Attachment styles: the four types, and can your attachment style change?

What are attachment styles?

“Attachment theory is the joint work of John Bowlby and Mary Ainsworth”^[5]. Bowlby, a British psychiatrist, set out the theory in papers from the late 1950s^[6] and in three books: Attachment (1969), Separation (1973) and Loss (1980)^[5]. Ainsworth added the idea of the parent as “a secure base from which an infant can explore the world”, and of a parent’s sensitivity to a baby’s signals^[5].

In the 1960s and 1970s, Ainsworth studied one-year-olds in the Strange Situation, a “20-minute miniature drama with eight episodes” in which a parent briefly leaves the room and comes back^[5][7]. Most babies went to their mother when she returned. Some were “surprisingly angry”, and others “seemed to snub or avoid the mother on reunion”^[5]. That gave the first three patterns: secure, avoidant and ambivalent (also called resistant). In the 1980s, Mary Main and Judith Solomon described a fourth, disorganised^[8].

Later, researchers applied the same ideas to adults. In 1987, Cindy Hazan and Phillip Shaver described romantic love as an attachment process, with three styles: secure, avoidant and anxious-ambivalent^[9]. In 1991, Kim Bartholomew and Leonard Horowitz proposed four adult styles, based on whether your image of yourself and your image of others is positive or negative^[10]. Those four are what most people mean by “the four attachment styles” today.

The four attachment styles at a glance

Each style is a mix of two things. Attachment anxiety is how much you worry that someone close to you will not be “available and responsive in times of need”^[11]. Attachment avoidance is how uneasy you are with closeness and depending on others: how much a person “distrusts relationship partners’ goodwill, strives to maintain independence”^[12].

Based on Bartholomew and Horowitz’s four-category model and the anxiety and avoidance scales^{[4][10][11][13]}. The examples are illustrations, not a checklist.

Anxious attachment

Anxious attachment means high attachment anxiety and low avoidance. You want to be close to people, but you worry they will not be there when you need them^[11], and you may depend heavily on others^[4]. In everyday life that might look like reading a slow reply as a bad sign, needing to hear that things are fine, or finding it hard to settle when someone you love is distant. Our post on dating anxiety looks at how this can feel when you are getting to know someone.

Avoidant attachment

Avoidant attachment means high avoidance and low anxiety. Researchers describe it as “difficulty relying on others”^[4]. Independence matters a lot, and it can be hard to trust that others mean well^[12]. You might keep things to yourself, prefer to sort problems out alone, or pull back when a relationship gets serious. In one long-running study, men scored higher in avoidance than women at every age^[2]. Our post on hyper-independence covers the “I’ll do it myself” side of this pattern.

Fearful-avoidant and disorganised attachment

Fearful-avoidant attachment means high anxiety and high avoidance at the same time. In Bartholomew and Horowitz’s model it goes with a negative image of both yourself and others^[10]. People often describe it as wanting closeness and fearing it at once.

You will often see “fearful-avoidant” and “disorganised” used as if they were the same thing. They are not. Disorganised attachment is a pattern seen in babies in the Strange Situation, and leading researchers say it is “not a validated individual-level clinical diagnosis”^[16]. In ordinary middle-class families, about 15% of babies show it, and frightening parenting “does not seem to be the only causal factor”^[17]. It does not reliably mean a child was maltreated, and it is not a fixed trait: it describes one relationship, at one point in time^[16]. In adults, the closest pattern is usually called fearful-avoidant.

Secure attachment

Secure attachment means low anxiety and low avoidance: “healthy interdependence with others”^[4]. It does not mean never worrying or never needing space. It means you can usually lean on people and let them lean on you, and closeness does not feel dangerous. Secure attachment is linked with better outcomes in therapy too: in a review of 36 studies, people who were more secure at the start tended to do better^[4].

Attachment types or dimensions?

The four styles are easy to remember, but they are not how most researchers measure attachment today. In 1998, researchers pooled 482 questionnaire items from older attachment measures and found “two essentially independent factors”: avoidance and anxiety. When they grouped people by their scores, the groups matched the four styles^[13]. Later tests found that differences between people look “more consistent with a dimensional rather than a categorical model”^[14].

Attachment also depends on the relationship. A study of more than 21,000 people found that people hold “distinct working models in different relationships”, so you might feel secure with a friend and anxious with a partner^[15]. A style is a summary of where you sit on two scales, with one person, at one time. It is not a box you live in.

