Roundtable Discussion: Electrolytes in Cardiovascular Disease and Hypertension

Abstract

During the meeting of the American Society of Hypertension in New York City in May 2003, a discussion of electrolytes in cardiovascular disease and hypertension was held. Dr. Jan Basile of the Ralph Johnson VA Medical Center, Charleston, SC, and Dr. Domenic Sica of Virginia Commonwealth University, Richmond, VA, were participants in the roundtable discussion, which was moderated by Dr. Marvin Moser of the Yale University School of Medicine, New Haven, CT.

DR. MOSER: Dr. Sica, there are epidemiologic data available in the literature to suggest that some electrolytes, specifically magnesium, potassium, and calcium, are related to hypertension. Are you convinced that deficits in any of these electrolytes contribute to elevations of blood pressure, or on the other hand, that replacing some of these elements has an effect on blood pressure? Is there any evidence that people who ingest a diet that appears to be deficient in calcium, potassium, or magnesium have higher blood pressures and more cardiovascular risk?

DR. SICA: The basis for a low calcium intake having a direct vasopressor effect is minimal. There is more convincing evidence that a low potassium intake is a contributor to a variety of cardiovascular events, including elevation of blood pressure and most probably stroke. Modifying a low potassium intake to a higher one has a proven protective effect on decreasing stroke rate as well as in lowering blood pressure. An increased potassium intake occurs as a concomitant to the Dietary Approaches to Stop Hypertension (DASH) diet, which is high in potassium and has been shown to reduce blood pressure.

It has been very difficult to specifically link magnesium to hypertension, although the occasional study has shown magnesium supplementation to marginally reduce blood pressure. More often than not, magnesium supplementation without evidence of a deficit has not proven beneficial. However, replacement of drug‐induced magnesium losses and/or losses as the result of intercurrent events like hyperglycemia will normalize levels and at a minimum has been shown to reduce ectopy.

DR. MOSER: Bottom line, if there is a population where the diet is deficient in potassium, this may contribute to more cardiovascular events and more hypertension. A diet deficient in magnesium may or may not have anything to do with blood pressure and may be related to cardiovascular events; the association with blood pressure has been difficult to evaluate.

DR. SICA: I would say that accurately captures my thoughts. However, we should remember that it is easy to measure and interpret plasma potassium values; the interpretation of the meaning of a serum magnesium value can be fairly tricky.

DR. MOSER: Even if plasma magnesium is measured, you don't learn much about total body magnesium.

DR. SICA: Yes. It is very difficult to accurately determine whether or not a magnesium deficit exists on the basis of a serum magnesium value alone. However, a clue to the presence of a total body magnesium deficit is that anyone who is being treated with potassium supplements for a potassium deficiency is likely to have some level of magnesium deficiency and should in many cases also be receiving magnesium.

There is another problem. Hypomagnesemia represents a level of total body deficiency severe enough to affect the serum determination. This is the exception to the rule. Total body magnesium losses are typically insidious, occurring in a progressive manner over time and developing in patients with multiple comorbid risk factors. This is the case in poorly controlled diabetes, where magnesium is continuously lost in the urine in conjunction with large amounts of urinary glucose. With the increase in the numbers of diabetics comes an expanding patient base characterized by conditions conducive to the development of total body magnesium deficiency.

DR. MOSER: Because magnesium is closely related to glucose metabolism.

DR. SICA: Yes. The relationship is mechanistically diverse in that urinary magnesium losses can occur with some regularity and like potassium, magnesium readily migrates across cell boundaries, prompted by insulin or β agonism.

DR. MOSER: One more question before we get to Dr. Basile. Are there any situations other than people with heart failure who are on diuretics or those with severe malnutrition who might be hypokalemic and have a deficiency of magnesium? Are there people in various parts of the world, for example, on limited diets of vegetables or nuts who might have these spontaneously?

