Parts work

Internal Family Systems and the body: what a pain clinician sees

Michael G. Hunter, L.Ac. · October 2026 · About 6 minutes

This site is mine, and UpSpiral is the company I founded. IFS is one of the pillars of our program, and my co-founder is IFS-certified, so I have a stake in this. Read accordingly.

Internal Family Systems is having a moment. Therapists are training in it by the thousands, it's on every podcast, and "a part of me" has entered ordinary speech. I use it. My wife Leela, a psychotherapist, is certified in it and teaches it inside our program. So I'm not a neutral observer, and this piece is going to be harder on the evidence than a fan would be, because the model deserves that.

The model in one paragraph

IFS, developed by Richard Schwartz in the 1980s, says the mind is naturally multiple. You have parts: a manager that keeps you organized and in control, a firefighter that grabs the wine or the phone when the pain gets loud, an exile that holds the old hurt the other two are protecting you from. Underneath the parts is what Schwartz calls Self, the calm, curious, compassionate center that was never damaged. Therapy is the process of getting the protectors to step back so Self can meet the exile and unburden it. No part is the enemy. Every part is trying to help.

What the body has to do with it

I spent 25 years with my hands on people in pain, and here is what I noticed long before I had the vocabulary: chronic pain is protective. A low back that locks up when the boss emails. A neck that seizes on the drive to the in-laws. The pain is doing a job, and the job is usually to keep something more frightening out of awareness. The pain research now says this plainly (the brain produces pain in response to perceived threat, not just tissue damage; see my piece on neuroplastic pain). IFS gives you a way to talk to the thing doing the protecting instead of fighting it.

That matters because fighting a protector makes it protect harder. Every pain patient who has tried to "push through" knows this. The IFS move is the opposite: get curious about what the pain is guarding, thank it, ask what it's afraid would happen if it stopped. Done well, in my experience, the body softens in a way that stretching never produced.

The evidence, honestly

There is exactly one randomized controlled trial of IFS that I'd call solid, and it's about the body. Shadick and colleagues at Brigham and Women's Hospital, published in the Journal of Rheumatology in 2013: 79 adults with rheumatoid arthritis, randomized to a nine-month IFS program (group and individual sessions) or an education control. At the end of treatment the IFS group had greater improvements in self-assessed joint pain, physical function, self-compassion and depressive symptoms. At one-year follow-up the pain and depression gains held. Objective disease activity (the joint counts, the inflammatory markers) did not differ between groups. In plain language: the disease didn't change; the experience of living with it did, and the change lasted.

That is a good result and a modest one. One trial, 79 people, a proof-of-concept study by the authors' own description, in a single condition. Beyond that, the IFS literature is pilots, case series, and a handful of small uncontrolled studies in PTSD and depression that report encouraging numbers without a comparison group. The model was listed in SAMHSA's evidence-based practice registry in 2015 on the strength of the Shadick trial; that registry has since been discontinued, which says more about federal budgets than about IFS. Larger trials are now in progress. Until they report, "evidence-based" is a stretch and "promising, with one good RCT" is the truth.

Where I think the model is right regardless

Three things I'd defend from the clinic even if every trial came back null.

Non-pathologizing works. Telling someone their rage, their numbness or their back spasm is a part trying to help, rather than a symptom to eliminate, changes how they hold it. Shame drops. When shame drops, the nervous system drops with it. You can measure that on a heart rate monitor.

Multiplicity is just accurate. Nobody is one thing. The person who wants to leave and the person who wants to stay are both real and both in the room. A model that admits that will beat a model that pretends you're a single rational agent with a bad habit.

Self is a state you can train. Whatever you believe about its origin, the calm, curious, compassionate stance is a reproducible physiological state, and people get better at entering it with practice. That alone is worth the price of admission.

Where to be careful

IFS is a therapy. Done with a trained clinician, with exiles that hold real trauma, it is powerful and occasionally destabilizing. The version that's safe to do on your own, and the version we teach in a group, stays with the protectors: noticing them, naming them, getting curious. Going after exiled trauma alone, from a worksheet, is not something I'd recommend, and no responsible teacher would either. If you have a trauma history, work with a clinician. If anyone tells you a course replaces that, they're selling.

Written by Michael G. Hunter, a licensed acupuncturist with 25 years of practice. Published on his own site. UpSpiral, his company, uses IFS parts work as one of three pillars alongside nervous system regulation and the neuroscience of habit; his co-founder Leela, a licensed psychotherapist, is IFS-certified and teaches it in the program. That's a disclosure, not a prescription. Not medical advice.

Sources

Shadick NA, Sowell NF, Frits ML, et al. A randomized controlled trial of an internal family systems-based psychotherapeutic intervention on outcomes in rheumatoid arthritis: a proof-of-concept study. Journal of Rheumatology. 2013;40(11):1831-1841.
Schwartz RC, Sweezy M. Internal Family Systems Therapy, 2nd ed. Guilford Press, 2020.
Schwartz RC. No Bad Parts. Sounds True, 2021.