In defense of aducanumab

The controversial new Alzheimer's drug probably doesn't work and represents everything that's wrong with science. The FDA was right to approve it.

Correction: a previous version of this post misspelled “aducanumab,” which is pretty pathetic because that’s what the post is about. Thanks to the redditor who pointed this out.

Correction 2: I misspelled it again. God damn it.

I. Introduction

On June 7, 2021, the FDA approved Aduhelm (aducanumab) to slow the progression of Alzheimer’s disease. Just one problem: it doesn’t work. As article after article describes, clinical trials showed virtually no efficacy, revealed dangerous side effects, and only made it to the FDA’s desk thanks to a laundry list of statistical shenanigans broadly under the “p-hacking” umbrella. Also worth noting: even in the contrived subpopulation for which it kinda sorta might have helped, it didn’t halt or reverse the disease—merely slowed its progression.

So why was it approved?

Well, speculation abounds, but probably a combination of industry and patient advocacy group pressure. Clearly, one of these groups is more sympathetic than the other. To be charitable, it’s awfully hard to tell millions of suffering fellow humans that their best shot at a normal life doesn’t work (especially when it sorta, kinda might, if you squint at the data).

The not so sub-subtext of all this is that, as the Washington Post writes, “it could carry a price tag as high as $50,000-a-year per patient, according to drug analysts, adding billions of dollars to the nation’s health tab,” putting the approval squarely in the domain of legitimate public discourse.


A few months ago, I wrote “Parler’s cancelling is an opportunity.” The jist was that the deplatforming, clearly a loss for conservatives, was an opportunity for liberals like me to “demonstrate sincerity and expand their political coalition.”

The same thing is happening here, albeit in a less culture-war coded context. Aducanamab’s approval is apparently a win for industry, government corruption, and statistical malfeasance, and a loss for objectivity, good government, and responsible scientific practice. All this makes it perfect opportunity for people who see themselves on the side of good government and responsible science to highlight why—contra the current regulatory regime—patients should be allowed to take dangerous, speculative medications.

My libertarian position on the matter is far from original. Though I haven’t read the book, Jessica Flanigan’s Pharmaceutical Freedom: Why Patients Have a Right to Self Medicate seems to present the case well. The Amazon blurb reads

If a competent adult refuses medical treatment, physicians and public officials must respect her decision. Coercive medical paternalism is a clear violation of the doctrine of informed consent, which protects patients' rights to make medical decisions even if a patient's choice endangers her health. The same reasons for rejecting medical paternalism in the doctor's office are also reasons to reject medical paternalism at the pharmacy, yet coercive medical paternalism persists in the form of premarket approval policies and prescription requirements for pharmaceuticals.

In Pharmaceutical Freedom Jessica Flanigan defends patients' rights of self-medication. Flanigan argues that public officials should certify drugs instead of enforcing prohibitive pharmaceutical policies that disrespect people's rights to make intimate medical decisions and prevent patients from
accessing potentially beneficial new therapies. This argument has revisionary implications for important and timely debates about medical paternalism, recreational drug legalization, human enhancement, prescription drug prices, physician assisted suicide, and pharmaceutical marketing. The need for
reform is especially urgent as medical treatment becomes increasingly personalized and patients advocate for the right to try. The doctrine of informed consent revolutionized medicine in the twentieth century by empowering patients to make treatment decisions. Rights of self-medication are the next step.

[emphasis mine].

To be clear, I’m not dogmatic about this. Convince me that a right to self-medication would increase suffering relative to the status quo, and I would change my mind. For now, though, I basically agree that the current system dictating who can get which medications and how is far too paternalistic and heavy-handed.

Look online, and there is a large constituency of patients struggling to find some doctor, any doctor, to prescribe them some drug they think that they need, whether prescription methamphetamine or plausibly life-extending metformin. Some of them are wrong. They are trying to get their hands on drugs that will do nothing, cause serious side effects, or worse.

But some of them are right, and are currently being prevented by the state from accessing substances that will seriously improve their lives.

Though I’m a utilitarian at heart, some common sense arguments from analogy are particularly convincing. We Americans can legally drink ourselves to death, smoke until our lungs turn black, and drink caffeine until our dopamine and adenosine receptors are fried. But the depressed patient needs pay a psychiatrist hundreds of dollars for the chance to convince her that a certain drug is worth trying? The office worker who has done her research and due diligence can’t get her hands on some modafinil?

