Glantz: A Selection of OSHA Comments on Lung Cancer

RESPONSES OF TOBACCO INDUSTRY WITNESSES TO QUESTIONS ON WHETHER ACTIVE SMOKING CAUSES LUNG CANCER OR OTHER DISEASES

This material is from Dr. Glantz's post hearing comment submitted to the OSHA dockett on indoor air. The page citations are to the hearing transcript of the OSHA hearing on the proposed indoor air rule that started in Sept 1994


Prepared by Stanton Glantz


Page 4359


MR. MOTLEY: For yourself, have you come to the conclusion based on the evidence that tobacco smoking 20 or 30 years will cause an increased occurrence of lung cancer among those who so smoke?
MR. HOLAN: Well, in my own family, I can't say because of past medical histories on both sides of my family, there's been non-smokers who have died of cancer, who have died of diabetes and other health-related diseases like that. So I have to throw in that hereditary aspect.
MR. MOTLEY: Is it fair to say that you, yourself, have not concluded to 100 percent of certainty that tobacco smoking by an individual for 20 or 30 years causes lung cancer, is that correct?
MR. HOLAN: That's correct.


Page 5201


MS. SHERMAN: Getting back to my original question, do you believe that active smoking causes lung cancer?
DR. IDLE: I think it's very strongly associated with lung cancer, in the same way that a number of other factors are. I've worked on a number of genes, for example, that myself and may others believe are risk factors in lung cancer, and they come out with the same sort of odds ratio in association with lung cancer, as does smoking versus non-smoking, i.e., around about ten.
MS. SHERMAN: Could you tell me which of the Bradford Hill criteria you feel that active smoking does not meet in terms of causality?
DR. IDLE: I'm not talking about active smoking. My comments in the...
MS. SHERMAN: But my question involved active smoking.
DR. IDLE: Active smoking. Well it presumably meets all of them. But as I say, I'm not trained in epidemiology. An answer to that question would best be given by an epidemiologist.


Page 5456-5457


MR. MOTLEY: Can you tell me, ma'am, if you have come to any opinions, conclusions as an epidemiologist as to whether direct smoking for 25 or 30 years, one pack a day or more, is a causal factor in cardiovascular disease of any kind?
DR. HUBERT: I believe that active smoking is a significant risk factor for heart disease.
MR. MOTLEY: Do you believe that it has been established according to the Bradford Hill, I don't know if I'll call it a hypothesis or whatever...
DR. HUBERT: It doesn't matter. There are lots of enumerations of causation. Don't work about Bradford Hill.
MR. MOTLEY: Whatever you apply, do you believe that causality has been established between long term active smoking and the occurrence of cardiovascular disease?
DR. HUBERT: I believe there is causality. I think the mechanisms are still being hotly debated.
MR. MOTLEY: In other word we know it occurs. WE just don't know how it occurs. Is that fair?
DR. HUBERT: Yes.
MR. MOTLEY: The same question with regard to lung cancer. Do you believe that long term direct smoking of cigarettes is causally related to the occurrence of lung cancer?
DR. HUBERT: Yes. Even though I do not work in the cancer area, yes, I do believe that.
MR. MOTLEY: You do believe that, you hold that view?
DR. HUBERT: Yes, I do.


