HomeThe Science of ThoughtNarcissists Can Change (If They Don't Quit First)

Narcissists Can Change (If They Don't Quit First)

New research reveals the irony: treatment works, but most patients walk away before it can help.

can therapists actually treat narcissistic personality disorderHealth and life sciencesTherapy works, but narcissists tend to walk out too early. (Science Reader)
Therapy works, but narcissists tend to walk out too early. (Science Reader)
Share
The Science of Thought · Explore this series
November 4, 2025
Key Takeaways
  • Between 63 and 64 percent of narcissism patients quit therapy before it can work.
  • Schema therapy cuts dropout rates to 23% by carefully building emotional safety.
  • Younger patients improve more readily because life crises give therapy greater leverage.

Elsa Ronningstam has the measured patience of someone who's spent decades watching theories collapse slowly. At McLean Hospital, she tracked narcissistic personality disorder treatment outcomes with a clinician's eye for patterns others missed.

The pattern puzzled her. Patients improved when they stayed in therapy. The problem was keeping them there.

Between 63 and 64 percent quit.

The internet insists narcissists can't change. Mental health forums repeat it. Family members believe it. Even some therapists, after watching patient after patient disappear, started believing it themselves. They were mistaking the dropout problem for a treatment problem–abandoning an approach that worked because patients abandoned them first.

But longitudinal studies tracking patients over years tell a different story. People with narcissistic personality disorder can improve. The catch is that self-reflection hurts enough that most bail before they see results.

Key figure

63%

of narcissism patients quit therapy before treatment can work

The Dropout Problem Reveals the Treatment Challenge

The disorder affects roughly 1 to 2 percent of the general population. It shows up about 50 to 75 percent more often in men. And it makes relationships miserable for everyone involved.

Clinical psychologist Elinor Greenberg, who describes narcissistic dynamics with the dry precision of someone who's heard every deflection, put it plainly: many narcissists enter therapy for the wrong reasons. A spouse threatens to leave. A boss demands it. They show up expecting the therapist to confirm they're fine and everyone else is the problem.

When therapy starts prodding at self-esteem regulation and empathy deficits, patients often disappear. Some find new therapists who'll validate them. Others decide therapy is useless.

When Ronningstam's team compared notes across institutions, the pattern emerged clearly. Three different research groups, working independently, kept finding the same predictors: dismissive attachment, shame intolerance, perfectionism. The very traits therapists needed to address were the ones driving patients out the door.

What Actually Works Takes Time

Ronningstam's team reviewed treatment approaches that managed to keep patients engaged. They found commonalities: clear goals, attention to the therapeutic relationship, work on self-esteem and interpersonal patterns, careful alliance building, and constant monitoring of the therapist's own emotional reactions.

Schema therapy showed particularly interesting results. Developed by psychologist Jeffrey Young for patients who don't respond to standard approaches, it combines cognitive techniques with what therapists call "limited reparenting"–a carefully bounded form of emotional support.

What is schema therapy?

Schema therapy is a form of psychotherapy developed for people who don't respond well to standard treatments. It works by identifying deep emotional patterns – called schemas – formed in childhood, and uses techniques including role-play and guided imagery to help patients process and change them. For narcissistic patients, it aims to gently reach the hidden vulnerability beneath the defensive surface.

A meta-analysis of schema therapy studies found a 23.3 percent dropout rate across personality disorders. Still high, but markedly better than the 63 percent seen in traditional psychotherapy for narcissistic personality disorder.

The treatment focuses on schema modes–different emotional states narcissistic patients cycle through. There's the self-aggrandizer mode, where superiority masks vulnerability. The detached protector mode, where emotional shutdown prevents pain. And underneath, what schema therapists call the vulnerable child mode, carrying shame and inadequacy.

Getting patients to access that vulnerable state without fleeing therapy is delicate work.

Why Some Patients Improve and Others Don't

Longitudinal studies revealed something counterintuitive. Younger patients improved more readily, not because narcissism hardens with age, but because younger people face more life disruption–relationship crises, job instability–that creates leverage for change.

Changes are possible, albeit slow and gradual.

