A post circulating on Facebook claims lithium helps with autism, citing a real PubMed ID (31308580). The citation is genuine. But what does the actual study say, and how far does it go?
The Real Study Behind the Claim
The paper is Mintz and Hollenberg (2019), published in Psychopharmacology Bulletin. The researchers conducted a retrospective chart review at a single clinical center, looking back at the records of 19 autistic adolescents and adults who had been prescribed low-dose lithium (extended or immediate release) added on top of their existing treatment, specifically for mood instability and aggression.
What They Actually Found
73.7% of patients (14 of 19) were rated as "improved" using the Clinical Global Impression-Improvement scale (a CGI-I score of 3 or lower), meaning a clinician judged their irritability and behavioral dysregulation to have gotten meaningfully better after lithium was added.
Where the Claim Overreaches
This is a real, promising result, not proof that lithium works for autism in general. Several limits matter:
- Retrospective chart review, not a controlled trial. There was no placebo group, no control group, and no blinding. Improvement was judged by clinicians who knew the patients were on lithium.
- Small and single-center. 19 patients from one clinic is not enough to generalize to autistic people broadly.
- Add-on, not standalone. Lithium was added to whatever treatment patients were already on, so the study cannot isolate lithium's effect from everything else happening in their care.
- A global impression score, not a specific behavior measure. CGI-I captures a clinician's overall sense of change, not a standardized, detailed rating of specific symptoms.
Lithium Is Not a Supplement
Lithium is a prescription mood stabilizer with a narrow therapeutic window: too little does nothing, too much is toxic. Prescribing it safely requires blood level monitoring and regular checks of kidney and thyroid function. This is a decision to make with a psychiatrist who can monitor bloodwork, not something to try on your own or give to someone based on a social media post.
What This Means for You
The signal is real enough that more rigorous research is underway. A randomized, double-blind, placebo-controlled trial (NCT04623398) is currently testing lithium against placebo specifically in Phelan-McDermid syndrome, a genetic subtype of autism caused by SHANK3 gene disruption, measuring effects on social communication. That is the kind of study that can actually answer whether lithium helps, and for whom.
If mood instability, aggression, or behavioral dysregulation is a real problem for you or someone you support, that is worth raising with a psychiatrist. Lithium may or may not be the right tool, but it is a medical decision, not a supplement aisle one.
If mood instability or emotional dysregulation is something you recognize in yourself, our free Emotion Regulation test can help put words to the pattern before that conversation.
Sources: Mintz & Hollenberg (2019), Psychopharmacology Bulletin 49(2):28-40, "Revisiting Lithium: Utility for Behavioral Stabilization in Adolescents and Adults with Autism Spectrum Disorder"; NCT04623398, "Effect of Lithium in Patients With Autism Spectrum Disorder and Phelan-McDermid Syndrome (SHANK3 Haploinsufficiency)", Assistance Publique - Hopitaux de Paris.