The Ne w E n g l a nd Jo u r n a l o f Me d ic i ne
Editorials
E XERCISE TO R EDUCE
C ARDIOVASCULAR R ISK —
H OW M UCH I S E NOUGH ?
I NCREASING levels of physical activity are associ-
ated with a decrease in cardiovascular events. Con-
trolled clinical trials suggest that exercise has benefits
in persons with coronary artery disease and in those
with glucose intolerance.1 Exercise produces improve-
ments in mood, blood pressure, insulin sensitivity, and
plasma lipoprotein profiles, but the amount and in-
tensity of exercise required in order to attain these
benefits and the underlying mechanisms are poorly
understood. In this issue of the Journal, Kraus et al.2
report on changes in plasma lipoprotein levels and par-
ticle sizes in an eight-month, randomized trial involv-
ing different amounts and intensities of exercise among
overweight men and women with dyslipidemia. They
found that low amounts of exercise at moderate or
high intensity (the equivalent of walking or jogging
12 mi per week, respectively) are associated with po-
tentially beneficial changes in the plasma lipoprotein
profile. However, higher levels of high-intensity exer-
cise (equivalent to jogging 20 mi per week) resulted in
more pronounced changes in lipoproteins and were
required to produce increases in the high-density lipo-
protein (HDL) cholesterol level. The graded response
of the plasma lipoprotein levels to increasing amounts
of exercise may help to explain the progressive decrease
in cardiovascular risk associated with increasing levels
of exercise.1
As compared with nonexercising controls, all exer-
cise groups had potentially beneficial changes in plas-
ma lipoproteins: decreases in total and very-low-densi-
ty lipoprotein (VLDL) triglycerides, an increase in the
size of low-density lipoprotein (LDL) particles, and a
trend toward decreased numbers of LDL particles.2
In the groups with a low amount of exercise, these
effects were independent of the intensity of exercise
and were unrelated to improvements in physical fitness.
However, an increase in HDL cholesterol levels and
particle size and the largest effects on LDL were seen
only in the high-amount–high-intensity group. These
changes included a significant decrease in the num-
ber of LDL particles, implying a decrease in the LDL
apolipoprotein B concentration, since there is one apo-
lipoprotein B molecule per LDL particle. In view of
the prominent elevation of HDL levels in long-dis-
tance runners,3 the findings regarding HDL may ap-
pear surprising, but they are concordant with other
studies showing that the mean HDL response in ex-
1522 · N Engl J Med, Vol. 347, No. 19 · November 7, 2002 · www.nejm.org
ercise studies is small (1.2 mg per deciliter of HDL
cholesterol) and that HDL levels are only increased at
higher exercise levels (the equivalent of running 10 to
15 mi per week).3,4
The exercise-induced changes in lipoprotein levels
and particle sizes were very similar to those previously
shown in subjects who exercised and lost weight or
who lost weight by dieting.5 In some studies, lipopro-
tein changes have been correlated with loss of body
weight and decreases in subcutaneous abdominal fat,3,4
suggesting that lipoprotein changes may in part reflect
decreases in adiposity. In the study by Kraus et al.,
subjects were encouraged to maintain their base-line
body weight, which led to the novel observation that
potentially beneficial lipoprotein changes occurred
with only small decreases in body weight (a mean de-
crease of 1.5 kg in the high-amount–high-intensity
group).
Despite the plethora of observations of exercise-
induced changes in the lipoprotein profile, there is
limited understanding of the underlying mechanisms.
Exercise conditioning is associated with an increase
in lipoprotein lipase activity in adipose tissue and
muscle.4,6 Increased lipoprotein lipase activity lowers
VLDL and chylomicron triglyceride levels and enhanc-
es clearance of cholesterol-rich VLDL and chylomi-
cron remnants (Fig. 1). VLDL triglycerides are ex-
changed for cholesteryl esters in LDL and HDL, a
process mediated by cholesteryl ester transfer protein,7
and the triglyceride in HDL and LDL is then hydro-
lyzed by lipases, causing a decrease in the size of par-
ticles.8 The decrease in VLDL triglycerides results in
the availability of less triglyceride for exchange and is
probably a major mechanism underlying the increas-
es in the size of LDL particles and in HDL cholesterol
levels and particle size. Exercise and weight loss also
reduce the level of cholesteryl ester transfer protein,
perhaps because a fraction of this protein is made in
adipose tissue.9 Another important factor underlying
the changes in HDL is likely to be a decrease in hepatic
lipase activity. Hepatic lipase degrades HDL phospho-
lipids and triglycerides, producing smaller HDL par-
ticles that are rapidly catabolized.
The effects of exercise on HDL cholesterol levels
are most clearly seen in overweight persons with high
triglyceride levels and low HDL cholesterol levels at
base line.4 In contrast, lean subjects with isolated low
HDL cholesterol levels had no significant increase in
HDL cholesterol with exercise, even though their lipo-
protein lipase activity was increased. This suggests that
the changes in HDL and LDL cholesterol may be sec-
ondary to improvements in hypertriglyceridemia, in-
sulin resistance, and adiposity. Many of the improve-
ments in lipoprotein variables and insulin sensitivity
that are associated with habitual exercise are also seen
after a single session of exercise. This finding could
 ED ITOR IA LS

