UPDATE (8/7/15; 9:50 am ET): Health professionals are noticing deep flaws in a recent University of Pittsburgh birth weight study. This week, the Magee-Womens Research Institute and Foundation, an organization focused on reproductive biology and women’s health released a statement on the study explaining that the researchers’ reliance on birth certificates is not scientifically rigorous – something that EID pointed out when the study was released. From the Magee-Womens Research Institute and Foundation release:
“Birth certificates do not record every pregnancy outcome. Therefore the researchers were able to evaluate only a limited number of outcomes, reflecting fetal development, growth, and preterm birth. The study found no significant and consistent association between residence near UGD wells and either preterm birth or fetal anomalies. However, it did show some association between residence near UGD wells and fetal growth. Fetal growth can be assessed using several criteria: (a) the incidence of low-birth-weight, which uses an arbitrary cut-off of newborn weight less than 2,500 grams, and is grossly misleading as it is not normalized to gestational age; (b) the difference in actual neonatal weight, which may uncover a differences in birth weight, but the clinical significance of these differences is questionable; (c) the incidence of small for gestational age (SGA) fetuses, which includes those newborns weighing less the 10th percentile, and is adjusted for gestational age but also includes healthy newborns with constitutionally small size, possibly reflecting normal variation or parental physique; and (d) the incidence of fetal growth restriction (FGR), the most stringent criterion that identifies only those SGA fetuses whose small size is associated with truly pathological growth and other markers of disease. While FGR (criterion d) is the best clinical measure of fetal growth abnormality, this definition requires additional measures that are not recorded in birth certificates. The investigators therefore measured a reduction in birth weight and the incidence of SGA newborns (criteria b and c), and found a small, yet consistent and statistically significant association between these parameters and residence near UGD wells.” (emphasis added)
Magee-Womens Research Institute and Foundation concludes that the University of Pittsburgh researchers’ data do not actually show that development is linked to low birth weights:
“As emphasized above, this is a preliminary association study, where negative results should not be interpreted as reassuring, and positive results should not be interpreted as alarming. The results do not rely on the most stringent criteria for clinically relevant fetal growth abnormality, and do not support a conclusion that the proximity to the UGD wells caused reduced birth weights or a higher incidence of SGA. The major conclusion from this study is that this area warrants further investigation in larger populations, with better estimates of exposure and more in-depth analysis of outcomes.” (emphasis added)
— Original Post June 3, 2015 —
Researchers from the University of Pittsburgh released a study this week claiming to have found a potential correlation between low infant birth weights and distance from well sites. Despite what you may read in the headlines, the researchers actually make it very clear that they did not find shale development to be the cause of low birth weights. As one of the researchers, Dr. Bruce Pitt, stated in the report’s press release:
“It is important to stress that our study does not say that these pollutants caused the lower birth weights.”
The study, which was funded by the anti-fracking Heinz-Endowment and published in a journal that does not require peer-review, looked at data from over 15,000 babies born in Washington, Westmoreland and Butler Counties from 2007 to 2010 that lived within 10 miles of well sites during their gestation period. The researchers base their claims on percentages of “small for gestational age” babies in four “quartiles” – the first quartile being farthest away from shale gas wells and the fourth quartile being the closest to shale gas wells.
The study has a number of glaring flaws, many of which the authors even pointed out in an entire page devoted to the study’s limitations. Here are the top seven important things you need to know while reading this study and the media coverage surrounding it:
Fact #1: Average birth weights in study don’t meet criteria for low birth weights; even highest numbers still well below the national average
“Birthweight rates are stratified into three categories: very low birthweight (less than 1500 grams or 3 1/3 pounds), low birthweight (less than 2500 grams or 5 1/2 pounds, includes very low birthweight) and not low birthweight (2500 grams or greater).”
The average weight for the babies born in all for quadrants in the study ranged from 3323.1 grams to 3370.4 grams – so that means the researchers completely disregarded well established medical definitions for low birth weights.
Further, according to the CDC in 2012, on average, 8 percent of babies were born with low birth weight in America. Even the researchers’ highest percentage of low birth weights was still well below the national average at 6.5 percent.
Fact #2: Researchers base their entire argument on “small for gestation age,” yet admit they have no way of telling what the gestational age is.
The authors’ entire premise lies not with the actual birth weights detected but with the weight associated with the “small for gestational age” (SGA).
That’s an interesting choice because anyone who has been through a pregnancy knows that gestational age is not exactly foolproof. According to Medscape:
“The 3 basic methods used to help estimate gestational age (GA) are menstrual history, clinical examination, and ultrasonography. The first 2 are subject to considerable error and should only be used when ultrasonography facilities are not available.”
Despite this, most doctors’ offices use menstrual history at least initially to determine the gestational age. But if a mother never goes to more than one pre-natal appointment, her due date may never have been adjusted for size of the baby and other factors.
The researchers state that they determined the gestational age at birth using birth certificates. From the report (emphasis added):
“Birth weight data are reasonably precise as derived from birth certificates, but such certificates appear less reliable for gestational age, so derived information like SGA may be spuriously affected.”
But determining gestational age at birth is also not reliable. As the National Library of Medicine has acknowledged:
“The baby’s developmental gestational age may not be the same as the calendar age. For example, an infant born at 36 weeks may actually have a developmental gestational age of 38 weeks. This baby may be more like a full-term infant than a baby born prematurely.”
In other words, the researchers have no way of knowing if their SGA numbers are accurate or if a baby born at 40 weeks was actually born early, on time or late. No wonder they couldn’t draw any conclusions from the data!