DR. SICA: Interesting question. People in various parts of the world tend to ingest perfectly adequate amounts of magnesium particularly when their diet is strongly vegetable‐based. This is typically the case for agriculture‐based societies—such as Borneo. Magnesium intake begins to decrease when meat intake goes up but not to a degree that diet alone produces a deficiency state.

Years ago Dr. Maurice Shils ¹ described a series of patients with head and neck cancer who had to be fed by nasogastric tube. He purposely reduced the magnesium content in the nasogastric feedings to examine the body's homeostatic response to the extremes of magnesium depletion in the diet. He found that with a low‐magnesium diet, urinary excretion still occurred despite a fairly rapidly evolving systemic deficiency state. This is different than is the case for sodium or potassium depletion; therein the response to dietary deprivation is a fairly sudden and profound reduction in urinary excretion (more so for sodium than potassium).

DR. MOSER: Magnesium loss continues regardless of your intake?

DR. SICA: Regardless of intake. Not only is there an ongoing pattern of unabated urinary loss but there also is an obligatory loss, which occurs by way of the gastrointestinal tract. This may only amount to a few milliequivalents of gastrointestinal loss of magnesium per day, but slowly and steadily magnesium depletion occurs in response to a diet low in magnesium.

DR. MOSER: So to summarize: Data on a low potassium intake suggest that this is a factor in possibly elevating blood pressure and increasing the risk of stroke. Data on magnesium and calcium are equivocal. Most people on a reasonable diet are probably not going to suffer from magnesium or calcium deficiency, but they may have a problem with a low potassium intake if sodium intake is high. Is that fair?

DR. SICA: Yes. I do believe, however, that in people with lactose intolerance it's much easier to become calcium depleted.

DR. MOSER: That might play a role in raising blood pressure, but that's extremely unusual.

DR. SICA: Well yes, it is unusual.

DR. MOSER: But we can't look at calcium without considering potassium and magnesium, and calcium metabolism is intimately involved with parathyroid hormone, vitamin D, and other factors. There are good data, especially from the South, about potassium‐deficient and sodium‐rich diets. Dr. Basile, will you comment about the effects of low potassium intake on blood pressure, on the use of potassium supplements, and the possible effects of calcium and magnesium on blood pressure? First the epidemiologic data on whether or not any of these minerals may be causative factors for hypertension.

DR. BASILE: The epidemiologic data are most sound for populations that are hypokalemic having an increased risk of stroke.

DR. MOSER: By hypokalemic you don't mean a low serum potassium, you mean a low intake.

DR. BASILE: Yes. Populations that have a low potassium intake in their diet, like the Rancho Bernardo, CA, population, have an increased risk for cerebrovascular disease and stroke. These are predominantly elderly folks who have a diet low in potassium. The data are stronger for cerebrovascular disease than cardiovascular disease.

DR. MOSER: And that's consistent with animal data?

DR. BASILE: Yes. There are also observational data that suggest that groups that have more calcium in their diet tend to have lower blood pressures. It doesn't necessarily translate into hard clinical end points as the potassium story does, but there are enough data so that we should recommend a diet high in calcium.

DR. MOSER: Does calcium supplementation decrease blood pressure?

DR. BASILE: Some work from Allender ² suggest that a reduction in blood pressure occurs with a high calcium intake.

DR. MOSER: How about extra potassium, does that lower blood pressure?

DR. BASILE: I believe it does.

DR. MOSER: Enough so you would recommend a high‐potassium diet like the DASH diet as definitive therapy?

DR. BASILE: No, I don't think that this should be recommended as definite therapy.

DR. MOSER: How about as adjunctive therapy? Would you tell your patients with hypertension that they ought to be on a high‐calcium diet in addition to a high potassium intake?

DR. SICA: We've got several problems. For a postmenopausal woman, recommendations would suggest that, but you have to be careful because high calcium intake translates into hypercalciuria, and with that the risk of renal calculi goes up.

