and the elderly which have once again been targeted for deep cuts that they cannot tolerate.
Mr Chairman, heavily taxing tobacco is an important step toward ensuring a healthier
America. Thank you for this opportunity to appear before the Committee. I will be pleased
to answer any questions that you may have.
76-304 0-94-20
604
Chairman Rostenkowski. You know, the Federal Government
can directly regulate an activity or discourage it by means of tax-
ation. Should the government use either of these tools to decrease
the aggressive advertising by tobacco manufacturers as well?
Dr. Goldstein. I think that it should be a multifaceted ap-
proach. One aspect of it is to remove the tax deductibility they get
for advertising. That would I think decrease their advertising, /aid
raising the excise taxes would — there is good data that show that
this would decrease the utilization of the product, especially by
younger users.
Chairman Rostenkowski. You mean just eliminate the deduct-
ibility for the tobacco industry and not anyone else? You think that
is fair?
Dr. Goldstein. Well, we worked with the major league ball play-
ers and we talked about removing the advertisements in the ball-
parks. And, you know, that is one issue that we couldn't touch, that
we wouldn't have the right to interfere with their ability to adver-
tise.
Whether removing the deductibility of tax for advertising is an-
other consideration.
Ms. NORTHUP. Mr. Chairman, too, I would point out that adver-
tising appeals almost only to children. All the new smokers, all the
statistics show come before the age of 21, mostly down around 13,
14.
Now, you will have to wrestle with whether or not you can deal
with advertising for tobacco products separately from other adver-
tising, but most people while they may claim that they believe peo-
ple have a right to make a choice in whether to smoke or not, do
not believe that that choice should be made when you are 13 or 14.
And it is hard to believe $10 million a day would be spent in ad-
vertising if you didn't know who the only new customers were you
would suck in and that is children. But whatever way you go,
whether it is excise tax or advertising, both of those two methods
have the most profound effect on smokers when they are children.
And I think that that is why the country is so in support of these
issues. Because they don't think a 14
Chairman Rostenkowski. Taxing increases the price of the prod-
uct.
Ms. NoRTHUP. And the group that is most affected by the in-
crease of the product is children.
Chairman Rostenkowski. Senator, you have recommended
strong agricultural education programs for tobacco farmers to con-
vert to other crops. How soon would such programs be effective,
and would they have to be run by the State governments, because
the content of the programs would differ from place to place?
Ms. NoRTHtrp. That is right, they have to be uniquely designed
depending on what the obstacles to conversion for those farmers
are. We have already an extension service in place. I might point
out there is a terrific connective link between farmers and a gproup
that could serve them and that is the co-ops.
And if a co-op bid — I would put the program on a State-by-State
basis, maybe phasing it out over 10 years, because I know you
don't want to pay for welfare for farmers forever. But during that
605
time, allow the co-ops to bid to provide the education to the farm-
ers.
Chairman Rostenkowski. You talked about the 3 cents per pack
set aside.
Ms. NoRTHUP. Yes, I am.
Chairman Rostenkowski. Would you be kind enough to provide
the committee with some additional detail on that?
Ms. NoRTHUP. I would be happy to do that.
[Ms. North up was unable to provide the information.]
Dr. Douglas. Mr. Chairman, over here, your fellow Chicagoan,
Mr. Rostenkowski, you asked a question, which I think raised both
a moral as well as ultimately a constitutional issue of whether you
in effect discriminate against one industry as differentiated from
others.
It is a firm belief that many of us have who are health conscious
and those of us who have a moral conviction about this so-called
legal substance called tobacco, that we should quote former Sec-
retary Louis Sullivan over and over and over again, and you have
heard the same statement I have, that tobacco is the the only legal
substance, or I would replace substance with drug, that if used as
intended, maims and kills people.
That makes the tobacco industry different from other presumably
legal industries. Hopefully, tobacco will ultimately be outmoded
and become a relic of the past, but in the meantime, I think elected
representatives of the people have an issue to deal with of the
moral facet of the use of tobacco.
We know that it maims and kills people. Therefore it is looked
upon differently. When billboards are up in public places, I am not
suggesting that we should throw paint on them, but I am suggest-
ing that we should view them diflferentlv than advertisements for
milk or eggs. It is different, it kills, and I think that makes it in
the view of the public a product that should be looked at differently
from other legal products.