The other war on drugs

Without implying false equivalence, the sequence of expensive (in time, money, and dignity) hoops that we have to jump through to access certain medications is not so different from the outright prohibition on “recreational” drugs like marijuana and cocaine. But many card-carrying progressives, liberals, and libertarians who enthusiastically endorse drug decriminalization or legalization don’t even have prescription drugs on their radar screen.

Why you should care

Ok, you might think, what is the big deal with having to go to a doctor to get a prescription? If you need a drug, you’ll get it prescribed, right?

Joe’s right. For one thing, more than 10% of American adults don’t have health insurance, and many who do have high deductibles and copays that make actually seeing a doctor quite difficult.

For another, patient and doctor incentives are not (entirely) aligned. I don’t want to be unduly cynical; in general, I’m sure that both patients and their doctors generally want the former to be “healthier” and feel better. Nonetheless, there are a few dimensions on which their interests likely diverge.

1. Risk tolerance

I can see this one going either direction, but in general patients are rationally more tolerant of risk than their doctors. The reason is simple; patients symmetrically internalize the positive and negative effects from some medical intervention, whereas doctors asymmetrically internalize the negative.

If taking metformin pays off and adds three years to my life, I actually get to live those three years. My doctor, on the other hand, doesn’t get a whole lot of personal benefit from prescribing a drug off-label for speculative reasons even if his patient ends up living a bit longer, especially because neither party will ever know that the drug in fact helped. If I am one of like seven people in the universe who develops lactic acidosis after taking metformin even though the risk is utterly trivial, though, my doctor has to watch out for malpractice action, professional censure, and (perhaps more nobly) the personal guilt associated with failing to “do no harm.”

In fact, I’d speculate that the whole hippocratic oath-inspired ethos has contributed to an inappropriate level of risk aversion (really, what is perceived as risk aversion in a the current socio-cultural context) in medicine. By the parenthetical, I mean that actions that seem to be “safe rather than sorry” aren’t always so safe. For example, only 1% of obese patients receive bariatric weight loss surgery even though the risks of surgery are minuscule in comparison to those of remaining obese. I’m not sure how many obese patients are being prevented from undergoing bariatric surgery by institutional medicine, but I’d be surprised if it isn’t very many.

This gets to my next point…

2. Inappropriately arbitrary degrees of tolerance for different “side effects.”

If you have any interest whatsoever in the depression, medicine as an institution, or psychopharmacology, I highly recommend the Slate Star Codex piece Prescriptions, Paradoxes, and Perversities. Long story short: the antidepressants that patients empirically like most and seem to work best hardly ever get prescribed because, theoretically, if you drink a liter of soy sauce while on them you, you could get a hypertensive crisis.

Now, I’m not saying that “risk of tyramine-induced hypertensive crisis” isn’t worth considering, but for some reason Big Psychiatry has decided that this risk is capital-B-Bad but other risks such as that of post SSRI sexual dysfunction are completely tolerable. Why? Once again, SSC has an answer: the undue deference assigned to flashy, rare side effects (highly recommended as well).

I’ll offer another reason: patients trust themselves more than doctors trust patients. Once Jim finally gets himself prescribed tranylcypromine (one of those antidepressants that never gets prescribed because soy sauce), the ball is in his court to not gorge himself on blue cheese (another potent source of tyramine).

Putting myself in Jim’s shoes, avoiding high tyramine foods like soy sauce and blue cheese seems trivially easy—at least in comparison to, you know, being depressed. Jim’s prescribing doctor, on the other hand, has to trust that Jim is going to follow instructions and lay off the all-you-can-drink soy sauce open bar.

3. Neglect of psychological harm

Suppose Jane’s is utterly convinced that a levothyroxine prescription is the only thing standing between her and a happy, satisfying life, but no one will prescribe it because her TSH is a smidgen inside the “normal range.” Suppose also that she is 100% wrong, and taking the drug won’t help her one bit. The mental anguish caused by not getting to do something she desperately wants to do is suffering every bit as real as that caused by genuine symptoms of hypothyroidism.

