A clinical reference for Theo's seizure-management protocol — with every ingredient list verified directly from animaleo.info, every constituent classified against independent peer-reviewed literature, and every recommendation tailored to his documented trigger pattern.
All ingredient lists verified against animaleo.info, May 2026. Constituents classified against Tisserand & Young 2014, Dr. Melissa Shelton ADR-II, and Mathew 2021 prospective EORS data. See evidence appendix below ↓ for the full reasoning behind each tier.
No constituents with documented seizure-threshold concern in independent literature. Includes NeuroBalance Diffusion + Calm-a-Mile RTU from Theo's active rotation. Use per AnimalEO guidance.
Fennel-containing blends with no other concerning constituents. Both primary authorities (Tisserand 2014 + Shelton ADR-II) say fennel is not directly convulsant. The remaining concern is pharmacokinetic — fennel inhibits CYP3A4, which metabolizes Theo's rescue meds (clorazepate, midazolam). Storm-day timing matters. Includes NeuroBoost RTU from Theo's active rotation.
Cineole/eucalyptus + camphor-bearing rosemary chemotypes. Genuinely contested evidence — Tisserand+Shelton dispute the convulsant mechanism, but Mathew 2021 prospective data implicates them, pediatric case series document seizures, and your existing avoidance practice is defensible for a 7-cluster/17-event seizure history. Verbenone CT is here on the camphor mechanism that both Tisserand and Shelton agree on.
Contains rue (Ruta graveolens) — documented neurotoxic compounds that both Tisserand and Shelton flag separately. Avoid.
No constituents with documented seizure-threshold concern in independent literature. Within this tier, blends are ranked by relevance to Theo's seizure-management protocol — from explicitly seizure-indicated formulas at the top, to specific-purpose blends at the bottom.
Designed by Dr. Shelton for seizures, epilepsy, and neurologic conditions. These are the AnimalEO line's primary tools for Theo's diagnosis. Both are in his active rotation.
Anxiolytic and emotional-balance blends. Storm anxiety and stress are documented seizure triggers — these blends address the trigger window directly. Strong anticonvulsant constituent profiles (Lavender, Roman Chamomile, Clary Sage, Bergamot).
Whole-body wellness blends. Not seizure-targeted, but Theo-safe and beneficial for overall health, immunity, and rotation variety.
Narrow-indication blends — joint support, dental, skin, GI, insect, cleaning. Theo-safe by ingredient profile but used for their specific purposes, not seizure management.
Cineole/eucalyptus + camphor-bearing rosemary chemotypes. The evidence is genuinely contested. Tisserand & Young 2014 primary toxicological analysis disputes the convulsant mechanism — 1,8-cineole is absent from their convulsant constituent table; eucalyptus species are absent from their potentially-convulsant essential oil table; they note Psidium guyanensis (40.5% cineole) and Laurus nobilis (40.8% cineole) are documented anticonvulsants. Dr. Shelton (ADR-II) does not include eucalyptus or cineole CTs in her epilepsy caution list either. However, Mathew 2021 — the largest prospective EORS study (350 patients across 4 hospitals over 4 years) — specifically implicates eucalyptus and camphor preparations. The pediatric case series (Spoerke 1989, Hindle 1994, Waldman 2011, Burkhard 1999) document seizures from eucalyptus exposure in children whose body mass profile is closer to Theo's than to an adult. Theo has a 7-cluster/17-event seizure history, and the user has been intentionally avoiding eucalyptus. Continued avoidance is defensible. Camphor-bearing rosemary chemotypes (Verbenone CT contains 12.5–17.8% camphor + isopinocamphone per Tisserand Table 10.2; 6.5% dermal max) are flagged on a separate mechanism that both Tisserand and Shelton agree on.
Fennel-containing blends with no other concerning constituents. Fennel deserves its own bucket because the evidence picture is genuinely different from the eucalyptus/cineole question. Both primary authorities (Tisserand & Young 2014 and Dr. Melissa Shelton ADR-II) agree fennel is not a direct convulsant: Tisserand explicitly states "fennel oil appears to be devoid of convulsant activity" and explains the Skalli 2011 case as a UGT enzyme drug-herb interaction with lamotrigine, not a convulsant effect. Shelton herself, the AnimalEO formulator, writes "In Tisserand & Young — there are no precautionary statements regarding use with seizures." What remains a real concern for Theo: fennel inhibits CYP3A4 (Subehan 2007 mechanism-based inactivation via bergapten). Theo's rescue meds — clorazepate and midazolam — are CYP3A4 substrates. Levetiracetam (his primary AED) is not. The pharmacokinetic concern is therefore around storm-day rescue dosing timing, not a direct seizure-risk mechanism. Includes NeuroBoost RTU from Theo's active rotation. This is the AnimalEO-recommended seizure protocol blend; under v5.0 framework, the recommendation is "use with awareness of CYP3A4 timing on storm days," not "avoid."
