Prenatal THC exposure yields hyperdopaminergic phenotype rescued by pregnenolone | Hacker News Reader

Prenatal THC exposure yields hyperdopaminergic phenotype rescued by pregnenolone

nature.com

31 pointsbookofjoe67 comments

My wife spent months in the hospital due to Hyperemesis Gravidarum (a disorder during pregnancy that causes continual vomiting and nausea. My wife vomited up to 40 times a day, eventually vomiting blood and breaking vessels in her eyes. She survived on a banana bag and lost 20% of her pre-pregnancy body weight. She said HG feels like being on an unending roller coaster while you're hung over and have the worst flu of your life). We lost the first baby. The second pregnancy was only survivable due to cannabis. It was risk we knew but 7 years later our child shows no signs of developmental disorders. I know, N=1, but still worth sharing.

fellow HG sufferer (survivor?) here. i was on multiple medications since week 5 trying to prevent hospitalization, and still on zofran six months in. careful use of medications can be literally lifesaving.

but... i would say due to easier access and state/local legalization, the perceived use of cannabis has skyrocketed in all my pregnancy groups. much much more than 3 years ago with my last pregnancy, i imagine worlds different from 7 years. most pregnant people seem to be using it casually, even more so than alcohol or tobacco. i do have to wonder what we’ll see from this trend, and i’m generally glad for any research to define safer amounts of consumption.

Yes when my wife was pregnant it was not common at all to use it during pregnancy, and was barely even medically legal outside of pregnancy. She became allergic to Zofran, Diclegis (?) didn't work, and neither did any of the other commonly used HG medicines.

I'm not a neuroscientist, I'm not a smoker, and I'm not a woman. The headline result seems like it could be plausible. However...

My first instinct reading the abstract is that the conclusion which finds an effect on male rats and not female rats is an unjustified subgroup analysis and likely a spurious result.

Given the sexiness of the topic (people think weed is harmless, but what if it's not), publication filters for stat. sig results, low a priori power, no adjustment for multiple comparisons, no pre-registration, and this bizarre subgroup contingent result... should we expect this to replicate?

Is there a strong theoretical basis for expecting the subgroup heterogeneity?

Livers of mammals are highly sexually dimorphic. After testes and ovaries, liver are probably the most sexually dimorphic organ in the body.

https://www.cell.com/cell-metabolism/pdf/S1550-4131(18)30631...

Evidence in mice :

You think that sexual dimorphism doesn't apply to .. brain development? Frankly, it's absurd to claim that's an inappropriate subgroup. There are certainly valid criticisms to this particular paper, but this sort of analysis is standard and eminently justified for mammalian development work.

I'm also not sure what 'low a priori power' means. Statistical power is well defined and depends only on the sampling involved. You don't need to make any assumptions about it, Bayesian or otherwise.

As for the statistics, if anything the low p-values they get are even more impressive given the modest power of these analyses. The effect size must be enormous. Granted the gross PPI response didn't give a dazzling p-value, the followup is quite convincing, showing obvious quantitative changes by a number of metrics, particularly the NAcS Dopamine time-response.

I'm curious where you think a multiple-comparison adjustment needs to be be made. The authors are quite focused on a narrow analysis of the systems implicated in observational PCE studies.

Where I agree is that I would like the authors to have discussed the sexual dimorphism to some degree. It would be handy to know if there is any pre-existing literature that suggests that PCE affects males more than females; all that I could find was one that found effects in females, not males. Someone in the field likely has some reasonable ideas about what might be going on. However, I'd hesitate to say the issue is avoided; Nature has strict length requirements and such a discussion would necessarily be somewhat meandering.

