Can We Prevent Psychosis With...Fish Oil?
A decades long back and forth that I try to sort out, desperately.
The Frontier Psychiatrists is a health themed newsletter written by a physician1. I'm Owen Scott Muir, M.D. I am the sometimes silly proprietor of this news-establishment.
Preventing psychotic illness would be awesome. It can be disabling—particularly whatever schizophrenia is! No one wants psychotic experiences they can’t control. I'll take the time to define what I mean by that. The premise of this article is to struggle with my readers to interpret data, together. I want to answer a question—is there a reliable and low enough risk way to prevent psychotic illness from developing in the first place?
Might it be as easy as supplemental fish oil?
Well, that possible option has been proved. Then disproved. Then proved again. It’s been decades of scientific back and forth. Let’s dig in.
What is psychosis?
The national institutes for mental health define psychosis thusly:
People with psychosis typically experience delusions (false beliefs, for example, that people on television are sending them special messages or that others are trying to hurt them) and hallucinations (seeing or hearing things that others do not, such as hearing voices telling them to do something or criticizing them). Other symptoms can include incoherent or nonsense speech and behavior that is inappropriate for the situation.
There is some degree of culturally-bound interpretation required to understand whether something is a bizarre belief, which is why we don't just use those symptoms alone to make a diagnosis of a disorder. Often times, the belief is queried by obtaining collateral information, from other people. Clinicians look for other symptoms like disruption in language, or physical exam findings, along with other neurocognitive deficits. People who are psychotic in the “schizophrenia” Or “psychotic mood disorder” sense of the term don't just believe strange things. Their brains work differently in other ways, too. History, timeline, and what someone was like before all play important roles in these assessments.
For example, Kanye West is …grandiose. He's also a rapper. Grandiosity is normative for rappers. Particularly in the context of rap songs, grandiosity is culturally normative. Kanye may have other problems, as I've written about previously. We get back to Kanye later, but not to address any diagnostic elements, we're just going to use his music to illustrate points about science—again (a previous hip-hop informed education piece is available to subscribers here).
What is “high risk for psychosis?”
There is a scale.2
They're actually a bunch of scales. There are whole programs dedicated to determining whether people are at high-risk for psychosis, but it's a combination of genetic loading—family members with schizophrenia and associated illnesses— as well as other factors like declining academic performance that we look at. I'm not gonna go into detail here, because it will derail us from the main thrust of this article. Suffice to say, there's a population where we can predict they will develop a psychotic illness soon, some of the time. Given that risk profile, it would be nice to prevent that bad outcome.
What is Fish Oil?
Our neurons (and all cells!) have a cell membrane made up of a phospholipid bilayer (charged phosphorous atom on the outside, facing the polar compound that is water, and also on the inside of the cell). A primer on neuron function is available here.
This is a graphic:
When we zoom in on the cell membrane, which matters for all of human physiology, we see the following:
To make that graphic a little bit more obvious, I've added an annotation:
Now, another to pull it together…
It turns out neurons already have oils, like fish oil, as in the graphic below, as part of those most healthy cell membranes:
The basic science? Since our cells are made up of these fatty acids, and we can augment people with more of the healthier fatty acids…might we help people who have a brain problem (psychosis) by making their brain cells healthier? This might be accomplished by dietary supplements of more good biochemical building block3? The fish oil we ingest has a ratio of these oils, and the ratio is thought to matter for the purposes of the health of one’s brain.4
I love organic chemistry, with background on the topic available here.
Why did we think fish oil could help?
As far back as 2004-2007, Paul Amminger and colleagues were running randomized controlled trials in Vienna. That early study had only 81 individuals, followed over 12 months, in which 2 of 41 individuals (4.9%) in the ω-3 group and 11 of 40 (27.5%) in the placebo group had transitioned to psychotic disorder (P = .007).5
This is a flashy finding, and they followed up this initial cohort for 6.7 years, and published the results in 20156, thanks to my prescient educational material for paid subscribers, we all now know how to interpret the following Kaplan-Meyer survival analysis:
Again, very impressive data, about a very small sample. More randomized trials were done, with mixed results. If only we had a tool, some kind of analysis of analyses, to make sense of this!?! Oh, right, the meta-analysis7, as published by Montazer et. al, makes it clear that no other group got the bonkers good outcomes of Amminger’s initial study, for psychosis…this plot only meta-analyses the psychosis rating scale, PANSS, data:
I will annotate it for you, dear readers, because I like you that much:
However, it appears when looking at the broader global assessment of functioning “GAF” measure—a way of globally assessing functioning, not limited to psychosis or not psychosis—some benefit was demonstrated in the studies that reported this measure:
Here is what the GAF score measures:
Can you imagine life’s problems never seeming to get out of hand?