Contains a constituent with documented direct neurotoxicity that both primary authorities (Tisserand 2014 and Shelton ADR-II) flag separately. Under the v5.0 framework, only rue (Ruta graveolens) meets this bar — its concern is independent of the cineole or fennel discussions.
Detailed reasoning, primary-source readings, and methodology notes. Click any panel to expand.
All 69 single oils sold by animaleo.info, evaluated for active-seizure use (post-ictal + cluster-window, with during-seizure as inhalation-only). Classified against published animal-model evidence, Tisserand & Young 2014, Shelton ADR-II caution list, and constituent mechanism data.
Coverage finding: Theo's existing Clean-tier blends already contain all 10 Tier A oils and 17 of 20 Tier B oils. NeuroBalance Diffusion alone contains every Tier A oil — it's the single most evidence-rich product across the entire AnimalEO catalog. Strongest rescue candidates by mechanism: Frankincense (anti-neuroinflammatory, multiple rat models) and Copaiba (β-caryophyllene CB2 agonist, intranasal nanoemulsion tested for seizures).
Animal-model evidence for acute anticonvulsant effect. Rescue / post-ictal candidates. All 10 are in NeuroBalance.
Address anxiety / agitation / sleep disruption that triggers seizure clusters. No acute anticonvulsant evidence but supportive role during storm-day windows. 17 of 20 are in Clean blends; the 3 not formulated into blends are marked with ⚪.
No clear seizure evidence either direction. Theo-safe by ingredient profile but not active for seizure protocol.
Includes the v5.0 framework cautions (cineole-bearing, fennel-class, rue) plus catnip (proconvulsant in 1996 mouse study, dietary route — see catnip note below) and the hot oils (cinnamon/clove/oregano/thyme phenolics). Note: Lemongrass, Clove, and Cinnamon Leaf have actual anticonvulsant constituent evidence (citral / eugenol) — caution is about dermal irritation, not seizure mechanism.
Catnip (Nepeta cataria) is a borderline case. The proconvulsant claim rests on one 1996 mouse study (Aydin et al.) using catnip as 10% of dietary intake for 1 or 7 days, finding increased susceptibility to picrotoxin and strychnine seizures. This is meaningfully weaker than the evidence base supporting the eucalyptus or fennel framework decisions: the study uses oral/dietary route at very high concentration (not topical/diffusion at typical 1–3% blend dilution), is a single source with no replication, and infers the anti-GABAergic mechanism rather than directly demonstrating it.
Importantly, Shelton uses catnip in 12 AnimalEO formulas including seizure-protocol blends, and her epilepsy caution list does not include it — that's a strong real-world clinical signal that topical/diffusion exposure doesn't replicate the dietary findings. Tisserand & Young 2014 also do not list catnip in Tables 10.1 or 10.2 as proconvulsant. The framework treats catnip the same way Shelton and Tisserand do: not concerning enough to flag at typical EO blend concentrations. The dataset is not reclassified for catnip.
Theo's current Clean-tier rotation (NeuroBalance Diffusion + Calm-a-Mile RTU) already covers the complete rescue-grade oil set and 17 of 20 calming/anxiolytic oils. Buying singles is only worth it for: (1) Frankincense or Copaiba if direct/concentrated inhalation during a seizure is desired (rather than diffused at low blend concentration); or (2) Neroli, Ledum, or Vitex (Chaste Tree Berry) if his TCVM vet specifically indicates them — these are the three Tier B oils not formulated into any Clean blend.