Page 5807-5809


MS. SHERMAN: Do you believe that cigarette smoking causes lung cancer?
DR. LeVOIS: I haven't taken the time to actually review every study in the literature, but I'm well aware of the Surgeon General's reports and report after report that I read every time I open a journal -- not every time, but frequently. I think that smoking is certainly the leading risk factor for lung cancer and increases the risk of lung cancer.
MS. SHERMAN: I'm just a lawyer. What's the difference between a risk factor and a cause?
DR. LeVOIS: A risk factor is what we observe. You can't really observe causes. You can observe associations and in this case where we're not actually familiar enough with the underlying mechanisms to know exactly what the causes are, you put together information from a variety of sources and draw an inference.
MS. SHERMAN: Maybe a layman would say cause, and a scientist would say risk factor?
DR. LeVOIS: You hear that all the time, yeah. From many factors.
MS. SHERMAN: Are there constituents in tobacco smoke that studies have shown cause cancer?
DR. LeVOIS: Yes, there are.
MS. SHERMAN: Do you believe that cigarette smoking causes heart disease?
DR. LeVOIS: Again, it is variously shown in studies to either be a risk factor or not. For instance, the Framingham study indicates about a 70 percent increase in risk for males, but when other factors are adjusted for, does not show a significantly increased risk in females, although it shows a point estimate of around 30 percent increase, but non-statistical.
In a variety of instances, for instance in French Normandy where they have the lowest heart disease in all of Europe, they have some of the highest smoking rates and smoke very strong cigarettes. The same is true in Japanese males. They have very low heart disease rates and very high smoking rates and smoke very strong cigarettes, and their heart disease rates tend to increase when they leave Japan and become second generation Americans such as is shown in the Honolulu heart study.
So the estimates for the increase in risk, while across the board tend to be elevated, they're not always statistically significant after adjustment for other confounders. I would still tend to believe that it is an important risk factor, and that it increases the risk of heart disease, but that's variable.


Page 5682-5683, 5685


MS. SHERMAN: Do you believe that cigarette smoking causes cancer?
DR. ASHFORD: I have seen a lot of data, almost all the data is from North America, mainly from the U.S. and Canada, and northern Europe, U.K., Sweden and so on. And where I differ from perhaps a lot of other people is I've also seen a fair amount of data from other parts of the world. I've done surveys myself, I know the data is good and all the rest of it.
So the information I have to interpret is perhaps different from the information that most people think about when they think about smoking.
On the basis of that, I am not prepared to accept the conventional interpretation of the relationship between smoking and lung cancer. I've seen too many anomalies in terms of data in which I have faith.
MS. SHERMAN: So in other words, then, you would not agree with the U.S. Surgeon General's report linking the two.
DR. ASHFORD: Yes. I'm an agnostic about this. I don't know the answer. I don't think that they know the answer either but that's another matter.

...


MS. SHERMAN: Do you think that cigarette smoking causes heart disease?
DR. ASHFORD: I am agnostic. I have seen no evidence which I would interpret as indicating that cigarette smoking causes heart disease. I mean, if I can comment on that just very slightly, I've been warned not to but I will. The causes of heart disease are not well understood in general.


Page 6089-6091


MS. SHERMAN: Do you believe that active smoking causes lung cancer?
DR. [Raphael] WITORSCH: I'm going to give you a response that you're not going to be satisfied with because it's going to go back to the old multi-factorial.
JUDGE VITTONE: Just give the response, doctor.
DR. WITORSCH: Okay. I believe that it is a risk factor for lung cancer.
MS. SHERMAN: As I said to Dr. LeVois, is this a semantic problem between a layman and a scientists...
DR. WITORSCH: No.
MS. SHERMAN: What is, in your understanding, the difference between causation and risk factor?
DR. WITORSCH: If you can tell me the cause of cancer, I'll give you a Nobel Prize, and if you follow the literature on a daily basis, the mechanisms behind the cause of cancer and the theories are changing eery day.
The latest theory, for example -- I'm not going to belabor it -- relates to a certain gene transformation that determines whether a cell lives or dies, so you have mechanisms that we don't know.
I would interpret cancer as a multifactorial phenomenon where numerous factors can increase the probability of cancer happening. If the conditions are optimal, the incidents of cancer will occur. And it's not a crap shoot, don't get me wrong. But for example, there are people who smoke for 50 years, five packs a day, and never get lung cancer.
I think as a risk factor it would increase the probability of lung cancer if a person has a genetic predisposition, if their dietary situation is appropriate...
An example of that is two things that came in the literature recently which are very interesting. One was a series of papers in Europe which show that housing pet birds increased the incidence of lung cancer six-fold. I know one paper authored by Holtz, but I understand there are others, that seem to show the same thing. That's one component that has really never been examined.
It's also my understanding there was a paper published last year which showed that high fat diets are an increased risk of lung cancer as well.
MS. SHERMAN: Let me see if I understand your position. If I understand your position correctly, there is no case where we know what causes cancer. It's not just a matter of tobacco smoke or environmental tobacco smoke. At most, all we can hope to know is an association?
DR. WITORSCH: No, I don't think that's appropriate. We can get some idea of what the contributory factors are...


