Content uploaded by Karine Spiegel
Author content
All content in this area was uploaded by Karine Spiegel
Content may be subject to copyright.
Brief Communication: Sleep Curtailment in Healthy Young Men Is
Associated with Decreased Leptin Levels, Elevated Ghrelin Levels, and
Increased Hunger and Appetite
Karine Spiegel, PhD; Esra Tasali, MD; Plamen Penev, MD, PhD; and Eve Van Cauter, PhD
Background: Total sleep deprivation in rodents and in humans
has been associated with hyperphagia. Over the past 40 years,
self-reported sleep duration in the United States has decreased by
almost 2 hours.
Objective: To determine whether partial sleep curtailment, an
increasingly prevalent behavior, alters appetite regulation.
Design: Randomized, 2-period, 2-condition crossover clinical
study.
Setting: Clinical Research Center, University of Chicago, Chi-
cago, Illinois.
Patients: 12 healthy men (mean age [±SD], 22 ± 2 years; mean
body mass index [±SD], 23.6 ± 2.0 kg/m
2
).
Measurements: Daytime profiles of plasma leptin and ghrelin
levels and subjective ratings of hunger and appetite.
Intervention: 2 days of sleep restriction and 2 days of sleep
extension under controlled conditions of caloric intake and phys-
ical activity.
Results: Sleep restriction was associated with average reductions
in the anorexigenic hormone leptin (decrease, 18%;
P
ⴝ 0.04),
elevations in the orexigenic factor ghrelin (increase, 28%;
P
<
0.04), and increased hunger (increase, 24%;
P
< 0.01) and appe-
tite (increase, 23%;
P
ⴝ 0.01), especially for calorie-dense foods
with high carbohydrate content (increase, 33% to 45%;
P
ⴝ
0.02).
Limitations: The study included only 12 young men and did not
measure energy expenditure.
Conclusions: Short sleep duration in young, healthy men is
associated with decreased leptin levels, increased ghrelin levels,
and increased hunger and appetite.
Ann Intern Med. 2004;141:846-850. www.annals.org
For author affiliations, see end of text.
See editorial comment on pp 885-886.
S
leep plays an important role in energy balance. In ro-
dents, food shortage or starvation results in decreased
sleep (1), and, conversely, total sleep deprivation leads to
marked hyperphagia (2). Leptin and ghrelin are peripheral
signals that contribute to the central regulation of food
intake. Leptin, a hormone released by the adipocytes, pro-
vides information about energy status to hypothalamic reg-
ulatory centers (3). In humans, circulating leptin levels rap-
idly decrease or increase in response to acute caloric
shortage or surplus, respectively (4). These changes in lep-
tin levels have been associated with reciprocal changes in
hunger (4). Ghrelin, a peptide produced predominantly by
the stomach, is also involved in energy balance regulation,
but, in contrast to the anorexigenic effects of leptin, ghrelin
stimulates appetite (5). It has been proposed that leptin
and ghrelin “represent the ‘yin–yang’ of one regulatory sys-
tem that has developed to inform the brain about the cur-
rent energy balance state” (6).
Over the past 40 years, sleep duration in the U.S.
population has decreased by 1 to 2 hours (7–10). The
proportion of young adults sleeping fewer than 7 hours per
night has more than doubled between 1960 and 2001–
2002 (from 15.6% to 37.1%) (7–10). The effect of sleep
curtailment on the control of appetite and food intake is
not known. Because of the well-documented associations
between sleep and food intake (1, 2), we sought to deter-
mine whether sleep duration influences the daytime pro-
files of leptin and ghrelin.
METHODS
Participants
Twelve healthy men (mean age [⫾SD], 22 ⫾ 2 years];
mean body mass index [⫾SD], 23.6 ⫾ 2.0 kg/m
2
) who
did not smoke or take any medications participated in the
study. All of the men were within 10% of ideal body
weight and had regular nocturnal time in bed of 7 to 9
hours. We excluded persons who had traveled across time
zones less than 4 weeks before the study.
Experimental Protocol
The Institutional Review Board of the University of
Chicago approved the protocol, and we obtained written
informed consent from all participants. During the week
preceding each study, we asked participants not to deviate
from a fixed time in bed (11:00 p.m. to 7:00 a.m.) by more
than 30 minutes. Naps were not allowed.