Triglyceride Lipase
HDL Small HDL Catabolism

CETP CE
VLDL
Lipase
LDL Small LDL

Increased VLDL Insulin

Glucose
secretion

Muscle
+ cells
Liver
Decreased apoB
degradation and
increased IRS1–
+
triglyceride PI3K
synthesis
-

+
Increased Increased mitochondrial
fatty acids oxidation

Exercise

Figure 1. Metabolic Abnormalities in Obese Persons and Persons with Insulin Resistance and the Potential Mechanism of Reversal
by Exercise.
In obese persons, increased delivery of fatty acids to myocytes leads to a defect in insulin signaling through insulin receptor sub-
strate 1 (IRS1) and phosphoinositide 3 kinase (PI3K) and a decrease in muscle glucose uptake by the glucose transporter GLUT-4.
The resistance to insulin action in muscle leads to more generalized insulin resistance and increased release of fatty acids from adi-
pose tissue. The liver synthesizes and secretes increased amounts of triglycerides and apolipoprotein B (apoB), in the form of very-
low-density lipoprotein (VLDL). Increased levels of triglycerides in VLDL are exchanged for cholesteryl esters (CE) in low-density
lipoprotein (LDL) and high-density lipoprotein (HDL). Subsequent lipolysis of HDL and LDL triglycerides results in decreased size of
particles. Exercise may reverse these abnormalities in part by diverting fatty acids in muscle toward mitochondrial oxidation. Other
mechanisms for changes in LDL and HDL include increases in lipoprotein lipase activity and decreases in hepatic lipase and cho-
lesteryl ester transfer protein (CETP).

indicate that short-term effects of exercise on insulin
signaling in muscle are a fundamental mechanism un-
derlying many of the observed changes in the lipopro-
tein profile.
Studies by Shulman and colleagues10 have demon-
strated that the acute delivery of fatty acids to muscle
results in insulin resistance. This effect is mediated by
a decrease in insulin signaling through insulin receptor
substrate 1 and phosphoinositide 3 kinase, leading to
decreased GLUT-4–mediated glucose transport into
muscle, and has been attributed to the accumulation
of fatty-acid metabolites in muscle (Fig. 1). It is tempt-
ing to speculate that exercise diverts the flow of fatty
acids toward mitochondrial oxidation, leading to im-
proved insulin signaling in muscle. This dynamic may
cause an overall improvement in insulin sensitivity,
leading to decreased release of fatty acids from adipose
tissue and an improvement in insulin signaling in the
liver. These changes may, in turn, lead to a decrease in
triglyceride synthesis and secretion in the liver and de-
creased hepatic secretion of apolipoprotein B.11
What is the evidence that exercise-induced lipopro-

