NSAID use and endurance running: foundations of a patient advisory.

Nonsteroidal anti-inflammatory drugs (NSAIDs) are the most common self-administered and prescribed drugs in the U.S. These agents are primarily used to reduce inflammation and pain, which makes them exceedingly popular with athletes. At an estimated economic impact of $2.5 billion a year, nonsteroidals are the largest drug class in the world. However, we annually spend an additional $4 billion treating their side effects. What are the current controversies in prescription and over-the-counter (o.t.c.) NSAID use among the exercising population and, in particular, endurance runners? When all the factors weigh in, how might clinicians make meaningful recommendations regarding NSAID use? At the Virginia Hospital Center in Arlington, VA, on Friday, October 29th, the 2004 AMAA Sports Medicine Symposium at the Marine Corps Marathon devoted a series of lectures to finding out.

**********

THE "NONSELECTS"

Traditional (a.k.a. nonselective) NSAIDs include o.t.c. medications such as ibuprofen (found in Advil[R] and Motrin[R]) and naproxen, which, though widely prescribed as Naprosyn[R], is also available o.t.c. in Aleve[R]. Less common traditional nonsteroidals include piroxicam and sulindac. Acetaminophen (found in Tylenol[R] and Excedrin[R]) is not considered an NSAID because it lacks anti-inflammatory properties; it is largely analgesic.

NSAIDs are cyclooxygenase (COX)-1 and COX-2 inhibitors. They control swelling, but by blocking the production of COX-1 enzymes, can cause or contribute to gastrointestinal (GI) problems. Even the widespread and principally analgesic acetaminophen, when ingested in amounts greater than 2 g per day, can lead to bleeding and perforations. At 650 mg per capsule, two capsules of Tylenol Arthritis Pain Extended Relief[R] need only be taken twice a day to exceed this dose. One study found, through endoscopic means, a 19% incidence of gastric ulcers in patients taking 1,000 mg of naproxen daily--at the end of one week. As Nicholas Scarpa, M.D., FACR, pointed out in one of his discussions, "There is no such thing as gastric adaptation. Those people who are at risk [of ulcers] are at risk in the first week or the first month, just as much as if they were taking it for six months."

Scarpa, who serves as Medical Director of Rheumatology at the Arthritis Research Center of New Jersey, went on to explain that approximately 30% of these bleeds occur in the lower GI tract--where proton pump inhibitors fail to work. Furthermore, by some estimates, up to half of all endurance athletes suffer GI symptoms anyway. Under certain biochemical and biomechanical stresses associated with certain types of exercise, the integrity of the GI mucosal block can become damaged, leading to an uptake of toxic substrates. So a more enzyme-selective, safer group of drugs, the COX-2 inhibitors (coxibs), was devised.

COXIBS AREN'T THE SAVIORS WE HOPED

Unfortunately, we now acknowledge a dilemma in prohibiting COX-2 enzymes also. In his discussion, Joseph Chorley, M.D., an assistant professor of Pediatrics and Director of Education of the Primary Care Sports Medicine Fellowship at Baylor College of Medicine, provided a nice overview of renal system changes that occur during distance running. This helped illuminate the role that all NSAIDs--even coxibs like celecoxib (Celebrex[R]) and valdecoxib (Bextra[R])--may play in hypervolemic hyponatremia among endurance athletes.

Hyponatremia risk. While COX-1 enzymes are vital to what Chorley calls "normal housekeeping functions" (including gastric regulation), COX-2 were always seen as "the inducibles"--enzymes responsible for colon tumors, pain and inflammation, undesirable cellular transformation, and even osteoarthritis (OA). But eliminating them is not a cure-all. For one thing, proper cellular regeneration from certain types of soft tissue injuries can be blunted by coxibs. And there are COX-2 receptors in the kidneys; whatever we once thought of them, we now recognize that COX-2 enzymes, prostaglandins (PGs) in particular, perform important regulatory functions as well.