The men participated in 2 studies that were conducted
in a randomized order, were spaced at least 6 weeks apart,
and were performed in the Clinical Research Center at the
University of Chicago, Chicago, Illinois. Six of the 12 men
first performed the study with restricted time in bed, and
the remaining 6 men first performed the study with ex-
tended time in bed. Average weight did not change over
the time period separating the 2 study conditions (75.2 kg
in the sleep restriction condition vs. 75.4 kg in the sleep
extension condition; P ⬎ 0.2). We obtained blood samples
at 20-minute intervals from 8:00 a.m. to 9:00 p.m. after 2
consecutive nights of 10 hours in bed (10:00 p.m. to 8:00
Article
846 © 2004 American College of Physicians
a.m.; sleep extension) and after 2 consecutive nights of 4
hours in bed (1:00 a.m. to 5:00 a.m.; sleep restriction).
Sleep was recorded every night. For both extension and
restriction conditions, each overnight stay began at 7:00
p.m. with a standard hospital dinner, and the first over-
night stay ended after breakfast, which was served at 8:00
a.m. We instructed the participants not to deviate from
their usual eating habits between breakfast and dinner, but
caloric intake was not otherwise monitored. Participants
were readmitted in the early evening and, after receiving a
standard hospital dinner at 7:00 p.m., remained at bed
rest. At 8:00 a.m. after the second night, the participants’
caloric intake was kept constant to avoid meal-related fluc-
tuations of hunger and satiety and consisted of an intrave-
nous glucose infusion at a constant rate of 5 g/kg of body
weight every 24 hours. There was no other source of calo-
ries. Every hour from 9:00 a.m. to 9:00 p.m., the men
completed validated visual analogue scales (0 to 10 cm) for
hunger (11) and appetite for various food categories (12).
To assess hunger, we asked participants to mark their re-
sponse to the question “How hungry do you feel right
now?” on a 10-cm scale (with “not at all hungry” on the
left and “extremely hungry” on the right). To assess appe-
tite, we asked participants to mark their response to how
much they would enjoy eating foods from 7 different food
categories on a 10-cm scale (with “not at all” on the left
and “very much” on the right). They were asked to provide
a score based only on their appetite at the moment, with-
out concern for calories, fat, or a healthy diet. The food
categories were sweets (such as cake, candy, cookies, ice
cream, and pastry); salty foods (such as chips, salted nuts,
pickles, and olives); starchy foods (such as bread, pasta,
cereal, and potatoes); fruits and fruit juices; vegetables;
meat, poultry, fish, and eggs; and dairy products (such as
milk, cheese, and yogurt).
Assays
We measured serum leptin levels in all samples by
using a human leptin radioimmunoassay kit (Linco Re-
search, St. Charles, Missouri) with a sensitivity of 0.5
ng/mL and an intra-assay coefficient of variation of 8.3%.
In 9 of the 12 participants, we also measured total ghrelin
levels at hourly intervals by radioimmunoassay (Linco Re-
search) with a sensitivity of 0.5 ng/mL and an intra-assay
coefficient of variation of 7.9%.
Statistical Analysis
We performed paired comparisons by using the Wil-
coxon matched-pairs signed-rank test. We calculated cor-
relations by using the Spearman coefficient. The mean rel-
ative changes in leptin, ghrelin, hunger, and appetite
between extended sleep (the reference category) and re-
stricted sleep were calculated by using the ratios of the
corresponding individual data and then deriving the mean
across all individuals.
Role of the Funding Sources
This work was partially supported by grants from the
National Institutes of Health, the European Sleep Research
Society, the Belgian Fonds de la Recherche Scientifique
Me´dicale, the University of Chicago Diabetes Research
and Training Grant, and The University of Chicago Gen-
eral Clinical Research Center. The funding sources had no
role in the design, conduct, and reporting of the study or
in the decision to submit the manuscript for publication.
