Five lessons from the Niacin failure

This week the New England Journal of Medicine published two studies on the role of niacin in preventing future cardiac events. The short story is that niacin was ineffective and associated with significant harm. A number of excellent summary pieces have been written, and I will reference them at the end.

The purpose of writing my thoughts on this matter is to put these unsurprising results into a larger context of health, and also to consider the changing role of the physician.

Lesson 1: Association does not equal causation

For years, medical experts have observed that patients with high HDL levels (good cholesterol) had lower rates of cardiac events. It made sense because HDL serves as a sort of cholesterol scavenger agent—the more HDL one has the less cholesterol available to get deposited in the artery. So it is mostly true that high HDL levels associate with good outcomes. That is much different than saying high HDL levels cause good outcomes.

The niacin failure, taken together with failures of many other potent HDL-raising drugs, strongly suggest that the relationship of high HDL and good outcomes is not causal. HDL may be a risk marker but it is not a risk factor. This theme will come up again in Lesson 3.

Lesson 2: Be careful with surrogate markers

You might wonder how niacin, a drug with such lousy results, got so well established in medical practice. The simple answer is that we thought moving numbers on lab tests would improve future outcomes. Niacin, like many other cholesterol-lowering drugs, is indeed able to change levels of cholesterol. This got it attention and approval, for it was assumed that these effects would be good.

The problem is that changing surrogate markers does not always change outcomes. Recall that the purpose of prevention of heart disease is not to lower cholesterol levels (or blood pressure for that matter) but to decrease future heart attacks, strokes and death.

I have long felt that the medical establishment fails to see the obvious surrogates of health: body weight, belt size, mobility and time spent smiling, for instance. Instead, we get bogged down in particle sizes of this and that molecule.

Before future drugs, procedures or surgeries get anointed, they should be shown to either safely and effectively relieve a symptom, or definitively reduce bad outcomes–not surrogates.

Lesson 3: Pills do not confer heart health – not even vitamins

Niacin is a vitamin. It failed. Essentially every study of every conceivable vitamin supplement has failed to show real health benefits. The vitamin D story is instructive. Patients with low vitamin D levels have higher rates of bad health outcomes. But supplementing vitamin D has not been shown to improve hard outcomes. That’s because patients with low vitamin D levels have such levels because they are ill, often immobile, overweight and frequently not outdoors playing. And treatment with a vitamin D pill does not change these important things.

But it’s not just vitamins like niacin that fail to improve health. I recently wrote a post expressing doubt about the role of statins in preventing future events. Millions of patients without heart disease are treated with statins in the hope that lowering cholesterol levels will extend life. But it doesn’t happen. Statin drugs, given to patients without heart disease, do not improve mortality. They do reduce the risk of a future heart attack, but by a measly 1 in 200, which means 99.5% of patients taking the drugs for primary prevention get no benefit (but all the risks).

The problem with using drugs to prevent heart disease is it distracts both patients and doctors from the obvious: that good health comes from making good choices. The defeatist attitude that people can’t help themselves is ridiculous. In the statin post I discuss the hypothesis that using drugs for preventing heart disease may interact negatively with lifestyle factors, which are most important for health. If patients on statin drugs move less and eat more, it becomes easy to see why the drugs do not confer significant long-term health benefits.

Lesson 4 – Do no harm

Nearly every day I see people with medical or surgical complications from therapies given for lifestyle-related problems. Maybe it’s a low potassium or sodium level due to a diuretic used for hypertension, or a fall due to low blood pressure from a blood pressure drug, or pneumonia due to immobility from a statin-induced myalgia. This list is endless.

When we intervene, we risk doing harm. No action is free. Tradeoffs are ever-present. And nowhere should this rule be more front and center than when a person tells you: I feel well. I have no complaints. Our goal here would be to not mess that up.

Lesson 5 — Learn from mistakes

Making mistakes is something all doctors try to avoid. It was awful that millions of patients were exposed to the risks and costs of niacin. But the upside of mistakes are what we learn from them. This one will teach us a lot–if we let it.

