Smell recovery with a blend
PEA plus luteolin, combined with smell training, improved recovery in a controlled post-COVID study. Luteolin alone was not tested. [1]
PLANT FLAVONOIDS · RESEARCH REVIEW
A plant flavonoid with interesting immune-cell research and promising findings in specific clinical blends. See where the human evidence is encouraging, where it is limited and why the formula matters.
Explore the findingsIngredient and formulation research. None of the studies below establishes results for a current Neurogan product.
01 / FINDINGS AT A GLANCE
Luteolin is a flavone found in plant foods. Its best-known immune and antioxidant mechanisms largely come from laboratory work. Human studies often test it alongside other ingredients.
PEA plus luteolin, combined with smell training, improved recovery in a controlled post-COVID study. Luteolin alone was not tested. [1]
In a small Gulf War Illness study, 200 and 400 mg/day did not outperform placebo on overall symptoms. [2]
Luteolin inhibited activation of cultured human mast cells and T cells. This does not establish allergy or mast-cell symptom relief from capsules. [6]
02 / A POSITIVE RESULT, IN CONTEXT
In this 90-day study, all participants also practiced smell training. The active supplement supplied 700 mg PEA plus 70 mg luteolin daily in a co-ultramicronized preparation. [1]
Percentage of participants with any increase in TDI score
The trial enrolled 185 adults with smell loss persisting more than six months after COVID-19. All completed follow-up. The intervention group was younger on average (42.1 versus 47 years), and allocation was unequal (130 versus 55).
The control was not an inert capsule: the methods describe multivitamins, vitamin D and/or alpha-lipoic acid. Both groups used olfactory training. The active group had additional interim testing, while the control group had baseline and day-90 testing.
TDI combines threshold, discrimination and identification. Here, 120 of 130 versus 23 of 55 had any score increase. The paper’s larger quoted “recovery magnitude” numbers describe selected responder subsets, not whole-group mean changes. They are not used as an overall effect-size chart.
This is a promising clinical combination result. It cannot identify the contribution of luteolin, establish generic brain benefits or validate a different liposomal formulation.
03 / HUMAN STUDIES IN DETAIL
There is more to learn from these papers than a headline percentage. Here are the methods, the limits and the studies that did not show a clear benefit.
The PEA-luteolin blend plus olfactory training outperformed control supplement plus training. [1]
Population and design: 185 adults with persistent post-COVID smell impairment; 130 intervention, 55 control. Randomized, participant- and assessor-blinded study, with limitations in allocation and assessment schedules.
Dose and duration: 700 mg PEA + 70 mg luteolin, co-ultramicronized, once daily for 90 days. Both groups performed structured olfactory training.
Results: 120/130 active-group participants versus 23/55 controls had any TDI score increase. Nine active participants worsened and one was unchanged. Between-group endpoint TDI scores favored the combination (p<0.00001).
What remains uncertain: The effect of luteolin alone, the role of PEA and training, durability beyond 90 days and results with different delivery systems. No inflammatory biomarkers were measured. No adverse effects were identified.
Source details: Full text verified. Control baseline TDI is inconsistent between Table 1 (18.3) and results prose (18.2), so no calculated mean-change graph uses it.
Combined treatment performed best; doubling the blend without training did not produce a clear responder advantage. [5]
Population and design: 250 recruited; 202 completed all three follow-ups and were analyzed. Four arms: training plus control, once-daily blend, twice-daily blend, and once-daily blend plus training. Attrition was 19.2% and differed between arms.
Dose and duration: Each sachet contained 700 mg PEA + 70 mg luteolin. Once or twice daily for 90 days, depending on arm. This study used the identification subtest, not the full TDI battery.
Results: Reported mean identification scores at baseline and day 90: control + training 6.5 to 9.0; once-daily blend 9.1 to 12.0; twice-daily blend 9.6 to 12.1; blend + training 8.2 to 12.8. Baseline differences complicate raw comparisons.
Numerical limitation: The text reports 76 combined-arm completers, but its responder statement gives 50 people and 89.2%, consistent with a different denominator. The combined-arm demographic table also differs from the narrative. This unresolved reporting inconsistency prevents a reliable responder chart.
Interpretation: This supports further study of the specific combination with training, not isolated luteolin. Participants could not be blinded to doing training. No adverse events were reported among study completers.
