April 6, 2004
Kathryn
McMurry
HHS Office of Disease Prevention and Health Promotion
Room 738-G
200 Independence Avenue, SW
Washington, D.C. 20201
Dear
Kathryn:
We applaud the
Committee for its determination to give further study to the questions raised by Dr. Appel
regarding recommendations for a Guideline on dietary salt.
The discussion raised a number of the key issues and suggested that the
Committee is, indeed, open to arguments about how its advice will influence health
outcomes which we continue to believe is the key scientific question before the Committee.
We offer these
further comments concerning salt/sodium to build on several of the key points raised at
the meeting March 30-31 in Washington. Here
is a selected summary of these remarks, followed by a longer discussion. In summary:
In further
elaboration of these points, then, these additional comments.
1. Dr.
Appel’s classic graph of “usual blood pressure” showing two similar-shaped
bell curves slightly offset, offered to illustrate the cardiovascular risk benefits of
“shifting the curve to the left,” is the clearest explanation of a mistaken
assumption that we challenge the Committee to consider.
Evidence is incontrovertible that usual blood pressures are directly related
to cardiovascular risk – the higher the blood pressure, the greater the chance that a
population will experience more heart attacks and strokes. The
mistaken assumption is that we can graph the shape of the (likely) curve representing
blood pressures that have been modified so they are no longer “usual.” Untreated blood pressure is a predictor of risk
and the shape of the curve can be accepted. Treated
blood pressure, be it by pharmacologic or non-pharmacologic intervention, is undertaken
with the assumption that the lowered blood pressure will reduce cardiovascular risk. Evidence of the efficacy of reducing BP is
voluminous, be the intervention diuretics, ACE inhibitors, beta blockers, weight loss,
exercise or, yes, salt reduction. The size of
the reduction and the heterogeneity of the response are critical considerations, but all
of these interventions “work” on populations to some degree. The question is whether the “unusual”
BP produced by the intervention achieves the hoped-for reduction in cardiovascular events. This has been tested for pharmacologic
interventions in the ALLHAT trial; it has never been tested for non-pharmacologic
interventions. It should be, of course, but
the only data we have on this point is observational.
Fortunately, those data are consistent.
All ten observational trials show that reduced sodium diets do not reduce
the incidence of heart attacks or strokes. Some
show an increased risk; others show no statistical change.
None show an improved risk profile even when blood pressures are lowered
substantially. Thus, the size and shape of
the “curve” representing the modified risk profile when the intervention
modifies “usual blood pressure” is unlikely to be represented by the simple
“shifting to the left” illustrated in the classic graph offered by Dr. Appel. Unfortunately, this profound insight was not
articulated in the Committee’s discussions and certainly was entirely absent from the
blood pressure-centric focus of the subcommittee’s report. Please review the attached “Health
Outcomes” graphs that illustrate the findings of some of the ten reported health
outcomes studies of salt reduction.
2. Fortunately,
Dr. Appel, in his discussion of potassium, made it clear that hypertension is, not a
disease end point, but rather “an intermediate variable” used in models to
estimate outcomes risks. We agree. In the case of sodium, hypertension is one of the
important intermediate variables affected by sodium intakes. So, too, are the “intermediate
variables” of plasma renin activity, sympathetic nervous system response and insulin
resistance. While we believe this should
underscore the imperative of assessing the health outcomes impact, we would also observe
that the discussion of “intermediate variables” was truncated. Unfortunately, the Committee focused its entire
discussion on blood pressure and ignored these other important and relevant
“intermediate variables.” Plasma
renin activity, for example has been shown to be highly correlated with risk of CV events
– a direct relationship representing a four-fold greater incidence of heart attacks
among those with higher PRA than those with lower PRA.
Plasma renin activity, of course, is stimulated by sodium reduction. The same dynamic characterizes sympathetic nervous
system response to sodium reduction and reduced insulin sensitivity on lower sodium diets. Some have argued for an adverse impact on lipids
as well, but the evidence is less clear-cut. In
any case, these other “intermediate variables” deserve full consideration in
addition to the “intermediate variable” of blood pressure. It is the sum total of the impacts of intermediate
factors that will determine the overall health impact of the intervention of lowering
dietary sodium. Failure to consider these
“intermediate variables” may produce poor science policy, but it will not change
physiology and, thus, will not affect the health outcomes of the intervention. Even in ignorance we cannot break the laws of
nature.