I think there are probably three things I want to follow up on:

1. I do not believe sexual dimorphism is necessarily irrelevant here. I do believe that any subgroup analysis is a researcher degree of freedom and an avenue for p-hacking or fishing. Ideally subgroups would be well specified in advance: particularly the expectation of an effect in male rats and a null in female rats. Given that this did not happen, I would expect in-text justification for the subgroup analysis. I assume there are many brain development processes which are sexually dimorphic, and many which are not. To trust the study I need to know that this is one such case without reference to the study's results. Your end-of-post comment suggests that dimorphism is valid to explore, but the dimorphism should favour effects in the opposite direction, which causes me to further doubt this study's results (in this statement only I am introducing a Bayesian notion of belief)

2. By "low a priori power" I mean that the number of units they chose would typically not be well-powered for modest effects in the subgroup. I am not talking about anything Bayesian. If I knew nothing about this subject, but picked a modest effect size on the quantity of interest and did a power analysis, I would find the study is well under 80% powered to detect the effect. That the effect was detected in the actual study should not lead us to say "wow, if it was underpowered but still detected, the effect must be huge", it should lead us to say "wow, this effect is very sensitive to any kind of confounding". This raises additional study integrity concerns and replicability concerns.

3. There are different points of view about multiple comparison adjustment. In the absence of pre-registration, I am inclined to believe that even closely related hypotheses that are all well justified theoretically should be adjusted for multiple comparisons. The authors report some null results and some significant results, and we are made to infer the null results are true nulls and the significant results are truly significant. The error rate for that joint set of inferences is much higher than the nominal error rate.

To reiterate, my concerns here are exacerbated by the fact that this is a topic likely to get public traction, and which speaks directly to political agendas on both sides of the marijuana policy aisle.

I have limited interest in following up on the neuroscience/biology stuff, since that is not my field. I want to speak only to responsible scientific practice. If neuroscience routinely publishes these kinds low-power multiple-comparison subgroup-analyses, I suspect there will be a replicability crisis in the near future.

> The increased legal availability of cannabis has led to a common misconception that it is a safe natural remedy for, among others, pregnancy-related ailments such as morning sickness.

Cannabis has been widely recognized among stoners as a natural miracle cure for everything from hangovers to stage IV cancer, and because it is produced by "nature" it is thought to be superior to "chemicals" found in ordinary pharmaceuticals.

If anything, legalizing cannabis will increase the visibility of studies like this and give us a clearer picture of its benefits and drawbacks.

> because it is produced by “nature”

Like cyanide and mercury

To make that a more accurate comparison than single compounds, one could say hemlock, datura and galerina mushrooms. But you'd also need to specify, more healthy compared to what? Almost all medicines have a poisonous dosage level. Cannabinoids are some of the least toxic therapeutically active groups of substances in existence.

I think it was Tim Minchin's Storm that brought the quip "You know what they call Alternative Medicine that works? _MEDICINE!_". This is an entirely true statement, but also seems to miss the R&D period. And I'm not specifically saying its a modern R&D period.

It was well known for centuries that willow and aspen tree bark had anti-inflammation properties, so R&D could simply be to test if chewing on branches or bark was what helped with inflammation, or does it need to be from a fresh tree or can a dead tree also provide benefit. Once it's been determine that it's the bark, then can one grind it up and drink it in tea? For centuries, drinking a tea made from some plant material was the cutting edge of drug delivery mechanisms. Then for modern times, what substance within that plant can be isolated that provides the specific function we've observed. Can it be synthesized? Can it be pressed into a pill, or placed in a capsule? Now it's what you and I may observe as modern medicine, but there is a fact that chewing on the bark, the so-called "alternative medicine" is just unrefined medicine, where the level of refinement is determined by your society's manufacturing and science communities. Some day, some culture will probably look back and think that it was uncivilized that we had to go find someone who we needed to convince to dispense physical chemicals that we had to consume in order to feel better, just as we might look at someone chewing on a stick to relieve a headache as wacky. That's what "Asprin" is for.

That said, I'm entirely in the camp that thinks that modern alternative medicine is a sham. Cannabis' biggest problem is obviously political, but it probably also has the biggest potential for being a candidate for research to isolate and properly identify the interactions with the human body today. It clearly has some effect on the body (and mind), but as far as I'm concerned it's flower is still a caveman era drug while CBD oils are essentially Victorian drugs. We might see some kind of modern medicine originate from it, but I'm skeptical for now.

