Telmisartan as a useful biological “nudge-therapy,” particularly in bumetanide-responsive autism

 

A nudge is usually better than the sledgehammer !

 

Today’s post is another one most appropriate for people living in autism treatment-friendly counties (Russia, Ukraine, India, USA, Italy, Poland etc). Others will likely see this as from an alternative reality! The post is a bit long, just skip through it. 

My trial dose continues to be 20mg in a 65kg person. Doses trialed in schizophrenia have been much higher. Low doses are always the safest.

The post started life not as a review of any peer-reviewed clinical trials, but rather as an observational report, showing that revisiting the basic science can pay off. I made my initial review several years ago for my own purposes, but shared it in my blog.

I see that in fact the research has partially caught up:

Feinstein Institutes’ scientists find common blood pressure drug could be beneficial in some cases of autism

Scientists at Northwell Health’s Feinstein Institutes for Medical Research have made a significant discovery in autism spectrum disorder (ASD): a widely used blood pressure medication, captopril, can restore healthy function to the brain’s immune cells and reverse ASD-like behaviors in a preclinical animal model. This invaluable research focuses on a specific type of ASD believed to be triggered by a mother’s immune system during pregnancy, and could better understand autism and autism-like symptoms.

 

The full paper is here: 

Captopril restores microglial homeostasis and reverses ASD-like phenotype in a model of ASD induced by exposure in utero to anti-caspr2 IgG

 

What this now means - research from 2025 supports Peter’s 2017 idea to use telmisartan for autism

In 2025, researchers at the Feinstein Institutes for Medical Research published a preclinical study showing that modulation of the brain’s renin–angiotensin system (RAS) can reverse autism-like features in a specific immune-primed mouse model. In this model, prenatal immune exposure led to persistent microglial activation, synaptic abnormalities, and altered social behavior — changes that were significantly improved by treatment with captopril, an ACE inhibitor capable of crossing the blood–brain barrier.

Importantly, the study demonstrated that central (brain) RAS signaling is biologically relevant to neurodevelopmental plasticity, and that immune-driven alterations are not necessarily fixed. The benefit was not seen with ACE inhibitors lacking brain penetration, highlighting the importance of central rather than purely peripheral effects.

While captopril was used as a proof-of-concept tool, the underlying mechanism strongly supports the rationale for angiotensin receptor blockers (ARBs) — particularly telmisartan — which offer several advantages. Telmisartan directly blocks AT1 receptors, preserves potentially beneficial AT2 signaling, has a long half-life, and exerts additional anti-inflammatory, metabolic, and mitochondrial effects that are highly relevant to common autism subtypes involving neuroinflammation, behavioral rigidity, fatigue, and impaired stress resilience.

Thus, although the 2025 study does not establish a clinical treatment for autism, it independently validates the systems-level reasoning behind using telmisartan as a chronic “nudge” therapy in carefully selected autism phenotypes. The research supports the mechanism Peter proposed years earlier: that gently modulating regulatory systems such as the brain RAS can restore function in plastic but dysregulated neurodevelopmental circuits.

 

The keep it simple approach

I set out a very simple of framework of classic (Level 3) autism many years ago in this blog. It is also in my book and some presentations.

 

Today’s post falls in to the “central hormonal dysfunction” category.

Renin and angiotensin are both hormones

For the brain, angiotensin (especially angiotensin II) is the one that really matters. Renin is mostly just the upstream trigger.

The brain has its own local renin–angiotensin system, partly independent of the circulating one.

 

A recap for the science lovers - Angiotensin II in the brain

Angiotensin II is the active signalling molecule that actually does things in neural tissue:

• Acts as a neuromodulator
• Shapes excitatory–inhibitory balance
• Influences dopamine, GABA, glutamate
• Regulates stress, threat detection, motivation
• Affects neuroinflammation, oxidative stress
• Alters plasticity and myelination

All of that happens via angiotensin receptors, mainly AT1 and AT2.

 

In the brain, which receptor is activated matters more than how much angiotensin is around:

AT1 receptor (problematic when dominant)

• Increases stress signalling
• Promotes neuroinflammation
• Increases sympathetic tone
• Worsens cognitive rigidity

AT2 receptor (generally protective)

• Promotes neurite growth
• Supports learning and repair
• Anti-inflammatory
• Pro-plasticity

This is why ARBs (especially telmisartan) are interesting neurologically:

• They block AT1
• They shunt signalling toward AT2
• They act inside the brain, not just on blood pressure

 

Back to the more readable stuff

When treating broader autism you can consider the 150-200 possible therapies as ranging from small nudges in the right direction, to a precise hit with a mallet that corrects a precise dysfunction (a specific ion channel dysfunction, a lack of folate in the brain) to a sledge hammer that affects the entire brain (potassium bromide, as an example).