That seems pretty sweet.
Keep in mind, that it is just “taking fish oil” that is leading to improvements in this measure for individuals at risk for psychosis.
There followed… tons of research for years!
Tons of studies—some of which even evaluated large cohorts with blood levels of fatty acids (This analysis is based on 4720 children who completed the Psychosis-LIKe Symptoms interview (PLIKSi) at approximately 18 years of age.):
There is no strong evidence that total plasma Ω-3 FA levels or Ω-6/Ω-3 ratios in childhood and mid-adolescence are associated with increased risk for PE's or psychotic disorder, but very marginal evidence that alterations in the Ω-6 pathway at developmental time points might influence risk8
Which is more data that fatty acid levels might matter, but only a little. It doesn’t answer the question “what should we do” yet.
What is the verdict? Is Fish Oil Good or Not?!?
We have a new article to help:
Network meta-analysis indicates superior effects of omega-3 polyunsaturated fatty acids in preventing the transition to psychosis in individuals at clinical high-risk
Oh my. A network meta-analysis! I don't know what that is!?!
This is the sort of science article that has definitions that don’t, by themselves, answer the question of what the heck they do:
Network meta-analysis is a technique for comparing multiple treatments simultaneously in a single analysis by combining direct and indirect evidence within a network of randomized controlled trials.9
In fairness, the level of abstraction is confusing, to me. I'm a smart enough person. You might be too. And highly abstract concepts are still hard for humans to have a gut sense of. Suffice it to say, the Cochrane Group has helpful training for my large-scale epidemiologist readers who want to brush up. It allows for multiple comparisons of things only comparable indirectly, and much like iterative approximations get us to the central limit theorem in calculus, a network meta-analysis is an approach that accepts the risk of baked-in bias in exchange for being able to more robustly compare treatments and create hierarchies when it comes to “rank-ordering” treatment options. They also feature snazzy diagrams that look like this:
And can be broken down to their simplest form thusly:
Now we have the idea: “It’s a study that compares studies directly and indirectly to infer when one treatment is better than another.” As a pop culture example, it’s the kind of inferential logic used by…
Kanye in “Golddigger”…the point of which is to help his audience infer, despite his protestations (“I ain't saying she’s a golddigger…”), that the woman in question? She is very much a “golddigger.” Mr. West is using the emotional force of inferential logic. He is insincere that he “ain’t saying” it,— he’s using a network meta-analysis framework to convince us:
If you [have something to do] with this girl, then you better be paid
You know why? It take too much to touch her
From what I heard she got a baby by Busta
My best friend said she used to [have a romantic relationship] with Usher
I don't care what none of y'all say, I still love her
If the above data were used in a network meta-analysis, we’d use
the prior child with presumably wealthy rapper Busta [Rhymes], and
the prior romantic relationship with Usher
to infer, given wealthy rapper B and wealthy rapper U, as prior relationships, even absent the direct comparison, the following data:
Cutie the bomb, met her at a beauty salon
With a baby Louis Vuitton under her underarm
Adds to the picture. That is to say, her love of expensive LV purses can be indirectly added to the data regarding her prior relationships with wealthy B and U, to draw the inference that the woman Kanye is describing might be interested in his money. I asked DALL-E to make a version of this chart, as it applies to the song, and we can see AI still has a long way to go.
Now that we have a sense of what’s going on, and we can “get down wit it,” let’s take a look at the network meta-analysis authored by Chen et al10
They compared many studies. Studies that tried to prevent psychosis with fish oil, studies that tried prevent psychosis with psychotherapy, studies that tried to prevent psychosis with medications, and with combinations of the above. In summary, they found a protective effect for fish oil:
Results: Out of 11,417 identified references, 21 studies were included, comprising 1,983 participants. CHR-P participants receiving omega-3 polyunsaturated fatty acids treatment were associated with a lower probability of transition compared to placebo/control at 6 months (OR 0.07, 95% CI 0.01 to 0.054), 12 months (OR 0.14, 95% CI 0.03 to 0.66), and ≥24 months (OR 0.16, 95% CI 0.05 to 0.54). Moreover, risperidone plus psychotherapy, was associated with a lower likelihood of transition at 6 months in comparison to placebo/control plus psychotherapy, but this result was not sustained over longer durations.