Methodology: Phase 1 evidence triage, May 2026. Sources: Koutroumanidou 2013 (PTZ mouse comparison of 8 oils); Goyal 2021 (Boswellia / β-boswellic acid PTZ-kindling); Tchekalarova 2018 (β-caryophyllene MES); Tchekalarova 2023 (BCP intranasal nanoemulsion for PTZ); Chindo 2021 (Melissa officinalis VGSC blockade); Skalli & Soulaymani Bencheikh 2011 (fennel case report); Aydin 1996 (catnip dietary mouse study); Shelton ADR-II Canine Conditions / Seizures protocol; Tisserand & Young 2014 Essential Oil Safety 2nd ed. Tables 10.1–10.2.
v5.0 is a fundamental reframing built on direct review of the primary toxicological texts: Tisserand & Young 2014 Essential Oil Safety 2nd ed. (Chapters 3 and 10) and Dr. Melissa Shelton's Animal Desk Reference II (the AnimalEO formulator's own textbook, including her dedicated seizure protocol page). Reading both primary sources side by side resolved several of v4.x's classifications and identified one error of the opposite kind.
The biggest structural change: fennel now has its own bucket, separate from the cineole/eucalyptus discussion. The two questions are evidentially different. Tisserand explicitly states "fennel oil appears to be devoid of convulsant activity" and explains the Skalli 2011 case as a UGT enzyme drug-herb interaction with lamotrigine. Dr. Shelton, the AnimalEO formulator, cites Tisserand as her authority for using fennel in seizure-protocol blends. The remaining concern around fennel for Theo is pharmacokinetic (CYP3A4 inhibition affecting clorazepate/midazolam rescue meds) — not a direct seizure-mechanism concern.
For eucalyptus and cineole-bearing oils, the evidence is genuinely contested. Tisserand and Shelton both consider them not directly convulsant. But Mathew 2021 (the largest prospective EORS dataset, 350 patients, post-Tisserand) specifically implicates eucalyptus, the pediatric case series document seizures from eucalyptus exposure in children whose body mass profile matches Theo's better than an adult's, and you've been intentionally avoiding eucalyptus. Continued avoidance is defensible, so cineole/eucalyptus stays in Conditional.
One v4.x error corrected: Rosemary Verbenone CT was previously rated as relatively safe. Tisserand 2014 Table 10.2 explicitly lists it with a 6.5% dermal max because it contains 12.5–17.8% camphor + isopinocamphone. Dr. Shelton agrees, noting "Due to its camphor content, [Rosemary] is less selected for use within Veterinary Aromatic Medicine." This moves AdrenoBalance and several other blends to Conditional under v5.0.
Although there have been some cautionary statements in regards to Basil with individuals who seizure or have epilepsy, we have not found this to be an issue when used properly. The KittyBoost has been a hallmark treatment for many of our patients, especially those with seizures.
animaleo.info / NeuroBoostOccasionally Fennel is reported as needing to be avoided for those with seizures — however clinically we have not found this to be true. Also within Tisserand's most current book Essential Oil Safety — there is little cause for concern for fennel oil to have convulsant activity.
animaleo.info / NeuroBoostA finding worth pulling out separately, given Theo's regimen of Levetiracetam (Keppra XR) 750mg q12h × 3 tabs:
Kumar et al. 2013 tested Ocimum sanctum (genus Ocimum, same as Ocimum basilicum) in PTZ-kindled rats and found that combining the Ocimum extract with Levetiracetam produced maximum protection against PTZ-induced seizures — additive anticonvulsant activity beyond either treatment alone. The mechanism appears to be GABAA modulation by linalool, eugenol, and estragole, complementing Levetiracetam's SV2A binding.
This is independent peer-reviewed evidence supporting what Dr. Shelton has observed clinically — the KittyBoost-based formulas (NeuroBoost, KittyBoost itself, the entire Body Boost line) are not just "tolerated" alongside conventional AEDs. The basil component appears to be actively contributing to seizure protection, with mechanistic synergy documented for Theo's specific medication.
Additional supporting basil literature: Costa 2024 (zebrafish PTZ + estragole, GABAA modulation); Khodabakhshi 2017 (mouse anticonvulsant + oxidative stress attenuation); Sakurada 2009 (linalool suppresses neuronal discharge in experimental epilepsy); Oliveira 2009 (dose-dependent latency increase in PTZ).
A finding worth surfacing, given the recent Royce 2024 integrative veterinary case report:
In Dr. Royce's 17-year-old Dachshund cluster seizure case — managed entirely without ASDs after they caused unacceptable side effects — bacopa was the single intervention that achieved complete seizure elimination. The protocol started with valerian, passionflower, and ziziphus (70% reduction) and added boswellia, skullcap, and L-theanine without further effect. The addition of bacopa took the patient from 1.5 seizures/month to zero. The patient remained seizure-free for two months until passing from cancer progression.
