There's a phrase that has become a ubiquitous fixture in mental wellness discourse. It's repeated in therapy offices, on social media, and in conversations between friends trying to do right by themselves and each other.
Heal first.
Don't enter a relationship until you've done the work. Don't bring your wounds to someone else's door. Get yourself right before you try to build something with another person.
The instinct behind this is sound: people who use relationships primarily as avoidance — who pursue connection to avoid self-examination, or to look for a partner to fix what they won't look at — tend to replicate the same damage they'd fled from before. That's a real pattern and a real problem.
But the literal instruction — heal first, then connect — contains an assumption so deeply embedded that almost nobody questions it. It assumes that healing is a sequential, neatly-compartmentalized process with an endpoint, where after so many days in therapy or so many journal entries, you'll wake up one day and say, "ah, and today I am healed enough." It assumes that you can become healed in isolation – with enough self-help study and meditation – and that this healing is both achievable and necessary before you're ready to take that next step towards romantic human connection.
This assumption is not supported by the science of how healing actually works.
What the Phrase Gets Wrong
Healing is not a prerequisite for relationships. Especially for a specific and significant category of wounds — attachment wounds — relationships are the mechanism that produce healing.
This is not just a semantic distinction; it changes the entire prescription.
Attachment wounds are injuries of relationships. They develop when early relational experiences with caregivers — characterized by inconsistency, unavailability, or harm — produce internal templates that shape how a person expects others to behave (Farina & Schimmenti, 2025). These templates aren't stored as memories you can think your way out of; they operate below conscious awareness, as automatic patterns of relational behavior that activate in interpersonal contexts, precisely when they're most difficult to observe (Şar, 2025).
If the wound is implicit and relational, insight alone cannot resolve it. You can understand, with considerable sophistication, that your anxious attachment comes from an inconsistent caregiver, or that your avoidant attachment comes from a caregiver you couldn't trust with your emotional needs. You can connect the dots and trace the lineage of the pattern with clinical precision. And then you can go on your next date, feel the familiar unnerved discomfort when someone doesn't text back immediately – or get "the ick" when they text too much – and watch the pattern execute itself as though you'd never thought about it at all.
Understanding the map is not the same as walking the terrain.
Neuroscience Doesn't Agree With "Heal First"
Here is what the research says about what relationships actually do.
Close relationships are a health determinant on the same order of magnitude as smoking, body weight, and physical activity — not as a metaphor, but as a measurable biological fact (Holt-Lunstad, Smith, & Layton, 2010). A meta-analysis of 148 studies involving over 300,000 participants found that social connection predicted mortality risk as strongly as these established health factors. The quality of intimate relationships produces health effects comparable in magnitude to diet and exercise on clinical endpoints (Robles, Slatcher, Trombello, & McGinn, 2014).
The mechanisms are specific. Coan, Schaefer, and Davidson (2006) demonstrated that holding a trusted partner's hand during an anticipated threat measurably reduced activity in brain regions responsible for threat processing — an effect not produced by holding a stranger's hand, and not available to someone facing the threat alone. The partner's presence functioned as a neurological safety signal. Not a symbolic one – a literal one, visible on a brain scan.
The implication is direct: the biological effects of secure human connection are not available to someone deferring connection until they are healed. You cannot access the mechanism from the waiting room.
If the studies bore you, consider a dose of common sense based on something you've surely seen already: a feel-good video of an abused or abandoned cat or dog who, when first rescued, shows heightened aggression and fear responses, snapping at its caregivers. Then, by the end of the video fast-forwarded some months later, it's happily cuddling with its caregivers. It's worth noting the obvious: these animals have exactly zero capacity to comprehend counseling to work through their trauma. They were simply exposed to consistent kindness, love, and safety until their brains rewired away from scarcity or fear responses and towards secure attachment and trust. The bajillion different books, online videos, peer-reviewed research, etc., tend to obfuscate – by way of overcomplication – the very simple truth that exposure to consistent safety is roughly all that's needed.
You Cannot Practice in Isolation
There is a clinical principle so well-established that it barely requires argument: exposure therapy. Without it, avoidance of the feared stimulus perpetuates the fear. You don't get over a fear of heights by reading a dozen books on post-mountain-climbing empowerment; you, yourself, incrementally climb higher until your brain rewires. This goes beyond attachment work as well: if the things you need to work on are communication skills and conflict resolution, you don't learn them by avoiding opportunities to communicate or resolve conflict. A football player doesn't become skilled by only studying playbooks in the safety their room – they get out on the field and practice what they're learning by challenging themselves. Nor does a ballroom dancer become skilled by not practicing and falling. The examples are virtually endless.
A person with social anxiety who avoids social situations doesn't heal. They practice avoidance, and avoidance becomes a more deeply grooved response. A person with avoidant attachment who latches onto excuses to avoid intimate connection doesn't resolve their attachment wounds. They practice avoidance of intimacy, and the programming that made the avoidance feel necessary never gets the corrective data it would need to update.