If you have epilepsy and severe aggression then a sledgehammer may well be what you need.

When I first trialed Telmisartan many years ago, I saw that it had an immediate effect, but back then I did not see it as being big enough. It certainly was a nudge, but I was still looking for that mallet, or indeed a sledgehammer. So I moved on.

Last year I revisited Telmisartan and now it is a core therapy. I am happy to include nudge therapies.  

If you have mild autism then a nudge or two maybe all that you need to overcome troubling issues.

If you follow my polytherapy approach for severe autism, then you might select a few nudge therapies and some stronger ones to create a personalized optimization.

 

Telmisartan

Telmisartan is an ARB (angiotensin II type-1 receptor blocker) commonly used to lower blood pressure. But, Telmisartan is thought of best understood not as a single-target drug, but as a system-level regulator.

Telmisartan is highly fat soluble (lipophilic) so it can penetrate the brain and even your bones. Bones matter for old people and all people with level 3 autism.  

Bones are a weak point in severe autism due to the side effects of drugs commonly used, poor diet, lack of exercise and specific genetic issues (in some monogenic autisms).

Bone is not inert. It has active RAAS signalling, telmisartan reaches bone tissue and blocks local AT₁ signalling, reduces inflammatory and oxidative tone in bone microenvironments. Via PPAR-γ, can influence osteoblast/osteoclast balance improving bone density

So an unexpected nudge towards stronger bones. 

The core actions

• AT₁ receptor blockade (RAAS modulation)
  Reduces chronic angiotensin II signalling, lowering background stress, sympathetic drive, and neurovascular strain.
• PPAR-γ partial agonism
  Improves metabolic efficiency, mitochondrial function, and lipid–glucose handling; contributes to anti-inflammatory effects.
• Autonomic calming
  Lowers sympathetic tone and stress reactivity without sedation. This a nudge effect towards better sleep, in some people
• Anti-inflammatory and antioxidant effects
  Indirectly reduces microglial activation and oxidative stress signalling. Microglia are the brain’s immune cells and can be in state of constant activation, which blocks them doing their basic housekeeping duties.
• Neurovascular effects
  Improves cerebral blood flow regulation and oxygen–nutrient delivery. In many types of severe autism, and also in dementia, the brain is unable to produce enough fuel (ATP). While there are many possible factors involved a key one is delivery of glucose and oxygen from your blood.

 

Indirect downstream effects (relevant to neurodevelopment)

• Improved cellular energy status
  Supports ion-pump function and transporter regulation. This is a nudge by improving the environment, Telmisartan does not force the ion-pumps directly
• Stabilisation of chloride homeostasis (indirect)
  Biases the NKCC1–KCC2 balance toward better chloride extrusion in vulnerable circuits, without forcing a gradient shift.

This is the big plus for bumetanide responders

Neuronal chloride levels are set by the balance between NKCC1 (chloride import) and KCC2 (chloride extrusion).

In some with autism the GABA development switch failed to activate after birth and so NKCC1 is overexpressed and KCC2 is under expressed

Stress, inflammation, and high activity increase NKCC1 influence and chloride loading.

KCC2 function is energy- and redox-dependent, and degrades under metabolic strain.

Telmisartan does not directly block NKCC1 or activate KCC2.

By reducing RAAS-driven stress signalling, it lowers pressure toward chloride accumulation.

Improved metabolic and redox conditions stabilise KCC2 membrane function.

Reduced autonomic overdrive lowers activity-dependent chloride loading.

The net effect is a bias toward more reliable chloride extrusion in vulnerable circuits.

This stabilises inhibition without forcing a chloride gradient shift, which KBr the sledgehammer would do. 

• Reduced excitability pressure
  Lowers the likelihood that inhibitory signalling becomes destabilised under stress.

 

Theoretical functional consequences (when it works)

• Lower baseline arousal and irritability
• Improved mood stability
• Increased behavioural flexibility
• Greater tolerance of sensory and cognitive load
• Enhanced availability for learning and interaction

 

My experience

When I conducted my review of “all autism” several years ago I did look at angiotensin. It looked to me that Telmisartan ticked many of the boxes for a cheap generic drug that could be repurposed for autism.

I did trial it and noted an immediate mood improvement with the strange effect of making Monty want to sing.

