Hacker News Reader

Comment by dan-robertson | Hacker News Reader

The methodology here is typical for testing antidepressants with animals but many people find it somewhat horrifying:

First the mice are made to be depressed by exposing them to constant sounds of birds screeching and other predators. Next some are given the intervention. The level of depression is then measured by putting the mice in a container of deep water from which there is no escape, and timing how long until the mouse gives up swimming. The mice are not left to drown but are fished out and killed in a presumably more-humane way (chronically depressed mice are not otherwise useful).

A second ethical issue with studies with mice is that occasionally the unfortunate phd student tasked with dispatching the mice will not have been taught any way to do so and will have to try to come up with something on the spot. This can be quite traumatic for both parties.

View on HN

I’m sorry but I’m having trouble believing the second part. Every university I’ve worked at has extensive training and monitoring for animal research, and there’s no way any of them would have approved what you’re describing. Screwing around with something like this could get the entire university’s federal funding cut off, and so they take compliance very seriously. Secondly, the brains are often collected as part of the experiment, and these samples would absolutely be destroyed by someone “coming up with something on the spot.”

I was so annoyed about 2nd part that I almost forgot to mention that the first part is untrue too. There’s no mention of predator sounds whatsoever in the paper. The standard forced swim test model just looks at how long the mouse actively tries to escape vs just floating too; there’s no drowning involved.

tl;dr: Lies

I understand the parent is wrong and that correcting them was needed, but I find all this hand-wringing about ethics in animal experiments kind of absurd, since it's so obviously fundamentally unethical in the first place (or at least I know of none that aren't unethical). I suspect it's mostly just a defense mechanism to shield the human mind from the trauma of continuously inflicting mass suffering and death, much like how the Nazis established gas chambers as a less traumatic way of perpetrating genocide compared to firing squad units going village to village and potentially developing symptoms of PTSD.

In both cases, zero or less value is attributed to the lives and minds that are being assaulted and extinguished on a loop (though the value is below zero in the Nazi case and sits roughly at zero in the animal experiment case). This isn't entirely fair, since I'm sure most researchers do attempt to prevent animal suffering from being the most utterly barbaric it can be due to ethical rather than selfish reasons, but I think anything that's moderately barbaric or below is still considered justifiable in their minds and I think attempts to "soften" it do mostly fall into the selfish defense mechanism category.

I understand the great benefit to scientific progress afforded by such experiments, but there was some long-term benefit from human experimentation during WWII, as well, and there probably could've been much more future benefit if those regimes had won.

Nonsense.

I don’t know any researcher that likes sacrificing animals though every single one is painfully aware of the moral weight. They also go out of their way to minimize suffering, for ethical reasons, for selfish reasons, and even to improve the quality of the data. The reason I jumped down the parent's throat is because it reinforces the mistaken idea that scientists are okay with “moderately barbaric” things and indifferent to suffering. We aren’t!

And what’s the alternative to animal research? Computational modeling isn’t there yet and in vitro experiments have pretty stark limitations too. In any case, both would need animal data to determine if they’re correct. Waiting for another Nazi regime[0], as you proposed(!), doesn’t seem like a viable—-or ethical—-option.

[0] Incidentally, I have read that most of that “data” was scientifically useless, not just because of its moral provenance, but also because the stuff that was done was more akin to torture than controlled research.

I'll grant that most of them don't like it, but there's a far too high fraction that uses animals even though they do not need to. This is a constant and ongoing point of contention between animal rights groups and large biologically-focused universities. Generally it comes down to budgeting concerns (ie it is cheaper to use animals) or time concerns (ie it is faster to use animals).

Whether or not one of those two arguments is compelling to you is another story.

Can you give some examples?

My impression is that animal work is literally never the cheap or fast option. It's slow, it's labor-intensive, and it's expensive, while career incentives favor doing things that are quick and cheap.

