Commentary: Salt, blood pressure and health

The Yanomamo Indians still lead a life very similar to the last million or so years of our evolution,1 and like primitive man eat a diet that is very low in salt and saturated fat and high in fruit, vegetables and roots.2 The Yanomamo Indians are not overweight, do not smoke and are very fit. Their blood pressure does not rise with age although they spend much of their time fighting and are under great stress.3 This tribe does not develop vascular disease, although many die of infection. However, when they migrate to a Venezuelan or Brazilian town and adopt a western lifestyle, they, like native Americans, become overweight and develop diabetes and premature vascular disease. They appear therefore, to be a group which, though predisposed to vascular disease, is protected by the way they live. There are other similar examples4 which clearly indicate that cardiovascular disease (strokes, heart attacks and heart failure) could be entirely prevented if we changed our diet and lifestyle. Approximately half of the population in the UK dies of cardiovascular disease5 yet very little is done to prevent this by changing lifestyle or diet. Clearly, much greater efforts are needed to try to understand what simple changes could have beneficial effects. To relate dietary variables such as salt, saturated fat, fruit and vegetable consumption to outcome is not easy because at present the intake of salt, fat, fruit and vegetables varies considerably from day to day, as does the blood pressure. Unfortunately it is not possible to set up dietary outcome experiments in which babies are randomized at birth, or ideally at conception, into one group that eats less salt or less fat and more fruit and vegetables, while another group continues on a ‘normal’ diet for life. Such experiments are never going to be done, we therefore have to rely on evidence from epidemiology, migration, intervention, treatment, animal and genetic studies. In spite of the difficulties of quantifying salt intake in an individual because of the variability of western diets and the variability of the blood pressure, the evidence that dietary salt is related to blood pressure, both in the rise that occurs with age and the number of people whose blood pressure is raised, is stronger than all other dietary variables. Alderman has been one of the main proponents of the dangers of reducing salt intake. He encouraged the National Heart, Lung and Blood Institute (NHLBI) in the USA to set up a large workshop on sodium and blood pressure in January 1999 in order to critically review all of the current scientific evidence on salt. For reasons that are unclear Alderman did not attend the meeting which concluded that ‘a high sodium intake is associated with higher blood pressure levels, and other cardiovascular and non-vascular conditions continue to increase‘, and that ‘a population wide strategy of reducing salt in the food supply is an important public health strategy that can lower blood pressure among populations‘.5 Alderman's current overview of the subject is highly selective. He ignores over 40 careful studies in unacculturated human populations who at one time consumed, or continue to consume, less than 3 grams of salt per day.4 In these populations blood pressure did not and does not rise with age. Furthermore, there are other unacculturated populations which demonstrate that it is the low salt intake which is responsible for the lack of rise in blood pressure, not some other aspect of unacculturation.6,7 Alderman points out that Kuna Indians appear to be an exception8 to the general experience that a high salt intake is associated with a rise in blood pressure with age but he does not mention that in this study 24-h urinary sodium excretion was not measured. The estimation of salt intake relied in part on the recollection of how many teaspoons of salt each person had added to their food. One of the reasons the large InterSalt study9 was set up with careful attention to methodology, including 24-h urine collections, was precisely because of the difficulty of estimating salt intake from dietary history. Alderman also refers to an observational study in Italian nuns where the blood pressure did not rise with age10 but he does not quote the investigators' own conclusions. It appeared to them that in order to avoid the customary rise in blood pressure with age it is necessary to live in a stress free environment, characterized by total silence, continuous meditation and isolation from society. It is noticeable that although the first account of these nuns appeared 10 years ago the findings have never been confirmed elsewhere.11 The InterSalt study9 was set up to study a wide range of salt intakes across the world and measure blood pressure and other variables under well controlled conditions. By the time the study was completed there were only four communities who ate less than 3 grams of salt a day. The others were in a very narrow range of salt intake, approximately 7–12 grams of salt a day. The majority of these communities were eating a western diet in which salt intake varies so much from day to day that individual variations in sodium excretion may be considerable, e.g. more than five-fold.12 It follows that in many of these communities salt intake in any one individual might vary from day to day by the same amount as the variability