sábado, 22 de noviembre de 2008


John D. Brunzell, M.D.

A healthy 45-year-old man is found on routine screening to have hypertriglyceridemia. He is a nonsmoker, has a reasonable diet, consumes one alcoholic drink per week, and exercises regularly. He takes no medications. His father died at the age of 55 years in an automobile accident; his mother is healthy at 67 years of age, and he has two healthy older brothers. His blood pressure is normal, his body-mass index (the weight in kilograms divided by the square of the height in meters) is 28, and his waist circumference is 96 cm. His fasting triglyceride level is 400 mg per deciliter, total cholesterol 230 mg deciliter, low-density lipoprotein (LDL) cholesterol 120 mg per deciliter, and high-density lipoprotein (HDL) cholesterol 30 mg per deciliter. His fasting blood glucose level is 90 mg per deciliter. Thyroid and renal function are normal. How should his case be assessed and managed?

The Clinical Problem

Hypertriglyceridemia is a common form of dyslipidemia that is frequently associated with premature coronary artery disease, which is generally defined by the occurrence of a myocardial infarction or the need for a coronary-artery procedure before 55 years of age for men and 65 years of age for women. These upper age limits might increase by 5 to 10 years in nonsmoking men and women. Whether hypertriglyceridemia causes coronary artery disease or is a marker for other lipoprotein abnormalities that cause premature coronary artery disease remains controversial. Specifically, hypertriglyceridemia correlates strongly with the presence of small, dense particles of LDL cholesterol and reductions in the HDL2 component of HDL cholesterol, both of which are known to be associated with premature coronary artery disease. Hypertriglyceridemia has been shown to predict coronary artery disease after adjustment for many traditional risk factors, but not after adjustment for LDL or HDL cholesterol subfractions.

Several common genetic disorders of hypertriglyceridemia cause premature coronary artery disease. These disorders include familial combined hyperlipidemia, the residual dyslipidemia in patients with well-controlled type 2 diabetes mellitus, and familial hypoalphalipoproteinemia. Each of these disorders shares features of the metabolic syndrome. Disorders associated with premature coronary artery disease are common: familial combined hyperlipidemia and familial hypoalphalipoproteinemia each affect about 1% of the general population, and type 2 diabetes affects more than 5%. Together, these three disorders have been purported to account for up to 50% of premature coronary artery disease events. These diagnoses therefore warrant aggressive interventions to reduce the cardiovascular risk. In contrast, one other inherited form of hypertriglyceridemia — monogenic familial hypertriglyceridemia — is not associated with premature coronary artery disease; this disorder also affects up to 1% of the population and must be distinguished from the other disorders in making treatment decisions.

Obesity (in particular, central obesity) is associated with increased triglyceride levels and decreased HDL cholesterol levels. Central obesity paired with insulin resistance is probably a major factor contributing to the dyslipidemia associated with type 2 diabetes, familial combined hyperlipidemia, and familial hypoalphalipoproteinemia. The metabolic syndrome is associated with an increased risk of coronary artery disease even among people who do not have diabetes, but it is unclear whether this is the case when familial combined hyperlipidemia and familial hypoalphalipoproteinemia are absent.

Other abnormalities leading to secondary hypertriglyceridemia include untreated or uncontrolled diabetes, treatment with several medications (Table 1), and alcohol consumption. More complex forms of secondary hypertriglyceridemia develop with hypothyroidism, end-stage renal disease, the nephrotic syndrome, and human immunodeficiency virus infection; these disorders are not covered in this article. Patients with triglyceride levels above 2000 mg per deciliter (22.6 mmol per liter) almost always have both a secondary and a genetic form of hypertriglyceridemia.

Tabla 1.

Strategies and Evidence


The evaluation of patients with hypertriglyceridemia should initially focus on whether there is a family history of the condition or a personal or family history of premature coronary artery disease; in addition, potential secondary causes (e.g., medications or untreated diabetes) should be identified. The presence of premature coronary artery disease in a first-degree relative (parent or sibling) or in a sibling of a parent suggests familial combined hyperlipidemia or familial hypoalphalipoproteinemia and indicates the need to consider drug therapy.

Xanthomas are usually not present in mild-to-moderate hypertriglyceridemia; when present, they do not help distinguish the various hypertriglyceridemic disorders. The body-mass index should be calculated, and waist circumference measured. The combination of an elevated triglyceride level and a large waist circumference may be a better marker of insulin resistance and the risk of coronary disease than hypertriglyceridemia alone; the presence of a large waist circumference (defined in European Americans as a value greater than 40 in. [101.6 cm] for men and 35 in. [88.9 cm] for women) may help to distinguish familial hypertriglyceridemia (which is not associated with central adiposity or an increased cardiovascular risk) from familial combined hyperlipidemia or familial hypoalphalipoproteinemia.

Usually, a fasting lipid profile is the only laboratory work needed to evaluate lipids in a patient with elevated triglyceride levels and premature coronary artery disease. Although nonfasting triglyceride levels have recently been associated with coronary heart disease, the measurement of nonfasting triglyceride levels is not currently recommended because no standard values have been developed and because most of the variation in postprandial triglyceride levels is determined by the fasting level. Patients with familial combined hyperlipidemia may have elevated or normal LDL cholesterol levels. HDL cholesterol levels are reduced in both familial combined hyperlipidemia and familial hypoalphalipoproteinemia.

