Letters to the Editor
Issues raised by psyllium meta-analysis
Dear Sir:
Anderson et al (1) recently combined 8 studies sponsored by the
Proctor and Gamble company, including 3 unpublished projects
from 1994, in a meta-analysis. They excluded research regarding
cholesterol lowering by psyllium in hypercholesterolemic adults if it
did not meet narrow criteria regarding the dietary lead-in period, the
dose of psyllium used, and the length of treatment. Studies in which
psyllium was provided in a cereal were excluded. Reports that appear
to otherwise meet their criteria but that had different industry spon-
sors were not included (2, 3). Furthermore, this report did not clari-
fy the quality of the clinical trials included in the analysis (4).
The inclusion of unpublished data in meta-analyses is contro-
versial. Authorities in the field believe that, at a minimum, all stud-
ies should undergo the same rigorous methodologic evaluation and
that results should be presented with and without the unpublished
material (5). This issue may be particularly important when report-
ing pooled data from research originating from a single industry
sponsor, which raises the question of whether there are additional
unpublished studies, not included, that failed to show a cholesterol-
lowering effect of psyllium. The possibility of publication bias is
an important concern in an area in which investigation is dominat-
ed by industry sponsorship (6). The inclusion of a funnel plot might
have also alerted the reader to the potential for unpublished nega-
tive studies (7).
The accompanying editorial by Jenkins et al (8) raises an addi-
tional concern. In considering the difference in evaluating thera-
peutic foods as opposed to drugs, they contend that “a health claim
must be related to a specific product based on the testing of that
product or accompanying specific line of products. Competing
companies wishing to make a specific product claim for a similar
product would have to give evidence of efficacy of that product.”
This raises an untenable specter of endless repeated clinical trials
sponsored by each company wishing to bring a product to market
that is similar to those already marketed. Surely, the better strategy
would be to demand standardization of food products in the per-
formance of clinical trials so that there can be some generalizabil-
ity of research findings. It should be possible, for example, to
undertake research on psyllium, which is fairly generic and would
allow for possible minor differences, rather than demand that
research be specific to any individual company’s psyllium product.
Are there any scientific data showing that psyllium products differ
significantly by manufacturer?
Guido MA van Rosendaal
Eldon A Shaffer
Alun L Edwards
Lloyd R Sutherland
Am J Clin Nutr 2001;73:653–64. Printed in USA. © 2001 American Society for Clinical Nutrition
Departments of Medicine and Community Health Sciences
The University of Calgary Faculty of Medicine
Health Sciences Center
3330 Hospital Drive NW
Calgary, Alberta
Canada
E-mail: gvrosend@ucalgary.ca
REFERENCES
1. Anderson JW, Allgood LD, Lawrence A, et al. Cholesterol-lowering
effects of psyllium intake adjunctive to diet therapy in men and
women with hypercholesterolemia: meta-analysis of 8 controlled
Downloaded from www.ajcn.org by on May 29, 2008
trials. Am J Clin Nutr 2000;71:472–9.
2. Davidson MH, Maki KC, Kong JC, et al. Long-term effects of con-
suming foods containing psyllium seed husk on serum lipids in sub-
jects with hypercholesterolemia. Am J Clin Nutr 1998;67:367–76.
3. MacMahon M, Carless J. Ispaghula husk in the treatment of hyper-
cholesterolaemia: a double-blind controlled study. J Cardiovasc
Risk 1998;5:167–72.
4. Juni P, Witschi A, Bloch R, Egger M. The hazards of scoring the qual-
ity of clinical trials for meta-analysis. JAMA 1999;282:1954–60.
5. Cook DJ, Guyatt GH, Ryan G, et al. Should unpublished data be
included in meta-analyses? JAMA 1993;269:2749–53.
6. Brown L, Rosner B, Willett WC, Sacks FM. Cholesterol-lowering
effects of dietary fiber: a meta-analysis. Am J Clin Nutr 1999;
69:30–42.
7. Yusuf S, Flather M. Magnesium in acute myocardial infarction.
BMJ 1995;310:751–2.
8. Jenkins DJA, Kendall CWC, Vuksan V. Viscous fibers, health
claims, and stategies to reduce cardiovascular disease risk. Am J
Clin Nutr 2000;71:401–2.
Reply to GMA van Rosendaal et al
Dear Sir:
Health claims for foods have value if they inform the public
of the exact nature of the proposed benefit, are accurate, and
provide encouragement for the producer or manufacturer to
produce food with added health benefits. In the United States,
as a result of the Nutrition Labeling and Education Act of
1991, an attempt was made to address this issue by the accep-
tance of 12 generic classes of health claims. A company can
therefore make a claim provided that it satisfies the criteria
related to the claim.
653
654 LETTERS TO THE EDITOR
Generic health claims may be excellent for getting a public health
message across, such as “eat more fruit and vegetables because it
may reduce the risk of. . .” but product-specific health claims may
be required if industry is to be encouraged to aim for maximal effi-
cacy of products that may form part of an effective cholesterol-low-
ering portfolio put together with dietary ingredients.
Product-specific claims may be particularly important in situ-
ations where active ingredients are processed in ways that reduce
effectiveness to enhance palatability. In these situations, consid-
erable manufacturing and research creativity may be required to
maximize both palatability and efficacy of the product simultane-
ously. These research data should be used to establish the valid-
ity of the health claim. The public is then protected by having evi-
dence that a specific product works. Taste and price they will
judge for themselves. The company will be protected from com-
petition from untested products for which the claim can only be
generic. A very clear difference must therefore be established
between claims made for tested compared with untested products.
For example, viscous fibers may be hydrolyzed to avoid a
gummy mouth feel, resulting in the production of more appealing
products. Nevertheless, viscosity may be a determining factor
because other physiologic functions, including the ability to flat-
ten postprandial glycemia, are greatly reduced as viscosity is lost
(1, 2) and viscous fibers such as gum acacia do not lower serum
cholesterol (3). On the other hand, in the case of guar gum, some
cholesterol reduction is retained if hydrolysis is incomplete.
Soy may be combined with other proteins such as gluten for tex-
ture in meat analogues or the soy protein component may be
washed with a solvent during extraction to remove isoflavones.
Again, the amino acid profile of soy (4) and its isoflavone content
(5) may be important to maximize the cholesterol-lowering proper-
ties of soy.
When health claim status is given to plant sterols, the effec-
tiveness of the product in lowering cholesterol is also likely to
be dependent on the sterol mix and its processing. There are dif-
ferences related to the sterol and the sterol ester, and possibly to
the medium (eg, the fatty acid profile of the margarine) into
which the sterols are incorporated, not to mention possible dif-
ferences between different mixes of phytosterols (eg, sitosterol
and campesterol) (6). Each plant sterol–enriched product is
therefore likely to be different in terms of palatability and effec-
tiveness. Therefore, for functional foods with demonstrable
therapeutic value, eg, the ability to lower cholesterol or blood
pressure, it is important to have evidence of the degree of effec-
tiveness of the product in question.
At present, Japan, by FOSHU (Foods for Specific Health Use)
regulations, has taken this approach; Europe is still considering
the approach to take; and Canada has recognized the potential
value of both generic and product-specific claims. Regrettably,
because of older regulations, it is uncertain what path Canada
can take in establishing product-specific claims. However, in a
recent meeting of industry, Health Canada representatives, and
academia, it was unanimously agreed that both generic and prod-
uct-specific types of health claims may play key roles in pro-
moting public health (7). An important question remains as to
how to stimulate industry to produce the products and how to
provide the incentives necessary for long-term development of
effective, palatable functional foods.
David JA Jenkins
Cyril WC Kendall
Vladimir Vuksan
G Harvey Anderson
Department of Nutritional Sciences
Faculty of Medicine
University of Toronto
Toronto, Ontario M5S 3E2
Canada
E-mail: cyril.kendall@utoronto.ca
REFERENCES
1. Jenkins DJA, Wolever TMS, Leeds AR, et al. Dietary fibres, fibre
analogues, and glucose tolerance: importance of viscosity. Br Med
J 1978;1:1392–4.