By inhibiting the production of PGs, as both nonselective NSAIDs and coxibs do, these agents can exacerbate--and theoretically cause--fluid imbalance, electrolyte disorders and kidney problems during a marathon. This is particularly true in older runners (who also happen to be more likely to rely on NSAIDs to get through a long race). Though marathoning may put you in a PG-dependent state--particularly via dehydration and exercise in high temperatures--most marathoners are not in such a state during the course of a race. Yet runners in their 50s and 60s will see renal changes sooner just by virtue of their age. Above 50% VO2max, urine production decreases by as much as 60%, half the amount of blood is filtered through the kidneys, and sodium excretion decreases by 40 to 60%. Chorley, who is assistant medical director of the hp Houston Marathon[R], tells us that marathoning at as slow as 12-minute mile pace puts a 20-year-old male above 50% VO2max; for a 30-year-old, 14-minute miles will do it; if a 60-year-old male runs faster than 18-minute miles, he is above 50% VO2max.

Most marathoners run between 75 and 80% VO2max. This means that it is far easier for these older runners to get to a place where the importance of PG function in their bodies is elevated. As vasodilators, PGs allow increased blood flow to the kidneys, which helps more water to be excreted through urine. The danger of fluid overload, then, is greater with a nonsteroidal present. Chorley says, "It's a lot harder to dilute your urine when you don't have the normal COX-2 prostaglandins that are supposed to be there."

Hypertension risk. The fluid-retention associated with NSAID use may contribute to another problem. There is in fact some evidence that NSAIDs can induce hypertension in normotensives. In any case, the kidneys play a significant role in blood pressure regulation, and it is estimated that 7 to 8% of the marathoning population is already hypertensive. Scarpa notes that in a hypertensive patient, any nonsteroidal--whether traditional or selective--will increase the risk of worsening the hypertension up to 8%. He is quick to point out, however, that the extent to which a drug will raise blood pressure is not a class effect. In fact, while other agents have been shown to raise blood pressure between 4 and 6% in hypertensives, one coxib in particular appears to raise it by some 25 to 30%. As it turns out, this agent, rofecoxib, better known as the notoriously-recalled Vioxx[R], offers even more serious trouble.

Risk to GI vs. Risk of MI. On September 30, 2004, Vioxx was taken off the market when it was discovered that in 25-mg doses it creates twice the risk of myocardial infarction (MI) as compared to placebo. The theory behind the so-called pro-thrombotic mechanism makes matters a little less easily solved than by simply pulling Vioxx off the shelves. Scarpa, who is a clinical physician at the Hospital for Special Surgery at Cornell Medical College, reminds us again that PGs dilate the vessels.

COX-1 enzymes cause platelet aggregation; clotting is therefore inhibited without them. COX-2 enzymes vasodilate. When you block both COX-1 and COX-2, you are creating a balanced state. A nonselective agent helps keep platelets from sticking, while at the same time blocking vasodilation. The smaller vessel is not a problem, because the platelets are not sticking. "But when I use COX-2 inhibitors, I'm not inhibiting COX-1, so I have vasoconstriction but I don't inhibit platelet aggregation," Scarpa says. When the sticky platelets try to pass through a smaller space, the theory goes, they increase the risk of strokes and heart attacks.

But as much as this is a coxib-specific problem, it too is not as of yet a class effect. Clinicians have observed drug accumulation so far only with rofecoxib. "When you give most NSAIDs, when you increase the dose or increase the duration, [the body] increase[s] the metabolic rate," Scarpa says, "and as a result you don't get drug accumulation." The data on Vioxx revealed a metabolic plateau, and after five days, an accumulation. "These drugs' molecular structures differ," he says.

In December, Pfizer found a 3.4 times greater risk of cardiovascular events in patients taking 800 mg of Celebrex daily, when compared to patients taking a placebo (see sidebar). And Bextra now comes with a warning label since it may cause heart and blood-clotting problems in some patients who have recently had heart bypass surgery. Until the molecular differences among these agents are better understood, no one is comfortable proclaiming that the pro-thrombotic mechanism only exists with this or that NSAID. But beyond the scope of the studies in question, Scarpa suspects there are other factors. The very act of marathoning can, after all, stress the cardiovascular system in ways that might, in some individuals, be considered pro-thrombotic.

