code for justifying millions of dollars in additional expenditures to feed a machine dependent on fear and anxiety.

Media plays a collaborative role in manufacturing the emotion and bias. The news reports were what one commentator notably described as “overwhelmingly sloppy” with headlines, promoting outcomes that weren’t explicit in the studies.¹ Many outlets conveyed conclusions not made in the studies, which is unethical, bordering on legally actionable. ​

Consider the following headlines appearing after the town hall presentation:

• “Study: Asthma severity, rare childhood cancer likelier near gas wells”
• “Research suggests link between fracking, rare childhood cancers”
• “A Pennsylvania study suggests links between fracking and asthma, lymphoma in children”
• “’Is it safe to live here?’: Questions loom at presentation of reports on fracking and health in southwestern Pa.”
• “Fracking is making Pennsylvanians sick. Lawmakers must act.”

Most headline readers come away with conclusions that the studies did not conclude. Which is the objective of the journalists writing the headlines. Yet compare those headlines to what the studies and researchers stated:

• “The researchers were unable to say whether the drilling caused the health problems, because the studies weren’t designed to do that.”
• “But the researchers said they found no association between gas drilling and childhood leukemia, brain and bone cancers.”
• “Limited evidence existed for a tie between gas extraction and central nervous system tumors.”
• “But no relationship was found between fracking and leukemia. Similarly, results did not show a link between rare bone cancer and shale gas development that was statistically significant.”
• “The studies used a retrospective model, which looks back in time at participants’ health instead of tracking patients in real time.”
• “If you are just looking at the studies and trying to demonstrate some kind of causation, and trying to say that this is the end-all-be-all of the situation, that is not what these studies are designed to do.”

Welcome to modern-day mainstream media, now devoid of ethical standards, in the fields of domestic energy and environment. No wonder that Gallup found only 18% of Americans today have “a great deal or quite a lot of confidence” in newspapers. ​

¹ My personal observation was that the media reporting was “calculatingly curated.”