Note from Fred, Brother in Law, Trucking company exec, interested in my battle with the EPA and the CARB on small particulates for many years.
Many people, including those in science who don’t have a personal, career stake in how this battle get’s decided, don’t care about what this is about, like Fred, who cares because it affects his life, his job. Fred just wants good science to guide good policy making and know that the enviros are not too much in control.
Fred is a trucking executive for a Nebraska Trucking Company who some would say, doesn’t have a concern, but he does, because he has to pay for stupid air pollution measures imposed by fanatics at the EPA. It compromises his ability to do business.
IF FRED DOESN’T CARE, NOBODY DOES. EPA is screwing with his livelihood.
Fred says
I have spoken to you about the RR factor for years John and never really quite grasped it. I like what you did here…it was not only interesting but it “slowed” down everything for me. I get it and I can see where you get, for lack of a better description, frustrated with what is going on in these so called studies. If there is no center point, no true control, everything goes all over the place and the real truth gets buried. It seems like pretty much all areas of the world today are falling down on this aspect.
Thanks again and keep up the good fight as they say. Fred
From John
Fred
The important thing is the EPA researchers in air pollution studies are looking for differences in the exposed population versus the non exposed populations. That means they must have two groups, and since everyone lives in the same world, the “exposed” group is defined in a strained way—it’s the group exposed to a higher level of air pollution, and then the researchers pick a short lag time to see if there is a higher death rate, for example at 3 days.
In the case of studies on asthmatics and increased ozone, the lag time picked is often one day and hospitalizations for respiratory problems might be the endpoint.
As you know, Milloy did a multi-year study of Southern CA hospital admissions for asthma and respiratory distress to see if there was a correlation with ozone, and he found no correlation.
Milloy Ozone study in CA
http://junkscience.com/2013/09/03/study-ozone-not-linked-with-asthma-hospitalizations-in-major-california-hospital-system/
For small particles, Milloy has found there is no California small particle air pollution correlations with deaths in the state of California for all the counties except the ones he couldn’t study because they were so remote and unpopulated that they don’t have air pollution monitors.
Milloy study on small particles in CA
http://junkscience.com/2013/12/26/epa-air-pollution-scare-debunked-by-best-data-set-ever-assembled-on-particulate-matter-deaths/
So you might say, how does the EPA have these stacks of studies that show small associations of deaths and small particle pollution?
Why are the Associations always in the less the 20% range? Well if you torture the data and the methods enough and you and the computer know what you are looking for, you will find an “association” that works for the EPA agenda. And you will get funding to do more studies, but you will also produce for the EPA mouthpieces an opportunity to claim that there is a small association and then you can project that small assocation to the population of the US–and voilayou get big numbers of people at risk.
The biggest reason they can keep on producing these small associations studies is that deaths vary from day to day—my point is they are harvesting noise and calling it signal. The Noise is created either by the tolerances of the measuring tool or the natural variability of the phenomenon being measured and the nature of epidemiological ecological studies, which have no control over the populations they study.
In the case of air pollution, they evaluate the air pollution levels and then they see if there is anything they can point to as an effect on people exposed to higher levels of air pollution. they pick two things, mostly, short-term hospitalization rates for asthma and ozone levels and “premature deaths” in populations compared to small particulate levels. Small particles are the component of air pollution they like the most and they have pretty much given up on the other criteria pollutant, carbon monoxide, sulfur oxide, nitrous oxide, ozone, volatile organic chemicals.
There are plenty of arguments that their studies are junk—they don’t have monitors for where people are—mostly indoors, so outdoor monitors don’t measure the real exposure, but most important, they deceive the public when they say they count premature deaths at a lag time of less than 5 days from an increase in air pollution on the monitors.
1. premature deaths are deaths before life expectancy, but they don’t count people who die before life expectancy because they don’t know what the life expectancy of those who died that day is—that would require a serious medical and actuarial assessment of each death.
2. they really mean they are counting as premature deaths the extra deaths above expected for the day in question that is has a lag time from the increase in air pollution—that means they are counting extra deaths for the day that match their lag time.
3. So they find places in time when they show a positive increase in deaths at the lag time chosen for air pollution increases. They do it for a bunch of geographic areas looking for the “positive” correlations. Then they measure those. So if, for a 20 city study, for example, the average death per day for non trauma, non cancer deaths is 100, the rate 100 represents zero effect. Any day when the death rate is more than 100 that fits the lag time chose, is a good day for the researchers.
4. If the good day increase for all the good days in the period of the study, over the whole group is 5 deaths more at the lag time chosen, then that is a 5% in increase and that translates to a RR of 1.05.
However, in the business of uncontrolled observational (ecological, some call them) populations studies there are many things that create false outcomes, false positives or false negatives, so the amount of the effect has to be big enough to more reliable—the amount chosen for an inference of proof of causation is 100 % increase in the increase in the rate of the endpoint measure, in air pollution studies that’s deaths above the average rate.
So why does that increase in deaths become so magical, because a 5 % increase in the death rate at the lag time can be projected to the population—and that’s when the panicky talk starts. 50,000 people die every day, and if the EPA researchers can sting those positive correlations together they can project a daily (that’s daily because of the conclusions of their studies about lag time deaths) to produce total death numbers that are as many as 300,000 annual or more. A Relative Risk of 1.05 for air pollution would project to 2500 deaths per day for the year—easy to get to a stunning number of 365 x 2500= 912,500.