N Engl J Med, Vol. 347, No. 19 · November 7, 2002 · www.nejm.org · 1523

 The Ne w E n g l a nd Jo u r n a l o f Me d ic i ne
tein changes are beneficial? There is substantial evi-
dence that VLDL and smaller remnant lipoproteins
(which are present in the VLDL or intermediate-den-
sity fractions) are atherogenic, and reductions in the
levels of these particles are likely to reduce atheroscle-
rosis.12 Small, dense LDL particles are somewhat more
susceptible to oxidation than larger LDL particles,
but it is questionable whether lipoprotein oxidation
in vitro predicts atherogenic responses in humans. The
size of LDL particles is inversely correlated with en-
dothelial reactivity, as measured in studies of blood
flow in the forearm,13 and endothelial reactivity ap-
pears to have predictive value for clinical events. In
population studies, the size of LDL particles has been
inversely correlated with the risk of myocardial infarc-
tion according to univariate analysis, but this effect was
not significant after adjustment for triglyceride and
HDL cholesterol levels.14
Thus, it is possible but unproved that there is a
benefit of changes in the size of LDL particles, and
such changes are likely to be less important than
changes in VLDL and LDL cholesterol and apolip-
oprotein B levels. The changes in HDL cholesterol
levels produced by exercise are small but are similar
to those that were associated with benefit in the Vet-
erans Affairs Cooperative Studies Program High-
Density Lipoprotein Cholesterol Intervention Trial.15
However, it is uncertain whether the primary medi-
ator of benefit is the change in HDL cholesterol. Al-
though exercise results in the accumulation of larger
HDL2 particles, HDL cholesterol measurements may
provide as much information about cardiovascular risk
as do measurements of HDL-subfraction responses.16
In summary, exercise is associated with a graded
response in a number of different lipoprotein variables.
The ensemble of changes is likely to be beneficial, even
though the role of each individual component is de-
batable. The study by Kraus et al. documents an effect
of exercise on lipoproteins with only minimal changes
in body weight and provides a ray of hope for those
who find it easier to exercise than to lose weight.
ALAN R. TALL, M.B., B.S.
Columbia University
New York, NY 10032
REFERENCES
1. Thompson PD. Additional steps for cardiovascular health. N Engl J
Med 2002;347:755-6.
2. Kraus WE, Houmard JA, Duscha BD, et al. Effects of the amount and in-
tensity of exercise on plasma lipoproteins. N Engl J Med 2002;347:1483-92.
3. Williams PT. High-density lipoprotein cholesterol and other risk factors
for coronary heart disease in female runners. N Engl J Med 1996;334:
1298-303.
4. Couillard C, Despres JP, Lamarche B, et al. Effects of endurance exer-
cise training on plasma HDL cholesterol levels depend on levels of triglyc-
erides: evidence from men of the Health, Risk Factors, Exercise Training
and Genetics (HERITAGE) Family Study. Arterioscler Thromb Vasc Biol
2001;21:1226-32.
1524 · N Engl J Med, Vol. 347, No. 19 · November 7, 2002 · www.nejm.org
5. Williams PT, Krauss RM, Vranizan KM, Wood PD. Changes in lipo-
protein subfractions during diet-induced and exercise-induced weight loss
in moderately overweight men. Circulation 1990;81:1293-304.
6. Thompson PD, Rader DJ. Does exercise increase HDL cholesterol in
those who need it the most? Arterioscler Thromb Vasc Biol 2001;21:1097-8.
7. Inazu A, Brown ML, Hesler CB, et al. Increased high-density lipopro-
tein levels caused by a common cholesteryl-ester transfer protein gene mu-
tation. N Engl J Med 1990;323:1234-8.
8. Deckelbaum RJ, Eisenberg S, Oschry Y, Butbul E, Sharon I, Olivecrona
T. Reversible modification of human plasma low density lipoproteins to-
ward triglyceride-rich precursors: a mechanism for losing excess cholesterol
esters. J Biol Chem 1982;257:6509-17.
9. Jiang XC, Moulin P, Quinet E, et al. Mammalian adipose tissue and
muscle are major sources of lipid transfer protein mRNA. J Biol Chem
1991;266:4631-9.
10. Shulman GI. Cellular mechanisms of insulin resistance. J Clin Invest
2000;106:171-6.
11. Ginsberg HN. Insulin resistance and cardiovascular disease. J Clin
Invest 2000;106:453-8.
12. Havel RJ. Remnant lipoproteins as therapeutic targets. Curr Opin Lip-
idol 2000;11:615-20.
13. Vakkilainen J, Makimattila S, Seppala-Lindroos A, et al. Endothelial
dysfunction in men with small LDL particles. Circulation 2000;102:716-21.
14. Kamigaki AS, Siscovick DS, Schwartz SM, et al. Low density lipopro-
tein particle size and risk of early-onset myocardial infarction in women.
Am J Epidemiol 2001;153:939-45.
15. Rubins HB, Robins SJ, Collins D, et al. Gemfibrozil for the secondary
prevention of coronary heart disease in men with low levels of high-density
lipoprotein cholesterol. N Engl J Med 1999;341:410-8.
16. Gaziano JM, Buring JE, Breslow JL, et al. Moderate alcohol intake,
increased levels of high-density lipoprotein and its subfractions, and de-
creased risk of myocardial infarction. N Engl J Med 1993;329:1829-34.
Copyright © 2002 Massachusetts Medical Society.
L EUKOTRIENE R ECEPTORS
AND A SPIRIN S ENSITIVITY
A BOUT 15 million people in the United States
have asthma. In a small percentage of them, vari-
ably estimated at between 3 percent and 10 percent,
acute, severe asthma accompanied by rhinorrhea and
sometimes associated with hives, flushing, or abdom-
inal pain develops after the ingestion of aspirin or non-
steroidal antiinflammatory drugs (NSAIDs). Patients
with this syndrome, known as aspirin-sensitive asthma,
often have severe rhinosinusitis and nasal polyposis.
It is important to recognize aspirin-sensitive asthma
clinically because severe episodes are life-threatening.
The common feature of drugs that provoke asthma at-
tacks in aspirin-intolerant persons is that they inhibit
cyclooxygenase-1 (COX-1); selective inhibition of cy-
clooxygenase-2 (COX-2) appears not to provoke such
a response.1 Cautious, incremental administration of
oral doses of aspirin can lead to a state in which per-
sons with aspirin-sensitive asthma can ingest aspirin
without untoward reactions, but daily administration
must be continued to maintain this state.
Persons with aspirin-sensitive asthma have increased
basal biosynthesis of the cysteinyl leukotrienes, as ev-
idenced by increased urinary excretion of leukotri-
ene E4 (LTE4). Leukotrienes are the enzymatic prod-