Small, dense LDL particles are also present in both familial combined hyperlipidemia and familial hypoalphalipoproteinemia and, as noted above, have been associated with premature coronary artery disease. In patients with elevated triglyceride levels in the absence of a personal or family history of clinical atherosclerosis, the measurement of apolipoprotein B levels may help to distinguish familial combined hyperlipidemia from familial hypertriglyceridemia. In both disorders, the level of apolipoprotein B can be used to estimate the total number of LDL particles (large and small). Apolipoprotein B levels are higher in familial combined hyperlipidemia and lower in familial hypertriglycieridemia. The nomograms for apolipoprotein B (Figure 1), developed from the third National Health and Nutrition Examination Survey, can be used to define elevated apolipoprotein B as an age- and sex-adjusted value above the 90th percentile. The current National Cholesterol Education Program recommends the measurement of non-HDL cholesterol (which consists of total cholesterol minus HDL cholesterol and includes triglyceride-rich lipoprotein cholesterol) instead of apolipoprotein B. Even though non-HDL cholesterol is a better predictor of cardiovascular risk than LDL cholesterol, several studies have suggested that apolipoprotein B may be an even better predictor than non-HDL cholesterol. Although data from an international study indicate that the ratio of apolipoprotein B to apolipoprotein A-I is highly predictive of early coronary artery disease, the added value of apolipoprotein A-I measurements in refining risk estimates is unclear; consequently, apolipoprotein A-I levels are not routinely measured in clinical practice.

Fig 1.

Likewise, measurement of the size or density of LDL particles is currently considered a research tool and is not recommended in routine care. Measurement of Lp(a) lipoprotein levels does not help to distinguish forms of hypertriglyceridemia, but it may be useful in assessing the relative risk of atherosclerosis among patients who have hypertriglyceridemia in combination with other lipid or nonlipid cardiovascular risk factors. However, data that support the routine measurement of Lp(a) lipoprotein levels in patients with normal lipid levels are lacking. Similarly, there are no data to support routine assessment for subclinical vascular disease (by means of coronary calcium scanning or other types of imaging) in patients with asymptomatic hypertriglyceridemia.


After treatment of secondary disorders and removal of offending medications, lifestyle modification and drug treatment should be considered in a patient with hypertriglyceridemia who is considered to be at risk for premature coronary artery disease (Figure 2). A major question in management is whether therapy should be directed solely toward reduction of triglyceride levels or toward the modification of associated abnormalities of intermediate-density lipoprotein, LDL, and HDL cholesterol. Triglyceride levels greater than 1000 to 1500 mg per deciliter (11.3 to 16.9 mmol per liter) require treatment with fibrates to reduce the risk of pancreatitis. The benefit of treating mild-to-moderate elevations in triglyceride levels is less clear

Fig 2.

Lifestyle Modification

Weight loss results in a mild-to-moderate decrease in triglyceride levels (about 22%) and an increase in HDL cholesterol (about 9%), largely because of an increase in HDL2 cholesterol (about 43%). The level of small, dense LDL particles may decrease by as much as 40%. Although losing large amounts of weight and maintaining that weight loss are difficult, even moderate weight loss may result in reductions in triglyceride levels and may be maintained with regular aerobic exercise.

Aerobic exercise of moderate intensity with high frequency (about 4 hours per week) has been associated with maintenance of improved cardiorespiratory fitness. It has also been associated with a decrease in intraabdominal fat, an increase in HDL cholesterol levels if those levels were low, and a small decrease in triglyceride levels.

Although a decrease in dietary fat can lead to weight loss (and associated reductions in triglyceride levels), diets that are low in fat but high in carbohydrates may result in reductions in both LDL and HDL cholesterol. Because replacing saturated fat with monounsaturated fat leads to a smaller decrease in HDL cholesterol than replacing saturated fat with carbohydrates, a reasonable approach is to reduce foods rich in saturated fat and replace them with complex carbohydrates and monounsaturated and polyunsaturated fats. It is also advisable to avoid simple sugars, particularly fructose, which has been associated with postprandial hypertriglyceridemia. Fructose is present in many carbonated beverages and fruit-juice mixes, and high-fructose corn syrup is added to many prepared foods as a preservative and sweetener. For patients with exercise limitations, the combination of a diet low in saturated fat and a regimen of walking on a daily basis is a lifestyle change that can usually be maintained.

Supplementation with n–3 fatty acids may lower triglyceride levels and, according to some data, reduce cardiovascular events. However, a meta-analysis of trials did not show a significant reduction in cardiovascular events or mortality with dietary or pharmacologic n–3 fatty acid supplementation.

Cigarette smoking is associated with an earlier occurrence of coronary artery disease, by about 10 years. Discontinuation of smoking is associated with improvement in lipid levels despite the weight gain that often follows cessation.

Alcohol intake is associated with a reduced risk of atherosclerotic cardiovascular disease but also leads to an increase in blood pressure and in the risk of hemorrhagic stroke. Modest alcohol intake (two drinks per day for men and one drink per day for women) is considered acceptable in people who do not have a predisposition for alcohol abuse. Patients with severe hypertriglyceridemia (triglyceride levels above 2000 mg per deciliter) associated with alcohol use should abstain. The limited effect in patients with triglyceride levels below 500 mg per deciliter (5.6 mmol per liter) should not preclude moderate alcohol intake.


As noted above, no set targets for triglyceride levels are clearly warranted other than to reduce the risk of pancreatitis. Generally, medications are considered in patients with hypertriglyceridemia who have a personal or family history of premature coronary disease. When medications are used, those that specifically decrease the level of small, dense LDL particles and raise the level of HDL2 particles are preferred (Table 2).

Tabla 2.

In patients with, or at risk for, premature coronary artery disease, statins are generally considered the first drug of choice to lower LDL cholesterol. Nicotinic acid therapy, often combined with a statin, may be an alternative first choice in patients at risk for premature coronary artery disease. Nicotinic acid can reduce the level of small, dense LDL particles and raise the level of HDL2 particles (Table 2). In the Coronary Drug Project, nicotinic acid resulted in a 15% reduction in the risk of myocardial infarction among men with hypercholesterolemia who had atherosclerosis and decreased total mortality by 10% at 15 years. In the Investigation of the Treatment Effects of Reducing Cholesterol (ARBITER) 2 trial, it was also shown to prevent the progression of carotid artery disease in patients with atherosclerosis who were already receiving statin therapy.