Page 6188


MS. SHERMAN: Do you believe that cigarette smoking causes heart disease?
DR. SPRINGALL: I don't know.


Page 6335-6340


MR. GERTLER: Maybe I'll ask it in a different way. Based on your review of the epidemiology, have you reached a professional conclusion or opinion as to whether direct or active smoking is associated with an increased risk of lung cancer in humans?
DR. SWITZER: First of all, I did not review the epidemiology for direct smoking. But I think what you're trying to ask me, if I can paraphrase this, is do I think smoking causes lung cancer. Is that what you're trying to ask me?
MR. GERTLER: Yes, I think that's better phrased. Go ahead.
DR. SWITZER: I want this on the record that I'm not responding as a statistician, because statisticians do not deal in causes. So let's put that here. I can only answer in a very personal way here, and not professionally. I know you want my professional opinion, but my professional opinion is not available with regard to causes as a statistician.
My personal opinion is I consider it a sufficiently important risk factor that I would not lightly undertake smoking as an activity.
MR. GERTLER: In your personal opinion you've reached the conclusion that direct smoking does cause an increase in lung cancer risk.
DR. SWITZER: No, I didn't say that.
MR. GERTLER: I'm groping, then, for what you said. As you phrased that question to me, I thought... Go ahead.
DR. SWITZER: I'm saying that I believe that the association between smoking and lung cancer is sufficiently strong that I would take that into account in a smoking decision.
MR. GERTLER: When you say take that into account, you mean take into account the fact that direct smoking is responsible for an increased risk in lung cancer?
DR. SWITZER: I'll take account of the fact that, let's say if I decided to do sky diving or something like that, even though I might enjoy it, I might not do it.
MR. GERTLER: So I think what you're saying is the evidence is clear to you at least, that smoking is responsible for an increased risk in lung cancer. You take that risk into account like you take other risks into account in your life in performing certain other tasks.
DR. SWITZER: Yes.
MR. GERTLER: That's correct?
DR. SWITZER: Um hmm.
MR. GERTLER: I think you have to verbalize. Did you say yes to the last question?
DR. SWITZER: Yes.
MR. GERTLER: Have you, and I don't know if I was clear on this, have you ever reviewed the epidemiology on the causal relation between direct smoking and lung cancer?
DR. SWITZER: No.
MR. GERTLER: Would you consider yourself sufficiently unfamiliar with that epidemiology to express an opinion as to whether it is a valid basis for concluding that lung cancer and direct smoking are causally related?
DR. SWITZER: Again, as a statistician, let me comment first as a statistician. This is all second hand, because I did not review those studies as I mentioned earlier. The relative risk factors are large enough that I wouldn't worry too much regarding some of the objections that have been raised with regard to the epidemiology involving ETS.
MR. GERTLER: So does that mean that you have reviewed the literature sufficiently to express an opinion about whether it's a valid basis for the conclusion that there's a relation between smoking and lung cancer?
DR. SWITZER: As I said, I did not review the literature at all. I'm just looking at the reported relative risk numbers.
MR. GERTLER: Do you have any plans to review the literature to express an opinion on the question of whether it supports the direct relation between active smoking and lung cancer?
DR. SWITZER: No.


Page 6945


MS. SHERMAN: Do you believe that active smoking causes lung cancer?
DR. STARR: I haven't really reviewed the literature on active smoking. I could only tell you what I know of through reports by others about this. I know, for example, the Surgeon General's report concluded that.