PRACTICAL DO'S AND DON'T'S

The Clearly No's and Probably Not's. Indeed, once certain other factors are introduced, the one thing that is safe is saying that nonsteroidals are not always a good idea. The PG-dependent state does not only elevate a runner's risk of becoming hyponatremic. Such a state means vasodilation is needed; cardiac patients are prostaglandin-dependent. Those with chronic heart failure are wise to avoid a nonsteroidal before a marathon.

Runners whose kidneys are maximally stressed already--cardiac patients, hypertensives, the salt-depleted--should not add an NSAID to the marathon mix, selective or otherwise. Additional factors that can create a PG-dependent state include hypovolemia, a salt-restricted diet, chronic liver failure, diuretics and possibly even high race-day temperatures. Chorley urges us to recognize that these all happen to be widely-accepted "bad marathoner traits"; his conservatism is reflected in a concern for those who would ignore this and compete anyway. He feels that one of the best things a race director can do to cut down on traffic in the marathon-day medical tents is to offer a half-marathon that day as well.

The Maybe Not's. From a hyponatremia-risk perspective, Chorley advises against taking an NSAID before a marathon if you are a first-time, female entrant expecting to finish in the four- or five-hour range. Women who drink the same amount of fluid per kg of body weight as men consistently lose less weight and retain more sodium during a marathon. Additionally, if you sweat heavily and your clothes cake with salt; if you have had heart disease in the past: or if you merely are an older runner, you are among those at risk for problems in endurance races. He adds to these factors anyone running in high heat and/or heavy humidity. So perhaps if a patient possesses two or three of these traits, it's best to steer him or her away from a nonsteroidal. If just one of these factors is in play, it might be okay.

Ill-advised drug interactions. Beyond heart failure, renal disease and high blood pressure in and of themselves, the medications these conditions necessitate can make NSAID use out of the question, or a bad idea. If an asthmatic patient takes an intranasal steroid, chronic nonselective NSAID use may increase ulcer and bleeding risk. Cardiac patients need to be aware that nonsteroidals attenuate the effects of beta-blockers and ACE-inhibitors. This can not only aggravate heart failure, but lead to uncontrolled hypertension if a patient is on these medications for high blood pressure.

NSAIDs do not lower MI risk as aspirin does. In fact, when taken with aspirin, regular (but not intermittent) use of ibuprofen appears--in some but not all studies--to negate aspirin's cardioprotective effect. Just one or two Advil a week have been shown to diminish the cardioprotective effect of a daily aspirin. Celebrex, however, seems to have no effect on aspirin's health benefit. In one study, one dose two to three times a week did not increase MI risk. Taking a nonselective NSAID with aspirin, though, quadruples the risk of ulcers. Note that Excedrin contains both acetaminophen and aspirin. And ingesting multiple NSAIDs increases this risk even further.

Dosage vs. regularity. In the Brigham and Women's Hospital rofecoxib study, as compared to doses less than 25 mg, doses greater than 25 mg posed a far higher MI risk in the 54,000 subjects, who were age 65 and over. The risk for doses over 25 mg was triple that of placebo. In the Celebrex trial that found increased cardiovascular event risk, patients were taking up to four times the regular dose for osteoarthritis. The normal dose for Celebrex is 200 mg.

How much a patient ingests, then, is an important factor--more important than for how long, perhaps. With rofecoxib, MI risk was elevated for the first 90 days of use but not thereafter. This is not good news; it indicates that, to a certain extent, chronic (vs. occasional) use puts people at a higher MI risk (the Celebrex patients had been consuming the drug twice daily for an average of 33 months), but regular, long-term use is not necessary. As noted earlier, the same is true for gastric ulcers.

In one study Chorley published on the hp Houston Marathon, a startling 85% of runners who wound up in the medical tents were taking NSAIDs 24 to 36 hours before the race. The ubiquity and convenience of the o.t.c. nonselective nonsteroidals make it important to look closely at both their dosage and frequency of use with an eye toward minimizing GI bleeds, especially given that 82% of ulcers from NSAIDs are asymptomatic.