Nicotinic acid in combination with other drugs has been shown to be effective in reducing the progression of atherosclerosis in patients with hypertriglyceridemia who are at risk for premature coronary artery disease. For example, in a randomized study involving patients with atherosclerosis, low HDL cholesterol levels, and borderline-high triglyceride levels, the combination of nicotinic acid and simvastatin was associated with a slight regression of coronary stenoses, whereas placebo or antioxidant vitamins were associated with progression. However, neither medication was studied alone, and there were few clinical end points. In another study, involving men with elevated apolipoprotein B levels, nicotinic acid in combination with colestipol reduced the frequency of progression of atherosclerosis as compared with placebo. Nicotinic acid is available in crystalline form and extended-release form. Flushing may be a bothersome side effect, but its frequency may be minimized by education about use.

Fibrates also lower triglyceride levels, but the results of randomized trials have been equivocal in terms of major outcomes, generally showing decreases in the rates of nonfatal myocardial infarction but not in the rates of fatal coronary events or total mortality. In the Veterans Affairs High-Density Lipoprotein Cholesterol Intervention Trial, the use of gemfibrozil resulted in a significant decrease in the primary outcome of coronary heart disease and nonfatal myocardial infarction, but not in the secondary outcome of fatal myocardial infarction or death from any cause. In a World Health Organization trial, the use of clofibrate in men with hypercholesterolemia resulted in a reduction in the rate of nonfatal myocardial infarction, but there was no decrease in the rate of fatal myocardial infarction or total mortality. Such findings raise questions about the use of a fibrate as a first-line drug for mild-to-moderate hypertriglyceridemia. Gastrointestinal side effects were also common with this therapy; similar findings were reported in another randomized trial of gemfibrozil.

Treatment in Patients with Diabetes

Use of statin therapy to lower LDL cholesterol levels is recommended for adults with type 2 diabetes mellitus who are considered to be at increased risk for coronary artery disease according to the criteria of the American Diabetes Association and the National Cholesterol Education Program. In patients with type 2 diabetes, the combination of statins and fibrates has often been used to treat hypertriglyceridemia. However, in the Fenofibrate Intervention and Event Lowering in Diabetes Trial, use of fenofibrate did not result in a significant reduction in the primary outcome — nonfatal myocardial infarction or fatal coronary heart disease — in patients with type 2 diabetes. A reduction in the rate of nonfatal myocardial infarction was offset by a slight increase in the rate of fatal myocardial infarction.

It has been suggested that nicotinic acid, which may interfere with glucose control, not be used as a first-line drug for the treatment of hypertriglyceridemia in patients with diabetes. However, several trials have demonstrated that nicotinic acid therapy can be used in patients whose diabetes is well controlled, with little effect on glucose levels.

Areas of Uncertainty

Without knowledge of the family history, it may be challenging to differentiate patients who are at increased risk for premature coronary artery disease from those who are not. Apolipoprotein B levels, and often LDL cholesterol levels, tend to be higher in familial combined hyperlipidemia than in familial hypoalphalipoproteinemia or familial hypertriglyceridemia, and levels of small, dense LDL particles tend to be lower in familial hypertriglyceridemia than in the other two conditions, but there is considerable overlap among all three. Until the basic biochemical defects for each of the genetic forms of hypertriglyceridemia are defined, decisions about the use of drugs to prevent premature coronary artery disease must be based on the presence or absence of a family history of premature atherosclerosis and dyslipidemia. Once premature coronary artery disease has developed, secondary prevention with lipid-lowering therapy is indicated, and there is no need to differentiate among the types of hypertriglyceridemia. The optimal pharmacologic approach in patients with premature coronary artery disease remains uncertain, including whether it is preferable to start with a statin or to begin with nicotinic acid and add a statin as needed. The role of fibrates also remains unclear.


The National Cholesterol Education Program has specific recommendations for target LDL cholesterol levels, but not for target triglyceride levels. Treatment with fibrates is recommended for patients with triglyceride levels over 1000 mg per deciliter in order to decrease the risk of triglyceride-induced pancreatitis. After the target level for LDL cholesterol has been reached, the program recommends lowering triglyceride levels if they are above 200 mg per deciliter (2.6 mmol per liter), although there are no data to support this recommendation.

Conclusions and Recommendations

The patient described in the vignette has an elevated triglyceride level and a low HDL cholesterol level, but his LDL cholesterol is not elevated. Neither the increased triglyceride level nor the low HDL cholesterol level helps to determine whether he has familial hypertriglyceridemia, familial combined hyperlipidemia, or familial hypoalphalipoproteinemia, which makes it difficult to assess the associated risk of premature coronary artery disease. There is no apparent secondary cause of his hypertriglyceridemia, and he has no other apparent risk factors for premature coronary artery disease. A first step in evaluating patients with hypertriglyceridemia is to obtain an extensive family history, sometimes having the patient do homework to establish the presence of atherosclerosis in one or more family members, the age at onset, and, if an affected first-degree relative has died, the age at and cause of death; a family history will often identify other relatives who might need lipid-lowering therapy. A family history of premature coronary artery disease would suggest familial combined hyperlipidemia or familial hypoalphalipoproteinemia.

If a patient has many adult relatives with hypertriglyceridemia but without clinical evidence of atherosclerosis, successful treatment of the hypertriglyceridemia and low HDL cholesterol level might be accomplished with lifestyle modifications alone, including a reduced-calorie diet that is low in saturated fat and a program of regular aerobic exercise. If the patient has or is at apparent risk for premature coronary artery disease on the basis of the family history and appears to have familial combined hyperlipidemia or familial hypoalphalipoproteinemia, pharmacologic therapy should be considered in addition to lifestyle modification. Although the optimal therapy is uncertain, in such cases I would favor combined treatment with nicotinic acid and a statin.