I know this one personally. Around seventh grade, my mother pulled me out of school due to chronic bullying that the school wasn't addressing. The intention was sound; the outcome wasn't. I emerged from five years of home schooling with zero social exposure — which made re-entry significantly more difficult, which made the social environment more hostile with even worse bullying, which produced a longer and harder recovery than the original problem might have. Protecting someone from a context does not prepare them for the context, it just delays the reckoning.
Attachment wounds are no different. You cannot practice secure attachment by avoiding attachment. You practice it by attaching — imperfectly, with fear, making mistakes, and ideally, with the support of someone who has enough clinical training to help you notice what you're doing as you're doing it.
The Therapy Room as Evidence
If you need proof that healing is relational, look at where healing happens most often: psychotherapy.
Research consistently shows that attachment security increases through psychotherapy across modalities and approaches — not just relational or attachment-focused therapies, but broadly, across the field (Taylor & Rietzschel). If internal insight – e.g., guided via self-help books or online videos – were the primary mechanism of change, the quality of the relationship between therapist and patient would matter less than the technique employed. The evidence shows the opposite.
Mikulincer, Shaver, and Berant (2013) found that the therapeutic relationship functions as a secure base — that the patient's experience of the therapist as a reliable, responsive relational figure mediates change in attachment-relevant domains more than the specific content of the interventions. Allen (2025) put it more directly: effective therapy for attachment trauma requires, above all, the therapist's capacity for genuine human engagement. Not technique. Not method. The relational quality of the encounter.
A qualitative study of therapeutic relationship formation by Lavik, McAleavey, Kvendseth, and Moltu (2022) makes this concrete. Examining the first five sessions of therapy across twelve patient-therapist pairs, the researchers found that clients who could not experience the therapeutic relationship as safe — who could not allow the relational contact to counteract their fear and shame — showed the least improvement regardless of the therapist's technical competence. Safety, produced through relational experience, was the prerequisite for the work. Not its product.
Pearlman and Courtois (2005) say it plainly: relational treatment is not simply one modality among many for complex trauma – it's the foundational condition for therapeutic change. The therapy relationship provides, through consistent attunement and repair, what the original attachment relationships failed to provide. The patient does not arrive at a safe internal state and then engage relationally; they experience relational safety first, and internal change follows. In the context of relationships, this moves the goalpost from "spend energy on healing first before getting into a relationship" to "spend energy on adjusting your people-radar to choose someone who provides relational safety."
The question this raises for "heal first" is uncomfortable: if, even in the context of individual psychotherapy — the paradigmatic setting for internal work — healing occurs through relational mechanism, what exactly is the solo preparation supposed to accomplish?
Who Gets to Heal First
There is something else embedded in the instruction to "heal first" that deserves a direct look.
"Heal first" is advice most often delivered from a position of relative stability. It assumes access to sustained individual therapy, sufficient economic security to defer partnership, and a social support network robust enough to sustain psychological development in the absence of intimate connection. It assumes, in other words, a safety net.
For people without those resources — who cannot afford extended therapy, for whom partnership carries practical as well as emotional stakes, whose support networks are thin — the instruction is written for someone else's life. It was, like most wellness advice, developed by and for people whose circumstances make it viable. The person without a safety net cannot afford to defer the thing that the research suggests is itself a mechanism of healing.
What the Phrase Is Actually Trying to Say
Again, "heal first" assumes a neatly-defined segmentation where healing ends and other pursuits begin; in other words, a finish line. There is no finish line. Psychological development is continuous; a direction, rather than a destination. Attachment security is not a state you achieve but a skill you develop and maintain through ongoing relational experience (Mikulincer & Shaver, 2016). The instruction to complete healing before entering a relationship prescribes an endpoint that clinical literature does not recognize. The person waiting until they're healed enough is waiting for something that isn't coming.
To be clear, "don't enter relationships if you're too toxic and could cause harm" is sound advice for people who should, in fact, heal first.
However, "complete your healing before connecting" is the common intent when someone says to heal first, which is something the research simply doesn't support.
What the research supports is this: bring enough self-awareness into connection that you can use it as a mechanism of growth rather than a mechanism of avoidance. Not heal, then connect; instead, heal, while connecting — with eyes open to what the connection is activating, with enough self-reflection to notice your patterns as they arise rather than only after the damage is done.
The semantic difference is significant. One version of "heal first" prescribes a finish line; the other prescribes a direction.
The Opportunity Cost of Healing First
A foundational concept learned in the business world is that of opportunity costs. Essentially, if you spend your time watching Game of Thrones, you've paid an opportunity cost with regards to all the other shows you could've watched instead. If you spend your time dating this guy or girl, you're paying opportunity costs in not dating other people. If you invest in one business idea, your opportunity cost are the other ideas you didn't invest in that could've yielded higher returns on investment. A decision for one thing almost always is a decision against something else.
The pragmatic reality in the context of relationships is efficiency. If your end goal is to retire early, your career decisions especially in the preceding decades matter a lot. If your end goal is to find your forever person, how you spend that time also matters a lot.