JMM

References:

Journalist Michael O’Riordan on theHeart.org — HPS2-THRIVE: As Niacin Fails, HDL’s Role Is Debated Anew

Cardiologist Harlan Krumholz in the New York Times — 3 Things to Know About Niacin and Heart Health

Journalist Larry Husten on Forbes (with an excellent comment from Dr. Sanjay Kaul) — New Evidence Fuels Concerns About The Safety Of Niacin

Comments

  1. Pete says

    Wow. Thanks for commenting on this. I’ve been waiting for someone from the medical community to step up and clearly state what most are missing. In less than 1000 words, you laid out probably the five most important considerations that doctors should think about when prescribing medicine (in my opinion as a patient and layman). What a great preface for the book you should be writing based on your writings in your blog.

    • jane says

      I second that – with all the bullying, fearmongering and nonsense being spread by cardiologists who couldn’t calculate an NNT to save their lives, a book by Dr. John would be an invaluable and possibly unique resource for patients. You really should write one!

  2. says

    Thanks John – very wise words. We are now at 20 large well designed studies (see below) in a row from the last 5 or so years showing no benefit from changes in surrogate markers. This whole thing really needs more press.

    Twenty plus studies in a row showing no benefit is not a trend; it is not an interesting signal. In my opinion, it should be an absolute wake up call to all of us that there is something wrong with at least some of our current approach to and beliefs around blood pressure, cholesterol and glucose.

    4 studies for blood pressure – ALTITUDE (aliskiren); VALISH, AASK, ACCORD
    (aggressive blood pressure lowering)

    4 studies for cholesterol – AIM-HIGH, HPS2-THRIVE (niacin); ACCORD (fibrates); dalOUTCOMES (dalcetrapib); STABILITY (darapladib)

    8 studies for type 2 diabetes – ACCORD, ADVANCE, VADT
    (aggressive glucose lowering); ROADMAP (olmesartan); ORIGIN (insulin); SAVOR-TIMI 53 (saxagliptin); EXAMINE (alogliptin); ALECARDIO (aleglitazar)

    3 studies that have an effect on different surrogate markers – ACTIVE (irbesartan); CRESCENDO (rimonabant); VISTA-16 (varespladib)

  3. says

    Not long ago a physician acquaintance went out on what I thought a dangerous limb when he voiced concerns about the approval process for HPV vaccines. Since so many of us have a reflexive distaste for the anti-vac crowd I was dubious. If certain strains of HPV are associated with cervical/oropharyngeal cancer, and vaccine demonstrably decreases infection rates, should that not be enough justification for the vaccine?
    But the vaccine has not (at this point) demonstrated a benefit to the endpoint, and it’s good to be reminded we’ve got to keep our eye on the ball.

  4. Jeff Patten says

    Dr. J, your writings ought to be read as a daily prescription. Always a fresh outlook. Almost always an affirmation of my own unthought inklings.

    One phrase today causes unease. For a new ‘whatever’ to be shown to “definitively reduce bad outcomes” means that enough research time must pass for a substantial percentage of thousands of research subjects to die.
    “End points” is a euphemism, right?

    OK. It’s what’s needed to be certain.
    It’s just that those of us who might benefit from the new stuff must watch the research proceed as the years, often decades, disappear.

    Time runs out.

  5. says

    Hi John! This article was excellent. I am sending Ken to his cardiologist with a copy of your article! We also took special note about how much a person smiles, I certainly believe that is the actual, best medicine. Ken stopped taking his Niacin a short while ago, but he is still on a statin. Thanks! Dara

  6. says

    I appreciate the sharing of this info. Recently, in my research, I was fascinated to learn that several safety measures, such as helmet and seat belt laws, did not actually change our fatality or injury numbers much (if at all… and in fact some were worse) because people were then taking more risks on their bikes or driving faster. I can now see this phenomenon in every aspect of our lives. These “solutions” are quite often motivating us to be irresponsible for ourselves.

  7. Angela says

    This is my absolute favorite paragraph out of the whole article:

    “Essentially every study of every conceivable vitamin supplement has failed to show real health benefits. The vitamin D story is instructive. Patients with low vitamin D levels have higher rates of bad health outcomes. But supplementing vitamin D has not been shown to improve hard outcomes. That’s because patients with low vitamin D levels have such levels because they are ill, often immobile, overweight and frequently not outdoors playing. And treatment with a vitamin D pill does not change these important things.”

    I just completed the certificate in Plant Based Nutrition course through ecornell and this right in with the curriculum. It does make me wonder whether there’s any other options for us on B12 though and whether we’re really benefiting from supplementation.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Notify me of followup comments via e-mail. You can also subscribe without commenting.