Transforming Growth Factor β Signaling, Vascular Remodeling, and Hypertension

Phyllis August, M.D., M.P.H., and Manikkam Suthanthiran, M.D.

The new england journal o f medicine

Hypertension affects more than 1 billion adults worldwide and is the leading cause of preventable death. Despite the plethora of drugs available, the disease is adequately controlled in only about one third of patients. Hormones and vasoactive peptides, like lead actors in a drama, have garnered much attention in the context of the pathogenesis of hypertension. A recent study by Zacchigna et al. has now brought to center stage a multifunctional cytokine, transforming growth factor β (TGF-β).

TGF-β is a ubiquitously expressed member of a superfamily of proteins critical to developmental processes. It regulates cell growth and differentiation, inflammation, and immunity. One mechanism by which its activity is regulated is the enzymatic cleavage of its inactive precursor, pro–TGF-β, by an enzyme called furin (Fig. 1).

TGF-β is linked to human disease primarily as a promoter of fibrosis; there is extensive evidence that TGF-β is overexpressed in chronic kidney disease. TGF-β has also been implicated in restenosis after angioplasty, atherosclerosis, angiogenesis, and cardiac fibrosis. The evidence associating it with hypertension has steadily been accruing: our group reported a correlation between serum TGF-β levels and blood pressure, and others have shown an association between polymorphisms in the TGF-β gene and hypertension.

Zacchigna et al. showed that an extracellularmatrix molecule associated with blood vessels, Emilin1, regulates the maturation of TGF-β. The observation that Emilin1 is highly expressed in the mouse cardiovascular system during development prompted the authors to investigate cardiovascular structure and function in mice with a deficiency of the molecule (Emilin1/− mice). Although the mice were morphologically normal, their blood pressures were significantly greater than those of wild-type mice. The cardiac output, vascular contractility, and stiffness of the blood vessels were similar in the two groups of mice, but blood-vessel diameter was smaller and the rate of proliferation of vascular smooth-muscle cells was lower in the Emilin1/− mice. These observations suggest that reduced blood-vessel diameter, resulting in increased peripheral vascular resistance, is the cause of hypertension in these animals.

Where does TGF-β fit into this model of hypertension? Using Xenopus laevis embryos and mammalian cell–culture techniques, Zacchigna et al. showed that Emilin1 inhibits TGF-β signaling by binding to the pro–TGF-β precursor in the extracellular space, thereby preventing its maturation into TGF-β (Fig. 1). Consistent with this mechanism is the observation that TGF-β signaling in the vascular tissue is greater in Emilin1/− mice than in wild-type mice. Furthermore, the vascular abnormalities associated with the deficiency of Emilin1 were not observed in mice that both lacked Emilin1 and had half the normal level of TGF-β. This finding provides support for the notion that a deficiency of Emilin1 increases the blood pressure by increasing the availability of TGF-β.

These experiments offer a new direction for the investigation of the mechanisms of hypertension in humans and, ultimately, for treatment, although some gaps need to be filled. Zacchigna et al. concluded that an excess of TGF-β decreases the blood-vessel diameter, which then causes increased peripheral vascular resistance and hypertension in Emilin1/− mice. However, it is possiblethat this decreased diameter represents vascular remodeling and is the consequence, rather than the cause, of hypertension. The development of a conditional mouse knockout model that represses the production of TGF-β late in life might help address this possibility. Other models of hypertension are linked to an excess of TGF-β, suggesting that TGF-β mediates damage to the renal parenchyma or vasculature, thereby increasing blood pressure. In some of these models, theantibody-mediated blockade of TGF-β has been shown to reduce blood pressure.

Can drugs that lower TGF-β levels be used to treat essential hypertension? Inhibitors of angiotensin-converting enzyme and antagonists of the angiotensin receptor have been shown to reduce TGF-β production by decreasing the levels of angiotensin II (which stimulates TGF-β); whether this reduction contributes to the antihypertensive effects of the inhibitors remains to be determined. New strategies to modulate the fibrogenic effects of TGF-β are being pursued in the hope of preventing the progression of renal disease. Clearly, the hemodynamic effects of such

strategies should be studied, as should the possibility that the maturation of TGF-β could be modulated by compounds with effects similar to those of Emilin1.

The role of the kidney must be considered in the context of this new finding by Zacchigna et al. Increased peripheral vascular resistance alone, without concomitant changes in sodium excretion,will not lead to sustained hypertension. Thus, we must assume that the reduction in blood-vessel diameter that causes hypertension in the Emilin1/− mouse also causes the constriction of renal vessels, resulting in increased vascular resistance in the kidney. Such increased resistance would shift the relation between sodium excretion and arterial pressure, so that for any given arterial pressure, less sodium would be excreted and hypertension would be sustained. Given the fibrogenic effect of increased TGF-β levels on the kidney, the study of the kidneys in Emilin1/− mice seems critical to a complete understanding of the relation of TGF-β to blood pressure.

The study by Zacchigna et al. underscores the power of mechanistic studies to address complex traits such as hypertension. It has opened the door to the exploration of TGF-β and related peptidesfor the individualized management of hypertension. A word of caution is in order, however, whenone is considering approaches to suppress the production of TGF-β: a deficiency of TGF-β may have adverse consequences, such as unmitigated inflammation.

Figura 1.

Propiedades Básicas de las Lipoproteínas


Sodium and Potassium in the Pathogenesis

The New England journal of Medicine

Hypertension affects approximately 25% of the adult population worldwide, and its prevalence is predicted to increase by 60% by 2025, when a total of 1.56 billion people may be affected. It is the major risk factor for cardiovascular disease and is responsible for most deaths worldwide.