3. Dr.
Appel correctly identified the different impacts of “responders” and
“non-responders” to changes in electrolyte intake. For sodium, we use the terms
“salt-sensitive” and “salt resistant.”
Of course, this is a continual gradation, but we strongly agree that there is vast
heterogeneity of response, including blood pressure response. In fact, the exceptionally modest population BP
impacts of even drastic sodium reduction pale in comparison with the relatively robust
expression of heterogeneity. Dr. Appel noted
that swings of 20 mm/Hg in blood pressures could be obtained in successive measures of
blood pressure in the same individual. Many
individuals will experience that magnitude of increase of blood pressure with extremely
high intakes of salt; others (many, though not quite as numerous) will have decreases in
blood pressure on those same elevated intake levels.
In this case, it would be well to consider whether the standard deviation is
more important than the mean. Fortunately,
quality work on these blood pressure impacts has been available for some time. Unfortunately, the discussion at the March 30-31
meeting focused entirely on the direction of the mean BP change and not on the magnitude
of that change in comparison with the vast heterogeneity of population BP response –
to say nothing of the absence of the discussion of the variations and magnitude of the
heterogeneous response of other variables like PRA, insulin resistance and sympathetic
nervous system response.
4. In a
related point, we salute Dr. Appel for calling the Committee’s attention to the fact
that while genetic factors are clearly part of the explanation for salt sensitivity, an
individual’s salt-sensitivity is a “modifiable risk factor.” We have made this same point in earlier
communications: consumption of
currently-recommended amounts of potassium, calcium and magnesium (as in the DASH Diet)
erases or substantially reduces salt sensitivity, even among the most salt-sensitive
subgroups. Dr. Appel usefully pointed out
that research by Curtis Morris demonstrates that “potassium blunts the pressor effect
of sodium.” In the DASH-Sodium trial,
the results proved that the DASH Diet entirely eliminated salt-sensitivity for subgroups
representing a majority of the population – normotensives, males and those under age
45; their salt-sensitivity disappeared. In
addition, the DASH Diet virtually eliminated salt-sensitivity for Caucasians. Salt-sensitivity was greatly reduced for
hypertensives, females and those over age 45. And, finally, even for the most
salt-sensitive subgroup, African-Americans, the DASH Diet was responsible for substantial
reduction in salt-sensitivity. Elimination of
salt-sensitivity has myriad benefits; this is not to be confused with any modification of
salt intake levels. I am attaching an article
by Dr. Alexander G. Logan making this point for your consideration.
5. The
Salt Institute has been an active supporter of the DASH Diet since 1997. It seems totally consistent with other accepted
advice to increase the intake of fruits and vegetables.
It addresses the recorded nutrient shortfalls in the American diet of
potassium, calcium and magnesium. While the
two DASH trials, each only 30 days in duration are hardly sufficient evidence upon which
policy advice for 290 million Americans should be based, these are well-done studies and
exciting results that deserve attention. They
highlight the broader issue that with regard to blood pressure and to incidence of disease
end point incidence, the evidence shows that higher quality diets play a major role in
modifying risks and promoting health. The
discussion of the DASH-Sodium trial made the point that the “DASH Effect” vastly
outweighs the “Sodium Effect.” But
the story is even stronger than articulated by Committee members in their discussion. In every subgroup the DASH effect dwarfed
the Sodium effect. In earlier communications,
we have commended the study by Kant et al in JAMA
in 2000. Attached are graphs representing
the improvements in risk exposure attributable to improved dietary quality. We particularly endorse the authors’
conclusion:
The results from this large cohort of
women with prospective follow-up suggest that dietary patterns characterized by
consumption of fruits, vegetables, whole grains, low-fat dairy, and lean meats are
associated with lower risk of mortality. Given the simplicity of the diet quality
score used in this study, increasing the intake of recommended foods – without
undue emphasis on learning about hidden fat, total amount and type of fiber, or individual
vitamins and minerals – may represent a practical recommendation for improving
health. (emphasis added)
6. Before moving
on from DASH-Sodium, we were pleased to learn of the new DASH-Sodium paper (“Bray
article”) that Dr. Appel presented (we were somewhat surprised, however, having been
informed on earlier occasions that only published studies could be considered). As you know, the Salt Institute has called for
NHLBI to make public the data upon which they interpret this study to demonstrate a
universal benefit for salt reduction. Perhaps
this article will contain the long-sought data. The
Salt Institute is so committed to securing release of these data so that they might be
available, for example, for your Committee’s use, that on March 31 we, joined by the
U.S. Chamber of Commerce, filed suit in federal district court under the Data Quality Act. Our success will enable you to use the DASH-Sodium
trial results. Unless we are successful,
reliance on DASH-Sodium as a basis for your decision becomes problematic since experts
have declared that the data released to date does not justify the conclusions advanced by
its authors. Absent compliance of the
DASH-Sodium trial with the Data Quality Act, federal agencies (including HHS and USDA) are
prohibited from making policy on the basis of such studies.