Bacopa's documented mechanisms include: presynaptic Ca²⁺ channel inhibition (similar to pregabalin), Na⁺ channel inhibition, NMDA receptor modulation, GABAA receptor enhancement, glutamic acid decarboxylase support (glutamine→GABA conversion), reduced neuronal inflammation, and astrocyte/glial cell support. It's also a documented cognitive enhancer and adaptogen.
Theo's Mycodog Clarity contains Bacopa monnieri as one of its key ingredients alongside lion's mane and other nootropic mushrooms. This means Theo is already getting some bacopa exposure as part of his current protocol. Worth discussing with the TCVM vet whether the current Mycodog Clarity dosage is therapeutic for the bacopa component, or whether dedicated bacopa supplementation might be warranted given the strong anti-seizure mechanism profile.
Royce 2024 (CIVT integrative veterinary case report); Walker & Pellegrini 2024 (Bacopa monnieri, StatPearls); Liu 2017 (herbal medicine and epilepsy review).
This is the most evidentially contested classification in the v5.0 framework. Worth documenting explicitly so the reasoning is transparent:
Tisserand & Young 2014 (Ch10 p136) argues 1,8-cineole is a CNS depressant, questions whether it could even cause seizures, and notes only 4 of 192 reviewed cases had seizures (2%) — all young children with substantial exposure. Anticonvulsant activity is documented for two essential oils with cineole as major constituent: Psidium guyanensis (40.5% cineole) and Laurus nobilis (40.8% cineole). Jenner 1964 LD50 testing of 1,8-cineole in rats showed CNS depression and coma at lethal doses with no convulsions.
But:
The right epistemic stance here is honest acknowledgment: the evidence is genuinely mixed. Tisserand's analysis is rigorous. Mathew 2021 is also rigorous. They don't agree on eucalyptus. Under v5.0, eucalyptus and cineole-bearing oils sit in Conditional rather than Clean, reflecting that continued avoidance for Theo specifically is appropriately weighted against the contradicting evidence. If a future TCVM consultation suggests reintroduction, that's a values-based decision; the data isn't going to settle it cleanly.
Mathew et al. 2021 Epilepsy Research 173:106626; Tisserand & Young 2014 Essential Oil Safety 2nd ed., Ch10 p135-136; Spoerke et al. 1989; Hindle 1994; Waldman 2011 Pediatr Neurol; Burkhard et al. 1999 J Neurol 246:667-670; Royce 2024 (CIVT integrative veterinary case report); Jenner et al. 1964 (LD50 1,8-cineole, no convulsions).
The Skalli & Bencheikh 2011 paper is the single primary source most often cited as evidence that fennel oil causes seizures in epilepsy patients. It's a clinical commentary published in Epileptic Disorders, and reading it directly reveals important nuance:
The reported case: A 38-year-old Moroccan woman with well-controlled epilepsy on lamotrigine 300 mg/day. Her last seizure had been three years prior. She ate "around five or six cakes containing an unknown quantity of essential oil of fennel" — homemade with commercially-acquired fennel EO. Two hours later, she had a generalised tonic-clonic seizure lasting 45 minutes, accompanied by involuntary diarrhoea.
What the paper actually establishes:
What Tisserand & Young 2014 add: Lamotrigine is metabolized by UDP-glucuronosyltransferase (UGT). (E)-anethole, the dominant constituent of sweet fennel (58.1–91.8%), significantly enhances UGT activity (Rompelberg et al. 1993). Tisserand's reframing — that fennel ingestion likely cleared the lamotrigine from her system too quickly, leaving her vulnerable to her underlying epilepsy — fits the case better than "fennel is a primary neurotoxin," because it explains why she was fine for three years on lamotrigine, then had a single breakthrough seizure two hours after fennel ingestion.
Implications for Theo specifically: Theo's primary AED is Levetiracetam (Keppra XR), NOT lamotrigine. Levetiracetam is metabolized by acetamide hydrolysis (~66% renal excretion unchanged) — neither UGT nor CYP-mediated. So the Skalli mechanism (UGT induction reducing AED levels) does not apply to Theo's primary medication. The CYP3A4 pharmacokinetic concern around clorazepate and midazolam is a separate issue, relevant only during rescue dosing.
Skalli S, Soulaymani Bencheikh R. Epileptic seizure induced by fennel essential oil. Epileptic Disord 2011; 13(3): 345-7. Read in primary text. Tisserand & Young 2014 reframing: Ch10 p136. Rompelberg et al. 1993 (UGT enhancement by anethole). Patsalos 2000 (Levetiracetam pharmacokinetics).