Primary hypertension, also known as essential or idiopathic hypertension, accounts for as many as 95% of all cases of hypertension.Primary hypertension results from the interplay of internal derangements (primarily in the kidney) and the external environment. Sodium, the main extracelular cation, has long been considered the pivotal environmental factor in the disorder. Numerous studies show an adverse effect of a surfeit of sodium on arterial pressure. By contrast, potassium, the main intracellular cation, has usually been viewed as a minor factor in the pathogenesis of hypertension. However, abundant evidence indicates that a potassium deficit has a critical role in hypertension and its cardiovascular sequelae. In this review, we examine how the interdependency of sodium and potassium influences blood pressure. Recent evidence as well as classic studies point to the interaction of sodium and potassium, as compared with an isolated surfeit of sodium or deficit of potassium, as the dominant environmental factor in the pathogenesis of primary hypertension and its associated cardiovascular risk. Our review concludes with recent recommendations from the Institute of Medicine concerning the dietary intake of sodium and potassium.

Dietary Sodium and Hypertension

Primary hypertension and age-related increases in blood pressure are virtually absent in populations in which individual consumption of sodium chloride is less than 50 mmol per day; these conditions are observed mainly in populations in which people consume more than 100 mmol of sodium chloride per day. The International Study of Salt and Blood Pressure (INTERSALT), which included 10,079 subjects from 32 countries, showed a median urinary sodium excretion value of 170 mmol per day (approximately 9.9 g of sodium chloride per day). Although individual sodium intake in most populations throughout the world exceeds 100 mmol per day, most people remain normotensive. It appears, then, that sodium intake that exceeds 50 to 100 mmol per day is necessary but not sufficient for the development of primary hypertension.

In an analysis across populations, the INTERSALT researchers estimated an increase in blood pressure with age over a 30-year period (e.g., from 25 to 55 years of age); mean systolic blood pressure was 5 mm Hg higher and diastolic blood pressure was 3 mm Hg higher when sodium intake was increased by 50 mmol per day. In an analysis within single populations, a positive correlation between sodium intake and blood pressure was also detected after adjustment for a number of potentially confounding variables.

Humans share 98.4% genetic identity with chimpanzees, and a landmark interventional study in chimpanzees showed that adding up to 15 g of sodium chloride to the diet per day increased systolic blood pressure by 33 mm Hg and diastolic blood pressure by 10 mm Hg; the increases were reversed after withdrawal of the sodium chloride supplement. In the Dietary Approaches to Stop Hypertension (DASH) sodium study, a reduction in sodium intake caused stepwise decreases in blood pressure. Levels of sodium intake studied in random order were approximately 150 mmol per day, 100 mmol per day, and 50 mmol per day.A meta-analysis of randomized controlled trials lasting at least 4 weeks concluded that reducing sodium intake by 50 mmol per day decreases systolic blood pressure by an average of 4.0 mm Hg and diastolic blood pressure by an average of 2.5 mm Hg in hypertensive subjects and decreases systolic blood pressure by an average of 2.0 mm Hg and diastolic blood pressure by an average of 1.0 mm Hg in normotensive subjects.

Potassium Content of Sodium-Rich Diets

As compared with diets based on natural foods, diets based on processed foods are high in sodium and low in potassium.For example, two slices of ham (57 g) contain 32.0 mmol of sodium and 4.0 mmol of potassium, and a cup of canned chicken noodle soup contains 48.0 mmol of sodium and 1.4 mmol of potassium. Conversely, diets containing abundant fruits and vegetables are sodium-poor and potassium-rich. For example, an orange (131 g) contains no sodium and 6.0 mmol of potassium, and a cup of boiled peas contains 0.3 mmol of sodium and 9.8 mmol of potassium. Isolated populations that eat natural foods have an individual potassium intake that exceeds 150 mmol per day and a sodium intake of only 20 to 40 mmol per day (the ratio of dietaryn potassium to sodium is >3 and usually closer to 10).By contrast, people in industrialized nations eat many processed foods and thereby ingest 30 to 70 mmol of potassium per day and as much as 100 to 400 mmol of sodium per day (the usual dietary potassium:sodium ratio is <0.4).

Hypertension affects less than 1% of people in isolated societies but approximately one third of adults in industrialized countries. Differences in the prevalence of hypertension among these populations have usually been attributed to differences in the amounts of dietary sodium consumed, but they could also reflect differences in potassium intake. The movement of isolated populations into more urban areas is consistently associated with age-related increases in blood pressure and a rise in the prevalence of hypertension as the dietary potassium:sodium ratio decreases in the new location.

Vascular Effects of Potassium Depletion

Early reports of the vasodilatory or blood-pressure– lowering properties of both potassium depletion and potassium supplementation delayed recognition of the effects of potassium depletion that are toxic to the blood vessels. These studies of the effects of a low intake of potassium on blood pressure, performed mostly in young rats, also involved a low intake of sodium and chloride. Potassium restriction causes a deficit in cellular potassium that triggers cells to gain sodium in order to maintain their tonicity and volume.The deficits of potassium, sodium, and chloride in the body imposed by those early studies contracted both the intracellular and extracelular compartments, thereby engendering a decrease in blood pressure. Subsequent studies in rats showed that the pressor effect of potassium depletion requires abundant consumption of sodium chloride (e.g., 4.5 g of sodium chloride per 100.0 g of dietary intake).