The thing about aspirin tablets vs. herbal tea, is not so much the delivery method itself. It's that the active compound in the tree bark has been discovered, isolated, and dosage-controlled so that you get a fixed amount in each tablet. Note that we still use drinks as a means of dispensing medicines -- consider cough syrup or Alka-Seltzer Plus Cold.

a level of accuracy useful in no context ever

I would have to say your original comparison is fairly useless. You're welcome for the improvement I offered.

Cannabis attenuates inflammation and inflammation is an underlying or contributing cause to just about every disease so the stoners are right, in a way.

https://www.health.harvard.edu/newsletter_article/Inflammati...

That article is from 2006. Is this still just a theory?

Cannabis is from dynastic Egyptian at a minimum.. If its anti inflammatory properties havent changed much since then... I think research from 13 years ago is still okay. We have a problem as a species of assuming anything older than 4 years is garbage..

It's not that it being 4 years old makes it garbage - it's that in domains where there is rapid research and new understanding, something being 4 years old may be inaccurate as that's 4 years new research may have been released.

As far as I can tell from a brief search [0] cannabis is definitely considered anti-inflammatory.

So obviously there's a huge lack of "decent" research in marijuana in general, so which specific chemicals and mechanisms cause this effect isn't widely understood.

Also research suggesting many diseases and ailments are somehow caused by or affected by inflammation (allergies are an obvious example - not that I'm claiming cannabis is a treatment for allergies) is fairly recent. So there may be _some_ merit there.

That said, cannabis is definitely not "a cure for everything" like many stoners like to claim. And I never understood why it being "from a plant" makes it better than other drugs - poison ivy and poison hemlock are both "perfectly natural" but I think you'll certainly find some negative effects from those.

[0] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6308289/

Inflammation is part of the immune response. It's not a cause, it's a symptom. And much like other symptoms like vomiting or fever, the symptom is actually an effective way for the body to fix root causes.

Being in a perpetual state of fever, vomit, or inflammation is clearly an unhealthy state, but that doesn't mean suppressing these symptoms is useful.

Yes and no.

I had inflammation issues due to a back injury that contributed to other health issues. Think of inflammation as a enhancement for other things.

Dysregulation of the inflammatory system is, to a growing extent, being considered as a cause of disease per se. So it's somewhat more complicated than that; it's plausible that the broad anti-inflammatory effects of cannabis could treat a variety of common diseases.

>Being in a perpetual state of fever, vomit, or inflammation is clearly an unhealthy state, but that doesn't mean suppressing these symptoms is useful.

A growing body of research disagrees. For example, this public-oriented article from Johns Hopkins entitled Fight Inflammation to Help Prevent Heart Disease

https://www.hopkinsmedicine.org/health/wellness-and-preventi...

What you're talking about are things like, when you break a bone it isn't necessarily appropriate to "treat" the inflammation because that inflammation is useful to the healing process. It helps bring blood supply, differences in temperature, etc.

It can also have the opposite of the "intended" evolutionary effect, inflammation pressure closing off blood access to the injured location because evolution is often stupid.

> inflammation is an underlying or contributing cause to just about every disease

Inflammation is correlated with many disease symptoms but the causal chain is unclear. In fact there is good research going back decades suggesting that MMPs are part of disease fighting and that suppressing them might make things worse.

This is sort of the inverse of people taking antioxidants in the belief that it would be cancer fighting -- oops.

Part of me suspects this is an over-correction against decades of prohibition and negative misinformation coming from D.A.R.E and other such programs. The legalization movement has been touting myriad medicinal benefits for a while, but that has bled over into the popular consciousness as the idea that weed is a miracle drug.

>Cannabis has been widely recognized among stoners

Well lets also not forget the medical and scientific community has also widely recognized the Endocannabinoid system/Cannabinoid Receptors in the human body.

And although they have generally been prohibited from conducting research and clinical trials with cannabis the potential functions of cannabis receptors in the body include: cognitive, memory, appetite, energy balance/metabolism, stress response, immune system, female reproduction, nervous system, thermoregulation, sleep, and the physiological/cognitive effects of physical exercises.

That’s a problem when Federal rules are held hostage. There’s no science because it’s difficult or impossible to get funding.

"miracle cure for everything" That's my main argument to why it can't be good for them, it's evidence how much the use makes them want to believe.