DR. MOSER: So you might worry in your hypertensive patients about giving too much calcium because of concerns about renal calculi.

DR. SICA: Right, and there are data suggesting that the risk of nephrolithiasis is higher in hypertensive patients.

DR. MOSER: Right, so we can be on firm ground if we say to most people that a high‐potassium diet is beneficial, but with calcium we have to be a little more careful because of concerns about renal calculi. What about magnesium? Does magnesium lower blood pressure or does a lack of it raise blood pressure?

DR. BASILE: The data are not quite as strong here. I look at magnesium as sort of a bystander, important in its role with potassium especially when we use agents and therapies that can deplete potassium.

DR. MOSER: But magnesium is a vasodilator and may act like a calcium channel blocker. Doesn't that suggest that it would be effective?

DR. BASILE: We don't have much data. If we could get our patients to adopt the DASH diet, which is rich in fruits, vegetables, and low‐fat products (that are high in calcium content), and in addition lower their sodium intake, we would be giving them the type of diet that would be very advantageous for furthering the control of blood pressure. In some studies, this has lowered systolic blood pressure by 8–14 mm Hg, as much as single drug therapy. I hesitate to recommend taking supplements if there's no evidence of specific deficiencies.

DR. MOSER: A recent meta‐analysis by a group from Hopkins ³ described a dose‐dependent effect of magnesium on blood pressure. With each 10 mmol/d increase in dosage there was a 4.3/2.3 mm Hg decrease in blood pressure. Many of the 14 studies in hypertensive patients were small and of short duration, but it does appear that with adequate supplementation, even in people who are not magnesium deficient, there may be an effect. What do we know about potassium supplementation and reduction in blood pressure?

DR. SICA: The one thing we've learned is that potassium intake has a remarkable palliative effect on many of the evils induced by high sodium intake. Research indicates that what constitutes salt sensitivity may relate to potassium intake and its relationship to sodium intake. The use of salt substitutes that contain potassium may make some sense if we are simultaneously looking to re‐establish a different balance between sodium and potassium. As a point of reference, a teaspoonful of most salt substitutes contains about 60 mEq potassium chloride.

DR. MOSER: Many of the vitamin supplements contain small amounts of potassium. Is that of any use?

DR. SICA: It's basically 1 or 2 mEq so it's not much. If you ask if I use potassium, yes, in some people I do. I'll use potassium therapeutically for people who are athletic with a lean body frame, who are sympathetically activated, and who don't want to go on a specific antihypertensive medication but still require some form of therapy for early stage 1 hypertension. Remember, any trial with potassium as a supplement automatically imputes placebo effect into what you're doing. Sometimes I will use potassium for its placebo effect and other times I'm presuming that you may get a specific vasodepressor benefit.

DR. BASILE: In salt‐sensitive populations, especially African Americans, the elderly, and perhaps diabetics, a given salt load increases blood pressure more than usual. You can ameliorate this sensitivity to salt by repleting potassium or going on a high‐potassium diet. That's an important point in groups where it is difficult to reduce sodium intake.

DR. MOSER: So there may be a reason to use potassium supplements.

DR. BASILE: This has not been well studied but I think there are certain populations where potassium supplementation may be advantageous and it may be for the group of patients who are more salt sensitive.

DR. SICA: In some African‐American adolescents whom we have studied with an average daily sodium intake of 275 mEq/d, about one fourth to one third of them were found to be nondippers—that is their blood pressure doesn't go down at night. These are otherwise normotensive kids. When we took a look at their potassium intake, it was about 35 mEq/d.

DR. MOSER: Instead of 60, 70, or 80 as it should be.

DR. SICA: Instead of 60 or 70. When we simply brought the potassium content in their diet to the more normal 55–60 mEq/d, most returned collectively to a normal nocturnal dipping status (a blood pressure drop of 10%–20% at night). These studies have taken on a provocative flavor because we don't know what nondipping means in a normotensive right now. It probably suggests residual sympathetic activity. This residual sympathetic activity and nondipping status may be driven by sodium intake.

