OSHA: Proposed Standard For Indoor Air Quality: ETS Hearings, September 23, 1994

UNITED STATES DEPARTMENT OF LABOR

OCCUPATIONAL SAFETY AND HEALTH ADMINISTRATION

PUBLIC HEARING

PROPOSED STANDARD FOR INDOOR AIR QUALITY

Friday,

September 23, 1994

Department of Interior

Washington, D.C.

The above-entitled matter came on for hearing,

pursuant to notice, at 9:36 a.m.

BEFORE: HONORABLE JOHN VITTONE

Administrative Law Judge

AGENDA

PAGE

Witnesses:

Judson Wells 951

Questions:

Pat Sirridge 988

Jeff Furr 1024

John Rupp 1072

James D. Woods 1104

Questions:

Michael Jawer 1139

George Benda 1164

Saneya El-MeKawi 1178

Debra Janes 1180

Lee Hathan 1181

EXHIBITS

EXHIBIT NO. IDENTIFIED RECEIVED

26 1020 1020

27 1072 1072

28 1103 1103 P R O C E E D I N G S

(9:36 a.m.)

JUDGE VITTONE: We finished last night with

Dr. Samet. We will begin this morning, and we have three witnesses on the schedule.

Ms. Sherman?

MS. SHERMAN: I would like to call first on

Mr. Judson Wells.

JUDGE VITTONE: Would Mr. Wells come forward, please. Would you come up here, sir?

Off the record a second.

[Discussion off the record.]

JUDGE VITTONE: Sir, would you identify yourself, please, for the record?

MR. WELLS: My name is A. Judson Wells, and I'm testifying this morning for OSHA.

JUDGE VITTONE: Who are you affiliated with, sir?

MR. WELLS: I guess I'm a freelance consultant, in a way.

JUDGE VITTONE: Okay.

MR. WELLS: I am a half-time volunteer, roughly, for the American Lung Association.

JUDGE VITTONE: All right. You've previously submitted your comments for the record?

MR. WELLS: Yes.

JUDGE VITTONE: Okay. Let me get out of your way, and you can begin your presentation.

MS. SHERMAN: Mr. Wells, let me ask you before you start. Every time you ask to have a slide shown, please identify it by name or number so that the transcript will make some sense.

JUDGE VITTONE: You may begin, Mr. Wells.

A. JUDSON WELLS

MR. WELLS: Yes. As I said, my name is A. Judson Wells. I graduated from Harvard College in 1938 with a summa cum laude degree. I went on at Harvard and finished a PhD there in physical chemistry in 1941.

I was employed in the chemical industry, specifically by E.I. dupont and deNemours Company, from 1941 until 1980.

I worked in chemical research, research management, and general management. From 1969 to 1980, I was director of a business division, with revenues at that time of $125 million. Since 1981, I have served as a volunteer consultant in the smoking and health area for the American Lung Association. For the whole of that time, I have studied the scientific literature on the health effects of passive smoking.

From 1989 to 1983, I was an unpaid consultant to Kenneth G. Brown, Incorporated, a subcontractor to the U.S. Environmental Protection Agency.

Their work leading up to the publication of their report, Respiratory Health Effects of Passive Smoking, Lung Cancer, and other Disorders, I am a co-author of that report.

More recently, I have consulted, again unpaid, for the U.S. Occupational Safety & Health Administration on Health Effects of Passive Smoking, and I am testifying on their behalf today.

In my testimony today I will be covering two topics. One is the Association of Exposure to Environmental Tobacco Smoke, which I will call ETS and heart disease; and, secondly, methods for observe, passive smoking relative risks for the effects of smoker misclassification.

I will also comment on the ETS part of OSHA's preliminary quantitative risk assessment as published in the Federal Register April 5, 1994.

On heart disease, exposure to environmental tobacco smoke, also known as passive or involuntary smoking, results in a number of adverse health effects. Although most of the attention for adults has been directed towards lung cancer, heart disease, at least in terms of probable deaths, is much more important.

The reason is that the percent increase in risk from ETS exposure is about the same for both diseases. Lung cancer among people who have never smoked, is a rare disease, while heart disease among never-smokers is many times more common.