Population studies have shown an inverse relation of potassium intake to blood pressure, the prevalence of hypertension, or the risk of stroke. After adjusting for potentially confounding variables, the INTERSALT researchers estimated that a decrease in potassium excretion by 50 mmol per day was associated with an increase in systolic pressure of 3.4 mm Hg and an increase in diastolic pressure of 1.9 mm Hg. The urinary potassium:sodium ratio in the INTERSALT study had a significant, inverse relation with blood pressure. This ratio bore a stronger statistical relationship to blood pressure than did either sodium or potassium excretion alone. As compared with whites, blacks have a higher prevalence of hypertension and lower potassium intake; sodium intake among whites and blacks is similar.For example, in the Evans County Study, 23% of whites and 38% of blacks had a diastolic pressure of 90 mm Hg or higher. The 24-hour urinary potassium excretion averaged 40 mmol per day for whites and 24 mmol per day for blacks.

In clinical studies, a diet low in potassium (10 to 16 mmol per day) coupled with the participants’ usual sodium intake (120 to 200 mmol per day) caused sodium retention and an elevation of blood pressure; on average, systolic pressure increased by 6 mm Hg and diastolic pressure by 4 mm Hg in normotensive subjects, and systolic pressure increased by 7 mm Hg and diastolic pressure by 6 mm Hg in hypertensive subjects.

Cardiovascular Effects of Potassium Supplementation

Studies have shown that increasing the potassium intake of hypertensive rats that were fed highsodium diets lowered blood pressure, reduced the incidence of stroke and stroke-related death, and prevented cardiac hypertrophy, mesenteric vascular damage, and renal injury. In one of the studies, these benefits were independent of the blood pressure–lowering effect of the diet.

Kempner’s rice–fruit diet, which was introduced in the 1940s, was rich in potassium and extremely low in sodium. This diet was widely used in treating hypertension and congestive heart failure. Subsequently, many studies examined the effect of potassium on blood pressure and most of them identified a salutary effect.

A meta-analysis of 33 randomized trials that evaluated the effects of an increased potassium intake on blood pressure concluded that potassium supplementation (≥60 mmol per day in all but 2 trials) lowered systolic pressure by an average of 4.4 mm Hg and diastolic pressure by an average of 2.5 mm Hg in hypertensive subjects and lowered systolic pressure by an average of 1.8 mm Hg and diastolic pressure by an average of 1.0 mm Hg in normotensive subjects. This effect was independent of a baseline potassium deficiency, and it was greater at higher levels of sodium excretion (≥160 mmol per day) and in trials in which at least 80% of the subjects were black.

Potassium supplementation can reduce the need for antihypertensive medication. One study showed that with an increased dietary potassium intake in hypertensive subjects, 81% of the subjects needed less than half of the baseline medication and 38% required no antihypertensive medication for blood-pressure control, as compared with 29% and 9%, respectively, in the control group at 1 year of follow-up.

In the DASH trial, a diet rich in fruits and vegetables, as compared with the typical American diet, reduced systolic pressure in the 133 hypertensive subjects by 7.2 mm Hg and diastolic pressure by 2.8 mm Hg, at a constant level of sodium intake. The potassium content of the diet of fruits and vegetables was more than twice as high as that of the typical American diet; therefore, its higher potassium:sodium ratio probably accounted for most of the observed reduction in blood pressure.

Sodium sensitivity, defined as an increase in blood pressure in response to a higher sodium chloride intake than that in the baseline diet, occurs in many normotensive and hypertensive subjects; in normotensive subjects, sodium sensitivity appears to be a precursor of hypertension. Dietary potassium has been shown to exert a powerful, dose-dependent inhibitory effect on sodium sensitivity. With a diet that was low in potassium (30 mmol per day), 79% of normotensive blacks and 36% of normotensive whites had sodium sensitivity. Supplementation with 90 mmol of potassium bicarbonate per day resulted in sodium sensitivity in only 20% of blacks; this proportion matched that of whites when they received supplementation with only 40 mmol of potassium bicarbonate per day. An increase in dietary potassium can even abolish sodium sensitivity in both normotensive and hypertensive subjects.

Lack of Adaptation of the Kidneys to the Modern Diet

Human kidneys are poised to conserve sodium and excrete potassium. Prehistoric humans, who consumed a sodium-poor and potassium-rich diet, were well served by this mechanism. With such a diet, sodium excretion is negligible and potassium excretion is high, matching potassium intake. The kidneys account for 90% or more of potassium loss, with the remainder exiting through the fecal route. This mechanism, however, is unfit for the sodium-rich and potassium-poor modern diet. The end result of the failure of the kidneys to adapt to this diet is an excess of sodium and a deficit of potassium in hypertensive patients (Fig. 1).

Aldosterone contributes to the retention of sodium by the kidneys. Evidence from the Framingham A low-potassium diet leads to a potassium deficit in the body as a result of inadequate conservation of potassium by the kidneys and the alimentary tract; with such a diet, fecal potassium losses can exceed even urinary losses. Furthermore, a high-sodium intake increases kaliuresis, especially when sodium reabsorption by the renal cortical collecting tubule (where sodium reabsorption and potassium secretion are functionally linked) is enhanced (as it is in primary hypertension). Offspring Study suggests that relative aldosterone excess, as defined by the higher aldosterone values within the physiologic range, predisposes normotensive subjects to hypertension. In animals and humans, a low-potassium diet itself causes renal sodium retention by means of several mechanisms.

Excess sodium and a deficit of potassium in hypertensive animals and humans have been described previously. Exchangeable sodium (measured by the isotope-dilution technique) is increased in hypertensive subjects and correlates positively with arterial pressure; this correlation is highest in older patients. Despite an excess of sodium, extracellular fluid volume, plasma volume, and blood volume are not increased in primary hypertension. Conversely, exchangeable potassium (measured by the isotope-dilution technique) correlates negatively with arterial pressure in primary hypertension. Skeletal-muscle potassium is decreased in untreated hypertension, but serum potassium, generally an unreliable index of potassium content in the body, is within the normal range. Systolic and diastolic blood pressures are negatively correlated with muscle potassium in normotensive and hypertensive subjects.