DR. MOSER: Dr. Basile, let's get to a major issue. We have millions of people now on agents that decrease potassium and magnesium. We're all fans of diuretics here, we know that their use decreases morbidity and mortality in hypertensive patients. Could we improve outcome still further if we paid more attention to potassium and magnesium? How often do you have to replace and how often should you replace these minerals in patients receiving diuretics? We'll get to heart failure and renal failure patients later. For noncomplicated hypertensives, are we doing them a disservice by allowing potassium to decrease even minimally because we then know that magnesium is also decreasing? What should we do about this?

DR. BASILE: That's a very difficult question to answer, but there are some observational analyses of clinical trials that may help us. First of all, the Antihypertensive and Lipid‐Lowering Treatment to Prevent Heart Attack Trial (ALLHAT) heavily influenced the Seventh Report of the Joint National Committee on the Detection, Prevention, Evaluation, and Treatment of High Blood Pressure and suggested that a clinician should consider a thiazide‐type diuretic as initial therapy in most patients with hypertension. And the use of a low‐dose thiazide‐type diuretic, which in the past was 12.5–25 hydrochlorothiazide, needs to be re‐evaluated since chlorthalidone, a longer‐acting thiazide, was used in the ALLHAT trial. The mean dose of chlorthalidone has still not been published, although the ceiling dose was 25 mg. We need to re‐evaluate the effects of this dose, which is probably not equal to the low doses that have been advocated. Thiazides may cause hypokalemia and they also cause obligatory hypomagnesemia, often unrecognized.

DR. MOSER: Does it always go together? Can you be hypokalemic and not suffer some magnesium deficit? I don't think so.

DR. SICA: It is quite uncommon for potassium losses to occur with a diuretic without concurrent magnesium losses.

DR. BASILE: In the Systolic Hypertension in the Elderly Program (SHEP) trial, in the 6.8% of the patients who had a potassium level less than 3.5 on a thiazide type diuretic, benefit was less than in patients who remained normokalemic.

DR. MOSER: It wasn't worse than the placebo, but…

DR. BASILE: It was no different than placebo, but there was not the same favorable effect as in those who were normokalemic defined by a serum potassium level of 3.5 or better. So that trial in elderly hypertensives with systolic hypertension suggested that more attention should be paid to potassium (and perhaps to magnesium) if the full benefits from a thiazide type diuretic are to be realized.

DR. MOSER: So would you suggest on the basis of those data and some other observations that elderly hypertensive patients on a thiazide should always have a potassium and a magnesium supplement?

DR. BASILE: Not necessarily. In ALLHAT, about 8.7% were hypokalemic. We were told within 24–48 hours if the serum potassium was less than 3.5, and we were alerted sooner if it was less than three. Clearly in this study clinicians were reminded if their patients were hypokalemic.

DR. MOSER: Did you use potassium supplements?

DR. BASILE: We did supplement. Although we did not use potassium‐sparing agents, we supplemented with potassium.

DR. MOSER: Did you use magnesium supplements?

DR. BASILE: We did not supplement magnesium.

DR. MOSER: Should you have?

DR. BASILE: Perhaps we should have, and that data needs to be looked at. Because magnesium follows potassium, I am now a believer in potassium‐sparing therapy for patients who are on a thiazide type diuretic and become hypokalemic because not only will you preserve potassium but you'll preserve magnesium as well. If you just give a potassium supplement, it may not be enough. The use of a substance such as magnesium oxide should also correct the magnesium deficit. The clinician needs to be aware in trying to improve outcome with a thiazidetype diuretic that potassium levels should be determined within a few weeks after a thiazide is started. If serum potassium decreases below normal, both potassium and magnesium deficiencies should be addressed by supplementation.