Therefore, the same small fraction of deaths, multiplied by a much larger number of total deaths, results in a large number of deaths attributable to passive smoking.

Over the past six years, there have been five papers in the peer reviewed literature that are known to me and that have assessed the available evidence on adult mortality from passive smoking and heart disease. Each of these papers is appended as part of my testimony, and the five papers are listed in Table 1.

The first is a paper of mine published in December 1988 in the Journal Environment International. The 7 epidemiologic studies then available on passive smoking and heart disease were reviewed.

The relative risks from the 7 studied were pooled to develop a combined risk, which indicates a 30 percent increase in risk for ETS exposure and a predicted 32,000 heart deaths per year in the U.S. among non-smokers that are caused by passive smoking.

The second important peer reviewed paper is that of Glantz and Parmley published in January 1991, in circulation, which is the leading medical journal of the American Heart Association.

They reviewed not only the epidemiologic studies and Wells' death estimate, but also provided a thorough review of the physiology and biochemistry that connects ETS exposure with heart disease. They conclude that, quote: "ETS causes heart disease," unquote.

The third important paper came from Karl Steenland of the National Institute for Occupational Safety & Health, that was in the January 1992 issue of the Journal of the American Medical Association.

He reviewed, again, the epidemiologic studies, now grown to 9, and the biologic plausibility. He then made a risk assessment based on the relative risk from the largest U.S. study, namely, Helsing, et. al, and concluded that if the epidemiologic results are valid, then there 35,000 to 40,000 ischemic heart disease deaths per year in the United States, that are associated with passive smoking.

Ischemic heart disease is that type of heart disease where there is obstruction in the coronary arteries so that part of the heart muscle becomes disabled because of lack of oxygen.

The fourth paper is a medical scientific position statement from the American Heart Association, published in the August 1992 issue of Circulation. The authors, Taylor, Johnson, and Kazemi, reviewed again the available evidence and concluded that, quote, "ETS is a major preventable cause of cardiovascular disease and death," unquote.

The fifth important peer reviewed paper is one of mine, published August 1st, 1994, in the Journal of the American College of Cardiology, another leading medical journal in the heart field.

This paper reviews the newer studies since Glantz and Parmley that support the biologic plausibility of a link between passive smoking and heart disease. Results of the epidemiologic studies, now grown to 13, are again pooled. They indicate a 37 percent increase in ischemic heart disease morbidity, and a 22 percent increase in heart disease death for those exposed to ETS at home versus those not so exposed.

Then, using the EPA's methods of calculating deaths from the relative risks, it is concluded that in 1985, there were 62,000 heart deaths per year in the U.S., caused by passive smoking.

These five papers, covering some 48 pages of medical journal text, cannot not be reviewed in detail here, but the main points will be summarized. There are several ways in which tobacco smoke can affect the heart, even though there is no direct contact between the smoke and the heart itself.

As with lung cancer, active smoking causes heart disease. Therefore, we would expect passive smoking also to cause heart disease but in lesser amounts.

There is one important difference between lung cancer and heart disease as far as ETS exposure is concerned. With lung cancer, the only effect is long-term; say, 20 years exposure before cancer appears. With ETS and heart disease, there are both long-term and very short-term effects, some apparent after only 20 minutes exposure.

Let's look at the short-term effects first.

The platelets in the blood are one of the factors that determine the blood's tendency to coagulate. Platelet sensitivity is a measure of this tendency. Active smokers have a lower platelet sensitivity than non-smokers, meaning that their blood is less resistant to clotting.

The evidence for decreased platelet sensitivity among non-smokers exposed to ETS, comes from the laboratories of J.W. Davis in Kansas City and

H. Sinzinger at Vienna, Austria.

For example, Davis found that non-smokers exposed to only 20 minutes in a hospital lobby, where smoking was allowed, lost about 60 percent of their platelet sensitivity advantage over active smokers.

However, their platelet sensitivity returned to normal shortly after the exposure ceased. Similar results from Burghuber, et al, and Sinzinger's group are shown in Figure 1.

