Biology of Autism —Questions & Answers

It is a way of saying that autism can come from a chain reaction in the body, rather than one single cause.

Several things can add pressure to the system — immune stress in pregnancy, gut problems, toxins, chronic stress — and they all push on the same control points that handle inflammation, energy, and learning signals. Once those control points are overwhelmed, the brain and body settle into a stuck pattern: inflammation stays on, the gut and energy systems struggle, and the brain becomes less able to adapt and learn from experience.

No. It says multiple pressures can lead to a similar stuck state.

Examples of pressures the model highlights:

• Immune stress during pregnancy
• Gut barrier problems and "leaky gut"
• Mitochondrial (energy) stress
• Toxin and oxidative load
• Chronic stress and sleep issues

Different children can arrive at similar brain and gut patterns through different mixes of these factors.

This is one of the most important questions the model needs to answer — and the AofA paper addresses it directly through what it calls the constitutional susceptibility architecture.

The founding conditions — C-section delivery, formula feeding, antibiotic exposure, glyphosate exposure, maternal zinc depletion — describe a large fraction of the general pediatric population. Most children exposed to these pressures do not develop immune-derived autism. The children who do are those who entered life with reduced biological headroom at multiple cascade steps simultaneously.

Seven tipping points determine who crosses the cascade threshold:

• Tipping Point 1 — Gastric acid production capacity: variants in ATP4A/ATP4B proton pump genes reduce how efficiently the stomach makes acid from birth.
• Tipping Point 2 — CD26/DPP-IV adenosine clearance: documented lower DPP-IV activity in ASD children means the adenosine clearance system is already running at reduced capacity before any dietary blockade arrives.
• Tipping Point 3 — Methylation cycle reserve: MTHFR C677T reduces methylfolate availability, narrowing the margin before methylation fails when adenosine accumulates.
• Tipping Point 4 — Inflammatory resolution capacity: ALOX pathway variants reduce the body's ability to actively terminate LPS-driven immune activation.
• Tipping Point 5 — Kynurenine pathway branch bias: KMO variants determine whether inflammation pushes more toward neuroprotective kynurenic acid or neurotoxic quinolinic acid.
• Tipping Point 6 — Mitochondrial reserve: documented abnormal mitochondrial reserve capacity in ASD cell lines means the interneurons are closer to State 2 exhaustion from the start.
• Tipping Point 7 — HLA autoantibody susceptibility: HLA class II alleles determine whether casomorphin and streptokinase peptides trigger autoantibody production through molecular mimicry.

Think of it as a funnel. At the wide end, millions of children experience common exposures. Most have enough reserve at each tipping point that the cascade never loads. The smaller group who do develop IDA are those whose constitutional profile left insufficient headroom at multiple steps — so the same pressures crossed thresholds that most children's biology absorbs. Same insults. Different starting point. Different outcome.

The updated model has a more precise answer. The central molecule is SST-14 — somatostatin-14, produced by inhibitory interneurons in the cortex, hippocampus, and hypothalamus. These SST-14 interneurons are the cascade's convergent node — everything upstream in the cascade arrives here, and everything downstream flows from here.

When SST-14 interneurons function normally, they coordinate the release of three neuropeptides the brain and body need simultaneously: oxytocin (social salience and bonding), VIP (sleep, sensory regulation, gut motility, immune balance), and secretin (digestion, cerebellar timing). When they are silenced, all three systems lose their coordinating signal at the same time — which is why social, sensory, sleep, and GI features appear together as a cluster rather than separately.

The cascade silences SST-14 interneurons from two directions simultaneously — and within the transcriptional arm, two distinct mechanisms operate at once:

• Arm A — Excitotoxic: Inflammation activates IDO1, which diverts tryptophan into quinolinic acid — an NMDA receptor agonist that depletes the energy SST-14 interneurons need for tonic high-frequency firing.
• Arm B — Transcriptional, Mechanism A: The same cytokines activate NF-κB, which outcompetes CREB for CBP — the co-activator both require to open the SST-14 gene's cyclic AMP response element (CRE). NF-κB also recruits HDAC enzymes to compact the chromatin around the CRE, making the gene physically harder to read.
• Arm B — Transcriptional, Mechanism B: Adenosine accumulation from CD26 blockade activates inhibitory Gαi-coupled receptors, suppressing adenylyl cyclase and depleting cAMP — starving CREB of the PKA phosphorylation it needs to be activated in the first place. Mechanism B operates independently of Mechanism A and removes the compensatory route that would otherwise partially overcome it.