2. Wood PJ, Braaten JT, Scott FW, Riedel KD, Wolynetz MS, Collins
MW. Effect of dose and modification of viscous properties of oat
gum on plasma glucose and insulin following an oral glucose load.
Br J Nutr 1994;72:731–43.
3. Jensen CD, Spiller GA, Gates JE, Miller AF, Whittam JH. The effect
of acacia gum and a water-soluble dietary fiber mixture on blood
lipids in humans. J Am Coll Nutr 1993;12:147–54.
4. Carroll KK. Review of clinical studies on cholesterol-lowering
Downloaded from www.ajcn.org by on May 29, 2008
response to soy protein. J Am Diet Assoc 1991;91:820–7.
5. Crouse JR III, Morgan T, Terry JG, Ellis J, Vitolins M, Burke GL. A
randomized trial comparing the effect of casein with that of soy
protein containing varying amounts of isoflavones on plasma con-
centrations of lipids and lipoproteins. Arch Intern Med 1999;159:
2070–6.
6. Weststrate JA, Meijer GW. Plant sterol-enriched margarines and
reduction of plasma total- and LDL-cholesterol concentrations in
normocholesterolaemic and mildly hypercholesterolaemic subjects.
Eur J Clin Nutr 1998;52:334–43.
7. Anderson GH, ed. An evaluation of Health Canada’s Consultation
Document Standards of Evidence for Evaluating Foods with Health
Claims: a proposed framework. Toronto: University of Toronto,
24–25, 2000.
Ross Conference had too narrow a
representation to speak about matters
of the world
Dear Sir:
The 17th Ross Research Conference on Medical Issues, “Physi-
ologically Active Food Components: Their Role in Optimizing
Health and Aging,” was a supplement to the June issue of the Journal
(1) and thus was provided to all subscribers. Such a venue should
make it fair game for an evaluative critique from the readership.
As a scientist beginning to work my way out of the exclu-
sive paradigm of undernutrition toward the coming epidemic
of chronic disease in the low-income countries of Central
America, I found the words of the conference chair in his sum-
mary provocative: “Current perceptions of functional foods
and components held by scientists and consumers in different
parts of the world were presented and discussed.” Were they?
I searched through the supplement to revisit the basis for this
claim of a global outreach for this rather exclusive gathering.
 LETTERS TO THE EDITOR
With a world defined in terms of diversity of its people, only
36% of the conference participants were women. On the basis of
birth origin, only 0.9% of the participants present were born in
developing countries. Edmonton, Canada, and Brussels, Bel-
gium, vied for the most northerly latitude of representation in
the conference, whereas the southern limit was defined by Tuc-
son, AZ, and San Antonio, TX. The percentage of the world’s
population represented by the combined populations of Bel-
gium, Germany, Canada, and the United States, the nations rep-
resented by the participants, is < 7%. So, the chair’s claim edged
into hyperbole.
Several hypotheses about this lack of geographic representa-
tion occurred to me. Is San Diego really that much further from
Bangkok, Rio de Janeiro, or Mexico City than it is from Brus-
sels or Stuttgart? Perhaps there is no interest in or experience
with functional foods outside of the Western world, although the
International Life Sciences Institute sponsored and published
results from a functional food seminar held in Southeast Asia
(2). Perhaps there is not the same degree of research expertise
and contribution from developing and tropical countries that
there is from the northern temperate zone. If this is the explana-
tion, it constitutes a serious wake-up call for those of us work-
ing in Latin America, Asia, and Africa to invest research
resources to earn our place on the platform of upcoming meet-
ings on this growing topic. We represent, after all, the majority
of the world’s potential consumers of functional foods. The
topic (and such products) impinges on our health, given that
commercial globalization has already brought outlets for so-
called physiologically active food components to Guatemala
and many other countries of Latin America. So even if
researchers into this topic are yet to be found in the academic
institutions of developing countries, at least the roster of the dis-
cussants could have gone beyond Arizona, Texas, and 3 sites
within the Washington, DC, Beltway.
A few other observations are worthy of mention. Despite its
title, relatively little evidence or experience was shared on aging
save for the associations of vitamin E and immune function (3)
and some passing references to Alzheimer disease (4). When
reading the supplement, one should focus on the words of John
Fernstrom, eg, “I am concerned at this moment more about
safety than efficacy,” because he alone cried out to orient the
priorities to the logical axis in any paradigm for interventions
for human health amid a wilderness of enthusiastic advocates.
The lessons I derived from receiving my Ross Research Con-
ference proceedings as an AJCN supplement were profound but
not all of them were of an academic nature.
Noel W Solomons
Center for Studies of Sensory Impairment, Aging and Metabolism
Guatemala City
Guatemala
REFERENCES
1. Harper AE, ed. Physiologically active food components: their role
in optimizing health and aging. Am J Clin Nutr 2000;71(suppl):
1647S–743S.
2. Proceedings of the 1st International Conference on East-West Per-
spectives on Functional Foods. Singapore, September 26–29, 1995.
Nutr Rev 1996;54(suppl):S1–202.
655
3. Meydani M. Effect of functional food ingredients: vitamin E mod-
ulation of cardiovascular diseases and immune status in the
elderly. Am J Clin Nutr 2000;71(suppl):1665S–8S, discussion
1674S–55S.
4. Fernstrom JD. Can nutrient supplements modify brain function? Am
J Clin Nutr 2000;71(suppl):1669S–73S.
Metabolic response to weight loss
Dear Sir:
Several years ago Leibel et al (1) reported data suggesting that
altered body weight produces changes in energy expenditure that
favor a return to original body weight. Additional data from this
group suggest that the compensatory changes in energy expendi-
ture might be related to changes in thyroid or catecholamine sta-
tus, or both (2). In this recent report, Rosenbaum et al (2) men-
tioned that “previous studies did not achieve the degree of weight
Downloaded from www.ajcn.org by on May 29, 2008
stability and control of nutrient intake of this study and did not
examine the same subjects during dynamic weight change as
well as during static weight maintenance at altered body
weights.” In fact, we reported the metabolic responses of sub-
jects during dynamic and energy balance phases of weight loss
under tightly controlled metabolic ward conditions (3, 4). Our
findings provide a potentially important contrast in results.
We measured resting metabolic rate (RMR), thyroid hormones,
and plasma and urinary catecholamines in 24 overweight post-
menopausal women (3). The metabolic variables were first assessed
in the static phase of energy balance in the overweight state after
10 d in the General Clinical Research Center (GCRC). Subjects
were then provided an energy-restricted diet containing 3347 kJ/d
(800 kcal/d) while they were outpatients of the GCRC until they
reached a normal body weight. They returned to the GCRC while
still in the dynamic phase of weight loss and after 10 d underwent
metabolic reassessment. After an additional 10 d in the GCRC, dur-
ing which time energy balance was restored, a final assessment was
made. Weight loss averaged 17% (12.7 kg; range: 74.0–61.3 kg) and
mean body mass index (BMI; in kg/m2) fell from 28 to 23. Through-
out the study, subjects received all meals from the GCRC so that we
could control the macronutrient composition of the diet and provide
a fixed content of sodium (174 mmol/d) and potassium (115
mmol/d) to avoid confounding effects on catecholamines.
Our findings during the dynamic phase of energy restriction
are similar to those of Rosenbaum at el (2), showing significant
reductions in RMR and triiodothyronine (T3) and an elevation in
reverse T3 (rT3). Rosenbaum et al also found that urinary norep-
inephrine fell during energy restriction, whereas we observed no
significant changes in 24-h urinary or fasting plasma concentra-
tions of epinephrine or norepinephrine. During the static phase
of weight-loss maintenance, Rosenbaum et al found that most of
the metabolic responses to energy restriction persisted, includ-
ing reduced RMR, reduced urinary norepinephrine, reduced T3,
and elevated rT3. By contrast, in our subjects, during stabiliza-
tion in the normal-weight state, all of the measured variables
returned to levels that were statistically no different from those
in the overweight state, including RMR adjusted for fat-free
mass (FFM) (5) and fat mass, T3, and rT3 (3, 4) (Figure 1).