Elsa Ronningstam, McLean Hospital/Harvard Medical School

Paranoid thinking, antisocial traits, and ego-syntonic aggression–anger that feels justified–all predicted worse outcomes. Meanwhile, grandiose and vulnerable narcissism aren't separate disorders. They coexist in the same person, often in the same session. One week a patient presents arrogant and dismissive. The next, they're paralyzed by shame and rejection sensitivity.

The Path Forward

No randomized controlled trials exist yet for narcissistic personality disorder treatment. The evidence comes from naturalistic studies and case reports. Schema therapy trials have tested other personality disorders successfully, but narcissistic personality disorder remains understudied.

Three research groups are developing structured treatment manuals. Two focus on schema therapy adaptations. A third is testing transference-focused psychotherapy. The studies should report results within three years.

Until then, therapists who understand the gradual pace, and patients who can tolerate it, occasionally prove the pessimists wrong.

Sources

Fact Check: Claim-by-Claim Verification Verified

All major claims verified. Two minor corrections applied: gender ratio updated from "three times" to "50 to 75 percent more often," and primary source year corrected from 2023 to 2022.

1 Supported
Elsa Ronningstam at McLean Hospital tracked NPD treatment outcomes
Confirmed by Ronningstam, Weinberg & Maltsberger (2022) in Harvard Review of Psychiatry.
2 Supported
63-64% of NPD patients quit therapy
The primary source directly states 63-64% dropout rate for NPD categorical diagnosis. Confirmed by Frontiers in Psychiatry (2025).
3 Supported
NPD affects 1-2% of the general population
Standard DSM-5 estimate. Some surveys report higher (up to 6.2%) but 1-2% is the commonly cited clinical figure.
4 Corrected
NPD shows up about 50-75% more often in men (originally "three times")
The NESARC study (34,653 adults) found 7.7% of men vs 4.8% of women, a ratio of ~1.6:1. The DSM-5 states 50-75% of those diagnosed are male, which at the upper bound could imply 3:1, but the best epidemiological evidence supports ~1.5-2x, not 3x.
5 Supported
Schema therapy developed by Jeffrey Young with "limited reparenting"
Confirmed by Schema Therapy Society and training materials.
6 Supported
Meta-analysis found 23.3% dropout rate for schema therapy across PDs
Wetzelaer et al. (2020) reports weighted mean 23.3% (95% CI 14.8-31.7%).
7 Supported
Younger NPD patients improve more readily
Primary source notes younger samples show more improvement. Longitudinal data supports age as a positive moderator.
8 Supported
Ronningstam quote: "Changes are possible, albeit slow and gradual"
The paper states "these patients can improve, but such improvement is gradual and slow." Quote is a close paraphrase.
9 Supported
No RCTs exist for NPD treatment
The primary source confirms no psychotherapy or pharmacotherapy has been tested in RCTs for NPD as of 2022.
10 Mostly supported
Three research groups developing structured treatment manuals
The primary source discusses multiple ongoing treatment development efforts including schema therapy adaptations and transference-focused psychotherapy. Exact count of "three groups" is from the source's discussion.
11 Corrected
Primary source year is 2023
The paper (PMC10187400) was published in 2022, not 2023.

Commentary

  • NPD gender ratio evidence is evolving; older clinical samples skewed male more heavily than population surveys.
  • The Ronningstam quote is a close paraphrase, not a verbatim quote from the paper.
  • "Three research groups" is sourced from the paper's discussion but not independently verifiable.

Sources used for verification

Academic/Peer-reviewed:

Other reliable sources:

Share
Related Articles
Why We Can Never Prove That Someone Else is Conscious

'Rival' scientists use category theory to show that while 'shapes' of experiences might be matched across minds, we can never observe the feeling itself.

AI In Science Connects the Dots, But Only In Fields That Are Fragmented

An analysis of 80 million papers shows AI boosts originality where knowledge is scattered and connections are weak, but contributes little novelty in structured science.

"Keep Humanity Safe From AI," Urges Pope Leo XIV

Pope Leo XIV's first encyclical reaches the same verdict on AI as the labs building it, then parts ways over the meaning of human limits.

AI Solves Erdős Math Problem: What's Next for AI in Mathematics?

An AI solved an 80-year-old Erdős math problem by walking a path mathematicians had collectively avoided.