RESULTS
Leptin levels were stable across the daytime period un-
der both sleep conditions, which was consistent with the
fact that calories were exclusively delivered in the form of a
constant glucose infusion. Average total sleep time was 9
hours and 8 minutes when the men spent 10 hours in bed
and 3 hours and 53 minutes when the men spent 4 hours
in bed (P ⬍ 0.01). When spending 4 hours in bed, the
participants had mean leptin levels that were 18% lower
(2.1 ng/mL vs. 2.6 ng/mL; P ⫽ 0.04) (Figure 1, part A)
and mean ghrelin levels that were 28% higher (3.3 ng/mL
vs. 2.6 ng/mL; P ⫽ 0.04) (Figure 1, part B) than when the
participants spent 10 hours in bed. The ratio of the con-
centrations of orexigenic ghrelin to anorexigenic leptin in-
creased by 71% (CI, 7% to 135%) with 4 hours in bed
compared with 10 hours in bed. Sleep restriction relative to
sleep extension was associated with a 24% increase in hun-
ger ratings on the 10-cm visual analogue scale (P ⬍ 0.01)
and a 23% increase in appetite ratings for all food catego-
ries combined (P ⫽ 0.01) (Figure 1, parts C and D, and
Table 1). The increase in appetite tended to be greatest for
calorie-dense foods with high carbohydrate content
(sweets, salty foods, and starchy foods: increase, 33% to
45%; P ⫽ 0.06) (Table 1). The increase in appetite for
fruits and vegetables was less consistent and of lesser mag-
nitude (increase, 17% to 21%) (Table 1). Appetite for
Context
Studies in animals and humans suggest that sleep duration
is an important regulator of metabolism.
Contribution
In this study, 12 young, healthy, normal-weight men ex-
hibited reductions in the satiety hormone leptin, increases
in the hunger hormone ghrelin, and increases in hunger
after 2 nights of only 4 hours of sleep compared with af-
ter 2 nights of 10 hours of sleep.
Implications
Inadequate sleep seems to influence the hormones that
regulate satiety and hunger in a way that could promote
excess eating.
–The Editors
ArticleLeptin, Ghrelin, Hunger, and Appetite during Sleep Loss
www.annals.org 7 December 2004 Annals of Internal Medicine Volume 141 • Number 11 847
protein-rich nutrients (meat, poultry, fish, eggs, and dairy
foods) was not significantly affected by sleep duration (Ta-
ble 1). When we considered the changes in ghrelin and
leptin in an integrated fashion by calculating the ghrelin-to-
leptin ratio, the increase in hunger was proportional to the
increase in ghrelin-to-leptin ratio (r ⫽ 0.87) (Figure 2). Al-
most 70% of the variance in increased hunger could be ac-
counted for by the increase in the ghrelin-to-leptin ratio.
DISCUSSION
We observed that sleep duration may affect the circu-
lating levels of neuroendocrine factors that regulate hunger
and appetite. Two days with 4 hours of time in bed each
night were associated with an 18% decrease in the levels of
the anorexigenic hormone leptin. By comparison, 3 days of
underfeeding by approximately 900 calories per day in
healthy lean volunteers has been reported to result in a
decrease of leptin levels averaging 22% (4). Sleep curtail-
ment was also associated with an almost 28% increase in
daytime levels of the orexigenic factor ghrelin. The recip-
rocal changes in leptin and ghrelin that we observed in
response to sleep restriction were associated with a 24%
increase in hunger and a 23% increase in appetite. Appetite
for calorie-dense nutrients with high carbohydrate content,
including sweets, salty snacks, and starchy foods, increased
by 33% to 45%. In contrast, appetite for fruits, vegetables,
and high-protein nutrients was less affected. The increase
in hunger during sleep restriction was strongly correlated
with the increase in the ghrelin-to-leptin ratio.
Our study involved intensive physiologic monitoring
under laboratory conditions in a relatively small group of
normal young men and will need to be replicated in a
larger sample. In addition, because age and sex may affect
neuroendocrine regulation of appetite (3, 5), our findings
may not readily apply to women and older adults. Recent
findings from a population study involving 1030 persons
are, however, in complete agreement with our observations
(13). In that study, restricted duration of sleep was associ-
ated with reduced leptin levels, increased ghrelin levels, and
elevated body mass index (13).
The alterations in appetite regulation that we observed
after sleep restriction may reflect a normal adaptation to
the increased caloric need associated with extended wake-
fulness. Our experimental protocol was designed to keep
energy intake and activity levels as constant as possible, and
the extra hours of wakefulness during sleep restriction were
spent lying in bed or sitting in a comfortable armchair.