Here, the ETS exposure is also for 20 minutes but at a somewhat higher level. As you can see, the non-smokers have lost about 80 percent of their platelet sensitivity advantage over the smokers. The before and the after, you can see, obviously.

Sinzinger's group also found that after repeated ETS exposures, nonsmokers baseline platelet sensitivity was reduced to a level near to that of smokers. Low platelet sensitivity is a known heart risk factor.

Experiments by Davis with sham cigarettes indicate that the lower platelet sensitivity may be related to the nicotine in this cigarette smoke.

Another short-term effect is that provided by the carbon monoxide in the smoke. Carbon monoxide is the odorless, colorless, toxic gas that also occurs in the exhaust from automobiles.

Carbon monoxide reacts with a hemoglobin in the blood and renders it incapable of transporting oxygen. It is the ability of the blood hemoglobin to carry oxygen from the lungs to the heart, brain and muscles, that sustains life.

Typical ETS atmospheres contain from 3 to 25 parts per million of carbon monoxide resulting in .4 to 5 percent of the hemoglobin being inactivated.

A typical situation would be ETS with 15 parts per million of carbon monoxide resulting in 2 percent of the blood hemoglobin being inactivated.

When 30 percent of the hemoglobin is inactivated, the result is death. With part of the blood hemoglobin inactivated by carbon monoxide from ETS exposure, the heart must work harder. That is, it must pump more blood with the same level of physical exertion.

The carbon monoxide also affects the ability of the heart to process oxygen by attacking some of the proteins and enzymes in the heart muscle that is essential to what is called myocardial/mitochondrial respiration.

Thus, ETS exposure reduces the effective blood supply of the heart, while at the same time reducing the heart's ability to process the blood that it receives.

This results in reduced exercise capability, both in healthy people and particularly in people with existing coronary disease.

Going on now to long-term effects, one theory about how heart attacks arise is, first, there damage to the coronary artery wall; then plaque builds up around the injury, eventually reducing the blood's supply to the heart, and thereby causing the heart attack.

In the platelet experiments that I described earlier, it was also noticed that short-term ETS exposure of nonsmokers results in an increase in endothelial cell carcasses in the blood. This indicates damage to the endothelium, which is the very thin lining of artery walls and blood vessels.

Once the initial damage is done, researchers have found that exposure to cigarette smoke accelerates the growth of these plaques. In other words, not more plaques form but bigger ones form sooner.

These experiments in mice, pigeons, chickens, rabbits and dogs, so far, indicate that it is the poly aromatic hydrocarbons in the smoke that are causing the effect.

The most recent of these papers, namely that by Penn, et al, show a significant increase in plaque development when cockerels were exposed to the smoke from only one cigarette over a 16-week period.

Moreover, one researcher has found through ultrasound experiments that the arterial walls in humans become thicker if the subjects smoke or are exposed to ETS.

Another, I'd say, medium-term effect is that ETS exposure appears to lower the high density or good cholesterol in the blood or to raise the total cholesterol high density cholesterol ratio. These are known heart disease risk factors.

Other researchers have found that ETS exposure reduces plasma ascorbic acid, which is Vitamin C, in nonsmokers by an amount that is 65 percent of the reduction experienced by active smokers.

Thus, we have ample biologic evidence that passive smoking can affect the blood in ways that injurious to the heart and that these effects are often stronger than one would expect considering the difference in dose between active and passive smoking.

To get some idea of the magnitude of the heart risk from passive smoking, one must turn to the epidemiology. In my 1994 paper, I reviewed 13 studies based on over 3,000 cases, where either heart disease or heart death was studied relative to ETS exposure.

This is an enormous data base compared to what is usually available in regulating work place toxins. Five of the studies are U.S. based and the others come from Australia, China, England, Japan, New Zealand, and Scotland.

Eight of the studies are mortality studies, and four of them have data for both males and females, resulting in 12 separate data points that are shown in Figure 2.

The 12 odds ratios or relative risks for prospective studies are platted on a log scale against the statistical weights for each data point. You must remember, the larger studies have larger statistical weight than the smaller studies.