But on a more cynical note, I get the impression that it might even help fight global warming: just consider how small the ecological footprint of a stereotype stoner failure lifestyle would be compared to a median middle class career. Find a solution to the packaging of spontaneous midnight snacks (sorry for reveling in stereotypes) and it would be about as green as it gets, except suicide. Could we capitalize (well, "ecologize") on that?

  (even more off topic ramblings from here on)

You could set up "fun monasteries", were people could self-admit to dedicate their lives to nonproductive but low-impact activities such as stonerism, abdicating from consumerism, career and procreation in exchange for minima food, shelter and the means to follow their low-impact hobby. Those "fun monasteries" could be set up not just for doing drugs but for all kinds of low-impact but sufficiently addictive activities. Video games, certain sports, artisanship, coding, even for nonreproductive sex, science, or, if you want to go full retro, for religion. I'm sure that they could all be cheaper and lower impact per head than excusing models of welfare that just treat everybody in need as temporarily embarrassed middle class. "fun monasteries" would not replace welfare, they would be an individual alternative to it. Just like the prison system already kind of is. Would there be a mass rush? I doubt it. Butt supporting people who might be willing to severely lessen their footprint in exchange for a nonmonetary stipend sounds like a no-brainer low hanging fruit.

(for those who read till the end and who were wondering if that isn't just a green spin on the "concents" in Stephenson's Anathem, sure, the idea grew when someone mentioned the concents in a different thread on the site. Before that it was just general disappointment in how much we still resent low-impact lifestyle choices, even in circles who outwardly agree on strong measures to fight climate change)

How much THC did they give the rats compared to human quantities? I read an article about monkey brains showing some type of symptom when they looked at white matter under a microscope. Turns out the poor monkeys were given the equivalent of an adult smoking a pound a day.

2mg/kg subcutaneous. I don't know the conversion between subcutaneous and other forms of ingestion, but this would be equivalent of ~150mg of THC for an adult male. I also don't the THC content of weed, but it doesn't sound like an unfathomably high dose. I think the other design concerns are more likely to be the culprit here than external validity worries about dose.

Just for reference, your average edible will come in a suggested dose of ~10mg THC per piece.

Normal weed "flower" is 10-30% THC. So 1 gram of weed is 100-300ish mg of THC. A single prerolled joint is usually 1 gram of flower. So this dose is around 1 joint.

Edit: importantly this info is for smoked flower, edibles and concentrates are a little different because of how they are absorbed.

Smoking an entire 1g preroll by yourself is a huge dose for a human who doesn't have a high tolerance.

I mean, it would depend on the THC concentration, right? A 1g pre-roll with 30% THC would be 3 times as potent as the one with 10%, so I can totally see someone with rather low tolerance smoking a full 10-12% joint by themselves and feeling just fine.

There is a scaling factor between mouse and human dosage that I can't remember off the top of my head, but it's not 1:1.

Looks complicated. https://en.wikipedia.org/wiki/Allometry#Drug_dose_scaling

Whoops, yeah. I meant linear, not 1:1.

Study says: "... we modeled PCE by administering THC (2 mg per kg, subcutaneously (s.c.) once daily) to rat dams during pregnancy (from gestational day (GD) 5 until GD20). This low THC dose does not recapitulate behavioral responses in the can-nabinoid tetrad assay or elicits cannabinoid tolerance after repeated administration, hence it represents a mild insult without any substantial direct impact on maternal behavior. ... In terms of human consumption, this dose is equivalent to THC content in mild cannabis cigarettes (joints) (5%), since the average THC content in illicit cannabis preparations has significantly increased in the last two decades (from ~4% to ~12%)."

The figure quoted is a concentration (5%), not a dose.

How much actual THC did the rats ingest? That's the question.

Fair enough, another commenter already gave 2mg/kg, but I should have quoted it too really. Edited with more detail.

What does this mean in plain terms?

You usually shouldn't draw such conclusions from single studies. This was a rat model studying a specific gene and a drug therapy for it, it might be one datum among many if you were surveying the literature to try to draw conclusions about plain terms conclusions to extract.