Where attachment styles come from

Attachment theory says children build “working models” of relationships: from experience, they “predict the attachment figure’s likely behavior and plan their own responses”. Supportive parents also signal that these models are “open to questioning and revision”^[5]. In Ainsworth’s studies, babies who avoided their mother or were angry at reunion tended to have a less harmonious relationship with her at home^[5].

Childhood matters, but it is only part of the story. Hazan and Shaver saw adult styles as “determined in part by childhood relationships with parents”^[9]. A review of 211 studies found that childhood maltreatment is linked with more attachment anxiety and avoidance in adulthood, especially emotional maltreatment, but the link is modest^[18]. Life later on counts too: in young people aged 7 to 19, a parent’s depression and rising stress between parent and child went with growing insecurity over 3 years, while the 12 genetic variants tested were not linked to attachment^[19]. If family patterns feel like they repeat, see our post on intergenerational trauma and our page on family conflict.

Can attachment styles change?

Yes, partly, and usually slowly. Your early experiences do keep some influence: one review found attachment security “moderately stable across the first 19 years of life”^[20]. But the link weakens with time. Pooling 127 papers on 21,072 people followed for up to 29 years, researchers found medium stability overall (r = .39), and “no significant stability was found in intervals larger than 15 years”. Hard circumstances can shift it too: children in at-risk groups who started out secure were less likely to stay secure^[1].

On average, attachment also eases with age. In a study that followed 628 people over 59 years, between the ages of 13 and 72, attachment anxiety fell, especially in middle and older age, and avoidance fell steadily across life. Being in a relationship went with lower anxiety and avoidance^[2].

But a relationship is not an automatic fix. In a study of 1,036 couples over 20 months, women’s anxiety fell a little, while avoidance stayed stable and change was “largely unaffected by relationship dynamics”^[21]. Researchers think a partner “can provide immediate reassurance, but the path to long-term security may hinge on addressing the individual’s insecure self-perceptions”^[22].

Can avoidant attachment change?

It can ease, but the evidence is thinner than for anxiety. Avoidance does fall with age on average^[2]. In therapy studies, though, “findings are unclear with regard to attachment avoidance”^[3], and in couples it looked stable over 20 months^[21]. One small couples therapy study did find that partners became less avoidant with each other^[23]. Researchers suggest avoidance may ease most through good experiences of depending on someone, but that idea has not yet been tested^[24].

Can you go from anxious to secure?

There is more evidence here. Across therapy studies, attachment anxiety tends to fall during therapy^[3]. One proposed model suggests anxiety eases most in situations that build confidence in yourself^[24]. It is not quick: people higher in attachment anxiety tend to do somewhat less well in therapy on average^[25], so be patient with yourself.

What is earned security?

“Earned security” is a research term for adults who describe a difficult childhood in a clear, reflective way and who look secure in attachment interviews. It grew out of the finding that secure parents “gave a clear and coherent account of early attachments (whether these had been satisfying or not)”^[5]. In a 23-year study of one high-risk US sample, people who were insecure as babies and secure as adults went on to do well in their close relationships. But people labelled earned-secure only from how they remembered childhood were no more likely than people who had always been secure to have been insecure as babies^[26]. So the idea is real, but more complicated than it sounds.

How to move towards secure attachment

There is no proven programme for changing your attachment style. These steps are reflection and practice, built on what the research suggests, with a note on how strong the evidence is for each. And do not wait for a partner to fix it for you: as above, relationships alone do not reliably change attachment^[21].

How attachment styles affect relationships

A review of 73 studies and 21,602 people found that both attachment anxiety and avoidance go with lower relationship quality, and that anxiety is more strongly linked with conflict^[28]. Attachment is also linked with mental health. Across 224 studies, both anxiety and avoidance were linked with more depression, anxiety and loneliness, and anxiety more strongly (r = .42, against r = .28 for avoidance)^[29]. These are links, not proof that attachment causes these problems. If worry is a big part of your life, see our page on anxiety and stress.

One caution: it is tempting to diagnose a partner. Try not to. Attachment differs from one relationship to the next^[15], so how someone acts with you is not a fixed style, and you cannot tell someone’s attachment from their behaviour.