As you can see, for the smaller studies, the ones to the left here, there is considerable scatter. When the statistical weight gets beyond about 20, the odds ratios become very stable in the 1.15, the 1.30 range.

The combined odds ratio for all of the studies, as shown at the right, is 1.22, and is dominated by the large, statistically significant Helsing, et. al study, which comprises 42 percent of the total cases.

Many of the other studies are too small to reach statistical significance at the 95 percent level, but they do indicate appreciable increased risk.

The results of these smaller studies can be pooled to see if, together, they indicate higher statistical significance. When this is done, we find that in mortality studies for both men and women, in the Helsing study is removed, the pool results for the remaining studies, show an odds ratio of 1.26, that is still highly statistically significant.

A new study from Xian, China, not included by OSHA in their proposal in the Federal Register, while too small, 59 cases, to show statistically significant adjusted relative risk, did investigate the heart effects on nonsmoking women exposed to ETS both at home and at work.

For exposure at work, there was an 85 percent increase in risk, with a highly statistically significant trend of increasing risk with increasing amounts of ETS exposure.

Only 8 percent of women in that part of China smoke, but most of the men smoke, and they smoked freely at work, so the study provides an excellent opportunity to judge the effects of ETS exposure at work on heart disease in nonsmokers.

In many of the ETS heart studies, accrued risks were adjusted for various other heart risk factors. The relative risks noted above are the adjusted relative risks after adjusting for such factors as age, blood pressure, cholesterol, personal or family history of heart disease, weight or body mass index, exercise, marital status, education or social status, and diabetes.

Thus, the potentially and most important confounders have been considered in at least some of the studies and found not to explain the observed increased risk.

Comments from the tobacco industry have emphasized the large number of potential confounders and how they might affect the epidemiologic results. Yet, in no study of passive smoking in either heart disease or lung cancer has adjustment for confounders reduced the relative risk to 1.00.

It is evident in my 1994 paper, although adjustment for confounders in some cases raises relative risk and in others lowers the risk, the overall effect is that those studies that adjusted for more potential confounders had higher relative risks than those that adjusted for fewer.

Also in the literature, there are indications that things don't always go in one direction.

For example, LeMarsh, et. al., found that beta carotene intake went down as exposure to ETS went up, and that's a possible confounder for lung cancer. The intake of cholesterol and fat also went down with ETS exposure, and that's a possible protective effect for heart disease.

Lower social class is thought to be a potential adverse confounder for both lung cancer and heart disease, but Humble, et al, and the Evans County, Georgia perspective study, found that higher social status whites had higher passive smoking relative risks for heart disease than did lower social status whites.

Diet is alleged to be an important potential confounder, yet in the EPA report, diet effects in 9 of the passive smoking studies on lung cancer, were investigated. This included all of the studies that contained data on diet effects, including Kalandidi,

et al., and Fontham, et al., that were specifically designed to measure possible confounding by diet. No appreciable effect on the observed relative risk was found.

The heat disease studies have not adjusted for dietary intake as such, but a number of them have adjusted for total cholesterol and high density cholesterol, which should, in part, reflect dietary habits.

Again, no confounding effect was found that nullified the observed passive smoking relative risk.

LeVois and Layard, in their August 10, 1994 submission to OSHA, have presented analyses of the data in the CPS 1 and CPS 2 studies of the American Cancer Society, and also a study based on the national mortality follow back survey, all three of which are purported to show no increase in heart disease relative risk from exposure to ETS.

In none of these analyses was there any adjustment for any heart disease risk factors, other than age and race. This is a serious shortcoming since, as was observed earlier, studies that adjust for more heart risk factors, tend to have higher relative risks.

In terms of the quality tier levels in my 1994 review of Heart Disease and Passive Smoking, there are three studies. These three studies would have been assigned to the lowest tier level, Level 4, and would not have been included in my combined relative risk that was used for the basis of the risk assessment.

The quality tier levels were independent of the size of the studies, which was taken into account by the statistical weight. The CPS 1 study had a rather controversial history as far as passive smoking is concerned.