656 LETTERS TO THE EDITOR
FIGURE 1. Mean (± SEM) resting metabolic rate (RMR) adjusted
for fat-free mass (FFM) and fat mass (FM) and the ratio of triiodothyro-
nine (T3) to reverse T3 (rT3) in 24 women. Values were assessed in the
stable overweight state, during the ongoing dynamic phase of weight
loss after subjects had reached a normal body weight, and in the stable
normal-weight state (3–5). *Significantly different from stable over-
weight and normal-weight states, P < 0.05.
More recently, we studied 32 overweight premenopausal
women under weight-stable conditions in the GCRC before and
after weight loss and found that RMR decreased after normaliza-
tion of body weight (6). Again, however, the new RMR was appro-
priate for the reduced FFM and fat mass. Urinary norepinephrine
and dopamine also decreased after weight loss, but the changes
were not significantly different after adjustment for changes in
FFM (R Weinsier et al, unpublished observations, 2000).
It is of more than academic interest to point out the similarities
and differences between the results of our studies and those of
Rosenbaum et al because both were conducted under tightly con-
trolled metabolic ward conditions with use of identical energy
restrictions (3347 kJ/d). Arguably, the outcomes should have been
similar. Both studies suggest that during negative energy balance,
RMR is reduced disproportionately to the decreases in FFM and
fat mass and that thyroid hormones change in association with
(and may possibly explain) the changes in energy requirements
(5). However, our data suggest that on adaptation to a weight-
reduced steady state, RMR and thermogenic hormones are appro-
priate for the subjects’ new body composition and do not predis-
pose the subjects to regain the lost weight.
Differences in the 2 studies might have contributed to the con-
flicting outcomes. The subjects studied by Rosenbaum et al were
severely obese, with an average BMI of 48, whereas our sub-
jects were overweight, with a BMI of 28. This weight difference
could account for the subjects’ different metabolic responses.
However, we believe that more plausible explanations exist. First,
Rosenbaum et al’s more obese subjects had an absolute RMR
value 50% higher than that in our subjects. Thus, the fixed
3347-kJ/d energy-restricted diet would have produced an energy
deficit almost twice that in our subjects. We confirmed in a sub-
set of our subjects that 10 d of stabilization in the weight-reduced
state was sufficient to produce a metabolic steady state (4). How-
ever, the considerably greater energy deficit in Rosenbaum et al’s
subjects could have necessitated a period longer than the 14 d
they used to establish a steady state. Second, different statistical
approaches were used to adjust RMR for changes in body com-
position after weight loss. We adjusted RMR for FFM and fat
mass by using analysis of covariance (5, 7). Rosenbaum et al’s
findings in the earlier report of reduced RMR after weight loss (1)
appear to have been based on the ratio of RMR to FFM, which
may sometimes lead to spurious results (8, 9). When they
adjusted RMR for changes in FFM and fat mass by using regres-
sion analysis, RMR values after weight loss were apparently not
significantly different from baseline values (1).
In summary, both studies indicate that during the dynamic
phase of energy restriction, thyroid hormones change in con-
junction with reductions in RMR, placing subjects in an energy
Downloaded from www.ajcn.org by on May 29, 2008
conservation mode. The findings of Rosenbaum et al suggest that
on return to energy balance conditions, weight loss is associated
with changes in thermogenic hormones which might, by virtue
of effects on energy expenditure, favor a return to usual body
weight. By contrast, our findings suggest that the energy-con-
serving metabolic changes persist only during the period of
energy restriction. Once energy balance is restored, metabolic
variables appear to normalize and become appropriate for the
new reduced body mass and do not explain the weight-regain
tendency of these subjects (7).
Roland Weinsier
Gary Hunter
University of Alabama at Birmingham
Birmingham, AL 35294
E-mail: weinsier@shrp.uab.edu
Yves Schutz
University of Lausanne
Lausanne
Switzerland
REFERENCES
1. Leibel RL, Rosenbaum M, Hirsch J. Changes in energy expenditure
resulting from altered body weight. N Engl J Med 1995;332:621–8.
2. Rosenbaum M, Hirsch J, Murphy E, Leibel RL. Effects of changes
in body weight on carbohydrate metabolism, catecholamine excre-
tion, and thyroid function. Am J Clin Nutr 2000;71:1421–32.
3. Weinsier RL, James LD, Darnell BE, Dustan HP, Birch R, Hunter
GR. Obesity-related hypertension: evaluation of the separate effects
of energy restriction and weight reduction on hemodynamic and
neuroendocrine status. Am J Med 1991;90:460–8.
4. Nelson KM, Weinsier RL, James LD, Darnell BE, Hunter GR, Long
CL. Effect of weight reduction on resting energy expenditure, sub-
strate utilization, and the thermic effect of food in moderately obese
women. Am J Clin Nutr 1992;55:924–33.
 LETTERS TO THE EDITOR
5. Weinsier RL, Nagy TR, Hunter GR, Darnell BE, Hensrud DD,
Weiss HL. Do adaptive changes in metabolic rate favor weight
regain in weight-reduced individuals? An examination of the set-
point theory. Am J Clin Nutr 2000;72:1088–94.
6. Weinsier RL, Hunter GR, Zuckerman PA, et al. Energy expenditure
and free-living physical activity in black and white women: compar-
ison before and after weight loss. Am J Clin Nutr 2000;71: 1138–46.
7. Weinsier RL, Nelson KM, Hensrud DD, Darnell BE, Hunter GR,
Schutz Y. Metabolic predictors of obesity: contribution of resting
energy expenditure, thermic effect of food, and fuel utilization to
four-year weight gain of post-obese and never-obese women. J Clin
Invest 1995;95:980–5.
8. Goran MI, Allison DB, Poehlman ET. Issues relating to normaliza-
tion of body fat content in men and women. Int J Obes Relat Metab
Disord 1995;19:638–43.
9. Allison DB, Paultre F, Goran MI, Poehlman ET, Heymsfield SB.
Statistical considerations regarding the use of ratios to adjust data.
Int J Obes Relat Metab Disord 1995;19:644–52.
Reply to R Weinsier et al
Dear Sir:
Weinsier et al raise important issues regarding our studies of the
effects of weight loss on systems of energy homeostasis, including
carbohydrate metabolism, catecholamine excretion, and thyroid
function (1). The nature and magnitude of the changes in energy
metabolism that accompany weight reduction are critical issues
from both physiologic and therapeutic perspectives. In their letter,
Weinsier et al describe points of similarity and distinction between
our respective studies of the effects on energy homeostasis of
maintenance of a reduced body weight. We found that both the
process of weight loss and the maintenance of a reduced body
weight are associated with significant declines in total 24-h energy
expenditure (TEE), resting energy expenditure (REE), and
non–resting energy expenditure (NREE) beyond those expected
from the changes in metabolic mass (1–3). In our recent article (1),
we reported that both weight loss and the maintenance of a
reduced body weight are associated with significant declines in
urinary norepinephrine and dopamine excretion and circulating
concentrations of triiodothyronine (T3). In earlier articles, we
described decreases in REE in some subjects under these circum-
stances (2, 3). Weinsier et al (4, 5) noted similar decreases in REE,
circulating T3 concentrations, and urinary catecholamine excretion
during weight loss, but found that these decreases did not persist
during sustained maintenance of a reduced body weight.