Exceptions of course exist, and some get lucky, but generally speaking, the difficulty in finding "your person" seems to increase, compounding exponentially, every year. Perhaps counterintuitively, this actually makes a strong case to date more, not less.
If you take a couple years to "work on yourself," three things will be guaranteed: one, the dating pool will be worse two years later, as good options quickly pair off with someone else and exit the dating pool; two, you'll have paid an opportunity cost where the best person for you may have only signed up to a dating app for a brief moment during those two years; and three, you'll have missed many opportunities to learn through trial-and-error.
For that last one, I'm reminded of a coaching client who was an online content creator. She would get thousands of messages per day from men, incidentally and rapidly learning every "trick" in the book they had to try to gain her interest: every pickup line, every strategy, and every manipulation attempt. Without intending to, she speed-learned a cumulative several years of men's approaches in just a few months. When she later used dating apps to find a partner, it became almost second nature to spot genuine people and weed out the "noise" of bad actors.
The earlier point that healing happens messily and concurrently is something else I think is mirrored in business, with the concept of failing fast. Failing fast is a lean entrepreneurship strategy that emphasizes "going for it," as opposed to excessive time and resources spent over-analyzing the perfect decision, in order to quickly reveal the strategy's flaws and determine what can/should be invested in or divested out of. It's messier, but the value proposition is that the mess is worth the time saved in the long run.
Note that I'm not suggesting to throw caution to the wind and jump into a relationship after only a date or two, but there are ways to intelligently optimize your time with the "just right" ratio of caution and progress. This is one of the things we help with.
Again: if you're someone for whom relationships are a distraction from facing some serious issues, then I'd say yes, heal first.
However, for most of you – especially now that you hopefully understand that the healing you're seeking happens within relational contexts, not without – ask yourself a hard question: Do you truly need to heal first, or are you scared and using the phrase as a socially-acceptable buffer to avoid intimacy until your loneliness outweighs your fear?
Are you using "heal first" as a distraction from facing intimacy?
The Name
One of our fundamental principles is challenging common mental wellness mantras, so it was by design to choose a phrase with nuance that could, and should, be challenged. Thus, branding ourselves "Heal First" (and writing an entire article on the phrase's weaknesses) symbolically takes us full-circle, tying back to our core principles. In other words, we question everything, even our own brand name.
But the name isn't just an edgy attempt at irony either. Heal First still signals what this practice is committed to: psychological health as the foundation – like the base of a pyramid, rather than a mile marker – of everything. Instead of "heal first, then do the thing," we simply argue and advocate for the healing to be integrated and concurrent with life.
The nervous system does not learn safety in isolation and then apply that learning to relationships. It learns safety through the experience of safe relationships. The instruction to heal before connecting asks for a skill to be developed outside the only context in which that skill develops — which is roughly equivalent to preparing to swim by avoiding water.
The more accurate version is simpler, if harder to put on a sign: heal while living your life. The healing doesn't precede the life. It happens simultaneously and within it.
Tavi Ambrose (Tavari A. Keel) is a master's student in Clinical Mental Health Counseling and the founder of Heal First. He writes about psychology, relationships, and the patterns we learned to survive that we're still carrying.
Sources
Allen, J. (2025). Conducting effective therapy for attachment trauma requires skill in being human. Clinical Neuropsychiatry, 22, 374–378.
Coan, J. A., Schaefer, H. S., & Davidson, R. J. (2006). Lending a hand: Social regulation of the neural response to threat. Psychological Science, 17(12), 1032–1039.
Farina, B., & Schimmenti, A. (2025). The psychopathological domains of attachment trauma. Clinical Neuropsychiatry, 22, 351–373.
Holt-Lunstad, J., Smith, T. B., & Layton, J. B. (2010). Social relationships and mortality risk: A meta-analytic review. PLoS Medicine, 7(7).
Lavik, K. O., McAleavey, A. A., Kvendseth, E. K., & Moltu, C. (2022). Relationship and alliance formation processes in psychotherapy. Frontiers in Psychology. PMC9301379.
Mikulincer, M., & Shaver, P. R. (2016). Attachment in adulthood (2nd ed.). Guilford Press.
Mikulincer, M., Shaver, P. R., & Berant, E. (2013). An attachment perspective on therapeutic processes and outcomes. Journal of Personality, 81(6), 606–616.
Pearlman, L. A., & Courtois, C. A. (2005). Clinical applications of the attachment framework. Journal of Traumatic Stress, 18(5), 449–459.
Robles, T. F., Slatcher, R. B., Trombello, J. M., & McGinn, M. M. (2014). Marital quality and health: A meta-analytic review. Psychological Bulletin, 140(1), 140–187.
Şar, V. (2025). From attachment trauma to traumatic attachment. Clinical Neuropsychiatry, 22, 417–422.
Taylor, P.J., Rietzschel, J., Danquah, A.N., & Berry, K. (2015). Changes in attachment representations during psychological therapy. Psychotherapy Research, 25, 222 - 238.