Then all that the EPA researchers do is back down by number of days estimated for “unsafe” air and they eliminate the deaths that don’t pass the smell test for attributing as an acute effect of air pollution.
This is the USA annual deaths for 2010 from the Centers for Disease Control
Number of deaths for leading causes of death:
Heart disease: 597,689 (claimed to be due to air pollution)
Cancer: 574,743 (soon to be claimed in another count as chronic air pollution deaths in many cases)
Chronic lower respiratory diseases: 138,080 (claimed to be due to air pollution)
Stroke (cerebrovascular diseases): 129,476 (claimed to be due to air pollution
Accidents (unintentional injuries): 120,859 Not
Alzheimer’s disease: 83,494 not yet
Diabetes: 69,071 not yet
Nephritis, nephrotic syndrome, and nephrosis: 50,476 not yet considered air pollution caused
Influenza and Pneumonia: 50,097 (count the pneumonia? not the flu, not counted yet)
Intentional self-harm (suicide): 38,364 not counted
Source: Deaths: Final Data for 2010, tables 1, 7, 10, 20 Adobe PDF file [PDF – 4.17 MB]
So the number of increased air pollution days is used to temper that number of almost a million a year.
Read this Fred, and tell me if it makes sense. If it does, I will put it up as a follow-up to the initial discussion on small associations and EPA cheating, along with the admission By Dr. Devlin of the EPA that they do human exposure experiments because they know the epidemiology is not reliable for assessing the effect of pollution.
If you want to look at the whole declaration, this is it. From the court records.
http://junksciencearchive2.files.wordpress.com/2013/12/declaration-devlin-highlighted.doc
the important bits that Devlin gives up are these quotes from his declaration under penalty of perjury.
These declarations were made, no doubt with advice and counsel from EPA lawyers. The declaration was clearly the product of serious effort and research.
Dr. Robert Devlin:
I am a Senior Scientist (ST) for the Environmental Public Health Division (EPHD), National Health and Environmental Research Laboratory (NHEERL), Office of Research and Development (ORD), U.S. Environmental Protection Agency.
I was acting National Program Director for ORD’s Air Research Program in 2000. This position is the lead for developing research plans related to air pollution for all of ORD and representing the program to groups outside the EPA. I hold adjunct faculty appointments at the University of North Carolina (Chapel Hill) and North Carolina State University. I have been engaged in performing controlled human exposure studies as an EPA investigator since 1986.
. . .
7. Epidemiological observations are the primary tool in the discovery of risks to public health such as that presented by ambient PM2.5. However, epidemiological studies do not generally provide direct evidence of causation. They indicate the existence or lack of a statistical relationship between ambient levels of PM2.5 and adverse health outcomes. Large population studies cannot assess the biological mechanisms (called biological plausibility) that could
explain how inhaling ambient air pollution particles can cause illness or death in susceptible individuals. This sometimes leaves open the question of whether the observed association in the epidemiological study is causal or whether PM2.5 is merely a marker for some other unknown substance.
8. Controlled human exposure studies conducted by EPA scientists and EPA funded scientists at multiple universities in the United States fill an information gap that cannot be filled by large population studies. In 1998 the Committee on Research Priorities for Airborne Particulate Matter was established by the National Research Council in response to a request from Congress. The committee was charged with producing four reports over a five-year period which describe a conceptual framework for an integrated national program of particulate-matter research and identified the most critical research needs linked to key policy-related scientific uncertainties. Excerpts from their most recent report (published in 2004) are attached as Exhibit to this Declaration.
On page 36 the Committee says:
Controlled human exposure studies offer the opportunity to study small numbers of human subjects under carefully controlled exposure conditions and gain valuable insights into both the relative deposition of inhaled particles and the resulting health effects. Individuals studied can range from healthy people to individuals with cardiac or respiratory diseases of varying degrees of severity. In all cases, the specific protocols defining the subjects, the exposure conditions, and the evaluation procedures must be reviewed and approved by institutional review boards providing oversight for human experimentation.
The exposure atmospheres studied vary, ranging from well-defined, single-component aerosols (such as black carbon or sulfuric acid) to atmospheres produced by recently developed particle concentrators, which concentrate the particles present in ambient air. The concentrations of particles studied are limited by ethical considerations and by concern for the range of concentrations, from the experimental setting to typical ambient concentration, over which findings need to be extrapolated.
Exhibit 1 at 36. Controlled human exposures studies have been conducted for decades on important pollutants such as ozone, particulate matter, nitrogen dioxide (N02),sulfur dioxide (S02), VOCs emitted in from new homes, and carbon monoxide (CO).
9. Controlled human exposure studies assess the biological plausibility of the associations observed in the large-population epidemiological studies. Controlled human exposure studies usually compare the response of an individual following exposure to clean air with their response following exposure to a pollutant that was generated or prepared under carefully controlled conditions, thus providing direct causal evidence that observed effects are related to the pollutant of interest. These studies are done under conditions that are controlled to ensure safety, with measurable, reversible physiological responses. They are not meant to cause clinically
significant adverse health effects, but rather reversible physiological responses can be indicators of the potential for more serious outcomes in susceptible populations identified in epidemiology studies. As such, controlled human exposure studies do not study individuals felt to be at significant risk; they almost always study healthy individuals or people with conditions such as mild asthma.
So there it is, Fred. See if it makes sense to you, what I have said.