Marine Omega-3 Fatty Acid Supplementation for Borderline Personality Disorder: A Meta-Analysis
JAMES PHELPS, M.D.
01/08/21 - Psychopharmacology Institute
-----------------------------------------------
Key Points:
- In this meta-analysis of fish oil for borderline personality disorder, across 4 randomized trials, the effect size relative to placebo was 0.54 (substantial) and for affective dysregulation, 0.74 (clinically very relevant).
Did you know that the guidelines for Britain's National Health Service recommend not using any pharmacologic treatment for borderline personality disorder? The guidelines go so far as to suggest that any existing pharmacotherapy should be carefully tapered off. In that context then, might you consider fish oil?
In this Quick Take, we'll look at a meta-analysis of studies of fish oil for the treatment of borderline personality disorder. I will start with an apology to all concerned about the label of borderline. It's an unfortunate historical artifact, as you know. I think it's bad PR for psychiatry that we who are so normal will say that you have a personality disorder. Well, to avoid stumbling over the term every time I utter it, I'm just going to call it BPD from here.
What of fish oil then, or more properly, marine omega-3 polyunsaturated fatty acids? The authors of the review collected 5 papers covering 4 randomized trials with a total of 137 patients. For a meta-analysis, that's not very big. So, right off, one could be a little suspect. On the other hand, we esteem the randomized trial for clinical guidance, and a meta-analysis is usually better even with a small N. So, we're listening.
Well, here's the result of that pooled analysis at effect size of 0.54. Remember for effect sizes, roughly 0.2 is small, 0.5 is medium, and 1.0 is large. An effect size of 0.3 is generally regarded as the minimum worthy of clinical consideration, meaning that anything smaller is not likely to be clinically visible. Pooled studies of antidepressants with a very large net sample size yield an effect size of about 0.3.
So, can it really be that fish oil pills have a bigger effect on BPD than antidepressants have on major depression?
Well, imagine the people who will enter a clinical trial for fish oil—probably a relatively less symptomatic person with BPD and likely very hopeful for an effect from fish oil. If participants were able to detect what they were on, fish or placebo, then their very positive starting point could make a big difference. You know, the fish burp thing with fish oil that you can get rid of by putting the pills in the refrigerator in between doses or the freezer if you're desperate? Well, that could be easily detected.
Then there's always the risk of systematic bias on the part of the review, with authors somehow selecting only trials more likely to show benefit. There's a method for detecting such bias called the funnel plot, where outlier studies that are unduly influencing the end result can be identified. So, it's relevant that in this meta-analysis, the 5 studies all fall within the acceptable funnel. Also relevant in most of these studies, people were excluded if they were taking any other psychotropic medication.
Finally, a clinical useful twist from the article is that it includes separate examinations of 3 BPD symptom domains—affective dysregulation, behavioral dyscontrol, and cognitive or perceptual symptoms, including things like dissociation and paranoid ideation. Remember that 0.54 effect size basically pools these 3 domains. Which of these domains would you think that fish oil might affect most strongly? Which domain is the biggest contributor to that effect size, or what kind of symptoms would you expect to show improvement if you and your patient embark on a trial of omega-3 fatty acids?
Well, the answer is affective dysregulation, with an effect size of 0.74. That's really quite large, big enough to be realistically something you can predict that your patient might really benefit from. Well, at the right dose that is. Previous meta-analyses have suggested that for major depression, a minimum of 700 mg of EPA is necessary. That's eicosapentaenoic acid, not DHA, or docosahexaenoic acid, which does not seem to convey benefit. Three of the 4 BPD studies reviewed here used 1000 mg of EPA or more, so 700 to 1000 mg of EPA.
Here's the thing. Fish oil does make a superb placebo if you're, for example, tapering an antidepressant and saw what you hope was a transient worsening and just need a placeholder treatment to avoid going back up on the antidepressant. But if you really want to avoid the placebo effect, then you need to lower the barrier to the right dose. Make it easy. In my hands, this meant going to the grocery store nearest our office and spending a half an hour with a calculator identifying the very few versions of fish oil that could deliver 1000 mg of EPA in 3 pills per day or less and relatively cheaply.
To summarize, in this meta-analysis of marine omega-3 polyunsaturated fatty acids across 4 randomized trials, an effect size of 0.54 was found, with 0.74 for affective dysregulation. According to at least 1 set of guidelines, if you prescribe anything for BPD, it should be fish oil—not antidepressants, not antipsychotics, not mood stabilizers. Well, perhaps maybe some low-dose lithium for suicide prevention, but that's for another Quick Take.
For more on this, the article introduction has a nice review of why fish oil could be considered for BPD, including mechanisms and prior relevant studies.