DR. MOSER: Playing the devil's advocate: Why worry? Why not say that every elderly person with hypertension who is on a diuretic should be given a potassium and a magnesium supplement and not worry about the serum levels?

DR. BASILE: Or, why not just put everyone on Maxzide or Dyazide (GlaxoSmithKline, Research Triangle Park, NC) instead of hydrochlorothiazide or chlorthalidone? Well, there may not be a reason to do this routinely. Many of the other agents that we give in addition to thiazides to control blood pressure have a tendency to elevate potassium. We have to be careful.

DR. MOSER: You mean if you use an angiotensin‐converting enzyme or an angiotensin receptor blocker, or even a β blocker…

DR. BASILE: Yes. So you have to be a clinician. I don't think we should make generic rules or give carte blanche authority to always do one thing or another.

DR. MOSER: What if they're just using a thiazide, blood pressures are controlled, and you don't have to worry about using a possible hyperkalemic agent such as an angiotensin‐converting enzyme inhibitor…what should we do?

DR. BASILE: Let me just give you some reference points. In the Multiple Risk Factor Intervention Trial (MRFIT), which used chlorthalidone, I believe 18% had serum potassium less than 3.5, and only 3% had potassium level less than three.

DR. MOSER: That's right.

DR. BASILE: So hypokalemia with chlorthalidone in MRFIT with higher doses of chlorthalidone than used in ALLHAT affected one in five patients with the type of serum potassium levels we're trying to prevent. We've got to watch potassium levels. It's an important issue.

DR. SICA: First we've got to be careful to determine what our definition is of hypokalemia based on serum values. Most people have relied on an arbitrary cut point, so if you're between 3.5 and 5.5, you're in a normal range.

But if I've got somebody who starts out at 4.5 and goes down to 3.9, they've sustained a significant drop although they've not crossed the boundary of a definition of hypokalemia. We've got to consider this. Of all the diuretics, chlorthalidone is probably the one that causes the most loss of potassium or magnesium. This relates directly to the duration of diuretic effect, which is much longer with this agent. Potassium loss can sometimes even be greater with chlorthalidone than with loop diuretics.

DR. MOSER: So you believe that potassium loss is clinically significant in the average hypertensive patient who is on chlorthalidone?

DR. SICA: I think that anybody on 25 mg chlorthalidone is going to be magnesium depleted, and they don't have to be potassium depleted. Although potassium and magnesium losses track together, there may be a drop in magnesium stores independent of a decrease in potassium.

DR. MOSER: You believe we should be paying more attention to magnesium deficiencies than we have.

DR. SICA: The longer you diurese with each dose, even though it may be with a less profound overall response, the more magnesium you will lose. With loop diuretics, most of which are short acting, there is a limited duration of time over which magnesium loss occurs. Conditions are then set into play to prevent additional magnesium loss.

DR. MOSER: So the longer acting diuretics may be more dangerous.

DR. SICA: Yes, absolutely.

DR. MOSER: But ALLHAT and other trials did demonstrate, in studies where electrolytes were monitored, that the use of these agents reduced morbidity and mortality. Now, what about people with heart failure? Should we be paying more attention to magnesium or potassium loss in these people, or are most of us aware of this problem?

DR. SICA: I think most cardiologists are aware of the problems that accompany magnesium deficiencies, though they may underestimate the magnitude of the magnesium deficit and may not use the available replacement supplements correctly. Many times there is confusion about how to mix and match potassium and magnesium in the right proportions. Unfortunately, we do not have an easy way to assure ourselves that body stores have been repleted once we begin a course of therapy.

DR. MOSER: Are there general rules as to how much you need to adequately replace these minerals?

DR. SICA: There are approximations. One of the things that should come out of this roundtable are maybe some words about what should be an appropriate intake in a heart failure patient subjected to adequate diuretic therapy.