A quote from the abstract "Emerging clinical evidence, however, indicates that prenatal cannabis exposure (PCE) predisposes offspring to various neuropsychiatric disorders linked to aberrant dopaminergic function." is about the best you can do. There is growing clinical evidence that cannabis exposure during gestation leads to problems with dopamine.

So, correct me if I am wrong, but to summarize: Smoking marijuana while pregnant is bad for the baby?

Possibly. There is growing body of evidence it might be.

Please confirm: can I counteract the effects of marijuana on the developing fetus with Cocaine? My drug friends tell me they act like opposites in the body. Thanks

Smoking anything during pregnancy is bad for the baby.

You really shouldn't sort things into "good" and "bad" bins. The world is more complicated than that.

You're wrong to assume so on the basis of this study. It might be true, but this study doesn't show that.

possible "neuropsychiatric disorders linked to aberrant dopaminergic function."

I took this to mean: possible problems regulating/utilizing dopamine which could mean a whole lot of things. increasinly predisposed to things like depression, difficulty breaking addictions that supply the user with dopamine, etc.

THC should not be used during pregnancy.

The longer version: there is evidence of a pathway for THC to be harmful during pregnancy. Much more study is needed. In the meantime, since THC is not needed during pregnancy, the risks clearly outweigh the benefits and pregnant persons should avoid THC.

Further research needed to say anything of significance, unless you're especially concerned about your pregnant rat's marijuana habit.

Probably don't smoke a lot of weed during pregnancy? But I think that was already more or less a given.

I hope "Don't smoke when pregnant" goes without saying. The mother's blood IS the baby's blood during pregnancy, anything the mother takes the baby gets. That's why 90% of medications are unsafe to take when pregnant.

I would hope this would be common knowledge.

There's more to it than that - 90% of medications are untested for their effects during pregnancy because doing studies on pregnant women is seen as unethical/too-risky. This is a huge problem since pregnant women have no idea what is safe to take. Some of it is _probaby_ safe, but we have no idea.

> The mother's blood IS the baby's blood during pregnancy, anything the mother takes the baby gets.

This is not really true[0]:

> In the villi, these vessels eventually branch to form an extensive arterio-capillary-venous system, bringing the fetal blood extremely close to the maternal blood; but no intermingling of fetal and maternal blood occurs ("placental barrier").

There is some pass-through effect, but it is not 100%, and varies depending on the particular chemical.

[0]: https://en.wikipedia.org/wiki/Placenta#Fetoplacental_circula...

Not the full article but a longer summary of it: https://medicalxpress.com/news/2019-10-thc-rat-pregnancy-res...

Only in male mice, according to the abstract. The arbitraryness of it makes me suspicious.

I might be misunderstanding you or the abstract but I think the effect occurs in only in male rats. It isn't that they only tested male rats.

Yes, the effect only occurs in male rats. It's extremely specific which is why I'm suspicious.

How did they manage to get male rats pregnant?

I think male children of pregnant mother rats.

Does it matter? The notion of biological sex is a big debate right now within certain communities.

It’s my understanding that many mental health issues impact men and women very differently.

> Here, we show that male, but not female, offspring of Δ9-tetrahydrocannabinol (THC)-exposed dams, a rat PCE model, exhibit extensive molecular and synaptic changes in dopaminergic neurons of the ventral tegmental area, including altered excitatory-to-inhibitory balance and switched polarity of long-term synaptic plasticity.

Very possible that this effect will not be observed in humans. Certainly not a good reason to cease cannabis use if it's part of your pregnancy or birth plan, coordinated with experienced prenatal and neonatal professionals.

The opening of the abstract costs this study substantial credibility in my opinion:

> The increased legal availability of cannabis has led to a common misconception that it is a safe natural remedy for, among others, pregnancy-related ailments such as morning sickness.

There is no source provided for the claim that increased legal availability has led to a change in perception, but the use of this phrase to open the abstract betrays an unambiguous agenda.

I'll wait until more human longitudinal studies (the current corpus of which includes many which appear favorable to cannabis as part of a prenatal regimen) before changing my mind.

Reply on news.ycombinator.com