Unlike the prior non-network meta-analysis, they found that beyond the general assessment of function improvements, there was a reduced risk of conversion from non-psychotic to psychotic individuals at high risk thanks to the larger sample size of all the data together. I'm not saying the abbreviations are confusing, but I promise they're not worth your time unless you're really into this topic. In that case, the full text is here.
The bottom line:
For now, fish oil probably helps reduce the risk of conversion to psychosis in high risk individuals.
Now, feel free to become a paid subscriber, so I keep writing these sassy articles!11
Prior articles about antipsychotic medicines include great reasons to subscribe to this amazing resource for your life and or practice:
Risperdal (with a podcast version)
Clozapine, Part 1.
A Review of Tardive Dyskinesia
Finally, some of my own journey with obesity treatment after antipsychotic treatment.
Thanks for reading, and consider becoming a paid subscriber, and sharing it your friends and enemies. They need engaging educational content, too.
With some guest authors.
Paolo Fusar-Poli, The Clinical High-Risk State for Psychosis (CHR-P), Version II, Schizophrenia Bulletin, Volume 43, Issue 1, 1 January 2017, Pages 44–47, https://doi.org/10.1093/schbul/sbw158
Farooqui, A. A. (2009). Beneficial effects of fish oil on human brain (pp. 151-187). New York: Springer.
Assisi, A., Banzi, R., Buonocore, C., Capasso, F., Di Muzio, V., Michelacci, F., ... & Garattini, S. (2006). Fish oil and mental health: the role of n-3 long-chain polyunsaturated fatty acids in cognitive development and neurological disorders. International clinical psychopharmacology, 21(6), 319-336.
Amminger GP, Schäfer MR, Papageorgiou K, et al. Long-Chain ω-3 Fatty Acids for Indicated Prevention of Psychotic Disorders: A Randomized, Placebo-Controlled Trial. Arch Gen Psychiatry. 2010;67(2):146–154. doi:10.1001/archgenpsychiatry.2009.192
Amminger, G., Schäfer, M., Schlögelhofer, M. et al. Longer-term outcome in the prevention of psychotic disorders by the Vienna omega-3 study. Nat Commun 6, 7934 (2015). https://doi.org/10.1038/ncomms8934
Montazer M, Ebrahimpour-Koujan S, Surkan PJ, Azadbakht L. Effects of Fish-Oil Consumption on Psychological Function Outcomes in Psychosis: A Systematic Review and Dose-Response Meta-Analysis of Randomized Controlled Trials. Adv Nutr. 2022 Dec 22;13(6):2149-2164. doi: 10.1093/advances/nmac083. PMID: 36166847; PMCID: PMC9879727.
Thompson AD, Jones HJ, Heron J, Hibbeln J, Sullivan S, Zammit S. Omega-3 and Omega-6 fatty acids and risk of psychotic outcomes in the ALSPAC birth cohort. Schizophr Res. 2020 Oct;224:108-115. doi: 10.1016/j.schres.2020.09.018. Epub 2020 Oct 14. PMID: 33067055; PMCID: PMC7738752.
Rouse B, Chaimani A, Li T. Network meta-analysis: an introduction for clinicians. Intern Emerg Med. 2017 Feb;12(1):103-111. doi: 10.1007/s11739-016-1583-7. Epub 2016 Dec 2. PMID: 27913917; PMCID: PMC5247317.
Chen C, Deng Y, Li Y, Zhang M, Yu T, Xie K, Bao W, Li P, Sun L, Zhang T, Zhu Y, Zhang B. Network meta-analysis indicates superior effects of omega-3 polyunsaturated fatty acids in preventing the transition to psychosis in individuals at clinical high-risk. Int J Neuropsychopharmacol. 2024 Feb 26:pyae014. doi: 10.1093/ijnp/pyae014. Epub ahead of print. PMID: 38408281.
Yes, this one took a while, but was completed between 7am and 1pm EST on March 18th, 2024.