Fact #3: The researchers’ actual birth weight data do not correlate with their conclusions
There seems to be a good reason the researchers focused on the less reliable SGA number because the actual average birth weight numbers don’t fit in with the declaration in their press release that “Pregnant women living close to a high density of natural gas wells drilled with hydraulic fracturing were more likely to have babies with lower birth weights than women living farther from such wells.”
Their data actually show that the average birth weight in the first quartile, which the farthest away from shale gas wells, was 3343.9 grams. That’s much lower than those in the second quartile (3370.4) and third quartile (3345.4), which were closer to shale gas wells. But, again, none of these birth weight averages are considered low by the actual medical definition.
Fact #4: Data did not include birth month and day, so researchers just assumed any well that year was developed during the pregnancy.
From the study:
“Other limitations in the birth dataset included the lack of a birth month and day; we were therefore only able to identify those wells drilled during the birth year of the infant. Active drilling of a well occurs over a period of only a few months…” (p. 11)
According to the Marcellus Shale Coalition, the average time it takes to drill a well is 15 to 30 days. It then takes about a week or so for completions processes like hydraulic fracturing to take place. Even when you factor in the time it takes to move, set-up and take down equipment, the time to let the cement dry on the steel and cement casing, and other factors that may arise, that’s around three months.
But the researchers had no way of determining when the women became pregnant, or if they were pregnant during the shale development phase.
Fact #5: Data offered no record of whether or not the address given was the same address for the entire pregnancy, but it was assumed it was.
Not only do the researchers not know if the wells were actually developed during the pregnancy, they also can’t tell from the data the duration the woman actually lived at the address given during her pregnancy. She could have moved in last weekend and had no exposure to any kind of development, but any wells developed that same year would be used to determine her quadrant. From the study (emphasis added):
“Moreover we assumed that the address on the birth certificate was synonymous with exposure during the entire pregnancy.” (p.11)
Fact #6: There’s a lot of information not available on birth certificates.
In addition to what’s already been described, there is a lot more information that just can’t be determined from a birth certificate, and the researchers acknowledge this in their study:
“In addition, there may be a number of unknown factors that led to our conclusion well density was associated with lower birth weight and greater odds of SGA. As in any epidemiological study, these associations do not imply causation and are hypothesis generating only. The observed associations could be due to a contaminant related to UGD [unconventional gas drilling], an unknown confounding factor we were unable to account for in our analyses, or chance.” (p. 11, emphasis added)
While the authors accounted for factors such as smoking during pregnancy, they did this only on the basis of the mother self-reporting the habit, which she may or may not have done. They also can only guess if she went to the multiple pre-natal appointments required, took vitamins, and eat properly during the gestation.
The study also did not account for additional pollutants, either. Maybe the mother who lives by wells also lives next to an airport or highway. The birth certificates don’t say, so the researchers couldn’t account for other potential causes. That’s actually something that one of the researchers admitted in the press release (emphasis added);
“Unconventional gas development is dynamic and varies from site to site, changing the potential for human exposure. To draw firm conclusions, we need studies that thoroughly assess the exposure of a very large number of pregnant women to not just the gas wells, but other potential pollutants.”
#7: Study was funded by the Heinz-Endowment and the authors used the same methods as previous activist work.
It’s important to note that the paper acknowledges that the Heinz-Endowment, which has given millions of dollars to anti-fracking groups like PennEnvironment, provided the funding this study. Does that mean the study is inherently biased just because Heinz has funded other activist work before? Not necessarily, but it is very important context to keep in mind while reading the report.
Further, the authors admit to using the same methodology as a deeply flawed report by Lisa McKenzie in Colorado. From the report:
“We adapted the epidemiological and geographic information systems (GIS) approaches of McKenzie et al. to explore the potential effects of UGD on infants born to mothers living in Southwestern PA where unconventional drilling of the Marcellus Shale has been rapidly expanding. The objective of the present study is to use readily available data on birth outcomes for Southwestern Pennsylvania to investigate the relationship of proximity to UGD and perinatal outcomes for 2007 to 2010.” (p. 3)
That’s a problem because Larry Wolk, the Chief Medical Officer of the Colorado Department of Health and Environment cautioned pregnant women against seeing this study as credible. From the Coloradoan:
“As Chief Medical Officer, I would tell pregnant women and mothers who live, or who at-the-time-of-their-pregnancy lived, in proximity to a gas well not to rely on this study as an explanation of why one of their children might have had a birth defect,” said Wolk. “Many factors known to contribute to birth defects were ignored in this study.”
And Dr. McKenzie herself conceded there were many flaws in her research at a discussion in Ohio in 2014:
“The cancer risks are cancer risks for both residents further from the wells and closer to the wells were greater than 1 in a million but well below the 1 in 1,000 range. So they are kind of in this area, this grey area. It depends on the state you are in, it depends on the community, it depends on a lot of things if this is acceptable.”
Anyone who has been pregnant, and especially those that have had a complicated pregnancy, know that there are many, many factors that can contribute to the development of the baby. Low birth weight can result from genetics, outside influences, premature birth, the mother’s decisions—and the list goes on. As mothers, we want to know what these risks are to try to protect our children.
Health studies should add to the growing dialogue and have the potential to improve best management practices for multiple industries. What’s disconcerting is when studies such as this one, which simply don’t have the science to back up their claims, are paraded as credible research, causing undue concern for pregnant women.