Overall symptom severity fell from baseline, but placebo did as well. [2]
Population and design: Ten male Gulf War veterans contributed to the luteolin primary symptom analysis. Twenty-one men were analyzed across the report’s separate botanical experiments; eleven contributed luteolin pain/fatigue outcomes.
Dose and duration: One month baseline, one month placebo, one month 200 mg/day, then one month 400 mg/day. Luteolin was split between morning and evening doses. Lower dose always preceded higher dose.
Results: Mean 0–100 symptom severity: baseline 31.2, placebo 25.3, low dose 26.0, high dose 24.6. Neither low dose (p=0.718) nor high dose (p=0.492) was significantly better than placebo. Pain and fatigue also did not improve over placebo.
Limits: Small screening study, fixed phase order, possible carryover and a nonvalidated single-item symptom measure. It does not rule out all other uses, but it is direct evidence against assuming broad symptom relief at these doses.
Safety: Mild-to-moderate events occurred in the broader study. Luteolin phases included reports of diarrhea, reflux and flushing. Small numbers prevent precise adverse-event rates.
No adverse events were identified over six months, but the cancer outcomes were mixed. [3]
Population and design: Five men with low- or intermediate-risk prostate cancer under active surveillance. Single-arm phase I study with safety as the primary endpoint.
Dose and duration: 50 mg oral luteolin daily for 180 days. Blood tests and safety assessments at one, three and six months; repeat prostate biopsy after treatment.
Results: Two men had a negative repeat biopsy, one had stable findings and two had disease progression followed by surgery. PSA and PSA density did not change significantly. No symptomatic adverse events were found.
Interpretation: A negative biopsy does not prove a tumor was eliminated, and no untreated comparator was included. This is not evidence to use luteolin to treat cancer or replace active surveillance. Five people cannot establish general or long-term safety.
Source detail: Full text verified. Abstract and Table 1 disagree on some baseline age and PSA summaries; those disputed values are not used here.
Adaptive-behavior scores improved during treatment, but the study lacked a placebo group. [4]
Population and design: 50 children aged 4–10 with autism enrolled; 40 completed a 26-week open-label study. Assessments at baseline, week 18 and week 26.
Formulation and dose: Each oil-based capsule contained luteolin 100 mg, quercetin 70 mg and rutin 30 mg. One capsule per 10 kg body weight per day. This is a study exposure, not a pediatric dosing recommendation.
Results: Reported gains in Vineland age-equivalent scores: communication 8.43 months, daily living 7.17 months and social domain 8 months (p<0.005). These are test-score equivalents, not a claim that development accelerated by those amounts.
Limits and safety: Maturation, concurrent support, expectations and dropout can contribute to change without placebo control. Transient increased irritability occurred in 27/50 children for 1–8 weeks. Findings cannot be assigned to luteolin alone or used to recommend unsupervised supplementation in children.
Source access: Primary abstract verified. A later cytokine report from this trial is not counted as an independent efficacy study.
04 / WHY THE PLACEBO COMPARISON MATTERS
The same veterans reported symptoms in successive phases. Lower scores mean fewer symptoms. Improvement from baseline alone can be misleading. [2]
Mean severity, 0–100; lower is better
05 / WHY SCIENTISTS ARE INTERESTED
Mast cells are immune cells that can release histamine and inflammatory signals. Luteolin has inhibited their activation in cell experiments, but oral symptom relief is a separate question. [6] [7]
Mast cells derived from cord blood, plus a T-cell line.
Use myelin basic protein to study inflammatory signaling.
Luteolin pretreatment inhibited activation and cell-to-cell stimulation.
Cells were pretreated with luteolin for 15 minutes. Myelin basic protein at 10 micromolar stimulated mediator release. Luteolin inhibited mast-cell activation (3–6 experimental replicates) and mast-cell-dependent Jurkat T-cell stimulation (3 replicates), with p<0.05.
The 10 micromolar number describes the activating protein, not an oral luteolin dose. The accessible primary abstract does not provide a complete luteolin concentration schedule. Replicate experiments are not people taking supplements.
No clinical allergy score, histamine intolerance, mast-cell activation syndrome symptom endpoint or healthy-adult cognitive result was measured. The 2025 review maps a much larger immune-mechanism literature, but calls for clinical translation rather than establishing these uses.