Mechanisms of Altered Sodium and Potassium Homeostasis

Reabsorption of filtered sodium by the renal tubules is increased in primary hypertension because of stimulation of several sodium transporters located at the luminal membrane, as well as the sodium pump, which is localized to the basolateral membrane and provides the energy for such transport (Fig. 2). A pivotal luminal transporter is sodium–hydrogen exchanger type 3, which resides in the proximal tubule and the thick ascending limb of the loop of Henle, where the bulk of filtered sodium is reabsorbed. The activity of this exchanger is increased in the kidneys of rats with hypertension. Moreover, potassium depletion enhances sodium–hydrogen exchanger type 3 by inducing intracellular acidosis and by stimulating the sympathetic nervous system and the renin angiotensin system. Dietary potassium supplementation has opposite effects. The sodium–chloride cotransporter in the distal tubule, the epithelial sodium channel in the collecting duct, and the sodium pump are activated by the aldosterone excess in primary hypertension, thereby promoting sodium retention and potassium loss. A high-sodium diet increases potassium excretion by increasing distal sodium delivery.

An endogenous “digitalis-like factor,” which is identical to ouabain or a stereoisomer of ouabain, is released by the adrenal glands and the brain in response to a high-sodium diet. There are high levels of digitalis-like factor in the plasma of approximately 40% of untreated patients with primary hypertension, and these levels correlate directly with blood pressure. Digitalis-like factor mediates sodium retention by increasing the activity and expression of the renal sodium pump (Fig. 2).

Contrary to its short-term effects, the longterm effect of potassium depletion is to stimulate the activity and expression of the renal sodium pump, thereby promoting sodium retention. Such stimulation has been shown in cultured renal cells (after a 24-hour incubation in a lowpotassium bath) and in rats fed a low-potassium diet for 5 weeks. This effect is similar to the response to prolonged incubation of renal cells with ouabain for 5 days or to infusion of ouabain into rats for 3 to 4 weeks; the latter maneuver raises blood pressure. The long-term stimulatory effect on the renal sodium pump (which mediates sodium retention) contrasts with the inhibitory effect of potassium depletion and digitalis- like factor on the vascular sodium pump.

Additional mechanisms of sodium retention in primary hypertension have been proposed, including a congenital reduction in the number of nephrons, diminished renal medullary blood flow, and subtle acquired renal injury due to ischemia or interstitial inflammation. It is likely that heredity contributes to primary hypertension through several genes involved in the regulation of vascular tone and the reabsorption of sodium by the kidneys. Such a polygenic effect could result from gain-of-function mutations and polymorphisms in genes encoding components or regulatory molecules of the renin–angiotensin system and renal sodium transporters in subgroups of the population (Fig. 2). Examples are activating polymorphisms in the genes encoding G protein–coupled receptor kinases (which regulaten dopamine receptors involved in sodium reabsorption in the renal proximal tubule) and α-adducin (a cytoskeletal protein modulating the activity of the renal sodium pump). Populationbased investigations of candidate genes for hypertension have not produced unequivocal results, however. Expression of hypertension-related genes might be strongly affected by environmental and behavioral interactions that differ within a population and across populations.

In rats, coadministration of sodium and mineralocorticoids results in sodium retention, potassium depletion, hypertension, and extensive tissue damage. These changes bear a remarkable similarity to the changes in rats with hypertension induced by a high-sodium and low-potassium diet, which suppresses endogenous mineralocorticoids (Table 1). In both settings, the consequences of an excess of sodium and a potassium deficit in the body could be largely responsible for the hypertension and associated tissue injury. Furthermore, in primary aldosteronism, potassium administration augments aldosterone levels andyet reduces blood pressure, normalizes the circulatory reflexes of increased sympathetic activity, and corrects baroreceptor hyporesponsiveness.

Sodium Retention, Potassium Depletion, and Hypertension

Effects on the Arterial Wall

Sodium retention, by means of the release ofdigitalis-like factor, and a potassium deficit or hypokalemia inhibit the sodium pump of arterial and arteriolar vascular smooth-muscle cells, thereby increasing the sodium concentration and decreasing the potassium concentration in the intracelular fluid (Fig. 3). Increased intracelular sodium stimulates the sodium–calcium exchanger type 1 in the membrane, driving calcium into cells. A deficit of potassium in the body or hypokalemia inhibits potassium channels in the cell membrane, depolarizing the membrane (the membrane potential shifts closer to 0). Because of its electrogenic nature, the inhibition of the sodium pump itself decreases the membrane potential. Membrane depolarization in the vascular smoothmuscle cells promotes a further rise in intracelular calcium by activating voltage-dependent calcium channels in the membrane, calcium channels in the sarcoplasmic reticulum, and the sodium–calcium exchanger. The increased cytosolic calcium caused by these mechanisms triggers contraction of the vascular smooth muscle. PST 2238 (rostafuroxin), a compound that antagonizes the effects of digitalis-like factor on both the vascular and renal sodium pump, and SEA 0400, a specific inhibitor of sodium–calcium exchanger type 1, have shown promise as new antihypertensive agents, validating the importance of digitalislike factor and sodium–calcium exchanger type 1 in primary hypertension.

The homeostasis of sodium and potassium plays an important role in endothelium-dependent vasodilatation, which is defective in primary hypertension. Sodium retention decreases the synthesis of nitric oxide, an arteriolar vasodilator elaborated by endothelial cells, and increases the plasma level of asymmetric dimethyl l-arginine, an endogenous inhibitor of nitric oxide production. Sodium restriction has the opposite effects. A high-potassium diet and increases in serum potassium, even within the physiologic range, cause endothelium-dependent vasodilatation by hyperpolarizing the endothelial cell through stimulation of the sodium pump and opening potassium channels (Fig. 4). Endothelial hyperpolarization is transmitted to the vascular smooth-muscle cells, resulting in decreased cytosolic calcium, which in turn promotes vasodilatation. Experimental potassium depletion inhibits endotheliumdependent vasodilatation.

