You teach a client a beautiful grounding practice. In the room, settled, they do it perfectly. Then the week happens — the shutdown at the dinner table, the panic in the parking lot — and the practice is nowhere. Not because they forgot it. Because the version of them who could do it wasn't there when it was needed.
This is a clinician-to-clinician look at nervous-system regulation in the 167 hours between sessions, for the somatic and polyvagal-informed among us. The core problem is specific, and once you see it, what you hand a client to carry out the door changes shape.
The body fires first
Start with the thing the somatic frame makes obvious and the cognitive frame keeps forgetting: the client doesn't think their way into a shutdown. They're in it before the story arrives. The sympathetic surge, the dorsal collapse — these are bottom-up events. The nervous system has already shifted state by the time the narrating mind catches up to explain what's happening.
This matters for between-session work because it locates the problem in time. The intervention has to be available before the thinking comes back online, not after. Anything that requires the client to first notice, label, and reason about their state is arriving a step late — the pattern has already taken the wheel.
The trap: a regulation exercise that assumes regulation
Here's the quiet contradiction in most between-session regulation homework. The five-minute breathing protocol, the full body scan, the "find your calm place" visualization — each of them requires a baseline of ventral capacity to even begin. They're lovely when the client is already regulated. They're nearly unreachable when the client is flooded or shutting down, which is precisely the moment they were prescribed for.
So we hand a dysregulated client a tool that assumes regulation, and then read its non-use as resistance or poor motivation. It's neither. It's the same state-access problem that explains why insight doesn't stick: a capacity encoded in the calm state isn't retrievable from the activated one. With regulation the stakes are just more physical — you can't reason a collapsed dorsal state into a box-breathing exercise.
Top-down vs bottom-up — and why it decides everything
The distinction worth holding is the route the intervention takes.
| Route | Works through | Needs | Available when activated? |
|---|---|---|---|
| Top-down | Thought — reframing, talking down, cognitive coping | Prefrontal capacity (online) | Rarely — it's the first thing offline |
| Bottom-up | Body — breath, posture, orienting, touch, temperature | Only the body, which is always present | Yes — reachable pre-cognitively |
Between sessions, when the client is most activated and least able to think, top-down tools are the wrong bet and bottom-up anchors are what's actually reachable. This isn't a knock on cognitive work — it's about sequence. You stabilize the body enough to get one degree of choice back, and then the thinking has somewhere to stand.
What actually holds: one degree of choice
Reset the goal. The aim of between-session regulation is not to get the client calm. It's to move the system one notch — from gone to slightly here, from frozen to a single breath, from fully blended with the surge to a sliver of space beside it. One degree of choice is the whole win. It's reachable; full calm-on-demand usually isn't, and setting that as the bar just manufactures another failure.
A few principles make a bottom-up anchor portable into the 167 hours:
- Build it in the body, in session. Feet pressing the floor, the weight of a hand on the chest, a long slow exhale, an orienting sweep of the eyes around the room. Concrete, physical, no narration required.
- Encode it while regulated, retrieve it while activated. Practice the anchor when the client is settled so the body knows it — then it's familiar enough to find when things go sideways. (Same logic as rehearsing retrieval, applied to the body.)
- Tie it to the body's cue, not the clock. The jaw tightening, the chest going hollow, the floaty pre-shutdown feeling — that early-warning sign is the trigger to use the anchor, because it's present exactly when regulation is needed.
- Keep the bar at one notch. Not "calm down." "Come one degree back." Achievable beats ideal every time in the activated state.
A between-session regulation anchor you can build today
In the last few minutes of session, co-build it rather than assign it:
- Find the anchor in a regulated moment. While the client is relatively settled, locate one concrete bodily anchor and have them feel it land — you're encoding it now.
- Pair it with a real cue. Tie it to the body's early-warning sign, not a time of day.
- Aim for one degree of choice. Name the goal as one notch back toward here, not full calm.
- Route it back to the room. Have them note what fired and whether the anchor reached them, to bring in — so it's shared work that deepens the next session.
This is the somatic spoke of a larger principle: between-session work has to meet the moment the pattern shows up, in the client's own body and language, and carry material back into the room rather than away from it. The full frame is in the pillar on keeping clients engaged between sessions, and the IFS-specific version of "notice the state, one degree of space" lives in IFS between sessions.
The line — an anchor the client reaches, not a regulating chatbot
The temptation is to hand the client something that "co-regulates" with them at 11pm. Be careful with the category. A chatbot positioned as a calming presence or a stand-in for your attunement is the wrong tool — general-purpose AI chatbots respond appropriately to people in distress below 60% of the time, against around 93% for licensed clinicians, and a dysregulated nervous system is exactly where that gap does harm.
What's defensible is narrower: a mirror that helps the client notice their own state and reach the anchor you built together, then routes that noticing back to you. It reflects; it doesn't co-regulate, treat, or replace your attunement, and it keeps you in the loop. Anything touching a client's material has to be HIPAA-compliant by design, and anything that goes around the clinician straight to the client has crossed into a riskier lane. I draw that line in AI as clinical tool, not replacement.
FAQ
Why don't grounding exercises work for my clients between sessions?
Usually because the exercise assumes the regulated state it's meant to create. A five-minute breathing practice or a body scan needs enough ventral capacity to start it — exactly what's missing when the client is flooded or shutting down. The body has already moved before the thinking comes back online, so a top-down, calm-state tool arrives too late. What works is bottom-up and small enough to reach mid-activation.
What nervous-system regulation tools actually work when a client is activated?
The ones that meet the body where it is and ask for one degree of change, not full calm. A single long exhale, feet pressing the floor, orienting the eyes around the room, the felt weight of a hand — anchors built in session so they're already familiar. The aim is to move the system one notch toward choice, which is reachable, rather than to manufacture calm, which usually isn't in the moment.
What's the difference between top-down and bottom-up regulation between sessions?
Top-down regulation works through thought — reframing, talking yourself down, cognitive coping. It needs the prefrontal machinery that goes offline under real activation. Bottom-up regulation works through the body — breath, posture, orienting, touch — and is available pre-cognitively, before the thinking returns. Between sessions, when the client is most activated and least able to think their way out, bottom-up anchors are what's actually reachable.
Is it safe to use an app for nervous-system regulation with clients?
It depends on the category. A consumer app that coaches the client around you, or a chatbot acting as a regulating presence, is the wrong tool — general-purpose AI chatbots respond appropriately to distress below 60% of the time versus around 93% for licensed clinicians. A HIPAA-compliant mirror that helps the client notice their own state and reach an anchor you built together, then routes it back to you, is a different thing. Safety lives in the design: HIPAA-compliant, clinician-in-the-loop, not a treatment substitute.