The data sought, I am certain you will agree, are both basic and easily
supplied: the means, standard deviations and
sample size for each of the subgroups. That’s
it. Simple.
We hope you, and more relevantly, the federal district judge, agree.
7. Dr. Appel
properly called the Committee’s attention to other alleged adverse impacts of dietary
salt. In fact, the list is virtually endless
– salt has been blamed for virtually everything imaginable. With regard to what Dr. Appel asserted is a
“potentially important” direct relationship between sodium intake and calcium
excretion, there is clear evidence that this relationship exists. The question remains, as Dr. Appel tutored, that
no evidence associates this physiological truism with an adverse health outcome. The answer here is clearly to consume recommended
amounts of dietary calcium. With regard to
left ventricular mass, the evidence is intriguing concerning this, another
“intermediate variable.” Further
research may be appropriate, but, again, the correct policy focus is on health outcomes,
not “intermediate variables.” Finally,
Dr. Appel asserted a “pretty consistent” and “strong association” of
sodium intake with incidence of gastric cancer. We
take issue with this characterization. It is
true that higher gastric cancer incidence has been documented in Japanese populations and
isolated coastal fishing villages where heavy consumption of salted fish and vegetables is
common. However, at sodium intake levels and
with consumption levels of salted fish and salt-cured vegetables in the U.S., these
studies have been dismissed as irrelevant by the American Cancer Society.
8. The Committee
discussion included a number of statements about the relative merits of
“practical” goals as opposed to goals that might be indicated from the
scientific literature. Clearly both are
important. We agree with Chairman King that
the recommendations must be faithful to the science.
We also agree with Drs. Lupton and Caballero that it is appropriate for the
Committee to consider food science as well as medical science since its recommendations
are in the nature of implementing a strategy to achieve improved health. Not only does salt, undeniably, serve multiple
functions in food, including enhancing food safety, but food technologists have been
trying diligently for the past generation to respond to public health officials’
encouragement to reduce the salt content of foods. They
have succeeded, but there has been no reduction in sodium intakes (or, for that matter,
over the last century for which we have been able to measure urinary sodium). I could speculate on this matter, but there are no
data to confirm why. Dr. Lupton stated that
she was unimpressed, in practical terms, with the 5 mm/Hg SBP reduction in TOHP II. Had she reflected on the full 36-month outcomes of
that trial she (and the Committee, we hope) would be even more unimpressed. TOHP II found that subjects who maintained a 40
mmol Na reduction over the full 36 months of the trial achieved only a 0.6 mm/Hg SBP
reduction (127.7 to 127.1 mm/Hg). Dr. Appel
presented data showing that only 5% of American men consume 109 mmol Na or less and only
5% of women consume 78 mmol Na or less. Thus,
virtually no men today consume what Dr. Appel would recommend as an “interim
goal” and no men and virtually no women are consuming the amount he feels should be
ingested. We may (in fact, we do)
disagree on this recommended amount, but the Committee certainly needs to consider the
“practical” implications of current consumption levels. Not only have millions of dollars been spent
already trying to reduce dietary sodium – with absolutely no effect on intake levels
– but the diversion of those resources (and time and public attention) to salt
reduction has blurred public recognition of the benefits of improving dietary quality. Any strategy must be grounded in science but
recognize and prioritize the resources available. If
everything is a “target,” nothing is being targeted.