The differences in our results may have been due to differences
in our subject populations and study designs. In our previous stud-
ies (2, 3), we found that maintenance of a reduced body weight
(≥ 10% below initial body weight) is associated with a significant
decline in REE adjusted for fat-free mass (FFM) in obese pre-
menopausal subjects but not in subjects who have never been
obese. Weinsier et al point out that the obese subjects we studied
were significantly fatter than the obese subjects they studied: the
mean (± SEM) fat mass in our subjects was 67 ± 3 kg compared
with 30–31 ± 1 kg in Weinsier et al’s subjects (4, 5). Our obese
subjects had significantly (10%) higher REE (adjusted for FFM)
657
at their usual body weights than did our never-obese subjects. We
concluded that this relative increase in REE was due to increased
cardiorespiratory work in the more obese subjects. After weight
loss, the REE (adjusted for FFM) of our reduced-obese subjects
was significantly lower than it had been in the same subjects
before weight loss, but was similar to that of our never-obese sub-
jects at their usual body weight. Thus, the adjusted REE of our
reduced-obese subjects was not significantly lower than that of the
never-obese subjects. The persistent decline in REE in our obese
subjects during weight maintenance at a reduced body weight may
reflect their higher REEs at usual body weight compared with
Weinsier et al’s “leaner” obese subjects.
The subjects in our inpatient studies were intentionally
restricted to an amount of physical activity designed to maintain
a degree of physical fitness equal to that on admission to the
study (1–3). In contrast, Weinsier et al’s subjects were not
restricted with regard to physical activity (4, 5). The weight loss
in Weinsier et al’s subjects was apparently due to both a hypoen-
ergetic diet and physical activity [which increased by 33% in the
weight-reduced subjects (4)], whereas the weight loss in our sub-
jects was intentionally achieved solely through a reduction in
energy intake. Several studies showed that the addition of exer-
Downloaded from www.ajcn.org by on May 29, 2008
cise to a weight-loss regimen will significantly blunt the decline
in REE that occurs during and after weight loss (6–12). Thus, the
lack of association of maintenance of a reduced body weight with
lowered REE in Weinsier et al’s subjects may also reflect effects
of increased physical activity during their outpatient weight loss.
We showed previously that maintenance of a reduced body
weight is associated with a significant decline in sympathetic
nervous system tone as measured by effects on heart rate vari-
ability of sequential pharmacologic blockade of the sympathetic
and parasympathetic branches of the autonomic nervous system
(13). Schwartz et al (14) reported that weight loss induced by
diet alone results in a 17% decline in urinary norepinephrine
excretion (similar to the values in our studies), but that the addi-
tion of exercise to the weight-loss regimen abolishes this
decline. Although we are aware of no studies that examined the
effects of exercise training on the thyroid axis during mainte-
nance of a reduced body weight, the observation that exercise
increases sympathetic nervous system output (as reflected in
blunting of the weight-loss-associated decline in urinary norep-
inephrine excretion) suggests that the decline in T3 might also
be mitigated with exercise through sympathetic nervous sys-
tem–mediated effects on the thyroid axis (1).
The observations by Weinsier et al (4, 5) indicate that the
decreases in catecholamine excretion and in circulating concentra-
tions of thyroid hormones that accompany weight loss by diet
alone may be ameliorated by increased physical activity during
weight loss or maintenance of a reduced body weight. However,
despite the prevention of these endocrine adaptations to body
weight reduction, the weight-reduced subjects studied by Weinsier
et al still needed to increase their physical activity by 33% to
maintain a reduced body weight without decreasing energy intake
(4). Thus, Weinsier et al’s subjects apparently experienced the
same decreases in total energy expenditure experienced by our
subjects and compensated for this decline by their increase in
physical activity. Weinsier et al’s observation that TEE does not
change significantly after weight loss in weight-reduced subjects
who significantly increase their physical activity agrees with our
observation that TEE declines significantly in weight-reduced sub-
jects who do not change their physical activity from baseline after
658 LETTERS TO THE EDITOR
losing weight. Furthermore, Weinsier et al's report of an increase
in time spent in physical activity after weight loss, without a cor-
responding increase in NREE, agrees with our finding (2, 3) that
the energy cost of NREE is significantly decreased after weight
loss and that NREE is the component of energy expenditure most
affected during maintenance of a reduced body weight.
A small, persistent excess of energy intake relative to expen-
diture will, over time, result in substantial weight gain. To avoid
regain of lost weight, reduced-obese subjects must, as in our
studies (1–3), significantly decrease their energy consumption
or, as in Weinsier et al’s studies (4), significantly increase their
physical activity.
Michael Rosenbaum
Rudolph L Leibel
Columbia Presbyterian Medical Center
Columbia University College of Physicians and Surgeons
Division of Molecular Genetics
Russ Berrie Medical Pavilion, 6th Floor
1150 St Nicholas Avenue
New York, NY 10032
E-mail: mr475@columbia.edu
REFERENCES
1. Rosenbaum M, Hirsch J, Murphy E, Leibel RL. Effects of changes
in body weight on carbohydrate metabolism, catecholamine excre-
tion, and thyroid function. Am J Clin Nutr 2000;71:1421–32.
2. Rosenbaum M, Ravussin E, Matthews D, et al. A comparative study
of different means of assessing long-term energy expenditure in
humans. Am J Physiol 1996;270:R496–504.
3. Leibel RL, Rosenbaum M, Hirsch J. Changes in energy expenditure
resulting from altered body weight. N Engl J Med 1995;332:621–8.
4. Weinsier RL, Hunter GR, Zuckerman PA, et al. Energy expenditure
and free-living physical activity in black and white women: compar-
ison before and after weight loss. Am J Clin Nutr 2000;71:1138–46.
5. Weinsier RL, James LD, Darnell BE, Dustan HP, Birch R, Hunter
GR. Obesity-related hypertension: evaluation of the separate effects
of energy restrictions and weight reduction on hemodynamic and
neuroendocrine status. Am J Med 1991;90:460–8.
6. Melby C, Scholl C, Edwards G, Bullough R. Effect of acute resis-
tance exercise on postexericse energy expenditure and resting meta-
bolic rate. J Appl Physiol 1993;75:1847–53.
7. Bryner R, Ullrich I, Sauers J, et al. Effects of resistance vs. aero-
bic training combined with an 800 calorie liquid diet on lean body
mass and resting metabolic rate. J Am Coll Nutr 1999;18:115–21.
8. Wadden T, Vogt R, Andersen R, et al. Exercise in the treatment of
obesity: effects of four interventions on body composition, resting
energy expenditure, appetite, and mood. J Consult Clin Psychol 1997;
65:269–77.
9. Thompson J, Manore M, Thomas J. Effects of diet and diet-plus-
exercise programs on resting metabolic rate: a meta-analysis. Int J
Sports Nutr 1996;6:41–61.
10. Ryan A, Pratley R, Elabi D, Goldberg A. Resistive training increases
fat-free mass and maintains RMR despite weight loss in post-
menopausal women. J Appl Physiol 1995;79:818–23.
11. Shinkai S, Watanabe S, Kurokawa Y, Torii J, Asai H, Shepherd R.
Effects of 12 weeks of aerobic exercise plus dietary restriction on body
composition, resting energy expenditure and aerobic fitness in mildly
obese middle-aged women. Eur J Appl Physiol 1994; 68:258–65.
12. Dale BV, Saris W, Hoor F. Weight maintenance and resting meta-
bolic rate 18–40 months after a diet/exercise treatment. Int J Obes
1989;14:347–59.
13. Aronne L, Mackintosh R, Rosenbaum M, Leibel RL, Hirsch J.
Autonomic nervous system activity in weight gain and weight loss.
Am J Physiol 1995;38:R222–5.
14. Schwartz R, Jaeger L, Veith R, Lakshminarayan S. The effect of diet
or exercise on plasma norepinephrine kinetics in moderately obese
young men. Int J Obes 1990;14:1–11.
