ETS
AND HEART DISEASE
AND (OH YES) THOSE
"53,000 DEATHS"
It's now another standard part of the mantra that ETS "kills" 53,000
Americans a year. Stanton Glantz began
pushing that number early on and it keeps getting repeated, even though
its source has been pretty well discredited by, among others, the US Congressional
Research Service.
We will quote rather extensively from the CRS Report: "Discussion
of Source of Claims of 50,000 Deaths from Passive Smoking," Gravelle
and Redhead, CRS (3/23/94):
"The approximately 50,000
number was mentioned [in congressional testimony] by the AMA. This
statement, in turn, appears to be ultimately traceable to an article published
in 1988 in Environment International: "An Estimate of Adult Mortality
in the US from Passive Smoking," A. Judson Wells.
"The article used existing
epidemiological (statistical) studies...to estimate [as part of this total]
37,400 deaths from heart disease....The editorial in the issue containing
the Wells article was directed at that article [and] indicated that it
had received mixed reviews from referees (two recommending extensive revision,
and the third recommending against publication on the grounds that it was
too speculative) but the editors chose to publish it, despite these reviews.
"In the following years
there were a series of critiques and rejoinders relat-ed to this paper.
The main criticisms related to two points: that the evidence, particularly
with respect to heart disease, was at odds with information on the physical
levels of [actual] exposure, and that there are a variety of serious problems
with epidemiolgic studies. The following discussion elaborates on
these issues."
[There follows a discussion of the unreliability of studies based on
questionnaires--as Wells' underlying statistical studies were--and the
fact that the studies did not, and could not, fully control for confounders--other
factors that might be responsible for heart disease.]
CRS continues:
"While the death estimates
from at least some epidemiological studies are significantly larger than
the estimates of [actual] physical exposure, these results are not magnitudes
apart. The same cannot be said, however, for the Wells' estimates
of deaths from heart disease.
"Using the same type of
linear physical extrapolation [that was used for the other studies] would
result in 700 deaths from coronary disease in never-smokers. ...The portion
of the Wells 53,000 estimate from the studies, even for several years ago,
is 37,400, a number enormously larger...
"The biological plausibility
of passive smoking effects on cardiovascular disease has been the subject
of some discussion...There are limited data both to support and refute
these hypotheses [based on the action of carbon monoxide and nicotine as
they affect smokers] but exposures from passive smoke are generally thought
to be at concentrations below those at which any physiological changes
would occur...
"In
sum, this analysis suggests the Wells' estimates are so high relative to
measures of physical exposure that they seem implausible."
In 1981 and again in 1988, the American Cancer Society did two major
US studies (CPS-1 and CPS-2) and "major" means major: the first studied
1 million people, the second. 1.2. In addition to looking into the
correlations of secondhand smoke and lung cancer, it also looked for links
to heart disease. And found no connection. In either study.
-"Environmental Tobacco Smoke And
Mortality," Lee; Karger, 1992
- also LeVois and Layard, "Publication
Bias in the Environmental Tobacco Smoke/Coronary Heart Disease Epidemiologic
Literature," Regul Toxicol Pharmacol, 1995; 21
Further, the subjects from CPS-1 continued to be followed through 1998,
for a total of 39 years. Focusing on a large (35,561 subject) subset
of California never-smokers married to smokers for the full period, a statistical
analysis, peer-reviewed and published in the British Medical Journal in
2003, repeated these results-- showing incontrovertibly (0.97 @ 95% confidence)
that there was no increased risk of coronary heart disease from lifelong
exposure to secondhand smoke.
And again, this conclusion held both before and after adjustment for
logical confounders, and again, showed no variance related to either years
or amount of exposure.
Similar non-associations with secondhand smoke were found for lung cancer,
asthma, and other allegedly tobacco-related diseases, leading the authors
to state in their conclusion:
"The
results do not support a causal relationship between environmental tobacco
smoke and tobacco related mortality."
-"Environmental
Tobacco Smoke And Tobacco-Related Mortality In A Prospective Study Of Californians,
1960-98," Enstrom & Kabat, BMJ 5/17/03
Need more?
Okay. Then there was the Monica study, also commissioned by the World
Health Organization. This, too, was a 10 year, 21-country study,
conducted mostly in Western Europe but also in Russia, China, Canada and
Australia. It reported no link between heart disease and smoke.