But is any doctor going to say “I’m prescribing this drug to relieve the suffering associated with not getting to try taking the drug?” Of course not, which gets us back to this post’s raison d'être, anucanumab.

Back to Aduhelm

Lets say I’m right and anucanumab doesn’t work at all at treating Alzheimer’s disease. Taking anucanumab still might be in some patients’ best interests merely for the associated psychological comfort.

Right now, though, we don’t know for sure that it doesn’t work, so it still might be rational for patients to take if they determine that even a small chance at mitigating the disease is worth the financial cost, opportunity cost, and risk of side effects.

Now, I realize that this puts patients in an exploitable situation. No a die-hard libertarian, I would fully support a ban on advertising Aduhelm, and a big black-box warning on the label saying “THIS DRUG HAS BEEN FOUND TO BE INEFFECTIVE BY THE FDA. TAKE AT YOUR OWN RISK.”

I’d even be in favor of translating Matt Levine’s proposed Certificate of Dumb Investment into the pharmaceutical realm. He thinks that anyone should be able to make reckless, stupid investments (which the SEC often limits to rich people) after the following:

  1. Anyone can also invest in any other dumb investment; you just have to go to the local office of the SEC and get a Certificate of Dumb Investment. (Anyone who sells dumb non-approved investments without requiring this certificate from buyers goes to prison.) 

  2. To get that certificate, you sign a form. The form is one page with a lot of white space. It says in very large letters: “I want to buy a dumb investment. I understand that the person selling it will almost certainly steal all my money, and that I would almost certainly be better off just buying index funds, but I want to do this dumb thing anyway. I agree that I will never, under any circumstances, complain to anyone when this investment inevitably goes wrong. I understand that violating this agreement is a felony.”

  3. Then you take the form to an SEC employee, who slaps you hard across the face and says “really???” And if you reply “yes really” then she gives you the certificate.

Now, replace “SEC” with “FDA” and “rich people” with “licensed medical doctors.” If you want to take a medication without doctor approval, you go to, sign a form that says “I am a reckless idiot who thinks I know better than institutional medicine,” print out a nice pdf, and take it to the pharmacy where they will give you whatever drug you want.

Yes, you’ll still have to pay, but in this dream world of mine we are be incentivizing pharmaceutical innovation with prizes instead of patents, so there will be market forces pushing prices down toward the marginal cost of production.

Obvious subtext: COVID

I’d like to emphasize the effect that these ideas would have had on the COVID pandemic. Some people, after carefully (or not so carefully) analyzing the literature would have decided to take one of the COVID vaccines even prior to FDA emergency use authorization. The motivations for this could be many: perhaps they’re particularly scared of COVID, perhaps they think they’re particularly good at understanding the scientific evidence, or perhaps they just aren’t worried about the risks of mRNA vaccines.

Either way, I think this would have prevented a lot of death and even more unnecessary psychological distress. In this matter, my libertarian side shines through once again; no, the government should not be using their monopoly on violence to prevent people from taking possibly-lifesaving vaccines in the middle of a pandemic. There still would have been supply constraints, but allowing pre-authorization vaccine administration with a “certificate of dumb self-medication” would have shifted the whole vaccine timeline up a few months, not to mention effectively adding valuable safety and efficacy data which other people could use to decide whether to self-vaccinate.


Concession: my subtitle was a little clickbaity

After writing all this, it’s clear to that the FDA wasn’t exactly right to do what they did: put their stamp of approval on anucanumab as a safe and effective medication. What they were right to do was allow anucanumab to be sold and taken by patients. Under the current regulatory regime, then, nothing the FDA could have done would have been appropriate: to call Aduhelm safe and effective is to mislead, but not to call it safe and effective is to enforce a paternalistic prohibition.

In the anti-paternalistic patients-rights libertarian paradise I’ve imagined, Jessica Flanigan’s argument in Pharmaceutical Freedom would come true, and “public officials [would] certify drugs instead of enforcing prohibitive pharmaceutical policies.” In other words, the FDA would keep on certifying drugs as safe and effective, but they would lose their current authority to prevent uncertified drugs from being sold.

Ultimately, even though its recent approval can be attributed to some combination of corruption and incompetence, the millions among us with deteriorating minds should have the right to try anucanumab. And the FDA was right not to stand in their way.