Page 6695


MR. McNEELY: Dr. Layard, do you think that active smoking causes cancer?
DR. LAYARD: I believe that active smoking is a risk factor for cancer. Smokers are more likely to get cancer than non-smokers. I do not use the word cause because it means different things to different people. There are various interpretations of the term cause and I think that it is best in this context to use the language that I have used and that's what I do. This, after all, is a probabalistic exercise, we're talking about statistics here.


Page 9192-9195


MS. SHERMAN: Do you believe active smoking causes lung cancer?
DR. ROTH: It's a risk factor.
MS. SHERMAN: Have you done any analysis of the studies dealing with lung cancer and active smoking?
DR. ROTH: Any formal analyses like you see here? No. I've looked at them.
MS. SHERMAN: Do you have any reason to disagree with the conclusion of the Surgeon General linking active smoking with development of lung cancer?
DR. ROTH: I don't know precisely which Surgeon General report you're talking about and I've never done a formal analyses of all these data. I just accept it at face value that it's a risk factor.
MS. SHERMAN: Have you read the Surgeon General's report on the subject from 1964 to 1989?
DR. ROTH: In 1965 I read the 1964 report.
MS. SHERMAN: And that was the last one?
DR. ROTH: Well, I've looked at tables, I'm sure. I'm sure I've looked at tables that appeared in subsequent Surgeon General reports, you know, I've looked at these tables but I have never done a formal analyses of them.
MS. SHERMAN: How would you go about identifying what we call a confounder?
DR. ROTH: There are a number of steps. Step number one is just look through the literature. Look at the health end point that you're investigating, heart disease, lung cancer, and look through the literature to see what's available, has this been shown to be a risk factor. And that's step number one.
Even a simpler step, go through the NIH reports. There are books on cancer that they publish and they list the risk factors for all the different cancer end points.
MS. SHERMAN: Now, are you equating the term risk factor with the term confounder?
DR. ROTH: Sure. It depends how the study is designed.
MS. SHERMAN: Well, can you explain to me -- well, are you saying that a risk factor and a confounder is the same thing?
DR. ROTH: No, I'm not.
MS. SHERMAN: Okay. Well, then, what's the difference?
DR. ROTH: We're talking about semantics. If you go through a study, okay? Let's say you have a specific health end point, let's say A, heart disease, lung cancer, asthma attack, whatever it is. And let's say also that there is a variable that's associated with that health end point. Let's call it B. So heart disease -- and let's accept for now that cholesterol levels is a predictor of heart disease, so B is a risk factor of A, okay?
Now, let's say in that study you did not control for a third variable. Let's call it C. And C is directly related, let's say you know that it's not related to A but it's highly related to B. In other words, people that have high cholesterol levels also have this. Well, introducing C into the analyses or not considering it could confound the results. C is a confounder.
A risk factor is something that statistically in epidemiological studies is associated with the health end point.
MS. SHERMAN: So it's a casual relationship.
DR. ROTH: That's a dangerous word, okay?
MS. SHERMAN: I'm not a statistician.
DR. ROTH: Okay. Whenever you say that, it sends shivers down my spine. Cause is not an epidemiological term. Associations, predictions, that's the terminology of epidemiologists. Statistical associations. Not cause.