The contributions of prostaglandins, endothelin, atrial natriuretic peptides, kallikrein, and eicosanoids, as well as alterations in calcium balance, to potassium-induced changes in arterial and arteriolar tone and blood pressure are not well defined. Experimental studies suggest that in addition to its effects on vascular tone, a potassium-rich diet decreases cardiovascular risk by inhibiting arterial thrombosis, atherosclerosis, and medial hypertrophy of the arterial wall.

The long-term antihypertensive effect of lowdose thiazide diuretics reflects not hipovolemia but mainly decreased systemic vascular resistance, probably caused by changes in the ionic composition of the vascular wall. Natriuresis triggers cellular sodium loss and the redistribution of potassium into cells. The activation of potassium channels contributes to thiazide-induced vasodilatation.

Effects on the Brain

Changes in the concentrations of sodium and potassium in the cerebrospinal fluid, acting on a sensing region of the brain probably located near the third ventricle, have substantial but obverse effects on blood pressure (Fig. 5). Increasing the concentration of sodium in the cerebrospinal fluid by the intraventricular administration of hypertonic saline raises blood pressure, whereas increasing the concentration of potassium in the cerebrospinal fluid by administering potassium chloride has the opposite effect. Increasing dietary sodium chloride in animals and humans elicits small but significant increases in serum sodium; limited data suggest that the resulting increases in the concentration of sodium in the cerebrospinal fluid contribute to an elevation in blood pressure.

The intraventricular infusion of aldosterone at a dose that is too small to raise blood pressure when infused systemically decreases potassium in the cerebrospinal fluid and causes hypertension. The administration of either potassium or prorenone, a mineralocorticoid antagonist, through the same route prevents the decrease in potassium in the cerebrospinal fluid and the pressor effect of aldosterone (Fig. 5). The salutary action of small doses of spironolactone or eplerenone in hypertension and heart failure may largely depend on the central effects of the drugs in preventing or minimizing a reduction in the extracelular potassium in the brain, thereby moderating sympathetic discharge.

The central actions of changes in the concentrations of sodium and potassium in the cerebrospinal fluid and of an excess of sodium and a deficit of potassium in the body are probably mediated by changes in the activity of the neuronal sodium pump and the renin–angiotensin system in the brain. These changes alter sympathetic outflow, which then causes directional changes in blood pressure. Baroreceptor sensitivity is depressed by potassium depletion and restored by potassium supplementation.

Effects on Metabolism

Potassium depletion inhibits insulin secretion and is associated with glucose intolerance, whereas potassium infusion and hyperkalemia increase the secretory rate of insulin by changing the membrane potential of pancreatic beta cells. Insulin triggers endothelium-dependent vasodilatation in skeletal muscle by causing the release of nitric oxide; this response is impaired in primary hypertension.

Thiazide-induced hypokalemia worsens glucose intolerance in type 2 diabetes mellitus and increases the risk of the disorder; correction of hypokalemia ameliorates the glucose intolerance. As compared with diuretics, angiotensinconverting–enzyme inhibitors and angiotensin II receptor blockers, which promote potassium retention, are associated with a lower risk of newonset type 2 diabetes.Treatment of thiazideinduced hypokalemia with potassium augments the antihypertensive effect of the diuretic.

Implications for Prevention and Treatment

A modified diet that approaches the high potassium: sodium ratio of the diets of human ancestors is a critical strategy for the primary prevention and treatment of hypertension. Weight loss with diets rich in fruits and vegetables has been attributed both to the low caloric density and to the high potassium content of these diets, which tend to increase the metabolic rate.

In its 2002 advisory, the coordinating committee of the National High Blood Pressure Education Program identified both a reduction in dietary sodium and potassium supplementation as proven approaches for preventing and treating hypertension. The Institute of Medicine recommends an intake of sodium of 65 mmol per day (approximately 3.8 g of sodium chloride per day) for adults 50 years of age or younger, 55 mmol per day (approximately 3.2 g of sodium chloride per day) for adults 51 to 70 years of age, and 50 mmol per day (approximately 2.9 g of sodium chloride per day) for those 71 years of age or older. The institute also advises adults to consume at least 120 mmol of potassium per day (approximately 4.7 g of potassium per day, which is about twice the current U.S. average). These targets would require modifications for special groups, including competitive athletes, persons working in hot environments, patients with chronic kidney disease or diabetes, and persons taking medications that affect potassium balance. Adoption of the institute’s recommendations would increase the dietary potassium: sodium ratio by a factor of 10, from approximately 0.2 to approximately 2.0, which is much closer to our ancestral standard.

The concern that sodium restriction might increase cardiovascular risk by activating the sympathetic and renin–angiotensin system and by adversely affecting blood lipids and insulin sensitivity appears to be groundless for the recommended sodium intake. Forms of potassium that do not contain chloride, such as those found naturally in fruits, vegetables, and other foods, offer larger cellular entry in exchange for sodium and greater antihypertensive effects.

Following these recommendations would require a comprehensive, culture-sensitive campaign targeting both the general public and health care professionals. Food processing drastically changes the cationic content of natural foods, increasing sodium and decreasing potassium. Only approximately 12% of dietary sodium chloride originates naturally in foods, whereas approximately 80% is the result of food processing, the remainder being discretionary (added during cooking or at the table). Apart from educating the public, an agreement by the food industry to limit the deviation of the cationic content of processed foods from their natural counterparts is essential.

Figura 1