Several years later trialing it again. Again mood improved, there was no singing, but there was a desire to dance.

It also makes him less rigid. The best example is that when he empties the dishwasher he very clearly now puts things back in different places. You could argue this is negative, or you could see that as expressing his will rather than robotically following a pattern. More on that in the basal ganglia section.

The Basal Ganglia

The basal ganglia is the part of the brain that drives conditions like Tourette syndrome and PANS-PANDAS.

In PANS-PANDAS the immune system temporarily hijacks basal ganglia signalling. This is reversable, with prompt treatment.

The basal ganglia do not generate behaviour.

They gate behaviour.

When basal ganglia inhibition is stable:

• unwanted actions are quietly suppressed  (no tics)
• chosen actions feel voluntary
• habits can be overridden (no autistic rigidity)
• novelty is possible (try new foods, or watch a different cartoon) 

When basal ganglia function is disrupted (by genetics, inflammation, chloride instability, dopamine imbalance, Purkinje cell loss):

• the repertoire of behaviours is still there
• the motor programs still exist
• the thoughts still arise

What is lost is control over which ones fire. This manifest in autism as

Rigidity and stereotypies

• Repetitive behaviours are not “added”
• They become locked in
• Alternative actions cannot pass the gate

The system defaults to what feels safe and known.

This links to 2 further subjects of interest:

·        Purkinje cell loss in severe autism
·        ARFID (Avoidant/Restrictive Food Intake Disorder)

 

Purkinje cell loss as one possible driver of basal ganglia dysfunction

Purkinje cells are the very large, energy-intensive output neurons of the cerebellar cortex, providing continuous inhibitory timing signals to the deep cerebellar nuclei.

During the first two years of life, rapid brain growth creates extreme ATP demand, and transient mitochondrial or metabolic shortfalls can cause brief “power outages.” Because Purkinje cells are among the largest and most metabolically demanding neurons, they are selectively vulnerable and may be lost early, in a patchy and permanent manner. This a classic finding in port-mortem brain studies of people who had severe autism.

Purkinje cell loss leads to clumsiness and dyspraxia, because the cerebellum’s output signal loses timing precision, causing movements to be poorly planned, sequenced, and adjusted despite normal muscle strength. It does not lead to paralysis.

The resulting noisy cerebellar output propagates via thalamocortical loops to the basal ganglia, where it destabilizes action-selection and gating, particularly in the context of immature chloride regulation and weakened GABAergic inhibition.

Although the original cerebellar injury cannot be reversed, downstream circuits remain plastic, allowing pharmacological “nudges” such as bumetanide, atorvastatin, and telmisartan to partially restore inhibitory precision, improve basal ganglia gating, and reopen a window of motor-cognitive flexibility.

 

ARFID (Avoidant/Restrictive Food Intake Disorder)

ARFID can be the feeding expression of the same underlying circuit problem.

The framework explains ARFID extremely well, especially the autism-associated form of ARFID. In fact, it explains it better than sensory-only models. 

ARFID through the basal ganglia “gate” lens

Repetitive behaviours are not added — they become locked in.
Alternative actions cannot pass the gate.
The system defaults to what feels safe and known.

That description fits ARFID almost perfectly. 

In autism and related neuroimmune states, ARFID often acts as a behavioural marker of basal ganglia gating dysfunction, reflecting loss of choice rather than loss of appetite.

ARFID is not always basal ganglia–driven.

There are other ARFID subtypes:

• trauma-based (choking/vomiting)
• primary sensory aversion
• gastrointestinal pain–avoidance
• appetite dysregulation from meds or illness

But in autism-associated ARFID, especially when it:

• fluctuates with stress or illness
• coexists with rigidity, tics, OCD traits
• improves alongside mood and flexibility

the basal ganglia model fits extremely well. 

Blunt pharmaceutical treatment (sledgehammer) of ARFID does not work. Nudges seem a better choice. Nudges can be behavioral, or biological. 

Beyond telmisartan, the most promising pharmaceutical nudges for ARFID are likely those that reduce immune and autonomic stress on basal ganglia circuits while preserving motivation and behavioural flexibility, rather than suppressing output.

Many autism interventions provide a nudge to better functioning of the basal ganglia (NAC, ALA, low dose clonidine, atorvastatin etc).

Indeed one notable immediate effect of atorvastatin on Monty 14 years ago, was that he starting to come downstairs from his bedroom by himself, and not get “stuck” at the top of the stairs awaiting instructions.

I used to call this cognitive inhibition, but perhaps the gating model explains it better.