It's certainly true that someone could have run a version of this study where patients with depression were "prescribed" different doses of caffeine and asked to fill out rating scales to quantify its effects on their symptoms. This experiment could reveal its effect on symptoms, the headline of this article, but it wouldn't suffice for understanding the pharmacological mechanisms (see Table 4). I don't think there are PET ligands available for any of those metabolites, and even if there were (and ignoring the animal work needed to develop them), injecting people with radioactive tracers also raises ethical questions.

> The reason I jumped down the parent's throat is because it reinforces the mistaken idea that scientists are okay with “moderately barbaric” things and indifferent to suffering. We aren’t!

I never said they were okay with it. Merely that they're still participating in it regularly, and need a way to justify it.

> A second ethical issue with studies with mice is that often the unfortunate phd student tasked with dispatching the mice will not have been taught any way to do so and will have to try to come up with something on the spot. This can be quite traumatic for both parties.

uhhhhhh are you sure about that?? This strikes me as a flagrant violation of policy. Animal studies are heavily regulated.

In my experience, (and mind you I've only ever worked adjacent to animal labs, not directly in them) there is always a canister of CO2 and a euthanization chamber on standby in experiments where there is a risk of injuring the mice.

While this is standard, it shouldn't be.

Mammals sense CO2 buildup, not oxygen depletion. By smothering with CO2, the mice spend their last moments desperate to breathe and unable to.

If they were to use nitrogen instead, the mice, having no way to know they've been deprived of oxygen, would just pass out and die.

There's a good chance I have that wrong and they use nitrogen. Like I said, I don't work in animal labs, just adjacent to them.

I knew someone who worked with lab rats; her job was to kill them and she had to use CO2.

afaik (second hand) CO2 is used for safety reasons, humans can sense a CO2 poisoning but not an excess of nitrogen.

As for methods here's the list: https://animal.research.uiowa.edu/iacuc-guidelines-euthanasi...

In my work, our approved methods were isoflurane and decapitation. In some types of experiments CO2 will interfere with your result.

Probably should substitute “often” for “occasionally”. I’ve edited the comment.

I suppose there are other factors too, eg the student not thinking to say they don’t know how to kill a mouse and their advisor not thinking to ask.

Ethical horrors aside, I'm a bit confused by their definition of depression here. I think there's a solid gap between 'constantly exposed to predators and thus feeling like there's no chance you'll survive' and 'feeling like life is just a black void of emptiness'. The latter being a very, very rudimentary description of depression that I've heard from friends.

Is there a scientist in the thread who could clear up why a mouse's will to persevere and live is equated with how depressed it is? Do mice just not perceive circumstance as well?

That description of the experiment isn’t correct. They’re not using predator calls or anything to induce depression.

Instead, the readout is how long the mouse spends scrambling to escape vs. floating immobile in the water, usually after a “training run” that demonstrates to it that it can’t escape. You can draw some vague parallel to “coping with adversity”, but the test’s value is mostly that historically, it has predicted drugs that seem to help human patients with depression: mice receiving anti-depressants tend to spend less time immobile.

Ah, okay, that makes sense, thank you for clearing up my confusion. Basically, they're looking for the most effective substances to boost activity, which is one of the things that anti-depressants are supposed to help with.

It’s more that substances boosting activity also help with depression: nobody really knows the causal direction, or even if there really is one.

That said, one way to botch these assays is to do something that just directly increases locomotion. They did look at that in Table 1 though.

This is so wrong I don't know where to start! All of this kind of work is highly scrutinized, you can check it out by starting your own work on reading up on IACUC.

Here are the accepted methods for "dispatching the mice": https://animal.research.uiowa.edu/iacuc-guidelines-euthanasi...

You can decide whether you think this is satisfying to you, or you can consult the bioethicists that helped develop these protocols.

As to your second part, that shouldn’t happen. Animal welfare is highly regulated and monitored. Every university has strict rules about animal tests and every experiment has to be evaluated by a panel to make sure the animals aren’t mistreated AND that any limited suffering is justified by the value of the data.

If any PhD is making up ways to kill their animals, then they either didn’t pay attention or work for a crappy lab.

Sounds like the effect of an open office

Reply on news.ycombinator.com