9. Finally, Dr.
Lupton usefully reminded the Committee that its charge is to produce an evidence-based
product. She then stated that the IOM report
would be the appropriate starting point for that scientific quest. While Dr. Appel described the IOM report, released
February 11, a week after the previous meeting of the Committee, he neglected to call the
Committee’s attention to the recommendations of the five leading Canadians agencies
dedicated to improving cardiovascular health that were published the day following the
Committee’s January meeting. Using the
same body of evidence available to the IOM, the recommendations issued on behalf of The
Canadian Hypertension Society, The Canadian Coalition for High Blood Pressure Prevention
and Control, The College of Family Physicians of Canada, The Heart and Stroke Foundation
of Canada, and the Chronic Disease Prevention Division, Centre for Chronic Disease
Prevention and Control, Health Canada, do not include restricting salt intake for a
majority of Canadians. (Canadian Journal of Cardiology;
http://www.pulsus.com/CARDIOL/20_01/touy_ed.htm). The Canadian recommendations state that patients
being treated for hypertension and “normotensive
individuals at high risk for developing hypertension” who are also considered
“salt sensitive” should limit sodium intake to 2,300 mg/day or less. The
high risk normotensive population is a subset of individuals with diastolic blood
pressures greater than 80 mm/Hg. Thus,
individuals with diastolic blood pressure below 80 mm/Hg, the majority of the population,
are excluded from any current need to restrict dietary salt. Also, individuals with diastolic blood pressure
80-89 mm/Hg who are not at high risk are also excluded from recommendations to restrict
salt. These groups of “experts”
clearly took a different path from the “experts” on the IOM panel. Their very different conclusions raise the useful
question of exactly what is “evidence-based” public health policy. The Committee has produced reams of documents on
the science. Does this mean its
conclusion are “evidence-based”? As
described by Archie Cochrane in his classic text Effectiveness
and Efficiency: Random Reflections on Health Services in 1972, randomized controlled trials
are vital in assessing the effectiveness of interventions.
Cochrane’s work led to establishment of the now-worldwide Cochrane
Collaboration dedicated to evidence-based recommendations.
The essence of “evidence-based medicine” for Cochrane and those
who defined the term is that the quality of the process determines the quality of the
science which, in turn, determines the appropriate application of the scientific results. No one would base patient care on the conclusions
of a high school science project. Likewise,
the hierarchy of expert commentary, ecological studies, studies of intermediate variables,
observational studies of health outcomes and randomized trials of outcomes reflects the
ascending scale of levels of evidence. The
IOM study is a Grade D expert report; the Canadian recommendations a somewhat higher level
of evidence based on the process employed. We
exhort the Committee to employ a true “evidence-based” approach to evaluating
the science. If you do, we expect you will
come out where the Cochrane Collaboration ended up – declaring that universal sodium
restriction is unjustified by the science. We
commend to your attention the Cochrane Review “Advice to reduce dietary salt for the
prevention of cardiovascular disease” (2006:
http://www.cochrane.org/reviews/en/ab003656.html). These
are the conclusions of the
Cochrane Review.
Intensive interventions, unsuited to
primary care or population prevention programmes, provide only minimal reductions in blood
pressure during long-term trials. Further evaluations to assess effects on morbidity and
mortality outcomes are needed for populations as a whole and for patients with elevated
blood pressure. Evidence from a large and small trial showed that a low sodium diet helps
in maintenance of lower blood pressure following withdrawal of antihypertensives. If this
is confirmed, with no increase in cardiovascular events, then targeting of comprehensive
dietary and behavioural programmes in patients with elevated blood pressure requiring drug
treatment would be justified.
The Cochrane Collaboration has also updated, on November 26, 2003, the analysis in
its 1998 Review “Effects of low sodium diet versus high sodium diet on blood
pressure, renin, aldosterone, catecholamines, cholesterols, and triglyceride” which
concluded:
The magnitude of the effect in Caucasians
with normal blood pressure does not warrant a general recommendation to reduce sodium
intake. Reduced sodium intake in Caucasians with elevated blood pressure has a useful
effect to reduce blood pressure in the short-term. The results suggest that the effect of
low versus high sodium intake on blood pressure was greater in Black and Asian patients
than in Caucasians. However, the number of studies in black (8) and Asian patients (1) was
insufficient for different recommendations. Additional long-term trials of the effect of
reduced dietary sodium intake on blood pressure, metabolic variables, morbidity and
mortality are required to establish whether this is a useful prophylactic or treatment
strategy.