Although several studies have indicated that differences in
energy expenditure between sleeping in bed compared with
quiet wakefulness are very small, if at all detectable (14,
15), it is not known whether sleep deprivation increases the
energy requirements of maintaining wakefulness. In studies
in rats in which the disk over water method achieved total
sleep deprivation, researchers observed a marked increase in
energy expenditure that resulted in overall weight loss de-
spite increased food intake (2). However, an increased en-
ergy demand is an intrinsic feature of this method of sleep
deprivation, which involves forced locomotion and re-
peated water immersions. Careful studies of energy balance
in sleep-restricted humans under comfortable sedentary
conditions will be necessary to determine whether in-
creased hunger will result in excessive food intake and
weight gain.
The causes of decreased leptin levels and increased
Figure 1. Effect of sleep duration on daytime leptin levels,
ghrelin levels, hunger, and appetite.
A. Mean (⫾SE) daytime (9:00 a.m. to 9:00 p.m.) profiles of leptin after
2 days with 4 hours in bed or 2 days with 10 hours in bed. Mean leptin
levels were 18% lower when sleep was restricted. B. Mean (⫾SE) day-
time (9:00 a.m. to 9:00 p.m.) profiles of ghrelin from 9 of the 12
participants after 2 days with 4 hours in bed or 2 days with 10 hours in
bed. Mean ghrelin levels were 28% higher in the afternoon and early
evening (12:00 noon to 9:00 p.m.) when sleep was restricted. C and D.
Ratings of hunger (C) (0- to 10-cm visual analogue scale) and overall
appetite (D) (0- to 70-cm visual analogue scale) after 2 days with 4 hours
in bed or 2 days with 10 hours in bed. When sleep was restricted, ratings
of hunger and overall appetite increased by 24% and 23%, respectively.
Article Leptin, Ghrelin, Hunger, and Appetite during Sleep Loss
848 7 December 2004 Annals of Internal Medicine Volume 141 • Number 11 www.annals.org
ghrelin levels in a state of sleep loss remain to be deter-
mined. We have previously shown that 6 days of sleep
restriction in healthy adults resulted in an increase in car-
diac sympathovagal balance (16). Some, but not all, studies
have indicated that sleep loss is associated with increased
sympathetic nervous system outflow (17). Because leptin
release is inhibited by sympathetic nervous system activity
(18), decreased leptin levels, in the presence of a sleep debt,
may result from an inhibitory effect of increased sympa-
thetic outflow. Increased cardiac sympathovagal balance
could also reflect decreased vagal activity, which could ex-
plain increased ghrelin levels. Several studies have shown
that the vagus has a negative influence on ghrelin (5, 19,
20).
Sleep loss due to voluntary curtailment of time in bed
has become a hallmark of modern society. Self-reported
sleep duration in the United States has decreased by 1 to 2
hours during the second half of the 20th century (7–10).
The proportion of young adults sleeping 8 to 8.9 hours per
night has decreased from 40.8% in 1960 to 23.5% in
2001–2002 (8–10). During the same time period, the in-
cidence of obesity has nearly doubled (21). Three epidemi-
ologic studies have found a relationship between higher
body mass index and shorter sleep duration (13, 22, 23).
This epidemiologic evidence, together with our experimen-
tal findings that sleep restriction affects leptin levels, ghre-
lin levels, hunger, and appetite, suggests that additional
studies should examine the possible role of chronic sleep
curtailment as a previously unrecognized risk factor for
obesity.
From University of Chicago, Chicago, Illinois, and Universite´ Libre de
Bruxelles, Brussels, Belgium.
Acknowledgment: The authors thank Paul Rue for performing the lep-
tin assays, Miho Yoshida for performing the ghrelin assays, and Kristen
Knutson for providing statistical assistance. The authors also thank the
volunteers for their participation in this demanding study and the nurs-
ing staff of the University of Chicago General Clinical Research Center
for their expert assistance.
Grant Support: In part by grants from the National Institutes of Health
(DK-41814, AG-11412, and HL-72694), from the European Sleep Re-
search Society, from the Belgian Fonds de la Recherche Scientifique
Me´dicale (FRSM-3.4583.02), from the University of Chicago Diabetes
Research and Training Grant (NIH DK-20595), and from The Univer-
sity of Chicago General Clinical Research Center (NIH MO1-RR-
00055).
Potential Financial Conflicts of Interest: None disclosed.