Attachment across cultures

The research here is limited, so treat it with care. A 1988 review of almost 2,000 Strange Situation assessments of babies in 8 countries found that the secure pattern was the most common in all of them. It also found that differences within a country were nearly 1.5 times larger than differences between countries. Most of those countries were Western, and the authors said more data from other parts of the world was needed^[30]. A 2004 survey of 17,804 people in 62 cultural regions found secure attachment was the norm in 79% of cultures, but the four-style model did not fit every culture in the same way^[31].

Some researchers argue that attachment theory “represents the Western middle-class perspective, ignoring the caregiving values and practices in the majority of the world”^[32]. Others point out that adult attachment interviews give similar results “largely independent of language and country of origin”^[33]. The debate is open.

What counts as “too close” or “too independent” differs between families and cultures. In a relationship between two cultures, a different way of showing closeness can look like avoidance or anxiety when it is not. Couples from different cultures often name family acceptance, talking in a third language and different ideas about love and feelings as challenges^[34], and talking openly about those differences is linked with growing together^[27].

Are attachment style quizzes accurate?

An online quiz can be a useful prompt for reflection, but it is not a diagnosis. The questionnaires researchers use were built for research: one review examined 29 of them^[35], and one of them, the Experiences in Close Relationships questionnaire, was developed for “attachment researchers who wish to use self-report measures”^[13]. Questionnaires and in-depth attachment interviews also overlap very little: their link (r = .09) is “trivial to small”^[36].

A quiz that puts you in one box oversimplifies two sliding scales^[14], and your answers can change depending on which relationship you have in mind^[15] and how you feel that week. Use a result as a question to explore, not an answer, and never as a way to label someone else.

Is attachment style a diagnosis?

No. An attachment style describes patterns in close relationships. There is a diagnosis with a similar name, reactive attachment disorder, but it is very different. It follows “grossly inadequate child care (e.g., severe neglect, maltreatment, institutional deprivation)”, and the child does not turn to their caregiver for comfort. It “can only be diagnosed in children, and features of the disorder develop within the first 5 years of life”^[37]. Correctly diagnosed attachment-related disorders are “in all likelihood very rare in the general population”^[16]. If you are worried about a child, talk to a doctor or a child mental health professional.

How therapy can help

A review of therapy studies found that “attachment security increases following therapy, whereas attachment anxiety decreases”, across different kinds of therapy and settings. But findings for avoidance were unclear, most studies had no control group, and the authors say “further controlled trials are required”^[3]. Gains in security during therapy may go along with better results overall^[4].

The relationship with your therapist matters. Across 295 studies, a stronger working bond went with better outcomes, and the link was about the same in internet-based therapy^[38]. If you tend to keep people at a distance, it may take longer to trust a therapist, and that is normal^[39].

For couples, the therapy most studied for attachment change is emotionally focused couple therapy (EFT). In a study of 32 couples, partners became less avoidant with each other, and some became less anxious^[23]. In the same couples, anxiety about each other kept falling, more slowly, over two years of follow-up^[40]. But that was one small group with no comparison group, studied by a team that included the therapy’s developer. A 2026 review of 40 studies found “no couple therapy models met the standards for strong empirical support”, and rated EFT, like six other couple therapies, as having modest support^[41].

Language matters too. Talking about closeness and fear is hard enough in your first language. In an older review of 76 studies, therapy in a person’s native language (where that was not English) was twice as effective as therapy in English^[42], and therapy adapted to a person’s culture tends to work better^[43].

Your options include:

• Reading and reflection, like the steps above: a place to start.
• Couples therapy, if you both want to work on the relationship.
• Individual therapy, to work on your own patterns, whatever your relationship status.
• Bliss Therapy: individual therapy with a therapist you choose, who speaks your language, by video, with no waitlist.

On Bliss Therapy, 45 psychologists and counsellors list relationships, with sessions from €30 (about £25).

Our therapists are based outside the UK and the US, so they are unlikely to be on UK registers or covered by US health insurance. Ask any therapist about their training and their registration in the country where they practise.

When to get help

Consider talking to a therapist if worries about your relationships take up a lot of your time, if the same painful pattern keeps repeating, or if fear of closeness or of being left is affecting your relationships, work or sleep. Check in on your mood too: attachment insecurity is linked with depression and anxiety^[29], so if you also feel low or anxious most days, talk to a doctor or a therapist. If you are afraid of a partner, contact a domestic abuse service.