The COX-1 group is responsible for the regeneration of gastric mucousa. If you are taking high doses of acetaminophen or ibuprofen three to five days a week, you are in what Scarpa calls "ulcer territory." But low doses can balance frequency; there is a big difference between swallowing four Advil a day and swallowing only one. From a GI bleeding-risk perspective, a coxib makes a better choice. The fact that the o.t.c.'s are milder has led clinicians down the wrong path at times. "We used to give Tylenol to people we thought couldn't handle the prescription NSAIDs," he says. But recent evidence that GI abnormalities due to o.t.c. Tylenol has made it clear that 4 g per day is simply too much. "This tells us what to do about the runner who wants to take it four times a week," says Scarpa. "He might be at a high risk of developing ulcers." He adds, however, that for a marathoner who might be PG-dependent, less than 2 g daily of Tylenol is permissible and perhaps preferable to a coxib or a traditional NSAID like naproxen.

Scarpa feels that a 50-year-old runner with a history of myobilateral knee OA with no co-morbidity factors may take Tylenol or Motrin regularly while training--perhaps even before and after long runs--and still run marathons, provided he keeps the dosage in check. If this runner suffers from acid reflux or a peptic ulcer, however, a COX-2 inhibitor suddenly looks best. A hypertensive might avoid taking her diuretic the day before and/or the day of a long run or marathon, using exercise instead to lower blood pressure. In this way, individual factors tip the scale toward one type of agent or another, or none.

Likewise, for the runner with a soft tissue injury who is choosing to run through, say, an inflamed ITB, the question of whether they should take an agent and proceed with a marathon hinges in no small way upon what, if any, co-morbidity factors they have. But this example leads to another potentially thorny issue. A young runner with a ligament strain who needs to be ready for competition in a hurry will frequently rely on a nonsteroidal to attenuate pain and swelling. This is a kind of band-aid approach and needs to be addressed in any discussion of the proper role of NSAIDs in endurance sports. An NSAID may get the athlete back on track in the short term but there is the issue of whether they can regenerate and heal that tissue permanently if it is constantly bombarded with anti-inflammatories. NSAIDs can inhibit necessary swelling and be counterproductive with certain injuries, possibly even underlying OA. Do nonsteroidals, OA and exercise really mix at all?

OA AND NSAIDS

Of the 16 million people who presently have some form of OA, Scarpa highlighted two categories of distance runners in his talk: those who have been running all their lives and those who have recently taken it up, perhaps after a health-conscious revelation.

The runners who start before the arthritis, he says, are helped by the running. Running in and of itself does not cause OA, and most likely staves it off considerably for these patients. "But if you've got it already and are a new runner, you can advance the OA," he says. The weekend warriors with the underpinnings of OA who run in the same direction over the same sloped asphalt roads are likely to get worse. Scarpa recommends that instead, these people walk regularly. "Don't get them just sitting," he cautions.

In one study, two OA groups received the same NSAID, but one group fitness-walked while the other went about their normal routines. After six weeks of supervised fitness walking, the active group reported a 50% decrease in pain, and a third of them were even able to walk without taking their nonsteroidal medication.

Though we now see evidence of an inflammation mechanism in OA, NSAIDs are in many ways a less than ideal treatment for it. Researchers are beginning to realize that some OA seems to be induced by PGs and some by cytokines. This may explain why some patients do well with PG-inhibitors and others don't: different enzymes appear to be responsible for the inflammation.

In their ongoing Glucosamine/Chondroitin Arthritis Intervention Trial (GAIT), the NIH is currently testing the conclusions of a meta-analysis that found glucosamine/chondroitin sulfate to be as effective as a nonsteroidal in attenuating stage-one OA over the long term; it even improved patients' x-ray scores. 1,500-mg doses of the nutraceutical improved cartilage function and strength in these people, regenerating destroyed cartilage as no NSAID could. Given the potential minefield of nonsteroidal side effects, this treatment, combined with regular load-bearing exercise and quadriceps strengthening, would appear to be desirable in the long run.