 

What the behaviour actually shows

“Stuck at the top of the stairs awaiting instructions”
is not a strength or balance problem.

It reflects:

• failure of self-initiated action
• dependence on external cueing (prompt dependence, in ABA terminology)
• intact ability, but blocked execution

That already points away from motor cortex and toward action-selection systems.

 

Basal ganglia explanation

Basal ganglia = action gating, not movement generation

The basal ganglia decide:

• when an action is allowed to start
• whether it is safe to proceed without prompting

In autism (and related neuroimmune states), this gate can become over-conservative:

• “wait”
• “don’t move”
• “need instruction”

So the child:

• knows how to go downstairs
• but cannot release the action independently

 

Why stairs are a perfect stress test

Descending stairs requires:

• motor sequencing
• balance prediction
• suppression of fear/uncertainty
• confidence in outcome

Basal ganglia dysfunction often shows up first in:

• transitions
• initiation
• descending movements
• unprompted actions

So “stuck at the top of the stairs” is a classic gating failure.

 

Why atorvastatin could change this quickly

Atorvastatin did not:

• teach a new skill
• strengthen muscles
• improve coordination

What it plausibly did was reduce a state constraint.

Immune / inflammatory relief

If there was:

• low-grade neuroinflammation
• immune-driven basal ganglia noise

Then dampening that can:

• lower threat signalling
• stabilise dopamine–GABA balance
• relax excessive inhibitory gating

When that happens:

actions that were already available suddenly “go.”

That can be rapid.

 

Reduced “protective inhibition”

In stressed systems, the brain sometimes actively prevents independence:

• to avoid risk
• to avoid uncertainty

Once the stress signal drops, the system stops applying the brake.

This feels like:

• confidence
• initiative
• independence

 

Why instruction dependence disappears

Needing instruction is a workaround:

• the external cue substitutes for internal gating
• the basal ganglia borrow cortical direction

When gating improves:

• the workaround is no longer needed
• behaviour becomes self-initiated

This is not just basal ganglia

Other systems likely contributed:

Cerebellum

• prediction of movement outcome
• timing and sequencing
• fear of misstep

Cerebellar function improves when:

• inflammation drops
• prediction error decreases

Autonomic system

• high sympathetic tone increases “freeze”
• calming allows movement initiation

Confidence loop

• once one successful descent occurs
• future attempts are easier
• habit loop updates

But the gatekeeper is still basal ganglia.

 

Conclusion

The basal ganglia are not “movement centres” in the simple sense.
They are action–selection and state–selection systems.

They decide:

• which action to initiate
• when to initiate it
• whether to repeat the same pattern or explore a new one
• how much reward, pleasure, and motivation is attached to action

They sit at the junction of:

• movement
• mood
• motivation
• habit
• flexibility

That is why basal ganglia changes show up as movement + emotion + novelty, all together.

 

Basal ganglia are especially sensitive in autism

Basal ganglia circuits:

• are GABA-heavy
• are chloride-sensitive
• rely on finely balanced inhibition
• are vulnerable to stress, inflammation, and metabolic strain

When inhibition in these circuits is unstable:

• action initiation becomes effortful
• behaviour becomes repetitive and rigid
• novelty feels unsafe
• mood flattens or becomes anxious

What changed biologically (without forcing anything)

When inhibition becomes more reliable (not stronger, just more predictable):

• neurons fire when they should, not erratically
• “gating” improves and actions can pass through more smoothly
• reward signals are no longer drowned out by noise

 

Why did Monty become happier

Mood is not just cortical thought, it is basal ganglia tone.

With better inhibitory stability:

• dopamine signalling becomes cleaner
• reward prediction improves
• the background “threat” signal drops

The result is:

• spontaneous positive affect
• relaxed facial expression
• joy without obvious cause

This is state change, not learned happiness.

 

Why the urge to dance appears

Dancing is a near-perfect basal ganglia readout.

It requires:

• effortless movement initiation
• rhythmic pattern generation
• reward linked directly to motion

When basal ganglia output is constrained:

• movement feels heavy
• initiation is delayed
• spontaneous rhythm disappears

When the constraint lifts:

• movement becomes intrinsically rewarding
• the body “wants” to move
• rhythm emerges without instruction

That is why dancing appears before language or cognition improves.

 

Why he emptied the dishwasher differently

Changing how a familiar task is done means:

• the brain is no longer locked into a single motor–habit template
• alternative action sequences are now selectable
• exploration feels safe

This is classic basal ganglia flexibility.

Nothing taught him that new method.

The system simply allowed another option to pass through the gate.