NCT07280520 describes 100 mg/day luteolin versus placebo for twelve weeks in male athletes aged 18–35, with estimated enrollment of 50. Outcomes include gene expression, performance and body composition. [8]
On September 10, 2026, the record still reported enrolling by invitation and no posted results. Its last posted update was December 12, 2025. Estimated completion dates that have passed are not evidence that results are available.
06 / FROM RESEARCH TO PRODUCT
The current Neurogan products do not reproduce the PEA-luteolin or luteolin-quercetin-rutin study formulas.
200 mg / serving
Supplement Facts: two tablets, 60 servings. Suggested use: up to two tablets daily with food.
This listed daily amount matches the lower dose in the null Gulf War screening study, not the 70 mg luteolin plus PEA combination. The image also prints 12,000 mg total despite a 120-tablet count and 200 mg per two tablets, so its total-content text is inconsistent.
100 mg / serving
Supplement Facts: one tablet with a sunflower-lecithin phospholipid complex; 60 servings. The description instead calls it a capsule and lists one to two daily.
The listing and label disagree about dosage form. Active luteolin versus complex-mass wording also needs clarification. Confirm the current bottle label before use. No verified head-to-head human absorption result for this SKU was found in the selected evidence.
Product records and label images checked September 10, 2026. Listed milligrams are not independent potency verification. “Liposomal” is not interchangeable with co-ultramicronized PEA-luteolin or the oil-based pediatric blend.
07 / SAFETY & OPEN QUESTIONS
The clinical safety database is small, formulation-specific and incomplete.
No adverse events were identified in the five-person prostate pilot or the 2022 smell trial. The Gulf War study included digestive and other mild-to-moderate complaints. The pediatric blend study reported transient irritability in 27 of 50 children. These are different preparations and populations. [1] [2] [3] [4]
Ask a clinician or pharmacist about use alongside medicines, particularly during cancer treatment or with complex chronic illness. Drug-interaction data and long-term high-dose safety remain limited. Pregnancy, breastfeeding and routine pediatric use are not established by this evidence.
Persistent smell loss or suspected mast-cell symptoms deserve evaluation rather than self-treatment based on laboratory mechanisms.
SOURCES & FURTHER READING
Focused evidence review, not an exhaustive systematic review. Primary full text, abstracts, conference reports and registrations are labeled separately. Research and product records checked September 10, 2026.
Current Neuropharmacology, 2022. PMID 35450527; DOI: 10.2174/1570159X20666220420113513. Primary results: 120/130 versus 23/55 had any increase in TDI score at 90 days. This is not the stricter clinically meaningful response endpoint.
Read source 1 ↗International Journal of Environmental Research and Public Health, 2021. PMID 33802381. Luteolin primary symptom analysis n=10. Other botanicals were tested separately, not combined.
Read source 2 ↗Prostate Cancer, 2025. PMID 40046611; DOI: 10.1155/proc/8165686. Five patients, 50 mg/day, 180 days. Primary endpoint safety. Results and repeat-biopsy findings verified; not proof of cancer control.
Read source 3 ↗Clinical Therapeutics, 2013. PMID 23688534; DOI: 10.1016/j.clinthera.2013.04.006. Fifty children enrolled, forty completed. A distinct oil-based three-flavonoid preparation, not luteolin alone.
Read source 4 ↗European Archives of Oto-Rhino-Laryngology, 2023. PMID 37380908; DOI: 10.1007/s00405-023-08085-8. 250 recruited; 202 completers analyzed. The article has inconsistent combined-arm denominators, so its headline responder percentage is not graphed here.
Read source 5 ↗British Journal of Pharmacology, 2008. DOI: 10.1038/bjp.2008.356. Luteolin inhibited activation in cultured mast cells and Jurkat T cells. No people took oral luteolin in this experiment.
Read source 6 ↗Frontiers in Immunology, 2025. PMID 41164200. Primary abstract and open full-text archive retrieved. Used as context for immune-cell mechanisms and the need for clinical translation, not as direct supplement efficacy evidence.
Read source 7 ↗NCT07280520. Registry API checked September 10, 2026: enrolling by invitation; hasResults=false. Last posted update December 12, 2025. Planned 100 mg/day for twelve weeks; estimated enrollment 50.
Read source 8 ↗