A post claims children with autism took a luteolin supplement, described as a flavonoid found in celery and parsley, and their parents reported improvements in speech, social skills, eye contact and gastrointestinal symptoms. A real study sits behind this claim. The details matter, and several of them do not match the viral framing.
The Real Study
Published in 2013 in Clinical Therapeutics, the study by Taliou and colleagues gave 50 children with autism spectrum disorder a dietary formulation for 26 weeks; 40 completed the study. The formulation was not pure luteolin. It combined luteolin (100mg, sourced from chamomile), the flavonoid quercetin (70mg) and its glycoside rutin (30mg) in each capsule, an anti-inflammatory flavonoid blend, not a single celery-derived compound. The children who completed treatment showed significant improvement on the Vineland Adaptive Behavior Scales and a reduction in Aberrant Behavior Checklist scores, covering areas like irritability, hyperactivity and communication.
A Real Mechanistic Follow-Up
A 2015 follow-up study in Translational Psychiatry, by some of the same researchers, measured blood markers in the children from the original trial and found that those who improved the most also had significantly reduced levels of two inflammatory molecules, TNF and IL-6. This adds a genuine biological thread: autism has been repeatedly linked to markers of brain and immune inflammation, and luteolin has documented anti-inflammatory and mast-cell-stabilizing properties in laboratory research. The correlation between behavioral improvement and lower inflammation markers is a real, interesting finding.
Why the Evidence Is Weaker Than It Sounds
The original trial was open-label, meaning there was no placebo group and no blinding. Parents knew their child was receiving the supplement. That matters enormously for a condition assessed largely through parent-reported behavior checklists, since expecting improvement can genuinely shape what parents notice and report, independent of any real underlying change. The study authors themselves list this as the trial's central limitation, not something the viral post mentions. There is also a declared financial interest worth knowing: the senior researcher associated with this line of studies has a commercial connection to this specific luteolin formulation, disclosed in the papers themselves. That does not make the results false, but it is exactly the kind of detail that belongs in the full picture.
What This Does Not Show
This is one small, uncontrolled pilot study and one follow-up analysis of the same cohort, not a large randomized placebo-controlled trial, and it has not been broadly replicated by independent research groups since 2013. It also does not mean that eating celery or parsley, which contain much smaller and less standardized amounts of luteolin than a 100mg capsule, would produce anything similar. The inflammation-autism connection is a legitimate, active area of research, but "a supplement showed a promising signal in an open-label pilot" is very different from "this treats autism."
The Honest Bottom Line
There is a real, published, peer-reviewed signal here, not nothing, and it points toward a genuinely interesting research direction linking inflammation and autism symptoms. But the evidence is preliminary, unblinded, and comes with a disclosed conflict of interest. Before trying any supplement, talk to your child's pediatrician, especially since supplements are not risk-free and are not regulated the way medications are.
If you want to explore your own neurotype, our free autism screening is a starting point for self-understanding, not a diagnosis.
Sources: Taliou, A. et al. (2013). "An open-label pilot study of a formulation containing the anti-inflammatory flavonoid luteolin and its effects on behavior in children with autism spectrum disorders." Clinical Therapeutics, 35(5), 592-602. Tsilioni, I., Taliou, A., Francis, K., Theoharides, T.C. (2015). "Children with autism spectrum disorders, who improved with a luteolin-containing dietary formulation, show reduced serum levels of TNF and IL-6." Translational Psychiatry, 5, e647. Educational summary, not a diagnosis or medical advice.