After Hiriyama and Trichopoulos papers on passive smoking and lung cancer were published in 1981, Lawrence Garfinkel wanted to add a comment on passive smoking to his paper, which was in the National Cancer Institute Journal.

This was a paper on time trends and lung cancer mortality among non-smokers.

Karl Hammond, the former research head of the American Cancer Society, and who had gathered the CPS 1 data in the first place, and his colleague, Ervin Salakoff, both, I believe on the ACS Board at that time, were very opposed to publishing the passive smoking result, so when it was finally published, they expressed their reasons for opposing the publication in a paper in Environmental Research, 1981, Volume 24, pages 444 to 452.

Hammond said that he, and I quote: "Would have liked to estimate lung cancer death rates in relation to the amount of passive smoking among female subjects who never smoked. He" -- and he's referring here to

himself -- "he refrained from attempting to do so for the following reasons.

"Since his perspective study was not designed for that purpose, no special information on the subject was obtained. Information was available on the smoking habits of the husbands of many of the married women in the study but not on the smoking habits of the former husbands of women who were widowed, divorced, separated, or married for a second time.

"More important, in America, at that time, women were not generally barred from public and social gatherings where men were smoking, and smoking husbands who smoked, generally did much, if not most of their smoking, away from home," unquote.

Dr. Hammond and Dr. Salakoff then go on to point out why studies in Greece and Japan are more likely to yield a meaningful result because of the traditional segregation of married women from men, other than their husbands, in those societies.

These criticisms of CPS 1 regarding passive smoking and lung cancer apply at least as much to passive smoking and heart disease. CPS 2 suffers from the same problems. The questionnaire for CPS 2 did include one simple question on ETS exposure that covered current exposure only, at home, work, and other.

However, LeVois and Layard did not avail themselves of any of this information in their analysis.

They relied, instead, as Garfinkel did, on matching spouses on the basis of their individually determined smoking prevalence. So all of Hammond and Salakoff's criticisms of CPS-1 apply to the LeVois and Layard analysis of CPS-2 as well.

Their analysis of the data for the National Mortality Feedback Survey, aside from its failure to adjust for the various heart disease risk factors, relies on the strange choice for controls, namely, persons who died of other diseases, not smoking related.

There is a substantial difference in age of death between the cases and control with the cases living longer.

This NMFS survey may not provide a good basis for making a meaningful estimate of risk of heart disease from passive smoking.

Summarizing the epidemiology that qualified for Tier 3 or better quality rating in my 1994 paper, there is a 20 to 30 percent increase in risk of ischemic heart disease associated with exposure to spousal or household ETS.

There's also strong evidence that the risk from ETS exposure at work are similar to those experienced at home. Also, these increases in risk cannot be accounted for by the other known hard risk factors, and, as I shall note later, they are not accounted for by the misclassification of smokers as non-smokers.

I'll go on now to the effects of smoker misclassification on the passive smoking relative risks.

This part of my testimony, I will be discussing the misclassification and current and past smokers as never smoked, the effects that such misclassification has on the observed relative risks of passive smoking studies.

In doing so, I will be defending Appendix B in the EPA report on ETS and lung cancer. Appendix B describes EPA's methods of dealing with this type of misclassification. I was the author of Appendix B, and the tobacco industry consultants have criticized our methods.

I will deal with the effects of smoker misclassification and heart disease relative risks later.

Why is smoker misclassification an issue? It is known that a small percentage of smokers, if asked if they ever smoked, will say no.

It is also known that smokers tend to marry smokers and nonsmokers tend to marry nonsmokers. Therefore, for a given level of misclassification, more of these real smokers will show up in the group of so-called "never smokers married to the smokers" than the group of "never smokers married to the never smokers".

In the case of lung cancer, the high relative risk among these few misclassified smokers, will raise the average lung cancer risk of the self-reported never smokers, and they will raise it more for those married to the smokers than for those married to the never smokers.

This, then, will create a perceived increase in risk that could be mistaken for a passive smoking effect. There is no real question that such a misclassification effect exists. The real question is how big is it.