CONCLUSIONS

For those taking nonsteroidals, is the very fact that someone logs 50 or 60 miles a week a risk factor? Scarpa acknowledges that this kind of rigorous endurance exercise places certain physiological stress upon kidneys and the cardiovascular system. But without the presence of co-morbidity conditions, the short answer is probably no. For the marathoner who wants to mitigate pain and swelling by taking an NSAID prior to running, Scarpa says something along the lines of a 200-mg dose of Celebrex is most likely okay. Chorley is not so sure: Is it a hot day? Is this an experienced runner? Part of the controversy lies in the scarcity of studies of nonsteroidals specifically within athletic populations. Do these drugs have a place in endurance events? Yes. However, they are clearly not benevolent in all situations.

The pro-thrombotic mechanism documented with the COX-2 inhibitor Vioxx is not presently a class effect, and its ultimate clinical relevance--apart from banning Vioxx--remains under investigation. Still, coxibs cause vasoconstriction and warrant much further study, especially in the light of the recently-suspended Celebrex trial. When co-morbidity factors are present, coxibs should be used only as needed, and in doses lower than previously thought. Furthermore, vasoconstriction can be dangerous for anyone in a PG-dependent state; these runners should avoid NSAIDs before a marathon.

If rofecoxib is any indicator, there may be no grace period, as it were, before cardiovascular complications from NSAID use manifest. Similarly, there appears to be no grace period (or gastric adaptation) for GI complications from NSAID use, which appear in 30 to 50% of those who take them. Coxibs are clearly the better choice for those prone to gastric bleeding and/or peptic ulcers. This may include older runners, because over the age of 60, the body produces less COX-1 enzymes, which are responsible for normal gastric function and platelet aggregation. As older runners are more likely to take additional medications and to see renal changes sooner during an endurance event, this population must use great caution when considering taking a nonsteroidal.

NSAID use alone could theoretically cause hyponatremia, renal failure or kidney disorders in an endurance runner, but these conditions would far more likely be due to other causes and exacerbated by NSAID use.

Among those most at risk:
the salt-depleted
hypovolemics
cardiac patients and those with chronic heart failure
hypertensives
renal disease patients
those with chronic liver failure

Accompanying medications as additional risk:
beta-blockers
ACE-inhibitors
diuretics
for asthma, intranasal steroids

Among those possibly at risk:
first-time runners
female entrants
slower runners
older runners
salty, heavy sweaters
those who have had heart disease in the past

Possible additional risk factors:
high race-day temperatures
heavy humidity

Presently, most doctors agree that after exercise, patients should take a nonsteroidal only as needed. But Scarpa notes some promising preliminary research in preemptive coxib ingestion (before surgery) to curtail swelling and pain afterward. He also tells of one NFL team physician who reported that running backs and linemen who took 40 mg of Bextra two hours prior to a Sunday game were ready for full contact at practice on Monday. Previously, they had needed to wait until Tuesday afternoon or Wednesday to resume full contact at practice.

Nevertheless, the complex matrix of side effects and drug interactions with nonsteroidals calls for other pain relief strategies. Swedish massage and certain topical NSAIDs are two promising alternatives for the relief of delayedonset muscle soreness (DOMS). Similarly, glucosamine/chondroitin sulfate may be as effective against the pain associated with stage-one OA, and better than nonsteroidals at regenerating cartilage--with fewer side effects. (As an unregulated nutraceutical, however, it is important to be sure that patients receive proper doses in o.t.c. products.) Additional strategies should include regular exercise and quadriceps strengthening in the weight room.

In the end and as always, it is vital for patients to know their own bodies, the risk factors in play with all NSAIDs and how they can combine, and to simply have realistic goals. It is not essential for everyone to complete a marathon, and for certain people under certain conditions, it is far from advisable to attempt to do so.

REFERENCES

Berg A, et al. The gastrointestinal system--an essential target organ of the athlete's health and physical performance. Exerc Immunol Rev, 1999 5:78-95.

