Autistic people do form attachments, often very deeply. But the way those bonds are expressed, and the way others read them, can differ enough from neurotypical expectations to create genuine confusion, in relationships, in therapy rooms, and in clinical assessments. If you are autistic, or think you might be, and have spent years feeling misunderstood by the people you were closest to, understanding how autism and attachment interact may help some of that experience make more sense.
If an autism assessment is something you are already considering, it may also help to understand this dimension of the autistic experience before you start.
Can autism cause attachment issues?
Autism does not cause attachment disorders, but it can shape how attachment is expressed and how easily it is misread by others. Reactive attachment disorder and disinhibited social engagement disorder are rooted in early severe neglect or insufficient caregiving. Autism is neurodevelopmental. The two are distinct conditions, even if they can look similar from the outside.
What autism can do is create friction in the attachment relationship. An autistic person may struggle to signal their attachment needs in ways a neurotypical caregiver recognises. The caregiver may consistently misread those signals. Over years, that accumulated mismatch can contribute to insecure relational patterns, not because of autism itself, but because the autism went unrecognised and the relationship was built around a fundamental misunderstanding.
Getting this distinction right matters in practice. It means the goal is not to treat autism as an attachment problem, but to understand where the two intersect, and where support can most usefully be directed.
Does autism affect how people form attachments?
Autistic people can and do form secure attachments. The older clinical view, that autistic children were essentially unattached, aloof, indifferent to whether their caregiver was in the room, has not held up to scrutiny.
A 2004 meta-analysis by Rutgers and colleagues found that attachment security is fully compatible with autism. Autistic children showed clear preference for their caregiver, sought proximity, and used them as a secure base. A 2017 systematic review by Kalyva confirmed this: autistic children can form selective emotional bonds, even where the expression of those bonds differs from neurotypical expectations.
More recently, a 2025 meta-analysis by Trottier-Dumont and colleagues found that around 46% of autistic children show secure attachment when measured using the Strange Situation Procedure, a figure broadly in line with general population distributions. Rates of disorganised attachment were somewhat higher in autistic samples, but secure attachment was still the most common classification.
The question, then, is not whether autistic people attach. It is how that attachment is expressed, and how reliably it gets recognised by the people around them.
Why does attachment look different in autism?
Autistic attachment can be harder to read because of three distinct factors, and they do not all point to the same underlying mechanism.
Sensory processing as a filter on closeness
Sensory sensitivities can change the form that attachment takes without reducing the underlying need for it. Physical closeness and touch, which are central to many neurotypical expressions of connection, may feel uncomfortable or overwhelming to an autistic person. A child who pulls away from a hug, avoids sustained eye contact, or prefers sitting beside someone rather than on their lap may still be strongly attached. The need is present; what has changed is the channel through which it can be met. Sensory avoidance and emotional distance are not the same thing, and conflating them leads to the wrong conclusions.
How masking obscures attachment needs
Masking can make an autistic person’s attachment needs effectively invisible to others. Autistic people who have learned to suppress visible signs of distress, or to perform expected emotional responses, may appear far more independent than they actually are. A 2022 study by Camm-Crosbie and colleagues found consistent links between high levels of camouflaging in autism and poorer mental health outcomes, including greater internalising symptoms. In attachment terms, this can mean someone looks composed in the session and falls apart at home. Their distress and their dependency are real. They are simply not visible in the moments when they would typically be observed, which matters enormously for clinical assessment.
This is particularly relevant for adults who have been masking for decades. Apparent functionality has a way of being taken as evidence of security when it is anything but.
The double empathy problem
Communication difficulties between autistic and non-autistic people run in both directions, and attachment is no exception. The double empathy problem, first articulated by autistic researcher Damian Milton in 2012, argues that neither party is straightforwardly to blame for the mismatch: each is simply reading the other through a different interpretive framework. When a caregiver consistently misreads their autistic child’s bids for connection, the interaction begins to look insecure from the outside, even though the relational need is genuine on both sides. The problem is the mismatch, not the absence of attachment.
Why do some autistic people attach intensely to one person?
Are autistic people more likely to get attached?
Autistic people are not uniformly more prone to forming attachments, but when they identify someone as truly safe, the bond that forms can be particularly intense and significant. When an autistic person finds someone predictable, reliably available, and matched to their communication style, that person provides something quite specific: reduced uncertainty, lower cognitive load, and a dependable source of emotional and sensory co-regulation. The attachment makes sense in exactly those terms.
The pattern of intense attachment to one person, often a parent, a partner, or a close friend, is sometimes read as disordered. It is usually better understood as a brain identifying, accurately, who is genuinely safe. Finding one person who meets those needs reliably is not pathology. It is pragmatic.
Why do autistic people love so deeply?
When the cognitive effort of decoding an unfamiliar social world is reduced, because someone is known, trusted, and does not require constant translation, autistic people can experience and express connection with considerable depth and intensity. The emotional capacity is not diminished by autism. What can differ is the threshold for reaching it, and the form it takes once it is there. That same depth often shows up as loyalty, consistency, and an investment in the relationship that does not waver the way more casual bonds might. Our piece on autism and relationships explores this further.