The reason the transcriptional suppression is so difficult to overcome is that Mechanism A takes away the co-activator and buries the gene, while Mechanism B prevents the transcription factor from being switched on at all. Both happening simultaneously removes every compensatory route.

The model moves beyond "gut–brain axis" as a slogan and focuses on barrier failure and LPS leak.

• If the gut lining is leaky, bacterial fragments (LPS) can slip into the blood and continually poke the immune system.
• This triggers inflammatory switches (like NF-κB), which in turn drive the tryptophan pathway away from serotonin and toward stress metabolites.
• Over time that contributes to low serotonin, low NAD⁺ fuel for SIRT1, and ongoing neuroinflammation.

Mitochondria are the cell's energy engines, and antioxidants (like glutathione) are the cleanup crew for the chemical "exhaust."

In this model:

• Mitochondria under stress make less energy but more "exhaust" (ROS), which signals danger and keeps inflammation going.
• SIRT1/PGC-1α problems reduce the creation of new healthy mitochondria.
• FOXO/glutathione problems weaken the cleanup crew.

Microglia and astrocytes are support cells in the brain.

• Microglia are like immune scouts and gardeners.
• Astrocytes help feed neurons and tell synapses when to form, strengthen, or disappear.

When inflammation stays high:

• Microglia shift into a more aggressive, "clean-up-by-destruction" mode.
• They send signals that flip astrocytes into an A1 state that is less nurturing and more synapse-removing.

The model proposes a dual pattern:

• Some circuits (especially long-range ones like thalamus→cortex and social networks) are under-connected, because hevin is low and SPARC is high — so fewer healthy synapses form and more are removed.
• At the same time, local pruning is impaired by autophagy failure, so too many local synapses remain in some regions.

This yields:

• Too many short-range connections (local over-connectivity)
• Too few long-range connections (reduced integration between distant areas)

Because thalamus→cortex relay synapses depend heavily on hevin, and hevin is suppressed, those pathways are weaker and fewer.

• The thalamus doesn't gate and shape sensory input as well.
• The cortex receives more "raw," less filtered information, which local circuits then amplify in an unbalanced way.

For social cognition, the issue is integration across distant areas:

• Regions that read social cues, emotions, and context don't communicate as smoothly because long-range connectivity is reduced.
• That makes tracking other people's thoughts, intentions, and subtle signals harder.

For cognitive rigidity and perseveration:

• CREB/BDNF problems mean the brain has trouble locking in new, more flexible patterns.
• Many synapses remain "silent" because glypicans are irregular, so there are fewer reliable pathways available for new strategies.

In this model, they come from the same core biology:

• GI: Leaky gut, dysbiosis, low serotonin, and vagal signaling produce dysmotility, pain, and barrier problems.
• Sleep: Tryptophan is diverted away from serotonin and melatonin, while SST and stress hormones disturb circadian rhythms.
• Mood: Inflammation, HPA axis changes, and low serotonin/BDNF make emotional regulation harder and anxiety more likely.

It means that after a certain point, the system can keep itself stuck:

• Inflammation activates IDO1, depleting NAD⁺ and generating quinolinic acid that damages SST-14 interneurons.
• SST-14 interneuron silencing removes the anti-inflammatory brake they normally provide — so the inflammatory environment deepens.
• Deeper inflammation compounds NF-κB-mediated CREB suppression — further silencing SST-14 gene expression from the transcriptional arm.
• The compromised gut barrier keeps feeding LPS into systemic circulation — sustaining the cytokine environment driving both arms simultaneously.
• A1 astrocyte polarisation, driven by microglial signalling that SST-14 loss can no longer restrain, withdraws BDNF and synaptogenic support — making recovery progressively harder without external intervention.