Fecal folate
Dear Sir:
After watching the movie Alien, I’m not sure I believe the
old dictum that “the only individuals who get to speak in first
person plural are the queen, a pregnant woman, and someone
with parasites.” In response to Boddie et al (1), I think the pos-
sibility should be considered that the denizens of a person’s
gut may influence how much folate is available for the jejunal
Downloaded from www.ajcn.org by on May 29, 2008
brush border pteroyl-poly[hyphen]-glutamate hydrolase (-
Glu-X carboxypeptidase) to hydrolyze. Generically speaking,
this is not a novel idea; vitamin B-12 deficiency secondary to
Diphyllobothrium latum infestation is well known.
I find Boddie et al’s data depicting lower urinary excretion of
labeled folate in women with neural tube defect–affected pregnan-
cies than in control women to be suggestive but not convincing evi-
dence of impaired absorption of polyglutamyl folate. Perhaps in
future experiments, Boddie et al could fractionate or centrifuge the
stool of their participants to see whether labeled folate that was not
taken up (as evidenced by urinary excretion) simply passed through
or was taken up by microorganisms.
Kate Hendricks
Texas Department of Health
Infectious Disease Epidemiology and Surveillance
1100 West 49th Street
Austin, TX 78756
E-mail: kate.hendricks@tdh.state.tx.us
REFERENCE
1. Boddie AM, Dedlow ER, Nackashi JA, et al. Folate absorption in
women with a history of neural tube defect–affected pregnancy. Am
J Clin Nutr 2000;72:154–8.
Reply to K Hendricks
Dear Sir:
Hendricks raises the question of whether in our recent study
of folate absorption the women who had given birth to infants
with a neural tube defect (NTD) may have malabsorbed folate as
 LETTERS TO THE EDITOR
a result of a parasitic infestation (of terrestrial origin, one would
suspect). She also incorrectly interprets our data as indicating a
selective malabsorption of polyglutamyl folates.
In fact, we reported reduced urinary excretion of labeled folates
derived from oral doses of both mono- and polyglutamyl folate,
which we interpreted as being the result of less efficient intestinal
transmural transport. We recognize that differences in in vivo reten-
tion or extent of catabolism may also have existed between the
women with NTD-affected pregnancies and the control women.
The protocol used in our study involved saturating the subjects with
folic acid to enhance excretion of the newly absorbed folate from
the test doses. It is highly probable that this protocol was a specific
means of testing for differences in the extent of absorption because
the saturation protocol would minimize any differences in postab-
sorptive processing of labeled folates.
Whether the women with NTD-affected pregnancies had a
parasitic infestation during our study cannot be determined;
however, such an infestation is unlikely because all women were
in good health and exhibited no evidence of gastrointestinal
problems. Thus, it is unlikely that intestinal parasites or
pathogens, whether bacterial, protozoan, or helminthic, were
present in our case subjects in amounts that would have signifi-
cantly impaired folate absorption.
Hendricks’s suggestion that fecal samples be fractionated and
analyzed is interesting to consider but would have been imprac-
tical in our study. Any unabsorbed, labeled folate would be sub-
ject to cellular uptake, metabolism, and potential catabolism by
colonic microorganisms. In addition, labeled colonic folates
would be diluted extensively with unlabeled folates produced by
the colonic microflora. All of these factors would reduce the fea-
sibility and complicate the interpretation of fecal folate analysis
in this protocol. In summary, we thank Hendricks for her com-
ments and recognize the need to investigate folate absorption and
metabolism more thoroughly in women with a history of NTD-
affected pregnancies.
Jesse F Gregory III
Lynn B Bailey
University of Florida
Institute of Food and Agricultural Sciences
Food Science and Human Nutrition Department
PO Box 110370
Gainesville, FL 32611-0370
E-mail: jfgy@ufl.edu
Calcium and vitamin D intakes and blood
pressure
Dear Sir:
Jorde and Bønaa (1), in reporting a negative association
between estimated calcium intake and blood pressure, com-
mented on the lack of a discernible effect of vitamin D intake on
blood pressure despite previous reports suggesting that increased
vitamin D status leads to a reduction in blood pressure. In dis-
659
cussing the shortcomings of the study, they mentioned that
although vitamin D status is determined by exposure to sunlight
as well as by diet, this determinant of vitamin D status “was not
evaluated.” There are, however, 2 obvious markers of exposure to
sunlight in a country such as Norway, as in the United Kingdom
and other European countries, that could properly be used as sur-
rogates for such exposure and that may be available.
First, physical activity was examined as a determinant of
blood pressure. If the records allow identification of the dura-
tion of outdoor as opposed to indoor activity, then that data
could be used in analysis (2). Second, the capacity of sunlight
to induce vitamin D synthesis in the skin varies markedly with
season of the year in northern Europe, as can the markers of
bone turnover (3). Indeed, review of the literature shows that
reduction in blood pressure in the summer in such countries has
been used as evidence to support the suggestion that increased
vitamin D status is associated with a reduction in blood pres-
sure (4). If, therefore, the dates on which subjects had their
blood pressure measured are available, then the authors should
be able to examine their data for seasonal variation in blood
pressure. The use of one, or ideally both, of these factors in
appropriate multifactorial analyses is required before conclu-
Downloaded from www.ajcn.org by on May 29, 2008
sions can be drawn about the relative roles of calcium intake
and vitamin D status in the determination of blood pressure in
healthy Norwegian subjects.
BJ Boucher
The Royal London Hospital
Whitechapel
London E1 1BB
United Kingdom
E-mail: bboucher@doctors.org
REFERENCES
1. Jorde R, Bønaa KH. Calcium from dairy products, vitamin D intake,
and blood pressure: the Tromsø study. Am J Clin Nutr 2000;71:
1530–5.
2. Baynes KCR, Boucher BJ, Feskens EJM, Kromhout D. Vitamin D,
glucose tolerance and insulinaemia in elderly men. Diabetologia
1997;40:344–7.
3. Woigte HW, Scheidt-Nave C, Kissling C, et al. Seasonal variation of
biochemical indexes of bone turnover: results of a population-based
study. J Clin Endocrinol Metab 1998;83:68–75.
4. Boucher BJ. Inadequate vitamin D status: does it contribute to
the disorders comprising syndrome ‘X’? Br J Nutr 1998;79:
315–27.
Reply to BJ Boucher
Dear Sir:
I am grateful for the appropriate comments from Boucher and
can supply the following information. In the Tromsø study, the
subjects were examined from September 1994 until June 1995.
However, the subjects who were examined in June 1995 were not
660 LETTERS TO THE EDITOR
comparable with the rest of the cohort because they did not
respond to the initial invitation. Thus, we do not have data from
the summer months. A look at the data from the rest of the year
showed that blood pressure values were the same during the
autumn, winter, and spring.
Unfortunately, we do not have data on outdoor compared with
indoor activity. Accordingly, we are not able to reassess the
blood pressure data as suggested by Boucher. In retrospect, I
regret that the above information was not included in the origi-
nal article (1).
Rolf Jorde
Department of Internal Medicine
University Hospital of Tromsø
9038 Tromsø
Norway
E-mail: medrj@rito.no
REFERENCE
1. Jorde R, Bønaa KH. Calcium from dairy products, vitamin D intake,
and blood pressure: the Tromsø study. Am J Clin Nutr 2000;
71:1530–5.
Dairy consumption and bone health
Dear Sir:
In a recent review article that examined the evidence of a link
between dairy foods and bone health, Weinsier and Krumdieck
(1) conclude that “the body of scientific evidence appears inade-
quate to support a recommendation for daily intake of dairy
foods to promote bone health in the general US population.”
They base their conclusions on a division of studies into those
that showed a favorable effect of dairy food intake on bone
health and those that did not. They prioritized studies according
to the strength of the evidence. Category A studies (the
strongest) were randomized controlled trials or longitudinal
cohorts with ≥ 3000 participants who were followed for > 5 y.
However, these 2 types of studies are not equivalent. Observa-
tional studies can accurately assess the outcome measures of
bone mineral density or fracture, but their ability to assess
dietary intakes is weak. Large size does not overcome that weak-
ness; it merely adds a spurious sense of accuracy. The number of
epidemiologic studies that did not show a significant relation is
not surprising considering the weak ability to determine the
independent variable.