Requests for Single Reprints: Eve Van Cauter, PhD, Department of
Medicine, MC 1027, University of Chicago, 5841 South Maryland
Avenue, Chicago, IL 60637.
Figure 2. Association between the change in hunger ratings
and the change in ghrelin-to-leptin ratio during the 12:00 noon
to 9:00 p.m. time period when sleep is restricted as compared
with extended.
The changes in hunger ratings and in ghrelin-to-leptin ratio were calcu-
lated as the values obtained after 4 hours in bed minus the values ob-
tained after 10 hours in bed. For each of these variables, negative values
were obtained when the variable measured after 10 hours in bed was
higher than when measured after 4 hours in bed. The Spearman coeffi-
cient was 0.87 and the P value was 0.01.
Table. Average Ratings of Appetite after 2 Days of Sleep Restriction or Sleep Extension
Food Category* Ratings for
10hinBed
(
n
ⴝ 12)
Ratings for
4hinBed
(
n
ⴝ 12)
P
Value Change,
%
Sweets (cake, candy, cookies, ice cream, and pastry) 5.4 6.6 0.03 33
Salty food (chips, salted nuts, pickles, and olives) 5.0 6.7 0.02 45
Starchy food (bread, pasta, cereal, and potatoes) 5.9 7.4 0.03 33
Fruits and fruit juices 6.4 7.2 0.07 17
Vegetables 5.6 6.6 0.02 21
Meat, poultry, fish, and eggs 5.9 6.9 0.11 21
Dairy (milk, cheese, and yogurt) 5.5 6.4 ⬎0.2 19
Overall appetite† 39.7 47.7 0.01 23
* Each category is rated on a 0- to 10-cm visual analogue scale.
† Rated on a 0- to 70-cm visual analogue scale.
ArticleLeptin, Ghrelin, Hunger, and Appetite during Sleep Loss
www.annals.org 7 December 2004 Annals of Internal Medicine Volume 141 • Number 11 849
Current author addresses and author contributions are available at www
.annals.org.
References
1. Danguir J, Nicolaidis S. Dependence of sleep on nutrients’ availability. Physiol
Behav. 1979;22:735-40. [PMID: 225752]
2. Everson CA, Bergmann BM, Rechtschaffen A. Sleep deprivation in the rat:
III. Total sleep deprivation. Sleep. 1989;12:13-21. [PMID: 2928622]
3. Ahima RS, Saper CB, Flier JS, Elmquist JK. Leptin regulation of neuroen-
docrine systems. Front Neuroendocrinol. 2000;21:263-307. [PMID: 10882542]
4. Chin-Chance C, Polonsky KS, Schoeller DA. Twenty-four-hour leptin levels
respond to cumulative short-term energy imbalance and predict subsequent in-
take. J Clin Endocrinol Metab. 2000;85:2685-91. [PMID: 10946866]
5. van der Lely AJ, Tschop M, Heiman ML, Ghigo E. Biological, physiological,
pathophysiological, and pharmacological aspects of ghrelin. Endocr Rev. 2004;
25:426-57. [PMID: 15180951]
6. Muccioli G, Tschop M, Papotti M, Deghenghi R, Heiman M, Ghigo E.
Neuroendocrine and peripheral activities of ghrelin: implications in metabolism
and obesity. Eur J Pharmacol. 2002;440:235-54. [PMID: 12007539]
7. Kripke DF, Simons RN, Garfinkel L, Hammond EC. Short and long sleep
and sleeping pills. Is increased mortality associated? Arch Gen Psychiatry. 1979;
36:103-16. [PMID: 760693]
8. National Sleep Foundation. 2000 “Sleep in America” Poll. Washington, DC:
National Sleep Foundation; 2000.
9. National Sleep Foundation. 2001 “Sleep in America” Poll. Washington, DC:
National Sleep Foundation; 2001.
10. National Sleep Foundation. 2002 “Sleep in America” Poll. Washington,
DC: National Sleep Foundation; 2001.
11. Stubbs RJ, Hughes DA, Johnstone AM, Rowley E, Reid C, Elia M, et al.
The use of visual analogue scales to assess motivation to eat in human subjects: a
review of their reliability and validity with an evaluation of new hand-held com-
puterized systems for temporal tracking of appetite ratings. Br J Nutr. 2000;84:
405-15. [PMID: 11103211]
12. Fernstrom MH, Krowinski RL, Kupfer DJ. Appetite and food preference in
depression: effects of imipramine treatment. Biol Psychiatry. 1987;22:529-39.
