Can You Actually Change Your Attachment Style?
What the research says, and what I actually see in the room.
It's late, a friend sent you one of those attachment quizzes, and now there's a word sitting on your chest. Anxious. Avoidant. Disorganized, if the internet was feeling dramatic that day. You read the description and it clocks you so precisely it feels a little rude. And somewhere in the next ten minutes, the label quietly hardens into a verdict. This is just how I am. This is why it never works.
Let me stop you right there.
Yes, your attachment style can change. It is a pattern your nervous system learned under real conditions, and patterns learned under conditions can be re-learned when the conditions change. It tends to be stable, which is a very different thing from permanent. And a solid stack of research says it can move toward security across your life, and specifically through the right relationships and the right therapy.
Now let me show you why the "this is just who I am" story is so convincing, and why it's also wrong.
Why the "this is just who I am" story feels so true
Attachment forms early. Before you had language for any of it, you were running experiments. You reached, and you learned what came back. Consistent comfort, or a coin flip, or a closed door. Out of those results your system built a working model of what closeness costs and whether people stick around. That model is not a personality flaw. It was the smartest available solution to the room you were actually standing in.
The kid who learned to get loud to be seen built something that worked. The kid who learned to go quiet and self-contained to keep the peace also built something that worked. Your system kept these strategies for the same reason you keep a good tool. They did the job.
And that model is genuinely sticky, which is the part the internet gets half-right. R. Chris Fraley's 2002 meta-analysis in Personality and Social Psychology Review found that attachment is moderately stable over time. Researchers still argue about the mechanism. One camp says the early template endures and quietly steers everything after it. The other says new experiences keep rewriting it. The honest answer lives between them: the template is real and it has staying power, which is exactly why change feels impossible when you're looking at it from the inside.
Sticky is not the same as stuck. Keep going.
What the research actually shows about change
Here's the thing people leave out. John Bowlby, the man who built attachment theory in the first place, believed psychotherapy could change a person's attachment status. The founder was not a fatalist. The pop version is.
And the data has caught up to him. A review of treatment studies looking at attachment before and after therapy found that most of them showed people moving toward secure, with meaningful drops in attachment anxiety. A 2018 meta-analysis in the Journal of Clinical Psychology, pooling 36 studies and 3,158 patients, found that gains in attachment security over the course of therapy tracked with better outcomes overall. That same analysis hinted at something practical: people who start out lowest in security tend to do better in therapy that actually works with relationships and closeness, rather than treating your patterns like a spreadsheet to optimize.
Couples work has its own version of this. Emotionally Focused Therapy, developed by Sue Johnson, has a research base linking the work to partners building a more secure bond with each other, on purpose, in real time.
None of this promises you a personality transplant by spring. Change here is real, uneven, and earned. But "earned" is a wildly different word than "impossible."
The annoying good news about getting older
Time is quietly on your side, which is the most irritating kind of good news.
Longitudinal research following attachment across the lifespan (Chopik and colleagues) found that attachment anxiety tends to decline as people age, especially through middle and older adulthood, and that avoidance drifts downward across the years too. Translation: for a lot of people, the raw edges soften on their own. Life hands you a few relationships that go better than your model predicted, and the model updates a little without you filing any paperwork.
Great. Except nobody wants to white-knuckle their entire thirties waiting for their forties to feel safe.
And that drift is an average, which is not the same as a promise. The same body of research shows the change is usually modest and slow, and that plenty of people move the other direction entirely. Attachment updates on whatever your relationships keep teaching it, so repeated patterns in adulthood can erode security just as reliably as they can build it. Land in the same painful dynamic enough times and your system will faithfully learn the lesson you never wanted it to. Your nervous system just follows the evidence in front of it, wherever it points. Which is exactly why the pattern you keep ending up in matters so much, and why "give it time" is not a plan.
So the real question is how to move it on purpose instead of hoping the calendar does you a favor.
What actually moves it
Three things do the heavy lifting, and none of them are affirmations in the mirror.
Relationships that keep proving the old rule wrong. Your system updates on evidence, not on pep talks. A partner, a friend, a therapist who consistently does the thing your earliest relationships didn't, and does it again next week when you were braced for the door to close, is how the working model gets rewritten. Repetition is the active ingredient. One good weekend doesn't do it. A hundred small "I'm still here" moments does.
Being able to tell the truth about your own story. This is where the "earned secure" research gets interesting, and more complicated than the internet admits. Studies on earned security suggest the thing that actually predicts a secure adult isn't a spotless childhood. It's whether you can describe the childhood you did have coherently, without either airbrushing it into a greeting card or drowning in it. The story stops running you when you can finally hold all of it and still make sense.
Therapy that goes at the pattern, not just the mood. You don't tear the old strategy out by the roots. You show your nervous system, slowly and with evidence, that it's allowed to stand down sometimes. The double-text and the shutdown were both protecting you from a version of pain that was once very real. The work is convincing the guard that the threat has changed, so it can finally take a break.
My honest take
The label was never supposed to be a personality. "Anxious attachment" is a description of what you learned under specific conditions, and conditions change. Treating it like a star sign is convenient, because in one move you get to feel understood and also let yourself off the hook. But that convenience quietly costs you the best part, which is that you get to be different than you were.
And here's the part people get backwards: secure people are not calmer, wiser, or more evolved than you. They didn't win a personality lottery. They just tend to walk in expecting that reaching for someone will probably go okay. That expectation is not a birthright. It's learnable, which means it's teachable, which means it's yours to build.
Feeling "too much," or "too closed off," was survival math you did brilliantly as a kid with limited information. You're allowed to redo the math now that you're the one holding the pen and the threat is not the same.
So, can you change your attachment style? Yes. Slowly, unevenly, and usually with other people involved, which is almost funny, because the original wound happened with other people involved too. That symmetry is not a coincidence. It's the entire mechanism.
You didn't get like this alone. You don't have to get out of it alone either.
Sources
Fraley, R. C. (2002). Attachment Stability From Infancy to Adulthood: Meta-Analysis and Dynamic Modeling of Developmental Mechanisms. Personality and Social Psychology Review, 6(2), 123–151.
Adult attachment as a predictor and moderator of psychotherapy outcome: A meta-analysis. (2018). Journal of Clinical Psychology, 74(11). doi:10.1002/jclp.22685
Roisman, G. I., Padrón, E., Sroufe, L. A., & Egeland, B. (2002). Earned-Secure Attachment Status in Retrospect and Prospect. Child Development, 73(4), 1204–1219.
Johnson, S. M. Emotionally Focused Therapy research base, International Centre for Excellence in EFT (iceeft.com/eft-research).
Chopik, W. J., Edelstein, R. S., & Grimm, K. J. (2019). Longitudinal Changes in Attachment Orientation Over a 59-Year Period. Journal of Personality and Social Psychology, 116(4), 598–611. doi:10.1037/pspp0000167
Chopik, W. J., & Weidmann, R. (2024). Attachment security and how to get it. Social and Personality Psychology Compass, 18(1). doi:10.1111/spc3.12808 (review of how attachment shifts toward and away from security in adulthood)