Observational studies of the effect of folate on neural tube
defects showed a pattern nearly identical to that found for cal-
cium. It was the randomized controlled trials that confirmed
the importance of increasing folate intake during the repro-
ductive years.
If only randomized controlled trials were assigned to category
A, then studies showing a favorable relation between dairy food
intake and bone health would number 5 and span age groups rang-
ing from 11 y through menopause. Studies showing an unfavor-
able effect of dairy food intake on bone health would have no cat-
egory A studies.
Thus, using the same evidence as that used by the authors, we
conclude that adequate dairy consumption is supportive of good
bone health. One of us recently reviewed this same literature and
came to just that conclusion (2).
Connie M Weaver
Robert P Heaney
Purdue University
Department of Foods and Nutrition
1264 Stone Hall
West Lafayette, IN 47907-1264
REFERENCES
1. Weinsier RL, Krumdieck CL. Dairy foods and bone health: exami-
nation of the evidence. Am J Clin Nutr 2000;72:681–9.
2. Heaney RP. Calcium, dairy, products and osteoporosis. J Am Coll
Nutr 2000;19:83S–99S.
Downloaded from www.ajcn.org by on May 29, 2008
Reply to CM Weaver and RP Heaney
Dear Sir:
We appreciate the comments of Weaver and Heaney, whom we
respect as being internationally recognized experts in the field of
calcium metabolism and bone disease. They raise a valid concern
that randomized controlled trials (RCTs), especially double-blind
trials, are needed to establish causal relations in clinical investi-
gations and that even large longitudinal cohort studies may not be
equivalent for this purpose. As they indicate, there are both
strengths and weaknesses in longitudinal observational studies,
notably, the ability of such studies to detect hard endpoint out-
comes such as bone fractures and their limited ability to accu-
rately assess dietary intake. However, there are also shortcomings
in the few RCTs on the effect of dairy foods on bone status.
In our review (1) we classified 2 RCTs (2, 3) in the favorable-
effect category because each study showed less bone loss in the
dairy-supplemented group. On the other hand, at the end of the
intervention neither trial resulted in greater bone mass in the
dairy-supplemented than in the nonsupplemented group. Four of
the 5 RCTs classified in the favorable-effect category were not
blinded and did not have a placebo control. Hence, confounding
variables were not always removed. This is evident in one RCT
of adolescent girls in which the dairy-supplemented group had a
50% greater energy intake than did the control group and in
which energy intake correlated significantly with greater bone
mineral content (3). A further potential shortcoming of the avail-
able RCTs was summarized in Heaney’s (4) recent review article
in which he states that “all controlled manipulations of calcium
intake produce a bone remodeling transient, generally expressing
itself during the first year of treatment.” The average length of
the 5 RCTs assigned to the favorable-effect category was 1.5 y,
with a range of 14 wk to 3 y; none included a baseline period of
 LETTERS TO THE EDITOR
adaptation to the intervention. Thus, there is the risk that the pos-
itive effects seen in these RCTs may be, as Heaney (4) pointed
out, inflated by being a compound of the remodeling transient
plus an improvement in bone balance. By contrast, observational
studies do not alter customary calcium intake, thereby avoiding
the confounding problems of the remodeling transient.
Nevertheless, assuming that observational studies should not
be given level A strength-of-evidence status equal to that of
RCTs, downgrading the one observational study in the unfavor-
able-effect category (5) to level B status does not change the
results. That is, the reported ratio of favorable to unfavorable
effects for the stronger-evidence categories A and B remains
low and unchanged at 2:1 (6 favorable and 3 unfavorable out-
comes). We believe that the more important point is that there
are too few carefully designed studies of the effects of dairy
foods on bone health.
Weaver and Heaney conclude that adequate dairy consump-
tion supports good bone health. We agree with their conclusion,
if qualifying terms are added to clarify that adequate consump-
tion of certain types of dairy foods appears to be supportive of
good bone health among select age, race, and sex groups. Two
important caveats are as follows. 1) A clear distinction must be
made between milk and its derived dairy products in discus-
sions of dairy foods and bone status. Many dairy foods, partic-
ularly processed cheese products and cottage cheese, have
markedly different ratios of calcium to potassium, sodium con-
tents, and renal acid loads than does milk, and may have
markedly different effects on the way calcium is metabolized.
2) There are too few studies in males and ethnic minorities for
conclusions to be drawn about the effect of any dairy food on
bone health in these groups, which together represent more
than one-half of the US population. Hence, the body of scien-
tific evidence is inadequate to support a recommendation for
daily intake of dairy foods to promote bone health in the gen-
eral US population.
Roland L Weinsier
Carlos L Krumdieck
Clinical Nutrition Research Center
Department of Nutrition Sciences
231 Webb Building
1675 University Boulevard
University of Alabama at Birmingham
Birmingham, AL 35223-3360
E-mail: weinsier@shrp.uab.edu
REFERENCES
1. Weinsier RL, Krumdieck CL. Dairy foods and bone health: exami-
nation of the evidence. Am J Clin Nutr 2000;72:681–9.
2. Chan GM, Hoffman K, McMurry M. Effects of dairy products on bone
and body composition in pubertal girls. J Pediatr 1995;126:551–6.
3. Chan GM, McMurry M, Westover K, Engelbert-Fenton K, Thomas
MR. Effects of increased dietary calcium intake upon the calcium
and bone mineral status of lactating adolescent and adult women.
Am J Clin Nutr 1987;46:319–23.
4. Heaney RP. Calcium, dairy products and osteoporosis. J Am Coll
Nutr 2000;19:83S–99S.
5. Feskanich D, Willett WC, Stampfer MJ, Colditz GA. Milk, dietary
calcium, and bone fractures in women: a 12-year prospective study.
Am J Public Health 1997;87:992–7.
661
Abnormal fatty acid status in patients with
Crohn disease
Dear Sir:
We read with interest the article by Jeppesen et al (1) that
assessed the influence of administration of enteral or parenteral
nutrition on plasma phospholipid essential fatty acid (EFA) concen-
trations in patients with malabsorption. This was a well-designed
study that included 4 groups of patients; 2 groups received par-
enteral nutrition (groups C and D), group A had fat malabsorption
of < 50% of fat intake, and group B had fat malabsorption of > 50%
of fat intake. Group C received parenteral nutrition containing lipids
and EFA and group D received fat-free parenteral nutrition. EFA
absorption was negligible in groups C and D and EFA supplemen-
tation in group C was not enough to completely reverse biochemi-
cal EFA deficiency. The authors concluded that EFA requirements
in patients receiving parenteral nutrition are higher than the amounts
recommended to patients with preserved intestinal absorption.
However, as McCowen et al (2) pointed out in a letter to the
Journal, a major concern with the study is that most of the
patients had Crohn disease. In fact, an abnormal fatty acid pro-
Downloaded from www.ajcn.org by on May 29, 2008
file has been described in patients with active (3) and inactive (4)
Crohn disease and in both plasma (3, 4) and intestinal mucosa
(5). Plasma long-chain n3 polyunsaturated fatty acids (PUFAs)
were significantly elevated in patients with Crohn disease (3)
and a similar but nonsignificant trend was observed for n6
long-chain PUFAs in patients with inactive disease (4). This
finding is consistent with results of the study by Jeppesen et al
that showed higher values of both 20:3n6 and 20:4n6 in all
the groups studied (including most patients with Crohn disease)
than in the control group (1).
In addition, elevated concentrations of long-chain n6 PUFAs
were observed in the intestinal mucosa of patients with Crohn
disease (5). Similar findings were observed in patients with ulcer-
ative colitis and in experimental colitis, suggesting that there is a
common biochemical response to intestinal inflammation (5).