If the Committee wants to produce an
“evidence-based” report and recommendations, we would recommend adhering to the
evidence evaluated by the world’s definitive “evidence-based medicine”
pioneer and exemplar, the Cochrane Collaboration.
In conclusion, we are pleased that the
Committee intends to recommend areas requiring further research. We have several we could suggest and would be
pleased to share those thoughts with the Committee if you have an interest. Paramount among them, however, is the vital need
to commence a clinical trial of the health outcomes of salt reduction. It is almost impossible to believe that we could
have created such a powerful focus on this far-reaching intervention in the absence of
such evidence. Unless and until evidence
exists that demonstrates that people will improve their health by reducing dietary sodium,
we should concentrate our constrained resources available to improve the American diet
towards upgrading the quality of the overall diet, including, importantly, increasing the
consumption of fruits, vegetables and dairy products.
Even with regard to blood pressure, the evidence offered to support the
asserted dose-response relationship fails. The
Luft paper cited by Dr. Appel found no relationship of sodium and blood pressure within
the 50 mmol – 400 mmol range that encompasses virtually all individuals and every
community in the civilized world. In the
Midgley and Cutler meta-analyses, their regression lines did not pass through origin;
something else is creating the blood pressure effect (a point mentioned by Midgley, but
not Cutler). The science suggests, rather,
that blood pressure is multi-factorial and salt-sensitivity, heterogeneous. In the absence of any outcomes data that shows a
health benefit from salt reduction, the emphasis in our national public health nutrition
strategy is properly to seek overall improvements in dietary quality.
We recommend that this strategy be pursued
with advice on foods rather than nutrients, an approach reflecting the DASH Diet. Please review the attached slides on Dr.
Kant’s study which emphasize our conclusion.
Sincerely,
Richard L. Hanneman
President
P.S. Chairman King inquired about iodized salt and the adequacy of iodine nutrition in the U.S. diet. We do not believe that the Committee needs to act on this matter, but Dr. Appel’s dismissive response saying that we had solved this problem in the 1920s or 1930s ignores a more disturbing trend that bears ongoing scrutiny and possible attention from future DGACs. From NHANES I to NHANES II, there was a four-fold increase in the number of expectant mothers who failed the minimum iodine intake standard established by the World Health Organization. The Salt Institute was among those leading the successful lobbying effort to persuade the Clinton Administration not to axe funding for ongoing iodine data collection. In fact, in NHANES III and IV, apparently, the downward spiral (evident as Chairman King observed, in other countries like Australia and Canada) has been checked; the area requires continuing monitoring. Further details are available from Joe Hollowell at the CDC in Atlanta. Likewise, Dr. Appel’s confident assertion that the threat of Iodine Deficiency Disorders (IDD), while problematic, is a problem for “developing nations” is wide of the mark. Western Europe is among the areas seriously challenged by IDD. I serve on the Board of Directors of the International Council for the Control of Iodine Deficiency Disorders and the President of Morton Salt is the vice-chair of the UNICEF-chaired global Network for the Sustained Elimination of Iodine Deficiency. Our members in China, India and Europe are also involved in this global campaign to achieve, by 2005, the goal of the 1990 U.N. World Summit for Children to “virtually eliminate” iodine deficiency. Demonizing salt, of course, undermines confidence is achieving universal salt iodization in threatened areas – the consensus preferred strategy among endocrinologists and micronutrient nutritionists. We are chagrined by Dr. Appel’s statement that fortifying salt with iodine is akin to fortifying lard with a vital nutrient. Millions of adults and children around the world today suffer mental deficiencies caused by inadequate iodine intakes during their mothers’ pregnancies. There is far more health outcomes evidence of the benefit of fortifying salt with iodine than there is on reducing salt intakes for any health reason.
[About Salt Institute] [About salt] [About the salt industry] [News] [SI Member Business (password required] [E-Mail Salt Institute]