Page 9562-9565


MS. SHERMAN: So would it be fair to say that you believe that active smoking causes lung cancer in humans?
DR. GORI: It's fair to say that I believe that active smoking is a risk factor.
MS. SHERMAN: What is the difference between a risk factor and a cause?
DR. GORI: Let me try to put it in non-scientific terms.
MS. SHERMAN: Please. Yes.
DR. GORI: If a person has influenza he must have the virus of influenza, otherwise he will not have influenza. That's a necessary cause. It's a cause. If a person has lung cancer, he could or could not have smoked and of course there are a number of other things that would have caused that particular lung cancer, so that we are not in a position to ascribe that particular lung cancer to any particular cause. That's why we call smoking a risk factor, rather than a cause. Is that sufficient explanation or differentiation between what the concept of cause is and the concept of risk factor is?
MS. SHERMAN: I must admit I am still a little bit confused.
DR. GORI: All right. Let me try to expand on this issue.
MS. SHERMAN: Okay.
DR. GORI: The Surgeon General claims that 90 percent of lung cancers in the United States are caused by smoking. We have about 140,000 lung cancer cases a year, so about 126,000, 127,000, according to the Surgeon General, are caused by smoking.
Then we have other claims by other agencies that I will not name contending that about 50,000 lung cancer cases a year may be caused by asbestos, as many as 35,000, 40,000 may be caused by radon exposures, as many as another 50,000 or 60,000 may be caused by occupational exposures. You have seen also the slide that I presented with many reported risk factors for lung cancer. There are simply not enough lung cancers available in the United States to satisfy all these claims.
MS. SHERMAN: Well, then in order for something in your definition, then, if more than one thing contributes to a disease, then you don't say it causes the disease?
DR. GORI: We call it a risk factor. Again, let me give you an example. If you have in front of you a person with lung cancer, you cannot say what was the cause of that lung cancer. You can't. Whether he smoked or not. If you have a person in front of you with tuberculosis, you can definitely say that the bacillus of tuberculosis is or was the cause of that particular disease. Is that clear?


Page 10088-10093


MR. MYERS: So that it's your view that we don't yet have enough scientific evidence to say that direct smoking is a cause of lung cancer, is that correct?
DR. BRIDGES: Not from the studies that have been done.
MR. MYERS: I'm assuming then, you tell me if I'm wrong, that you would also say that if we don't have it for lung cancer, we certainly don't have enough scientific evidence to say that direct smoking is a cause or an initiator of cardiovascular disease. All we know is there is a statistical association.
DR. BRIDGES: At this stage, yes.
MR. MYERS: Have you read the study of the Royal College of Physicians or any of our Surgeon General studies that have concluded that there is enough evidence?
DR. BRIDGES: Of course, yes.
MR. MYERS: And you simply disagree with their conclusions.
DR. BRIDGES: I don't think they've looked at the mechanistic data sufficiently. I think they've used the word risk factor and then jumped to, well that means a cause. I'm just, as a toxicologist, rather more cautious because I rely on the direct experimental data.
MR. MYERS: So that it's your view that we simply don't have enough evidence about a substance like tobacco smoke to say that there's a causative relationship or an initiator relationship until we fully understand the mechanism by which the substance actually causes the health-related problem.
DR. BRIDGES: Let me take you to my position as a toxicologist. I have no ax to grind at all about...
MR. MYERS: I'm just trying to understand it.
DR. BRIDGES: In terms of the animal studies which have been done with tobacco smoke, it's generally been the case that cancer hasn't been shown. So for a toxicologist, if you can't show that tobacco smoke is causing an effect in animals, you've got a problem in identifying how the thing works. That's why I can only consider it at this stage a risk factor.
MR. MYERS: Would you also conclude then, that we simply don't yet have enough scientific evidence to conclude that direct tobacco smoke is a cause or an initiator, to use your word, I believe, of chronic obstructive lung disease? Would the same be true?
DR. BRIDGES: From the experimental animal data, that is certainly --
MR. MYERS: That's true. You're not prepared to reach that conclusion yet?
DR. BRIDGES: Not from a mechanistic point of view, no.
MR. MYERS: Well, but I'm asking your professional scientific point of view. That's what I'm getting, I'm hoping, your best professional scientific viewpoint is that you don't yet believe we have enough scientific evidence to conclude that direct smoking causes lung cancer, cardiovascular disease or chronic obstructive lung disease.
DR. BRIDGES: If you're asking me on the basis of the data, that's the case. I could speculate but I don't --
MR. MYERS: I'm not asking for speculation. I'm asking for your views.
That makes a couple of other questions I was going to ask difficult. If your conclusion is that there's no proven causal relationship, then it goes without saying that you also would say for direct smoking that we don't have any evidence about what the threshold is for a link between tobacco smoke and these diseases because we don't have any evidence that it actually causes them, is that fair to say?
DR. BRIDGES: Well, in terms of risk factors, you can look, of course, in human studies at a dose-response relationship for that risk factor. You don't need a mechanism.
MR. MYERS: Okay.
DR. BRIDGES: But in animal studies, then we can't get off first base because we haven't got a decent animal model that mirrors what appears to be happening in man.
MR. MYERS: Okay. Then looking at some threshold data for the purpose of risk factors, as you say, have we identified for direct tobacco smoke a threshold above which it's likely to cause harm and below which it's not?
DR. BRIDGES: I'm not a specialist in mainstream tobacco smoke and it's not part of my testimony so I would prefer not to get too --
MR. MYERS: Have you looked at the scientific data?
DR. BRIDGES: Some time ago but not recently on mainstream at all. No.
MR. MYERS: So you simply don't know, is what you're saying to us? As a scientist, you're not prepared to give us a professional opinion?
DR. BRIDGES: When I give a professional opinion, I like to take into account all the latest data and I haven't seen the latest data.
MR. MYERS: So in forming the opinions you've given us here today, you haven't gone back and looked at any of that data.
DR. BRIDGES: Not on mainstream. No.
MR. MYERS: Not on mainstream at all.
DR. BRIDGES: No.
MR. MYERS: And I asked that with regard to lung cancer but I'm assuming the same is going to be true for coronary heart disease and chronic obstructive lung disease.
DR. BRIDGES: I haven't gone back over that data recently for these hearings.
MR. MYERS: So you're not prepared to give us an opinion one way or the other as to whether or not there is data that would indicate that there is or there is not a threshold with regard to mainstream smoke?
DR. BRIDGES: I would need to look at the data again before I made that opinion. Yes.