-"WHO
Multinational Monitoring Of Trends And Determinants In Cardiovascular Disease"
Kuulasmaa Et Al
We'll stipulate this: Epidemiology--all epidemiology (even the stuff
we like) is, at best, iffy. (We've detailed some of the reasons in
How
To Read A Study, which, again, we urge you to read.) But don't
take our word for it:
Take, instead, the word of Charles Hennekens, a co-author of the Kawachi
study, below. Here's what he confessed to the New York Times (10/11/95):
"Epidemiology is a crude
and inexact science. 80% of cases are almost all hypothesis. We tend to
overstate findings either because we want attention or more grant money."
It becomes interesting, then, that of
the 31 extant epidemiological studies (at least that we know of) that
have attempted to link heart disease to secondhand smoke, 24... haven't
been able to, and only 7 have.
And though we bet you've never heard about the 24 that haven't, you
may well have heard of the publicized and hyped Kawachi study, the
most recent one of those 7.
The Kawachi study, also known as the Nurse's Study, was an ongoing study
of the general health of 122,000 nurses. It was not a study of either smoking
or exposure, nor was it a study of heart disease in specific. At
inception it asked the nurses to fill out a questionnaire.
Among a long list of questions was "are you exposed to smoke?"
This was asked only once, in 1982, and was never asked again. Between
1982 and 1992 (the parameters of the study) most hospitals in America went
entirely "smoke-free," ending workplace exposure for the subjects (all
nurses) while exposure from other sources was increasingly ended too.
But here's the real kicker: according to study results, it just happened
that the non-smoking women who claimed exposure to smoke, were ALSO:
"more likely to have reported
a diagnosis of hypertension, hypercholesterolemia [high cholesterol] and
diabetes meillitus compared with those not exposed. They were also less
likely to engage in vigorous exercise, consumed more saturated fat, and
has a lower intake of Vitamin E."
They were also more obese, less likely to use post-menopausal hormones,
came from lower socioeconomic backgrounds, and-- most crucially
"...had a higher parental
history of miocardial infarction before the age of 60."
Each of these is an undisputed risk for coronary heart disease. And
all these factors were found in quite notably higher percentages
in the subjects exposed to smoke. (And nobody's making the case-- or at
least not yet-- that secondhand smoke leads to laziness, obesity, a greater
consumption of fat, or a retroactive history of family misfortune.)
Yet Kawachi et al concluded that the higher rate of heart disease among
this particular group is caused by secondhand smoke (!?)
Nonetheless, this study was given national headlines. Proving (if nothing
else) that Kawachi's co-author was right when he said to the Times
"we tend to overstate findings...because we want attention."
-"A Prospective Study Of Passive Smoking And Coronary
Heart Disease," Kawachi Et Al, Circulation, May 20, 1997
Please note too. One of the most interesting and most important
findings here confirms what's been suspected as the monkey-wrench in all
these studies:
People exposed to smoke are notably different--and
in many crucial ways -- from the people who aren't.
This monkey-wrench may also account for the findings of the Howard study:
that non-smokers exposed to smoke showed a 20% greater increase of artheriosclerosis
in the course of 3 years than did the non-smokers who weren't. This added
"increase" by the way was 7 microns, defined as "the width of one red blood
corpuscle" and, further, as was subsequently pointed out in JAMA, "The
measuring technology used could not detect changes [even] 30 times as large."
Another point made was that measurements of artery linings can vary from
day to day in the same individual, and that two measurements taken three
years apart are hardly conclusive of anything.
The researcher never
adequately responded to these charges, but noted that his critic had
"ties to the tobacco industry." The Siegel and Glantz principles
once again at work.
Interesting too, that in the course of another
heart study, Kawachi showed the effects of (absolutely smoke-free)
"anger and grumpiness" present even higher risks of artheriosclerosis by
creating an "added stickiness of blood platelets." Mere "grumpiness"
was shown to have tripled the risk, and "moderate anger," such as experienced
during traffic jams, to "more than double the risk" of a heart attack within
the 2 hours following" it. Perhaps we can soon have a law against
traffic?
And then there's this: A much-hyped Japanese study, which "exposed"
15 (!) healthy twenty-somethings to an unspecified amount of smoke, showed
"no effect in basal coronary flow velocity, heart rate or blood pressure"
but a minute, reversible reduction in "coronary flow velocity reserves
(CFVR)"
Otsuka
Et Al, JAMA, Vol 286, Jul 25, 2001
Yet the same kind of moderate, reversible, vascular
changes were observed in healthy people who had just eaten dinner--and
of course, that's dinner in a smoke-free lab. Actually, the changes brought
on by bacon and eggs took longer (4 hours) to reverse.