between individuals. It is not surprising therefore that, in the InterSalt study, if one excludes the few communities who eat less than 3 grams of salt, and only considers those communities which consume between 7 and 12 grams of salt a day, there is no clear relationship with blood pressure. Alderman fails to point out that if all the communities are considered, there is a clear relationship between salt intake and blood pressure and the rise in blood pressure with age (Figure 1). The only intervention study that has successfully managed to reduce salt intake remains the Portuguese study where two villages were studied.13 One village was given information on how to reduce salt intake, particularly in relation to processed foods, and it was given processed foods with less salt. Salt intake was reduced by approximately 50%, as judged by 24-h urinary sodium. Over two years this resulted in substantial differences in blood pressure, compared to the control village where no reduction in salt intake was made (Figure 2). Perhaps it is understandable that Alderman all but ignores all the animal work that relates salt intake to blood pressure. He reduces it to one dismissive sentence. There are now numerous studies in the rat, dog, chicken, rabbit, baboon and chimpanzee, all of which have shown that when there is a prolonged increase in salt intake there is an increase in blood pressure.14 Furthermore in all forms of experimental hypertension, whatever the animal model, a high salt intake is essential for the blood pressure to rise. A recent study was carried out in chimpanzees,15 the nearest relative to humans (98.8% genetic-homology). The normal salt intake of a chimpanzee, which weighs up to 50 kg, is less than 0.5 gram a day. When it was increased to 15 grams a day the blood pressure rose slowly and the rise became significant after one year when it was still rising. Blood pressure returned to normal when the salt intake was reduced (Figure 3). These results show that if the animal species most closely related to humans, which normally consumes a diet as low in salt as the one it (and the human race) is genetically programmed to eat, increases its salt intake into the same range as that of present day humans, they, like humans, develop hypertension. Alderman now concedes that when salt intake is reduced there is a fall in blood pressure in both normotensive and hypertensive humans. He even concedes that a fall of ‘a few millimetres of mercury if sustained, assuming the method of its achievement induces no harm, could produce more reductions in morbidity and mortality than is currently achieved by treating high blood pressure’. His ensuing comment that this ‘possibility energises advocates of sodium restriction’ presumably indicates that nevertheless Alderman wishes to distance himself from these findings. It could be put forward that this is a reasonable uncritical reaction to meta-analyses which have included studies of extremely large changes in salt intake over periods of less than one week.16 Such manoeuvres have been known to stimulate the sympathetic nervous system.17 The inclusion of such short term studies, particularly in normotensives, is inappropriate when the recommendations for public health are for a modest reduction in salt intake from 10 grams to 5 grams a day over a lifetime, not a few days. A more recent meta-analysis that only included studies of modest and longer term reductions in salt intake, showed that the fall in blood pressure in normotensives was greater18 than in the previous meta-analyses. In the most rigorous trial (DASH Sodium Study)19 in which there were several hundred participants, and the daily sodium intake was well controlled, the fall in pressure was even greater (Figure 4). There is evidence that the full effect of salt restriction may not be seen within a month so that with longer term reductions in salt intake there may well be greater falls in blood pressure. In other words, there is every indication that the recommendations of a modest reduction of salt intake from 10 grams to 5 grams a day over a prolonged period of time has a pronounced effect on blood pressure, not only in the hypertensives but also in the normotensive population. When discussing the other effects of ‘sodium restriction‘, Alderman forgets to mention that in addition to, and independent of, raising blood pressure, a high salt intake increases the mass of the left ventricular wall,20 stiffens conduit arteries21 and thickens and narrows resistance arteries,22 including the coronary23 and renal arteries.24 A high salt intake is also directly related to the number of strokes,25 severity of cardiac failure,26 adhesiveness of platelets,27,28 carcinoma of the stomach29 and, to bone demineralization.30 Alderman misunderstands the relationship of sodium intake to vascular compliance; he conveniently reverses it—what has been found is that a reduction in sodium intake increases vascular compliance.31 He rightly but ominously points out that salt intake is well documented to relate closely to the renin angiotensin system. It is true that the renin angiotensin aldosterone system is one of the major compensatory mechanisms that maintains blood pressure and reduces the excretion of sodium, and thus when extracellular volume is reduced with diarrhoea or diuretics there is a rise in plasma renin. This is an entirely normal physiological