[PMID: 3580430]
13. Taheri S, Lin L, Austin D, Young T, Mignot E. Short sleep duration is
associated with reduced leptin, elevated ghrelin, and increased body mass index
(BMI) [Abstract]. Sleep. 2004;27(Abstr Suppl):A146-7.
14. Seale JL, Conway JM. Relationship between overnight energy expenditure
and BMR measured in a room-sized calorimeter. Eur J Clin Nutr. 1999;53:107-
11. [PMID: 10099943]
15. Zhang K, Sun M, Werner P, Kovera AJ, Albu J, Pi-Sunyer FX, et al.
Sleeping metabolic rate in relation to body mass index and body composition. Int
J Obes Relat Metab Disord. 2002;26:376-83. [PMID: 11896493]
16. Spiegel K, Leproult R, Van Cauter E. Impact of sleep debt on metabolic and
endocrine function. Lancet. 1999;354:1435-9. [PMID: 10543671]
17. Irwin M, Thompson J, Miller C, Gillin JC, Ziegler M. Effects of sleep and
sleep deprivation on catecholamine and interleukin-2 levels in humans: clinical
implications. J Clin Endocrinol Metab. 1999;84:1979-85. [PMID: 10372697]
18. Rayner DV, Trayhurn P. Regulation of leptin production: sympathetic ner-
vous system interactions. J Mol Med. 2001;79:8-20. [PMID: 11327106]
19. Sugino T, Yamaura J, Yamagishi M, Kurose Y, Kojima M, Kangawa K, et
al. Involvement of cholinergic neurons in the regulation of the ghrelin secretory
response to feeding in sheep. Biochem Biophys Res Commun. 2003;304:308-12.
[PMID: 12711315]
20. Heath RB, Jones R, Frayn KN, Robertson MD. Vagal stimulation exagger-
ates the inhibitory ghrelin response to oral fat in humans. J Endocrinol. 2004;
180:273-81. [PMID: 14765979]
21. Flegal KM, Carroll MD, Ogden CL, Johnson CL. Prevalence and trends in
obesity among US adults, 1999-2000. JAMA. 2002;288:1723-7. [PMID:
12365955]
22. Kripke DF, Garfinkel L, Wingard DL, Klauber MR, Marler MR. Mortality
associated with sleep duration and insomnia. Arch Gen Psychiatry. 2002;59:
131-6. [PMID: 11825133]
23. Vioque J, Torres A, Quiles J. Time spent watching television, sleep duration
and obesity in adults living in Valencia, Spain. Int J Obes Relat Metab Disord.
2000;24:1683-8. [PMID: 11126224]
Article Leptin, Ghrelin, Hunger, and Appetite during Sleep Loss
850 7 December 2004 Annals of Internal Medicine Volume 141 • Number 11 www.annals.org
Current Author Addresses: Dr. Spiegel: Laboratoire de Physiologie,
Centre d’Etude des Rythmes Biologiques (CERB), Universite´ Libre de
Bruxelles, Campus Hoˆpital Erasme–CPI 604, 808, Route de Lennik,
B-1070 Brussels, Belgium.
Drs. Tasali, Penev, and Van Cauter: Department of Medicine, MC
1027, University of Chicago, 5841 South Maryland Avenue, Chicago, IL
60637.
Author Contributions: Conception and design: K. Spiegel, E. Van Cau-
ter.
Analysis and interpretation of the data: K. Spiegel, E. Tasali, P. Penev, E.
Van Cauter.
Drafting of the article: K. Spiegel, E. Tasali, E. Van Cauter.
Critical revision of the article for important intellectual content: E.
Tasali, P. Penev, E. Van Cauter.
Final approval of the article: K. Spiegel, E. Tasali, P. Penev, E. Van
Cauter.
Provision of study materials or patients: E. Van Cauter.
Statistical expertise: E. Van Cauter.
Obtaining of funding: K. Spiegel, E. Van Cauter.
Administrative, technical, or logistic support: E. Van Cauter.
Collection and assembly of data: K. Spiegel, E. Tasali, E. Van Cauter.
www.annals.org 7 December 2004 Annals of Internal Medicine Volume 141 • Number 11 W-157