-
And that change, by the way, (in CFVR) would be rapidly reversed by
drinking a glass of wine. -Lancet,
354 (1999) also Lekakis et al, European Society of Cardiologists Congress,
Vienna, 2003.
Point? There's an awful lot of grant money and headlines available to
anyone who wants to finger secondhand smoke. What isn't so available,
is a way to put all of these studies into context and judge them in perspective.
If Dr. Whelan (see "Tactics")
disapproved of the Cherner
ad on the grounds that it was "alarmist, hyperbolic" and "without scientific
basis," one wonders what she'd think of an ad recently sponsored by the
city's Department of Health (New York Times, June 25, 2002.) Under the
scare-mongering headline, "Second-hand smoke Kills," it lists a series
of (hyperbolic, scientifically-challenged) Warnings. To wit:
"WARNING: Just 30 minutes of exposure to secondhand smoke can
greatly increase your risk of a heart attack."
We wrote to Dr. Freiden, asking him for the backup to this truly outrageous
statement (which is clearly meant to imply that having tea with a smoking
friend can lead to Mount Sinai before supper.) In our original
letter we said that we were sure that he wasn't basing "this bizarre distortion
on the Japanese study which proved nothing of the kind" as we were sure--
and we remain quite sure --that he clearly knew.
Yet when we got an answer, from a Dr. Colin McCord, it offered a bibliography
of 5 (count em) studies on ETS and the heart, which forced us to
then conclude that indeed "the
Japanese study" (Otsuka) was the basis of this canard. It's the
only one of the 5 that offered a theme about "half an hour."
That study has been discussed herein and--as you see for yourself--
it
showed nothing of the kind. If you doubt that, go to the source.
The other 4 studies in Dr. McCord's bibliography are meta-analyses of
exceptionally weak epidemiological studies. So said an editorial in the
New England Journal of Medicine. (see below)
But since the city's Department of Health apparently offers the stuff
as Proof, we ought to give it a closer look. Glantz et al (1991); Law et
al (1997); He et al (1999).:
After tossing together 14 weak studies, Mr. Glantz
can only produce an RR of 1.2; Law, of 1.23; He, of 1.25.
Further, these studies have serious technical flaws. The studies that
underlay them were not compatible. Most did not control for even
the most obvious and well-known confounders for heart disease-- hypertension,
high cholesterol, diet, family history, or status of ex- as opposed to
never- smoker, etc. etc.)
Then too, all the studies cherry-picked their data from other, previous
studies carefully culled from an internet search-- excluding, at will,
any study they didn't "like."
Thus, Law admits excluding the largest studies ever done: the American
Cancer Society's CPS-1 (with 1 million subjects) and CPS-2 (with 1.2 million
subjects) "because," he said, "they were inconsistent with other
studies."
As indeed, they were. Both CPS's showed no effect whatsoever
from secondhand smoke on heart disease. The statistical results were
1.0 (95%, 0.97 - 1.04, and 95% 0.90 - 1.07)
This is exactly the reason why they should have been included--
in any meta-analysis that purported to be objective.
And the underlying studies that Law choose to use were extraordinarily
weak. Of the 19 he picked, 11 lacked statistical significance, and
7 produced results of less than 2.0. And the study populations were
of course much smaller than the CPS studies-- itself a red flag.
Nor were any of these studies (or their underlying tiers) about a half-an-hour's
exposure. All were about lifetime spousal exposure.
As for Mr. Glantz's own scientific objectivity (in his own words) please
see his Biography in our section
on
Politics.
The final of the 5 studies-- the only one we haven't read-- is by A
Judson Wells-- the source of those "implausible"
(per the CRS) deaths.
When Mr. Wells was confronted with defending his own numbers, he backed
away with remarkable speed. And though his body-count gets repeated as
though his numbers were etched on tablets, he himself brushed them off
with this:
"If people are
looking at this estimate as a proven number, that is not the case."
-"The Scientist," October, 1989
(For more about Wells and Glantz, again see Politics.)
Perhaps the best summary of all the available evidence on ETS and the
heart came from Dr. John Bailar (U. of Chi). Writing the lead editorial
in the New England Journal of Medicine (the volume that published He et
al--a study he ripped apart) Dr. Bailar confided this:
"I regretfully
conclude that we still do not know with any accuracy how much--or even
if--exposure to ETS increases the risk of coronary heart disease."
- Editorial: "Passive
Smoking, Coronary Heart Disease And Meta-Analysis," NEJM, 3/25/99