response, and to try to intimate that it is abnormal, and harmful, for plasma renin to rise in response to a reduction in extracellular volume, is extraordinary. The Yanomamo Indians, who still live an evolutionary form of life and have a very low salt intake (<0.5 g per day), have the highest renin and aldosterone levels measured. But they do not develop vascular disease. According to Alderman's hypothesis, such high levels of angiotensin and aldosterone should precipitate an acceleration of vascular disease at an early age. Animal experiments, however, have shown that it is the appropriateness of the level of renin and aldosterone to the blood volume or extracellular volume that may cause damage, not the absolute level of plasma renin activity angiotensin II or aldosterone.32,33 There are experiments in animals that demonstrate that salt loading per se has identical effects to that of giving aldosterone or angiotensin II.22,34,35 In other words, the deleterious effect of aldosterone on cardiac fibrosis, etc, appears to be due to the associated retention of an excess amount of salt and water. In view of Alderman's belief in the dangers of an increase in renin and aldosterone, it is surprising that he supports the use of diuretics for the treatment of hypertension.36 Diuretics have been shown to be very effective in reducing cardiovascular disease, i.e. strokes, heart failure and coronary artery disease, particularly in the elderly, but they cause increases in plasma renin activity and aldosterone which, according to Alderman, should be harmful. A modest reduction in salt intake lowers blood pressure in an identical way to diuretics and increases both renin and aldosterone to the same levels. Alderman needs to explain why he considers diuretics to be so beneficial, and yet according to him salt restriction is not. Whilst it is true that non-acculturated societies have shorter life spans, this is not due to an increase in the incidence of cardiovascular disease, which they almost totally avoid, but to the greater risk of infection, particularly when exposed to western populations. It is true that the Japanese are an example of a society that has a longer life expectancy than many of those in the west. As the Japanese have a high salt intake, Alderman suggests that this illustrates that salt intake is not related to life expectancy. But the Japanese did not, and to a large extent still do not, eat much fat and they have a low plasma cholesterol37 which protects them from the development of vascular disease and atherosclerosis. Alderman fails to point out that the major causes of death in Japan are cerebral haemorrhage38 and cancer of the stomach,29 both of which are due to the high salt intake. Indeed, he appears to be unaware that in the 1960s there was a government campaign in Japan to reduce salt intake. It was successful in lowering blood pressure and causing large reductions in stroke mortality.39 It is very likely that if the Japanese were to reduce their salt intake further, whilst continuing to eat a minimum amount of saturated fat, they would live even longer than they do now. Alderman fails to mention the North Karelia project in northern Finland where the incidence of cardiovascular disease was very high. In co-ordination with the food industry, a government-backed campaign has been successful in reducing salt and fat intake, increasing fruit and vegetable consumption and cutting cigarette smoking. This has resulted in a reduction in blood pressure and fat intake and significant falls in stroke and coronary heart disease mortality.40,41 Three-quarters of the fall in coronary heart disease and two-thirds of the fall in stroke mortality were due to the change in risk factors. This study clearly indicates that if we were prepared to change our lifestyle, particularly our diet, major reductions in cardiovascular mortality could ensue. In a final attempt to obfuscate, Alderman once more revives the myth42 that a moderate reduction in salt intake is harmful. He again quotes completely inappropriate experiments in which the sodium intake of rats was so low that it stunted growth.43 The intake of sodium was lower than occurs naturally in mammals eating only fruit and vegetables. Therefore when discussing the possible effects of reducing salt intake from approximately 10 grams to 5 grams a day such experiments are irrelevant. For instance, in comparative terms, a reduction in salt intake to 5 grams a day, which has been suggested for humans at present, is about a thousand times less of a reduction than that used in the animal studies he is quoting. Alderman's other attempts to suggest that a low salt intake is dangerous in humans are based on two of his own publications in which a claim is made that there is a relationship between habitual salt intake and myocardial infarction.44,45 The data on sodium excretion in the first and of sodium intake in the are so however, that the claim is In the first study, 24-h urine were made in hypertensives after 5 of salt in order to stimulate the system so that the could be into normal and low renin these 24-h urinary sodium excretion no relationship to the habitual salt intake. Alderman continues to claim that the urine was when the were on their habitual salt intake. In addition to this on the of the diet the were to be Alderman and his of the dietary sodium that the were which was based on