Garcia Rodriguez LA, et al. Nonsteroidal anti-inflammatory drugs and the risk of myocardial infarction in the general population. Circulation, 2004 Jun 22; 109(24):3000-6.

Kurth T, et al. Inhibition of clinical benefits of aspirin on first myocardial infarction by nonsteroidal anti-inflammatory drugs. Circulation, 2003 Sep 9;108(10):1191-5.

Richy F, et al. Structural and symptomatic efficacy of glucosamine and chondroitin in knee osteoarthritis: a comprehensive meta-analysis. Arch Intern Med, 2003 163(13):1514-22.

Schlienger RG, et al. Use of nonsteroidal anti-inflammatory drugs and the risk of first-time acute myocardial infarction. Br J Clin Pharmacol, 2002 Sep;54(3):327-32.

Shoor S. Athletes, nonsteroidal anti-inflammatory drugs, coxibs, and the gastrointestinal tract. Curr Sports Med Rep, 2002 Apr:1(2):107-15.

Stollberger C and Finsterer J. Nonsteroidal anti-inflammatory drugs in patients with cardio- or cerebrovascular disorders. Z Kardiol, 2003 Sep;92(9):721-9.

Whelton A, et al. Effects of celecoxib and rofecoxib on blood pressure and edema in patients >+65 years of age with systematic hypertension and osteoarthritis. Am J Cardiol, 2002 Nov 1;90(9):959-63.

RELATED ARTICLE: The New View of Celebrex

The news in December that Celebrex caused a 3.4 times greater risk than placebo of cardiovascular events in patients taking 800 mg daily has resulted in a widespread call for the drug's removal from the market. But is such pressure premature?

At 33 months, the average duration of treatment in the trial was certainly long-term. Notably, for those taking 400 mg daily, the risk was 2.5 times greater; however, in another cancer study involving the long-term use of Celebrex, there was no increased risk in patients taking the 400-mg daily dose.

While the National Cancer Institute (NCI) has suspended the first study, Pfizer has halted its advertising of the drug to consumers, and the FDA plans to convene in February to review all COX-2 inhibitors; as of this writing there are no plans to pull Celebrex. Indeed, the doses in the study were two to four times the usual dose for arthritis patients. Celebrex has not been shown to be dangerous when taken at the regular dosage. Pfizer feels that the drug is still an appropriate option for many people.

Though Scarpa says that even one cardiac event certainly demands a careful evaluation of the drug, it's important to avoid wholesale panic. He offers, "When you look at the true numbers, there were 20 cases out of 2,000 patients on the 800-mg dose, and 10 cases out of the 400-mg dose. Remember the normal dose for Celebrex is 200 mg for osteoarthritis. Even at four times the recommended dose of Celebrex, the cv events occurred for two out of every 200 patients. That means that 198 out of every 200 patients taking four times the recommended dose had no problems."

Another recently suspended NSAID study, a three-year exploration of potential Alzheimer's cures, serves to illustrate the complexity of present controversies. Researchers in that study, comparing 400-mg daily of Aleve to Celebrex, had long observed a slightly elevated risk of heart problems and stroke with Aleve, but decided as late as December 10 that the phenomenon was only possibly of statistical significance, and not troublesome enough to stop the study. Only after news broke about the suspended NCI Celebrex study did the NIH halt the Alzheimer's trial--and according to University of Washington researcher John Breitner, M.D., M.P.H., largely because too many patients dropped out of the study. The other point of interest here is that an o.t.c. nonsteroidal proved riskier--if marginally so--than the prescription-only Celebrex, which presented no increased cardio- or cerebrovascular risk in the trial.

Nevertheless, the controversy further demonstrates that these drugs, while safe and effective for many people, can sometimes bring with them serious side effects. Scarpa feels that marathon runners with no previous cardiac abnormalities intermittently using Celebrex at the recommended dosage are most likely not at any high risk. But as always, patients should scrupulously discuss with their physicians what is best for them.

--Jeff Venables

Jeff Venables is the editor of Running & FitNews and a regular contributor to the AMAA Journal.