Chairman Rostenkowski. Doctor, you mentioned that you have
been a consultant to the World Health Organization. Are there any
policies other than higher taxes in place in other countries that
have been effective in preventing young people from taking up to-
bacco products?
Dr. Douglas. I want to emphasize today that raising excise
taxes is easily the single most effective measure for proteccting
children against tobacco addiction.
The taxes on tobacco in most other industrialized countries of the
world are far in excess of what they are here.
And in addition to that, in some of those countries there is also
a great deal of public health pressure in the schools and in other
places to get children not to start in the first place. But most im-
portantly, they are ahead of us in raising taxes on tobacco. Look
north of us. The country immediately north of us is setting prece-
dents that I beheve in many respects is making the United States
of America conscious of the fact that, yes, raising taxes on tobacco
does indeed prevent kids from starting in the first place, and it
does indeed also give a great deal of money for a health care deliv-
ery system which is far superior to our own.
606
Chairman Rostenkowski. Dr. Goldstein, you cite a figure of
3,000 deaths per year as a result of passive smoking. Other studies
have been unable to document such a link.
Could you familiarize us with the reliable evidence on this issue?
Dr. GoLDSTED^. Yes, that was from the EPA report that was re-
leased just very recently.
Chairman Rostenkowski. That is a pretty high number, isn't it?
Dr. Goldstein. Well, it is a reasonable number. The EPA report
classified secondhand smoke as a group of (known human) carcino-
gen and reports that secondhand smoke causes 3,000 lung cancer
deaths. Other reports estimate the number of passive smoke relat-
ed deaths as high as 53,000 taking into account deaths from heart
disease in addition to lung cancer.
Chairman Rostenkowski. Mr. Payne.
Mr. Payne. Thank you very much, Mr. Chairman. Representa-
tive Northup, let me — I am very interested in the agricultural issue
that you have raised because I have 5,000 tobacco farmers in my
congressional district.
^d let me explore just a minute a couple of the things that you
have brought up. First, from an agricultural perspective, you men-
tion that there are crops and you mention tomatoes specifically,
that the return was equally as good as tobacco.
What is the return per acre when you raise tomatoes?
Ms. Northup. It is over $4,000 per acre.
Mr. Payne. Over $4,000?
Ms. Northup. If there is a processor that is purchasing it. The
example that I looked at that was written up was I believe in
Owensboro, Kentucky. And these farmers were supplying a plant
in Owensboro.
There is also, I would give you another example, we have grapes
that are being harvested right now and they bring in I believe, if
I am remembering this right, double what tobacco, I believe it was
$6,500 per acre.
Now, of course, if we attract a winery into an area, as Virginia
has successfully done into an area of theirs, they make even more
than that per acre, and it serves far more farmers.
Mr. Payne. Well, generally speaking, the areas where we grow
tobacco are different from the areas where we grow grapes. Also,
let me say that in Virginia we have done a number of the things
that you mentioned.
For instance, in Danville, Va., we have Carnation there,
Contadina Foods are there, and they buy a lot of tomatoes. And yet
yesterday we had the testimony, the mayor of Danville, Va., very
concerned about the tax issue and what the impact would be cer-
tainly on that city and on the surrounding area.
There is no evidence that we have seen that because we have the
kind of infrastructure that you mention that we have been able to
convert from a tobacco product to a tomato product.
I think Mr. Jenkins, who was here, lives in that area and farms
in that area, mentioned that he thought the yield would be sub-
stantially less. Yesterday, too, we had Congressman Scotty Baesler
who was here who is from your State of Kentucky who is unique
in the Congress because he is not only a Congressman, but he also
is a tobacco farmer, family farmer. And he commented, too, on this
607
notion that we need to look about alternatives and stated that he
felt that in order for the areas which are now very dependent on
tobacco to be held harmless or returned whole, that what would
need to happen is that there would have to be almost a total relook
at rural areas, and how is it that you might go into a rural area
and do a substantial amount of investing in not only in agriculture,
but in many other kinds of things having to do with education,
having to do with infrastructure, with roads and so forth, having
to do with the rebuilding, I guess, of some rural areas.
And while we didn't talk about it, the expense of those, it seems
that there are certainly some major expenses involved. And I am
not saying all this — I applaud what you are looking into and this
notion of alternative crops, I think, is a good one if it works.
But I don't want the committee to be left with the idea that this
is an easy solution to a problem that a lot of people have in the
southeast in terms of how it is that they might be able to find
something else to do that even approaches the kind of profitability
and consequently the living that they would be able to make as
they do raising tobacco.