a 24-h urine was also The were into four on the 24-h urinary sodium It was noticeable that the sodium excretion which suggested that there had been with urine But as excretion had not been given it was not possible to be however, Alderman the data on This should be and by but it rose with the which confirmed that there had been with urine This was by the urine volume and excretion which also In other words, because of urine collections, many and their results had been into sodium excretion and not into true sodium intake One can therefore that the claim that an habitual low salt intake in hypertensive is related to myocardial be In Alderman's a 24-h dietary on intake was used to the habitual salt intake of sodium and in more than 10 between and The were years The of sodium intake, which was based on dietary no account of salt intake, i.e. salt added by an individual at the or in his own which, would have for approximately half of salt intake. The of salt intake in this study is therefore A simple of the results of the also that these are For instance, out that in the of daily salt intake in both and and grams intake was lower than the daily dietary and that they it that on such a diet there were so many They it should have even Alderman, that on this low intake were on than those in the higher sodium who were eating as many The of habitual salt intake which resulted from urine in the first and the data on sodium and intake in the any claim that a low sodium intake is related to myocardial evidence from Finland that there is a clear relationship between mortality and salt intake (Figure Unfortunately for Alderman's it that the higher the salt intake the greater the Salt was only by humans about years ago when it was found to have the of It therefore became of great in the development of communities and however, with the development of the and salt is no longer for with the development of processed foods, salt has once again become of great not only to the salt and but also to the food and of the processed foods are only with the addition of large of salt, the When people are exposed to foods which high of salt the salt are The individual therefore to this of which increases the for highly processed Salt is also very important to the processed industry, for a higher salt increases the In this way the of the can be by 10 to at no to the salt intake is an important to and therefore reduction in the salt intake have a large effect on and It is not therefore, that which the salt and e.g. the Salt Institute in the and the industry, with many of the food industry, have in the that salt is not in hypertension. They have also suggested that dietary salt only effects a number of people and that therefore it is not for the normotensive population to reduce its salt intake. They also the that reducing salt intake can be Alderman has as a of the for the Salt in the which the salt and not only in the but The Institute his in which the claim was made that salt intake was related to myocardial as a the same time the Institute put out a which that this study that hypertensives low sodium diets had increased of heart This evidence the main of the Salt to the to try to change the for low salt Alderman was one of the of this When all of the evidence is from migration, intervention, treatment genetic studies in humans and animal studies that relates salt intake to blood pressure and other the evidence is very It is stronger than evidence for other dietary variables that are also important in cardiovascular disease, e.g. saturated fat intake and fruit and vegetable all government and who have the evidence have a reduction in salt intake from an consumption of grams to grams per day. This was made in the in and in the UK in salt intake is increasing due to the much greater consumption of processed and prepared foods which are very high in salt, to that of In the UK there is a strategy to i.e. 10 to reductions in the salt of all processed These reductions not be by the and as the salt used to the lower reductions can be made in four or years In this way salt intake could be reduced the population the to reduce their salt intake However, the reduction would be greater if added less salt to the food both in and at the In this way it should be possible to reduce salt intake in the world by half over the all the dietary to reduce the intake of salt, this is the to the food is prepared to In the by manoeuvres which those by the industry, such are by the Salt and the and an the the of which are by the and other food The relation of salt excretion to the of the rise in blood pressure with age in in the from Blood pressure changes with time in two Portuguese one of which was on how to reduce salt intake and given processed foods with a reduced salt the other had similar of blood pressure but no on diet. the significant differences in blood pressure at year and continuing differences at from Blood pressure in who on their diet sodium per or were given an increased salt intake sodium per the of the study, the salt were and blood pressure to that of the control from in blood pressure and urinary sodium excretion with the reduction in salt intake in all on the normal diet control and on diet. from The for coronary heart disease cardiovascular disease and mortality associated with a increase in salt intake as judged by urinary sodium from The of salt in processed food