If you look at some of the supplements on the market to illustrate how much magnesium you needed to replace the loss, it might be helpful. For example, the Blaine Pharmaceuticals tablet has 400 mg magnesium oxide and about 240 mg elemental magnesium. In most heart failure situations about two of these taken daily would be adequate if you're not actively diuresing the patient. On the other hand, if there are 20 lb of extra fluid and you're actively diuresing, the odds of a significant deficit occurring are quite real, and greater supplementation would be required.

DR. MOSER: Any danger about too much magnesium?

DR. SICA: Yes and no. In the large majority of the population, no. In heart failure where mild‐to‐moderate renal insufficiency is common, one has to be ever mindful of the fact that this can happen. Usually magnesium levels don't create a problem until you're above about 5 mEq/L. (normal range is laboratory dependent but generally 1.8–2.4 mg/dL.

DR. MOSER: And then would you get some central nervous system symptoms?

DR. SICA: Fewer central nervous system problems, but cardiac conduction defects can occur, primarily in the atrioventricular node. If it goes far enough there could be refractory hypotension, which interestingly is responsive to parenteral calcium. If you use a potassium‐sparing diuretic, you also end up having a magnesium‐sparing diuretic. But, for example, with spironolactone we don't really know how much magnesium sparing occurs at the low doses that are given according to the current practice guidelines in heart failure.

DR. MOSER: Does eplerenone spare magnesium?

DR. SICA: Eplerenone should but eplerenone has not been studied nor has it been reported as to epleronone's effect on magnesium.

DR. BASILE: I'd like to make one point. I think we should look at magnesium loss from one or two mechanisms, either thiazide induced, which is an obligatory loss, or secondary to hyperaldosteronism that occurs with heart failure and cirrhosis. There are many patients with alcoholic liver disease who are profoundly hypomagnesemic, recognizable by serum magnesium determinations. We can either spare potassium with a potassium‐sparing agent or antagonize aldosterone mechanisms with or without replacement of magnesium depending on the physiologic mechanism. I think what I would conclude from this roundtable is that we need to start paying more attention to magnesium as well as to potassium if we are going to recommend more thiazide‐type diuretics. We need to be aware that when we use an angiotensin‐converting enzyme inhibitor and a thiazide, we have less hypokalemia than if we just use a thiazide, but Dr. Sica, do angiotensin‐converting enzyme inhibitors spare magnesium with potassium?

DR. SICA: They'll spare magnesium but no one has done total body magnesium studies so we're not exactly sure of this. Dr. Basile's point on liver disease is an important one. Average doses of spironolactone in a cirrhotic with ascites are anywhere from 100–400 mg/d. That is clearly in a range that confers a significant magnesium sparing benefit, but most of the patients still remain magnesium depleted as the result of gastrointestinal losses secondary to the effects of lactulose, an osmotic cathartic used to purge the gut and decrease nitrogen absorption. Stool magnesium losses in a patient with ongoing diarrhea can approach 30–35 mEq/d. A true total body magnesium deficit is about 2 mEq/kg body weight. So with 3 or 4 days of unchecked diarrhea, a definite magnesium deficiency could develop. It can happen quickly. On a diuretic like chlorthalidone, anywhere from 10–25 mEq/d magnesium can be lost based on the amount of sodium in the diet, which fuels diuretic‐related magnesuria.

I'd say we need to measure magnesium more frequently and we need to understand the wider prevalence of its deficiency. We need to supplement more with magnesium salts as an adjunct maneuver in patients on a potassium‐wasting diuretics. If we are caught up with the notion of treating hypokalemia we're way off base if we believe we have to absent ourselves from, at the same time, treating magnesium deficits in the body.

DR. BASILE: I would like to say one last thing.

DR. MOSER: Sure.

DR. BASILE: In patients with liver disease you often use a magnesium compound every day depending on what other therapies you're giving them. There are, on the other hand, older patients who will need both potassium and magnesium replaced in some form.