Ms. NORTHUP. Well, I would dispute a few of those things. First
of all, I would tell you that I think the more hopeless you make
it, the more you present the committee with only one choice.
You either lower the tobacco tax, or you put the farmers out of
business. The fact is your farmers and my farmers, and I know
that your market is already opened, they are in trouble, and we
know they are in trouble, and there is no decline in cigarette con-
sumption today or over the last couple years.
Even before the President, this President was elected, the to-
bacco companies were drastically reducing what they are purchas-
ing from your farmers. What this excise tax would do, by the to-
bacco companies' own estimates of how much the decline in smok-
ing would, 12 percent at $1 a pack, pales in comparison with what
they are doing every year now in their purchase intentions. And
they are up here, it would be hilarious to be up here and hear them
talking about the poor farmer, if it weren't so dire for the farmer,
they are the ones putting our farmers out of business.
Mr. Payne. Well, now, the people I am alluding to are not to-
bacco companies. These are the tobacco farmers that you just
heard, your own Congressman, or a Congressman from your State.
And it was just this year that we have passed a law talking
about domestic content and the fact that tobacco companies in the
future will be required by law to purchase 75 percent of their to-
bacco as domestic tobacco.
I think those things have gone a long way to dealing with some
of the issues that you cite as the reasons why we ought to do some
of these things. I am not here to say that what you are saying does
not make sense.
I am here to say that what we need to think about on this com-
mittee is that the solutions are not simple solutions and I am not
here to be a nay sayer, saying there is no way we can do anything
like this.
But I just don't want the committee and the record to be left in
such a way that we think this is an easy solution to what is a very
608
difficult situation that has been an income in my district and my
State for over 400 years, almost 400 years.
Thank you, very much.
Chairman Rostenkowski. Dr. Kaesemeyer, you mentioned the
famous Framingham heart study.
Dr. ICaesemeyer. Framingham, yes, sir.
Chairman Rostenkowski. Farmingham, is it?
Dr. Kaesemeyer. Framingham.
Chairman RosTEhfKOWSKl. Didn't that study include men only?
Dr. Kaesemeyer. Excuse me?
Chairman Rostenkowski. Didn't that study include men only?
Are there any studies of the relationship between smoking and car-
diovascular disease in women?
Dr. Kaesemeyer. That study included women also.
Chairman Rostenkowski. It did?
Dr. Kaesemeyer. Yes. And it showed in the form of the age-
related incidence of heart attack between men and women, and
that is on exhibit I.A.3. that I have submitted to you.
What it showed was that women manifest heart attack at a later
age than men. Women were included in that study.
[The following was subsequently received:]
Effects of cigarette smoking on fasting
triglyceride, total cholesterol, and
HDL-cholesterol in women
We examined the relationships of cigarette smoking with fasting triglycerides, total cholesterol,
and high-density lipoprotein cholesterol (HDL-C) levels among a group of 191 white women aged
20 10 40 years. The mean triglyceride level among current smokers was 100.0 mg/100 ml and
among nonsmokers was 68.4 mg/dl (p < 0.005). Mean total cholesterol values among current
smokers and nonsmokers were, respectively, 197.0 and 189.1 mg/dl (p < 0.1). Mean HDL-C
levc'S were 45.0 mg/dl among women who were smoking and 52.1 mg/dl among nonsmokers
lp< C.005). Simultaneous adjustments for the effects of age, weight, height, blood glucose,
resting pulse, and oral contraceptive use did not materially alter these relationships. A modest
portion of the effect of cigarette smoking on risk of coronary heart disease may be explained by
an adverse effect of cigarette smoking on blood lipids. (Am Heart J 105:417, 1983.)
Walter Willett, Charles H. Hennekens, William Castelli, Bernard Rosner,
Denis Evans, James Taylor, and Edward H. Kass Boston, Mass.
A 1 - to threefold incresise in risk of myocardial
infarction (MI) has generally been noted among
current cigarette smokers,' and in young women who
smoke this increase may be as high as tenfold.' The
mechanism by which cigarette smoking causes MI
remains obscure, but smoking acutely elevates blood
carbon monoxide to levels which produce cardiac
Ischemia in patients with pre-existing angina,^ and
smoking has been associated with an increased
nun ber of ventricular premature beats.* These
acute eflFects are compatible with the results of
epidemiologic studies, which generally show little
effect of past smoking or duration of smoking.*
However, autopsy studies among men have found an
increased prevalence of atherosclerosis among
smokers,'^ suggesting that smoking may accelerate
atherogenesis, possibly by altering other coronary
'isk factors.