The model does not claim a proven cure, but it does provide a biomarker-stratified, state-specific intervention logic, supervised by qualified clinicians.

The cascade identifies a three-state clinical framework based on how deeply SST-14 interneurons have been silenced:

• State 1 (transcriptional suppression): The interneuron's structure is intact — the gene has been silenced. Infrastructure repair (sulforaphane, NAC, magnesium) + immune clearance (IMIG/IVIG) to reduce the NF-κB burden and clear autoantibodies.
• State 2 (metabolic exhaustion): Energy substrate depleted. State 1 interventions plus NAD⁺ precursors (NMN or NR) and mitochondrial support to restore tonic firing capacity.
• State 3 (structural loss): Partial interneuron loss and A1 astrocyte polarisation. State 2 interventions plus MSC trophic restoration to rebuild the BDNF and synaptogenic environment.

The model is presented openly as a theoretical framework, not as a finished, clinically proven map.

• Many individual pieces (e.g., maternal immune activation, gut permeability, mitochondrial issues, some supplements) have human or animal studies behind them.
• The full integrated "cascade" as drawn has not yet been tested as a single, unified treatment model in large trials.

The roughly 4-to-1 male-to-female ratio is one of the most consistent findings in autism research — and one of the least mechanistically explained. The cascade framework provides a specific molecular answer, built from two papers published decades apart.

Paper 1 — Montminy et al. (PNAS 1986) established that the somatostatin gene contains a cyclic AMP response element (CRE). When cAMP rises, it activates PKA, which phosphorylates CREB, which binds the CRE and drives SST-14 gene expression. Anything that reduces cAMP reduces SST-14 production — including the adenosine accumulation from CD26 blockade described in the cascade.

Paper 2 — Aronica et al. (PNAS 1994) established that estradiol activates adenylyl cyclase through a non-classical membrane pathway — independently of the Gαi suppression that adenosine uses. Membrane-associated estrogen receptor alpha (mERα) couples to Gαq, activates PKC, upregulates adenylyl cyclase VII specifically, and generates cAMP through a route that adenosine cannot block.

The connection neither paper made: females with active estradiol signalling retain partial SST-14 transcriptional drive through this compensatory cAMP route — even while adenosine is suppressing the conventional pathway. Prepubertal males, with no meaningful estradiol, have no access to this compensatory route. The same cascade burden that crosses the SST-14 silencing threshold in a male may not cross it in a female with equivalent biology.

At puberty, aromatase converts rising testosterone to estradiol in the brain — giving males their first access to the mERα-AC VII compensatory pathway. This is the likely molecular explanation for the spontaneous improvements in social communication, flexibility, and adaptive function that many families report in autistic boys at puberty. The improvement is not behavioural maturation — it is partial SST-14 transcriptional rescue through a pathway that finally became available.

Both — but in that order, and with an important caveat. The cascade is not a cure map. It is a sequencing map: a framework for understanding what needs to happen, and critically, in what order, before the biology can begin to shift.

The logic is built around three waves:

• Wave 1 — Reduce the burden. Before anything else can work, the upstream inflammatory and oxidative drivers must be reduced: the gut barrier leak feeding LPS into circulation, the NF-κB-driven cytokine environment suppressing CREB and driving IDO1, and the oxidative load overwhelming the antioxidant systems. Adding plasticity-supporting interventions while the inflammatory driver is still fully active does not work.
• Wave 2 — Restore the infrastructure. Once the inflammatory burden is reducing, the intracellular prerequisites for SST-14 recovery can be restored: NAD⁺ precursors for mitochondrial energy, methylation cycle support (hydroxy-B12, folinic acid), magnesium for the adenine nucleotide pool, and cAMP support where appropriate. Circadian rhythm and sleep stabilization belong here — without them, the inflammatory load maintains its suppressive burden regardless of other interventions.
• Wave 3 — Rebuild plasticity. Only when inflammatory load is reduced and the SST-14 interneuron infrastructure is being restored can synaptic remodeling meaningfully occur. The interventions that support A1-to-A2 astrocyte shift (microglial calming, fisetin, butyrate), hevin/SPARC rebalancing, and BDNF restoration belong in this wave — not before it. For State 3 patients, MSC trophic restoration belongs here.