Under the v5.0 framework, fennel is no longer treated as a direct convulsant — both Tisserand 2014 and Dr. Shelton agree on that. But there's a real pharmacokinetic concern worth keeping on the radar:
Subehan et al. 2007 demonstrated that methanolic fennel extract is a mechanism-based inactivator of CYP3A4 — meaning it permanently disables the enzyme until new enzyme is synthesized, not just competitive inhibition. The active constituent is 5-methoxypsoralen (bergapten), one of the coumarins in fennel. IC50 of 18.3μM at zero preincubation, dropping to 4.6μM after 20-min preincubation. The 2025 Pharmaceuticals comprehensive review (PMC12655285) confirms fennel inhibits CYP3A4 and may reduce metabolism of multiple drug classes metabolized by this enzyme.
This is a separate concern from the seizure-mechanism question Tisserand addresses. Fennel can be both "not a direct convulsant" AND "a CYP3A4 inhibitor" — the two findings don't conflict.
Why this matters specifically for Theo's regimen:
The most clinically meaningful scenario: storm-day rescue dosing. If a Kp ≥5 storm is forecast and you're applying NeuroBoost on storm day AND Theo needs intranasal midazolam during the storm, the cumulative CYP3A4 inhibition (fennel + CBD) could meaningfully alter midazolam kinetics — potentially deeper or longer sedation than expected. This isn't necessarily a reason to avoid the rescue med (rescue is rescue), but it's worth knowing for dosing decisions. This is the actual reason fennel-containing blends sit in their own bucket rather than fully Clean.
Subehan et al. 2007, Biol Pharm Bull (mechanism-based CYP3A4 inactivation); Zahi et al. 2025, Pharmaceuticals 18(11):1761 (comprehensive fennel review); Patsalos 2000 (Levetiracetam pharmacokinetics — non-CYP metabolism); Keppra FDA label (NDA 21-035).
A note on how this dataset weights its evidence sources, so future-you (or your TCVM vet) can see exactly what's behind each tier assignment:
Primary toxicological reference: Tisserand & Young 2014, Essential Oil Safety 2nd ed. Chapters 3 (Toxicity) and 10 (The Nervous System) provide rigorous toxicokinetic and toxicodynamic analysis with documented dose ranges, NOAEL values, and constituent-level reasoning. Specifically: Tisserand's Table 10.1 lists convulsant constituents (camphor, methyl salicylate, pinocamphone, β-pulegone, thujone — not 1,8-cineole, not fenchone, not anethole). Table 10.2 lists potentially convulsant essential oils with dermal/oral maxima — fennel and eucalyptus species are NOT in this table; Rosemary Verbenone CT IS (6.5% dermal max due to camphor + isopinocamphone).
Primary clinical reference: Dr. Melissa Shelton, ADR-II. The AnimalEO formulator's own textbook. Her dedicated seizure-protocol page lists Basil, Fennel, Hyssop, Rosemary, Sage, Tarragon, Wintergreen as oils with "general recommendations to avoid" — but her individual oil entries explain that for basil and fennel, she uses Tisserand 2014 as her authority and treats them as safe in epilepsy. She agrees with the broader caution on sage (thujones), wintergreen (methyl salicylate), and rosemary's camphor-bearing chemotypes. Eucalyptus and cineole-bearing oils are NOT on her caution list.
Contradicting clinical evidence: Mathew 2021, Epilepsy Research 173:106626. The largest prospective EORS dataset to date — 350 patients, 4 hospitals, 4 years (2014-2018), post-Tisserand. Specifically implicates eucalyptus and camphor preparations. This is genuine contradicting evidence to Tisserand's analysis. Tisserand has not published a response.
Pediatric case series (Spoerke 1989, Hindle 1994, Waldman 2011, Burkhard 1999): document seizures from eucalyptus exposure in children. Tisserand reviews and dismisses these as low-incidence (4 of 192 cases, 2%) and idiosyncratic, but the children's body mass profile is closer to Theo's than to an adult human's.
How v5.0 weights these sources for tier assignment:
This is values-aware classification, not algorithmic. The Conditional tier specifically reflects "evidence is contested and you have a 7-cluster/17-event dog" — it's the right amount of caution for a real seizure history, not theoretical worry.
Tisserand & Young 2014 Essential Oil Safety 2nd ed.; Shelton ADR-II; Mathew et al. 2021 Epilepsy Research 173:106626; Burkhard et al. 1999 J Neurol 246:667-670.