'"m the Channing Laboratory. Departments of Medicine and Preventive
"d Social Medicine, Harvard Medical School and the Brigham and
*"mens Hospital; the Framingham Heart Study, National Heart. Lung,
'"<! Blood Institute, National Institutes of Health; and the Department of
'•Semiology. Harvard School of Public Health.
â– â– 'PPorted by a contract (HD 2832) from the National Institute of Child
"♦•aUh and Human Development, and Research Career Development
*«"d HL 0286 (Dr. Hennekens) from the National Heart, Lung, and
f'l"«l Institute.
^â– ^ei' .d for publication May 26, 1981; revision received Aug. 4. 1981;
'â– â– 'Pvd Aug. 10, 1981.
*l>rini requests: Walter Willett. M.D.. Charming Laboratory, 180 Long-
""•i Ave., Boston, MA 02115.
In our recent study of the relationships of oral
contraceptive (OC) use with plasma lipid fractions,'
the level of high-density lipoprotein cholesterol
(HDL-C) was lower in cigarette smokers than in
nonsmokers. This inverse relationship between
smoking and HDL-C has been previously noted in
men*^" but has been less well described in women.
Because HDL-C is inversely related to risk of MI,"
the finding of lower levels of HDL-C in cigarette
smokers suggests that the smoking-MI relationship
may be due, in part, to an effect on blood lipids. In
this report we examine in further detail the relation-
ship of smoking with fasting triglyceride, total cho-
lesterol, HDL-C, and total cholesterol to HDL-C
ratio (total-C/HDL-C) in a group of young women.
In addition, we examine the effect of potential
confounding variables on these observed relation-
ships including the possible interaction between
smoking and OC use.
METHODS
Population. A group of 287 white, predominantly work-
ing-class women, aged 20 to 40 years, who were premeno-
pausal, nonpregnant, and not taking antihypertensive
medication, was identified in a community-based, house-
to-house survey in E^t Boston, Massachusetts. Of these
women, 191 (67%) of those eligible agreed to attend a
study clinic after a 14-hour fast Further details are
reported elsewhere.'
Subject characteristics. Information concerning cur-
rent and past cigarette smoking history, OC use, age, and
stated weight and height was obtained by interview. Pulse
417
610
418 Willett et al.
Table I. Mean levels of fasting lipid fractions among current smokers and nonsmokers ( ± 1 SE)
Current smoking status
Lipid fraction
Crude
difference
; Other differences are not statistically significant.
Table II. Relationship of number of cigarettes smoked per day with mean levels of lipid fractions ( ± 1
Number of cigarettes per day
Adjusted
difference
Triglyceride (mg/100 mJ)
100.0(±6.0)
68.4(±3.3)
31.6*
49.5*
Tola! cholesterol (mg/100 ml)
197.0(±3.9)
189.1(±2.9)
7.9
7.9
HDL cholesterol (mg/100 ml)
45.0(±1.0)
52.1(±1.2)
-7.1*
-7.3*
Total cholesterol/HDL-C
4.62(±0.15)
3.80(±0.11)
0.82'
0.96«
Lipid fraction
(n = 23)
(n = 36)
(n =
39)
95.2(
tl0.0)
196.2(
t6.2)
45.2(
tl.3)
4.3(
tO.2)
Triglycerides (mg/100 ml)
ToUl cholesterol (mg/100 ml)
HDL cholesterol (mg/100 ml)
Total cholesterol/HDL-C
119.6(±14.9)
198.5(±8.5)
48.1(±2.8)
4.1(±0.3)
92.6(±7.7)
196.8(±6.5)
42.8(±1.7)
4.6(±0.3)
rate was measured after a 5-minute period of rest. In a
subsequent clinic visit, fasting blood specimens were
drawn by antecubital venipuncture with the women sit-
ting. The specimens were collected between 8 and 10 am in
tubes which contained sodiimi EDTA. The tubes were
inverted eight times and centrifuged, and plasma was
decanted. In addition, a fasting blood glucose was col-
lected in a tube containing fluoride-oxalate and measured
by the glucose oxidase method.