Enteral nutrition has a primary effect on the course of Crohn
disease, inducing remission in 60% of patients (6). It is not known
which component or components of the diet are responsible for the
control of inflammation, but there are some data suggesting that
the lipid composition of the diet may play a crucial role (6, 7).
Thus, the authors should be cautious before recommending
high doses of parenteral EFAs in patients with Crohn disease
because these patients have increased amounts of products of
these fatty acids, especially in intestinal mucosa (5). This maneu-
ver might exacerbate the activity of the disease (8). This nutri-
tional recommendation would probably be more useful in patients
with short-bowel syndrome and EFA deficiency due to another
non-immune–mediated disease.
Maria Esteve-Comas
Miquel Angel Gassull
Department of Gastroenterology
Hospital Universitari Germans Trias i Pujol
Ctra de Canyet, s/n
08916 Badalona
Catalonia
Spain
E-mail:mgassull@ns.hugtip.scs.es
662 LETTERS TO THE EDITOR
REFERENCES
1. Jeppesen PB, Høy C-E, Mortensen PB. Differences in essential fatty
acid requirements by enteral and parenteral routes of administration in
patients with intestinal malabsorption. Am J Clin Nutr 1999; 70:78–84.
2. McCowen K, Ling PR, Bistrian BR. Arachidonic acid concentrations
in patients with Crohn disease. Am J Clin Nutr 2000;71:1008–9.
3. Esteve-Comas M, Ramírez M, Fernández-Bañares F, et al. Plasma
polyunsaturated fatty acid pattern in active inflammatory bowel dis-
ease. Gut 1992;33:1365–9.
4. Esteve-Comas M, Núñez MC, Fernández-Bañares F, et al. Abnormal
polyunsaturated fatty acid pattern in non-active inflammatory bowel
disease. Gut 1993;34:1370–3.
5. Fernández-Bañares F, Esteve-Comas M, Mañé J, et al. Changes in
mucosal fatty acid profile in inflammatory bowel disease and in
experimental colitis: a common response to bowel inflammation.
Clin Nutr 1997;16:177–83.
6. Fernández-Bañares F, Cabré E, Esteve-Comas M, Gassull MA. How
effective is enteral nutrition in inducing clinical remission in active
Crohn’s disease? A meta-analysis of the randomized clinical trials.
JPEN J Parenter Enteral Nutr 1995;19:356–64.
7. Gassull MA, Fernández-Bañares F, Cabré E, et al. Fat composition
is the differential factor to explain the primary therapeutic effect
of enteral nutrition in Crohn’s disease. A double blind randomized
multicenter European trial. Gastroenterology 2000;118(suppl):
A117 (abstr).
8. Takahashi K, Kato T, Schreiner GF, Ebert J, Badr KF. Essential fatty
acid deficiency normalizes function and histology in rat nephritis.
Kidney Int 1992;41:1245–53.
Reply to M Esteve-Comas and MA Gassull
Dear Sir:
We thank Esteve-Comas and Gassull for their interest in our
study. As we emphasized in the article (1), it is difficult to corre-
late essential fatty acid status with clinical symptoms. Therefore,
we have been cautious in promoting clinical recommendations
based on our findings.
Our physiologic study assessed the relation between par-
enteral and enteral supplementation of essential fatty acids and
the status of these fatty acids in plasma phospholipids. We
showed that larger amounts of parenteral linoleic acid are nec-
essary to normalize linoleic acid concentrations in plasma
phospholipids in patients with fat malabsorption compared
with the amount provided enterally (1). However, we did not
recommend that attempts should be made to normalize fatty
acid status. We actually stated that “. . . the clinical benefits and
cost-effectiveness of lipid supplementation should be consid-
ered before attempting to normalize the fatty acid status of
HPN patients” (1).
In fact, it has been the clinical practice in our center to give
smaller amounts of parenteral lipids to patients receiving home
parenteral nutrition (HPN) than are given in other centers in the
world (2, 3). The feasibility of recruiting patients to group C in
this study illustrates this (1).
In clinical practice, the pros of restoring essential fatty acid
status and the cons of a reactivation of Crohn disease must be
considered. In most patients with both short-bowel syndrome as
well as Crohn disease, an extinction of disease activity is seen
over time; the clinical problems in these patients are more often
related to nutritional status than to disease activity.
As stated above, other centers in the world are less restrictive
in providing essential fatty acids to these patients (3). However,
our study was not an attempt to establish recommendations for
lipid supplements in patients with fat malabsorption. Rather, our
article simply describes differences in requirements when fatty
acids are given parenterally rather than enterally. We await with
great interest controlled clinical trials that address the pros and
cons of lipid supplementation in these patients.
Palle Bekker Jeppesen
Carl-Erik Høy
Per Brøbech Mortensen
Rigshospitalet
University of Copenhagen
Department of Medicine
Section of Gastroenterology
CA 2121
Blegdamsvej
Downloaded from www.ajcn.org by on May 29, 2008
Copenhagen 2100
Denmark
REFERENCES
1. Jeppesen PB, Høy CE, Mortensen PB. Differences in essential fatty
acid requirements by enteral and parenteral routes of administration
in patients with fat malabsorption. Am J Clin Nutr 1999;70:78–84.
2. Jeppesen PB, Hoy CE, Mortensen PB. Deficiencies of essential fatty
acids, vitamin A and E and changes in plasma lipoproteins in
patients with reduced fat absorption or intestinal failure. Eur J Clin
Nutr 2000;54:632–42.
3. Jeejeebhoy KN, Langer B, Tsallas G, Chu RC, Kuksis A, Anderson
GH. Total parenteral nutrition at home: studies in patients surviving
4 months to 5 years. Gastroenterology 1976;71:943–53.
The salt controversy at the turn of the
century: no to prejudiced thinking, yes
to concerted action
Dear Sir:
The articles by McCarron (1) and Kaplan (2) on the US
dietary guidelines for sodium effectively synthesize the oppo-
site views of detractors and supporters of the “salt hypothesis.”
The evidence in favor of each respective thesis is presented with
vigor and brightness. Having said that I personally share
Kaplan’s conclusive remarks, my concern is whether this type
of antagonistic approach to the problem offers a true perspective
to health professionals and to the population at large or, rather,
perpetuates a long-lasting scientific conflict that, I believe, has
been itself a major obstacle to large-scale implementation of
lifestyle modifications for prevention and treatment of hyper-
tension. Indeed, as Kaplan brilliantly pointed out in his textbook
on clinical hypertension (3), nonpharmacologic measures are
 LETTERS TO THE EDITOR
proposed as initial therapy for most patients by all official
guidelines; practitioners recommend them more now than ever
before but implement them as poorly as ever. What is the reason
for this discrepancy? How much of the continued medical edu-
cation highlights the merits of pharmacologic therapy and how
much does it promote the implementation of nonpharmacologic
measures recommended in the guidelines? Would it not be bet-
ter to recognize that pharmaceutical companies are probably as
interested in the implementation of dietary salt reduction as is
the lobby of salt producers?
Convinced as I am of the importance of dietary salt in the etiology
of hypertension, I believe that it is indeed time to waive the con-
flict-based approach and have interested scientists look at the sev-
eral issues on the table. All should be respectful of each other’s posi-
tions and bring their own experience to the discussion with the
purpose of having a sincere interchange and taking a step forward (4).
Having read McCarron’s and Kaplan’s articles with as unbi-
ased an attitude as I was able, I have concluded that at least a few
points in their conflicting positions could be good starting points
for positive action.
1) McCarron stated, “. . . although dietary salt does play a role,
it is not the archenemy of normal blood pressure regulation.”
May I presume that if salt is no longer identified as the “arch-
enemy” of blood pressure control, McCarron would be eager
to sit at the table and concentrate on the good reasons for salt
playing a role?