Page 10776, 10779-10780


MS. SHERMAN: Let's start with active smoking. Do you believe that active smoking causes low birth weight?
DR. HOLCOMB: I don't know. I know that there is literature on that. I have not reviewed that literature on direct smoking.

...


MS. SHERMAN: Do you believe that active smoking causes cancer in humans? Excuse me, lung cancer in humans.
DR. HOLCOMB: Believe me, I am not an expert on direct smoking. From what I can see in the literature, and I have not read all the Surgeon General's reports. I have read parts of them, or I have looked at a document now and the, but from what I can gather, I would call it an increased relative risk for those that are direct smoking compared to those that are not.
MS. SHERMAN: So the answer is yes?
DR. HOLCOMB: Certainly not anything that would indicate to me that it is the causal mechanism, but certainly an increase in relative risk.
MS. SHERMAN: Would you have the same answer if I asked you if you believed active smoking resulted in the development of heart disease in humans?
DR. HOLCOMB: I would say the same thing, except that the increase in relative risk there is much, much smaller.


Page 10643


MS. SHERMAN: Do you believe active smoking causes lung cancer?
DR. [Philip] WITORSCH: I think active smoking is a risk factor for lung cancer. I don't think you can reach a causal conclusion in that sense.


Page 10985


MS. SHERMAN: Do you believe that active smoking causes lung cancer in humans?
DR. NEWELL: I believe that smoking as such has a high risk factor for causing cancer in humans. As a causation aspect that has not been demonstrated to my satisfaction.


Page 11633


DR. SAMET: Does active smoking cause lung cancer?
DR. COGGINS: I don't know because the evidence here is much less, weak. We certainly have the epidemiology but we do not have corroborating animal evidence. There are numbers of studies, dozens upon dozens of inhalation studies that have been performed with mainstream smoke, many different species, durations of many years, and uniformly these studies have not produced an increase in the background rate of lung cancer as a result of those exposures. So we have the epidemiology, we do not have the corroborating animal evidence and I do not believe that we have any estimate of mechanism. So my answer is no, I do not think so, we just don't know based on the evidence we have today.


This document's URL is: http://www.tobacco.org/Documents/osha/oshaglantz.html
Return to OSHA Index Page


***********************
Go To: Tobacco BBS HomePage / Resources Page / Health Page / Documents Page / Culture Page / Activism Page
***********************

END OF DOCUMENT