Lipid determinations. Lipid determinations were per-
formed in the laboratory of the Framingham Heart Study,
which participates in a quality control program through a
cooperative arrangement with the Lipid Research Clin-
ics.'^ The high-density lipoprotein fraction was separated
by precipitating the other lipoproteins with heparin-
manganese chloride. Cholesterol concentrations in the
total plasma and HDL fraction were measured by the
method of Abell-KendaU and triglycerides were meastired
by a modification of the Kessler-Lederer method.
Subject groups. Initially, to evaluate the crude rela-
tions between current smoking and each lipid fraction, we
divided the population into those who did and those who
did not smoke. The difference in mean values for lipid
fractions was calculated and compared by means of a t
test. In addition, women who did not currently smoke
were divided into those who had never smoked and those
who had smoked in the past, and lipid fractions in these
two groups were compared.
Multiple regression analyses. Multiple regression anal-
yses were performed to determine whether the relation-
ships between smoking and lipid levels were independent
of the effects of other variables. Separate analyses were
performed with each lipid fraction as a dependent variable
and current cigarette smoking (yes vs no), age, OC use,
weight, height, blood glucose, and resting pulse as pndi^^
tor variables. The coefficient for smoking thus represents
the difference in mean hpid levels between smokers and
nonsmokers adjusted for the effects of other variables.
Cross-product terms were added to the multivariate mod-
els to examine the interactive effects of specific vari-
ables.
Dose-response analysis. A possible dose-response
effect was examined by categorizing current smokers into
those smoking 1 to 14, 15 to 24, and more than 25
cigarettes per day and examining the lipid levels in each
category, imadjusted for the effects of other variables. In
addition, the nimiber of cigarettes smoked per day (as a
continuous variable) was added to the regression models
already containing the current smoking variable.
Because the distributions of Upids were slightly skewed
toward higher values, all computations were also per-
formed with the use of log, transformations. For simplicity
only untransformed results are presented, becatise all
relationships were very similar to those using transformed
values. ii^
RESULTS ^
Blood lipids. Women who were currently smokers^
exhibited HDL-C values which were, on average, 7.0^
mg% lower than those in nonsmoking women (Table^
I), a difference unlikely to result from chance^
(p < 0.005). Triglycerides, total cholesterol, and^
total-C/HDL-C were higher in smokers than in^
611
Cigarette smoking effects on blood lipids in women 419
^gble III. Mean lipid fractions among smokers and nonsmokers by oral contraceptive use ( ± 1 SE)
OC users
OC Nonusers
Lipid fraction
Smokers
Nonsmokers
Difference
Smokers
Nonsmokers
Difference
(mg/100 ml)
(n = 50)
(n = 39)
(n = 48)
(n = 54)
Triglyceride
Total cholesterol
HDL cholesterol
109.0(±6.5)
204.0(±5.5)
44.5(±1.4)
75.7(±5.2)
193.3(4 4.5)
51.5(±2.1)
33.3(±8.7)
10.7(±7.3)
-7.0(±2.4)
90.5(±10.2)
189.6(±5.6)
45.4(±1.5)
63.0(±4.2)
186.1(4 3.8)
52.4(±1.6)
27.5(4 10.6)
3.5(±6.7)
-7.0(4 2.2)
nonsmokers, although the diflFerences in total cho-
lesterol may well have been due to random variation
When the differences between smokers and non
smokers were adjusted for the eflFects of age, OC use
weight, height, resting pulse, and blood glucose
these differences were not materiaUy altered and
were in fact slightly larger.
The 93 women who were not currently smoking
were further subdivided into 31 past smokers and 62
who had never smoked. All lipid fractions were
similar (p > 0.3) for these two groups. For this
reason, past smokers and women who had never
smoked are considered together throughout further
analyses.
Degree of smoking. Among current cigarette smok-
ers, triglycerides, total cholesterol, and HDL-C were
slightly lower in heavy smokers than in light smok-
ers (Table 11). However, when the number of ciga-
rettes smoked daily was included in the multiple
regression model, the relation of this variable to
each Upid fraction could be explained by random
variation.
Oral contraceptives. As current OC use is a deter-
minant of lipid levels,' we examined the possibiUty
of an interaction between cigarette smoking and OC
use. The difference in mean lipid values between
smokers and nonsmokers was similar among users
and nonusers of oral contraceptives, indicating that
the effects of smoking and OC use are additive
(Table III). This consistency with an additive rela-
tionship was confirmed in a multivariate analysis by
adding a cross-product term (current smoking X OC
use) to the previously described multiple regression
equation which already included terms for OC use