2) McCarron also admitted that “salt-sensitivity has now been
shown to be a reproducible phenomenon. . ., although as yet
there is not a specific definition of what constitutes a salt-
sensitive response. . . .” The acceptance of salt-sensitivity as a
reproducible phenomenon is in accordance with a role for salt
in blood pressure regulation. It is difficult to disagree on the
lack of a practical way to detect salt sensitivity. Therefore,
could all of us try and find new avenues for improving our
knowledge of such a crucial aspect of the problem? Could we
focus on the very fact that most of the monogenic forms of
hypertension so far identified imply a deficit in renal
sodium handling (4)?
3) If there is agreement on the lack of evidence of substantial
benefit for normotensive people from short-term trials of
NaCl restriction (1), would McCarron turn his attention less
reluctantly to the evidence of long-term benefit from salt
reduction in chimpanzees (5) and consider its relevance to
hypertension in humans?
4) May we expect that recognition of the limited applicability of
the INTERSALT study results in predicting the large-scale
effects of lifestyle modifications (1) will be followed by the
acknowledgment of the major limitations that make Alder-
man et al’s study on sodium intake and cardiovascular mor-
tality inconclusive (6)?
5) Finally, how do we challenge McCarron’s contention that the
potassium and calcium contents of Western diets are defi-
nitely too low in the same way that their NaCl content is too
high? Having realized that dozens of guidelines by scientific
societies and hundreds of authoritative expert recommenda-
tions over the years have not convinced people—or general
practitioners—to reduce NaCl intake to any extent, would it
not be wise to consider a different approach to the problem
and speak in terms of complex dietary changes that include,
of course, substantial NaCl restriction?
663
At the turn of the century, we can no longer afford to go on
with the sterile dilemma “to salt or not to salt.” Let us give up the
conflict, abandon prejudiced thinking, agree on what is already
clear, and build up valuable new knowledge based on unpreju-
diced observations and well-designed experiments.
P Strazzullo
Department of Clinical and Experimental Medicine
Federico II, University of Naples Medical School
Naples
Italy
REFERENCES
1. McCarron DA. The dietary guideline for sodium: should we shake
it up? Yes! Am J Clin Nutr 2000;71:1013–9.
2. Kaplan NM. The dietary guideline for sodium: should we shake it
up? No. Am J Clin Nutr 2000;71:1020–6.
3. Kaplan NM. Clinical hypertension. 7th ed. Baltimore: Williams and
Wilkins, 1998.
4. Siani A, Guglielmucci F, Farinaro E, Strazzullo P. Increasing evi-
dence for the role of salt and salt-sensitivity in hypertension. Nutr
Downloaded from www.ajcn.org by on May 29, 2008
Metab Cardiovasc Dis 2000;10:93–100.
5. Denton D, Weisinger R, Mundy NI, et al. The effect of increased salt
intake on blood pressure of chimpanzees. Nat Med 1995;1:1009–16.
6. Alderman MH, Madhavan S, Cohen H, et al. Low urinary sodium is
associated with greater risk of myocardial infarction among treated
hypertensive men. Hypertension 1995;25:1144–52.
Reply to P Strazzullo
Dear Sir:
A cease-fire in the salt war is long overdue. I am in complete
agreement with Strazzullo’s assertion that it is time to set aside
prejudicial thinking and work together to find resolutions to the
sodium controversy. Such has been my own contention for 20 y.
In our analysis of the first National Health and Nutrition
Examination Survey dietary data published in Science in 1984
(1), we concluded that health measures to reduce hypertensive
cardiovascular disease are not best served by focusing on isolated
dietary components, but rather by “the consumption of a diet
balanced in all the essential nutrients and appropriate for the indi-
vidual’s level of physical activity.” My commentary in the pro-
ceedings of the 1991 NIH Workshop on Salt and Blood Pressure
(2), entitled “A Consensus Approach to Electrolytes and Blood
Pressure: Could We All be Right?” speculated that salt was but
one of many dietary components contributing to high blood pres-
sure, and that “our future understanding of the effects of dietary
NaCl on blood pressure will only be advanced if we approach the
task from the position that salt’s action on blood pressure regula-
tion must be viewed in the context of the whole diet.”
In 1998, in an editorial in Science entitled “Diet and blood
pressure—the paradigm shift” (3) that followed the publication of
the first DASH Study, I concluded that “emphasis on sodium as
the single dietary culprit is counterproductive to our significantly
664 LETTERS TO THE EDITOR
reducing cardiovascular risk. . . and diverts attention from the
issues we need to address.” The DASH-Sodium Study, along with
several other studies and analyses [described in the article to
which Strazzullo is referring (4)], have proven this to be the case.
My stand that we should shake up the dietary guideline is
based on the fact that, as it presently exists, the guideline does
not represent the most current evidence. Rather, it is narrowly
based on data from one side of the sodium controversy and
merely iterates the status quo that has dominated this area for the
past 20 y. We would not still be debating this issue if the evi-
dence were as solid as the guideline proponents allege it to be.
If we are to find common ground in the salt war, as Strazzullo
and most of the rest of us would like, then the data on both sides
of the argument must be put forth, considered as objectively as
possible, and integrated into a clear and unified message to pol-
icymakers, practitioners, and the general public.
Does Strazzullo also call on the proponents of the sodium
guideline to set aside their prejudices in this issue? A recent
press release of the DASH-Sodium Study (5) offers a prime
example of how advocates of the current policy promote only
one side of the data. In this release, the National Heart, Lung,
and Blood Institute’s director Lenfant focused only on the effects
of sodium restriction in specific subjects, ignoring the most com-
pelling results from this federally supported trial: that the blood
pressure–lowering effect of improved dietary patterns were far
greater and more uniform than were those of sodium restriction.
What could have been an opportunity both to promote the
DASH-Sodium Study’s full findings and potential population-
wide benefits and to advance efforts to bring unity to this con-
flict became another presentation of incomplete and unbalanced
data, serving only to further inflame this issue.
By continuing this argument, proponents of the current
sodium guideline in the nutrition research and policy communi-
ties do the public a great disservice. Strazzullo suggests, and I
fully concur, that the continued conflict and acrimony among the
scientific community with regard to sodium and blood pressure
is a major deterrent to widespread and effective implementation
of nonpharmacologic management of high blood pressure by
both patients and physicians.
In addition to weakening the public’s confidence in nutritional
advice, a national health policy that promulgates an unresolved
recommendation further jeopardizes public health by diverting
attention away from areas where it can be beneficial. Public
health emphasis must be aligned with the evidence—it must be
placed where it can actually effect change.
We know unequivocally that both obesity and alcohol are lead-
ing risk factors for hypertension and numerous other medical con-
ditions. We now know that dietary patterns are more important in
blood pressure regulation than are modifications of any single
nutrient intake. Today, we have the information necessary to design
public health policies that can truly benefit public health. It is in
these areas that our emphasis and our efforts should be targeted.
Although Strazzullo states that his views are in opposition to
mine, we share the desire to find an equitable peace in the salt
war. I welcome Strazzullo’s voice of reason in this controversy
and join him in calling for an enlightened approach to its resolu-
tion, a position I have maintained for nearly 2 decades.
David A McCarron
Division of Nephrology, Hypertension, and Clinical Pharmacology
Downloaded from www.ajcn.org by on May 29, 2008
Department of Medicine
Oregon Health Sciences University
Portland, OR
E-mail: dmccarron@academicnetwork.com
REFERENCES
1. McCarron DA, Morris CD, Henry HJ, Stanton JL. Blood pressure
and nutrient intake in the United States. Science 1984;224:
1392–8.
2. McCarron DA. A consensus approach to electrolytes and blood pres-
sure: could we all be right? Hypertension 1991;17(suppl):I170–2.
3. McCarron DA. Diet and blood pressure—the paradigm shift. Sci-
ence 1998;281:933–4.
4. McCarron DA. The dietary guideline for sodium: should we shake
it up? Yes! Am J Clin Nutr 2000;71:1013–9.
5. Blood pressure falls along with sodium intake, researchers find.
Medical Industry Today 2000 May 18 (press release).