None of this means the intensity never creates difficulties. It can place real pressure on the attachment figure, strain other relationships, and leave the autistic person poorly resourced when that person is unavailable. The more useful question is not “what is wrong with this attachment?” but “what is this attachment doing, and how can we build on it rather than disrupt it?”
What do attachment styles look like in autistic adults?
What is the most common attachment style in autism?
Secure attachment is the most common classification in autistic samples, found in roughly 46% of autistic children studied using the Strange Situation Procedure (Trottier-Dumont et al., 2025). Disorganised attachment is more elevated than in general population norms, and the adult picture shows higher rates of insecure attachment overall, but the assumption that autistic people cannot be securely attached is not supported by the evidence.
Attachment theory describes four broad patterns: secure, anxious-preoccupied, dismissive-avoidant, and fearful-avoidant. Each has a distinct shape in relationships. Our attachment styles blog covers all four in detail.
Anxious-preoccupied attachment in autistic adults
Anxious-preoccupied attachment involves a heightened vigilance to signs of rejection or abandonment, and it appears with particular frequency in autistic adults. For someone who has spent years being misread, having their needs dismissed, and masking constantly just to maintain relationships, that hypervigilance is not irrational. It is a learned adaptation to a specific relational history, and it makes clinical sense once you know what sits behind it.
Fearful-avoidant attachment in autistic adults
Fearful-avoidant attachment involves both a longing for connection and a fear of it, and it also appears in autistic adults, particularly those with a history of rejection or difficult early relationships. The pull toward closeness and the pull away from it operate simultaneously, in a way that is exhausting to carry and often hard to articulate to the people who are on the receiving end of it.
Attachment, camouflaging, and burnout
Autistic adults show higher rates of insecure attachment than the general population, and that insecurity is associated with poorer mental health outcomes. A 2025 systematic review by Pérez-Lautz and colleagues found this pattern consistently across the available adult literature. The direct link to autistic burnout is still being researched and it would overstate the evidence to draw a simple causal line. What is well established, however, is the association between sustained camouflaging, internalising symptoms, and the kind of chronic exhaustion many autistic adults describe. Performing security you do not feel, across decades of relationships, has a cost.
Can autism be mistaken for an attachment disorder?
Yes. The behavioural overlap between autism and attachment disorders like reactive attachment disorder (RAD) and disinhibited social engagement disorder (DSED) is significant enough that misdiagnosis in either direction is a genuine clinical risk. Difficulties with emotional regulation, unusual proximity behaviour, inconsistent social engagement, apparent trouble forming close relationships: all of these can appear in both presentations.
The reverse confusion also occurs. Children with genuine early maltreatment histories and attachment difficulties are sometimes read as autistic, while their actual relational history is not adequately explored.
Key differentiators between autism and attachment disorders
These are the factors most useful in distinguishing between the two:
- Developmental history: Autism is present from birth and does not require adverse caregiving. RAD and DSED require a history of severe neglect or insufficient care.
- Restricted and repetitive behaviours: Characteristic of autism, and not a feature of attachment disorders.
- Pattern across contexts: Attachment disorders tend to affect social behaviour uniformly. Autistic presentation is often more context-dependent.
- Quality of social reciprocity: The nature of the social difficulty differs between the two in ways a trained clinician can assess directly.
- Co-occurrence: The two can and do co-exist. A diagnosis of autism should not automatically rule out attachment difficulties, and vice versa.
A clinical tool called the Coventry Grid is commonly used when the picture is ambiguous and a structured differential is needed.
For late-diagnosed autistic adults, this confusion can carry real consequences. Some will have spent years in therapeutic contexts framed around attachment difficulties or personality presentations that obscured an underlying autism diagnosis entirely. The appropriate support for each looks quite different, which is why getting the formulation right matters.
Can autistic people develop secure attachment?
Yes, and the evidence points clearly to what supports it.
For children: caregiver attunement matters most
Secure attachment in autistic children is most strongly supported by caregiver sensitivity and consistent, predictable responsiveness, even where the child’s communication style differs from neurotypical expectations. The focus in family and parenting work is on improving the fit between caregiver and child: reducing communication mismatches, building the child’s capacity for co-regulation, and responding to the child’s actual signals rather than the signals a neurotypical child might send. Pressing for neurotypical attachment behaviours that are simply not accessible tends to increase distress rather than security.
For adults: autism-affirming therapy over attachment-specific models
The evidence for attachment-specific therapies tailored to autistic adults is not yet strong enough for firm recommendations. What the research does support is autism-affirming therapy, trauma-informed approaches where there is adverse history, work on emotional regulation, and a reduction in the environmental pressure to mask. Therapy that starts from genuine acceptance of how an autistic person actually experiences and expresses connection, rather than working to reshape it toward neurotypical norms, tends to be the more defensible and more effective starting point.
How we can help
If any of what you have read here feels familiar, whether that is a pattern of intense attachment, a long history of relationships that never quite made sense, or a sense of being fundamentally misread by the people you were closest to, it may be worth exploring whether autism is part of the picture. The Private Therapy Clinic offers a thorough autism assessment for adults seeking a clear clinical picture, as well as psychiatry for those whose presentation involves significant anxiety, mood difficulties, or other concerns that may be connected to their neurodevelopmental profile. A free 15-minute consultation is available if you would like to talk through your options before committing to anything.