The model does not promise outcomes. But it does offer something most ASD frameworks do not: a mechanistic reason why certain things need to happen before others, and a coherent way to understand why a single intervention rarely shifts things deeply on its own.

Sound sensitivity is among the most consistently reported experiences across the autism spectrum, and the cascade explains it through a specific breakdown in the brain's sensory gating system.

Normally, the thalamus acts as a gatekeeper between the outside world and the cortex. Its job is to filter, prioritize, and shape incoming sensory signals before passing them on — so that background noise stays background, and the brain is not overwhelmed by every sound in the environment at equal volume.

In the cascade, the disruption of thalamocortical connectivity means that this gating function is weakened:

• Sound signals arrive at the cortex less filtered and more raw than in a typical nervous system.
• Local over-connectivity in the cortex then amplifies these already-unfiltered signals further.
• The brain also loses its normal ability to habituate — to learn over time that a repeated, harmless sound can be safely ignored. Every occurrence can feel as intense as the first.

Sound is the most commonly reported sensory challenge partly because the auditory cortex is particularly affected by the local over-connectivity pattern that the cascade produces — dense local connections with weakened long-range integration. This creates a system that amplifies detail while struggling to place it in context.

This is one of the most striking apparent contradictions in autism, and the cascade resolves it through a single concept: interoceptive filtering dysfunction — the same mechanism that causes sensory over- and under-reactivity in the external senses, applied to the body's internal signals.

The never-full group — those who eat continuously and seem to have no off-switch — have under-reactive internal satiety signals. The gut sends a "I am full" message, but it does not register at threshold in the cortex. The signal is there; it simply does not arrive clearly enough to produce the sensation of fullness. Dysbiotic gut bacteria compounds this further by sending distorted signals along the vagus nerve before they even reach cortical processing.

The no-appetite / restricted diet group is actually two overlapping issues happening together:

• Sensory over-reactivity applied to food. Texture, smell, temperature, and even the color or visual appearance of food are all processed through the same amplified, unfiltered sensory system. For a child with acute olfactory sensitivity — one of the most consistently reported sensory differences in autism — a food's smell alone can be overwhelming enough to prevent eating. The restricted diet is often the child's rational, self-protective response to a genuinely difficult sensory experience, not a behavioral choice.
• Chronic gut discomfort suppressing appetite. Gut dysbiosis, low digestive enzyme output (suppressed by elevated SST), and intestinal inflammation make eating physically uncomfortable or painful. A child whose gut consistently hurts after eating will learn to avoid eating — and their appetite signals will reflect that learned association over time.

This question deserves a complete, honest answer — not a dismissal.

But the parent observations are real — and the cascade gives them a biological explanation.

Many parents describe a child developing typically, who received vaccinations, ran a fever, and then appeared to change. The timing is real. What the cascade framework challenges is the interpretation, not the observation.

The question "Did the vaccine cause the autism?" may be the wrong question. The better question is: what happens when a normal immune challenge meets a biological system that was already under significant stress?

• Research shows many children who later develop autism carry measurable biological differences long before any behavioral symptoms appear — including oxidative stress, glutathione depletion, and elevated inflammatory markers. The cascade was already loading.
• Any significant immune challenge — vaccination, viral illness, gut flare — triggers cytokine release that activates IDO1, diverts tryptophan away from NAD⁺ production, and can tip a system already close to threshold.
• The vaccination did not build the cascade. In a vulnerable child, it encountered one already underway.

One more piece of evidence worth noting: roughly one in six children with ASD show measurable behavioral improvement during febrile episodes. If inflammation were simply toxic to the autistic brain, fever should worsen symptoms — not improve them. This paradox points toward a specific, directional immune dysregulation rather than a general toxicity — and it maps directly onto the cascade model.

For a deeper discussion, see the Vaccine Question page.