The EPA method offer estimating the bias introduced by smoker misclassification was developed by Dr. Walter Stewart of the Johns Hopkins School of Public Health and myself.

Dr. Stewart is an expert in occupational health epidemiology. The method is basically his which I have adapted to passive smoking. Our method involves dividing the misclassified smokers into three groups:

One: The current, regular smokers, with cotinine levels and body fluids, like average, self-reported, current smokers.

The second group are current occasional smokers, and the third group is classified ex-smokers.

This subdivision of the misclassified smokers was first suggested by Peter Lee. Misclassification rates are then developed for each class of misclassifieds, using cotinine data for the current smokers, discordant studies for the ex-smokers.

Cotinine is a longer-lived metabolite of nicotine which has become the biomarker of choice for determining relative levels of ETS exposure.

Proportionate distributions of controls and cases by smoking status of subjects and spouses are developed using demographic data and smoking relative risks. Misclassification rates have been applied to the various subclasses of smokers to estimate the proportions of each category misclassified.

These are then subtracted from the proportions of observed never smokers, yield a proportion of true never smokers among exposed and unexposed cases and controls. In these numbers, the corrected relative risks can be calculated.

Our method has been peer reviewed by EPA's Science Advisory Board. In late 1991, Peter Lee, a consultant to the tobacco industry, came up with an alternative method for estimating the smoker misclassification bias.

As shown on EPA's Appendix B and shown here in Table 2 in condensed form -- this may be a little hard to read -- but basically what it shows is that when the same inputs are used Lee's methods and ours give essentially the same results.

However, Lee's estimates are different, so that his results are different; namely, they come to a much higher bias. There are four important inputs for these calculations. They are, as shown in Table 3, the misclassification rates, the prevalence of smoking among the subjects. The more smokers in the cohort the more will be misclassified.

Three: The relative risks assumed for the misclassified smokers; and,

Four: The marriage concordance among smokers and nonsmokers; that is, the relative degree to which smokers marry smokers and the nonsmokers marry nonsmokers.

Regarding the first important input, namely, the misclassification rates, we used for current smokers, all of the literature data on females, where we could get individual cotinine measurements for each individual from the authors of the papers.

This allowed us to tell exactly how many of the false negatives among the never smokers were really regular smokers and how many were only occasional smokers. We are the only group that has gone to this extent to develop accurate current smoker misclassification rates.

Similar care was taken with the ex-smoker surveys.

We concluded that 1.09 percent of regular smokers would say "never" if asked; also, 24.2 percent of occasional smokers and 11.7 percent of ex-smokers.

Lee used a single, overall misclassification rate which, in effect, was about 30 percent larger than ours. We used smoking prevalence and smoker relative risks that were appropriate for each study. Lee used values that, in some cases, were twice as high as indicated in the studies themselves.

On marriage concordance, he and we used similar values. The overall effect, when these overages were all multiplied together, was that his estimated bias, as I think I showed there on that previous table, was, in some cases, several times as large as ours.

There is evidence that, even in our evidence of the misclassification effect, our corrections are too high. Our data come from some epidemiologic studies but mostly from community survey type studies.

For the regular smokers and the ex-smokers, the two most important classes, the misclassification rates from the community type surveys tend to run about 7 times as large as those from the few epidemiologic studies that we have.

For example, Fontham, et al., is one of the best lung cancer epidemiologic studies. After they went through all of their normal questionnaire procedures, they ran cotinine levels on the remaining subjects who said they never smoked. Among the cases, they found no regular smokers misclassified versus our estimate of 1.09 percent, and about 7 percent of occasional smokers versus our estimate of 24 percent.

Another indication that our bias estimate is high is found in the male data. Here, smoking prevalence are high. When we used male misclassification rates, derived as above for the female rates, namely, from this mixture of community surveys and epidemiologic studies, we found in some studies that we were making corrections for, that the number of smokers misclassified as never smokers, exceeded the number of self-reported never smokers that we had to begin with.

This is evidently impossible; again, indicating that our misclassification rates are too high.

Another reason that our estimates of smoker misclassification bias in the EPA repor

Tobacco.org