Codependency describes a pattern of caring for someone else at your own expense — putting another person’s needs, feelings, and problems first so consistently that your own needs, boundaries, and identity start to disappear. The word is everywhere in popular psychology, and many people recognise themselves in it. But it is also one of the most contested terms in relationship psychology: it is not a diagnosis, it does not appear in the Diagnostic and Statistical Manual of Mental Disorders (DSM-5; American Psychiatric Association, 2013), and researchers disagree about what exactly it means, how to measure it, and whether it is a useful concept at all.
This article explains what the term means, where it came from, the signs people usually describe, and what research does and does not support. It is educational, not a diagnosis — reading about codependency cannot tell you whether you or anyone else “has” it, because there is no agreed test or set of criteria that could settle that question. If a relationship is causing you serious distress, or you recognise patterns here that feel hard to change on your own, a qualified therapist is the right person to talk to.
What “Codependency” Means
At its simplest, codependency is used to describe a one-sided relationship pattern in which one person organises their life around another person’s needs — the caretaking, the fixing, the accommodating — and keeps doing it even when it costs them their health, friendships, finances, or sense of self. The other person may be struggling with addiction, illness, or emotional difficulty, or may simply have grown used to being looked after. What matters for the label is not the other person’s problem but the pattern: caring that has become compulsive, self-neglecting, and hard to stop.
Three features show up in most descriptions of the pattern. The first is self-worth tied to being needed: the person feels valuable mainly when they are helping, rescuing, or being useful to their partner, and feels empty or anxious when they are not. The second is difficulty with boundaries: saying no feels selfish or frightening, so the person agrees to things they do not want to do and then resents it. The third is fear of abandonment: the person worries that if they stop caretaking, the relationship will end — so they keep giving, even when giving is hurting them. The site’s guide to boundaries in relationships looks at the boundary side of this in more depth.
It is also not simply “caring a lot” — caring deeply for a partner, including through illness or hard times, is normal and healthy. The term is meant to describe caring that has become compulsive and one-sided, where the caretaker cannot stop even when they want to. Whether that distinction can be drawn reliably is one of the things researchers argue about.
Where the Term Comes From
Codependency began in the world of addiction treatment. In the 1970s and 1980s, therapists working with the families of people with alcohol dependence noticed that spouses and children often organised their whole lives around the drinker’s behaviour — covering up problems, managing crises, and sacrificing their own needs to keep the family functioning. The term “co-alcoholic” was used first, then broadened to “codependent” as therapists saw similar patterns in families affected by other kinds of addiction and dysfunction.
In 1986, psychiatrist Timmen Cermak proposed formal diagnostic criteria for codependence, arguing that it should be recognised as a personality disorder in its own right (Cermak, 1986). His proposal did not succeed — the diagnosis was never adopted — but the idea spread far beyond clinics. Through popular self-help writing in the late 1980s and 1990s, “codependent” became a household word, applied to almost any relationship in which one person gave too much. That popularity is part of the problem researchers now point to: a clinical term became a popular label, and in the process it stretched to cover so many different things that its meaning blurred.
The research response came later. In 2000, Dear and Roberts published the Holyoake Codependency Index, one of the first attempts to measure codependency with a standard questionnaire (Dear & Roberts, 2000), validated in a follow-up study in 2005 (Dear & Roberts, 2005). These were genuine steps toward making the concept testable — but one measure does not settle a definition, and researchers still disagree about what the concept includes.
Signs Commonly Described
Because there is no agreed diagnostic checklist, the signs below are best understood as commonly described patterns, not criteria for a diagnosis. They come from clinical writing and from the measures researchers have built, and they are worth knowing — but recognising some of them in yourself does not mean you “have codependency”, any more than recognising one symptom of anything else would settle a diagnosis.
- Putting the other person’s needs first, always. Your own plans, rest, and friendships are routinely set aside when your partner needs something, and this feels normal rather than chosen.
- Difficulty saying no. Refusing a request brings guilt, anxiety, or fear of conflict, so you agree even when you do not want to — and then feel resentful.
- Self-worth that depends on being needed. You feel most valuable when you are helping or fixing things for your partner, and lost or anxious when there is nothing to fix.
- Trying to control through caretaking. You manage your partner’s problems, moods, and choices — not by demanding things directly, but by doing so much for them that they depend on you. The control is real even when it looks like kindness.
- Trouble naming your own feelings and needs. When asked what you want, you draw a blank — or you answer with what your partner wants. Your own preferences have gone quiet.
- Staying in one-sided relationships. You remain in relationships where you give far more than you receive, and the idea of leaving feels impossible, even when the relationship is making you unhappy.
A note of caution: several of these signs — low self-esteem, fear of rejection, difficulty with boundaries — also appear in anxiety, depression, and other relationship patterns. A list of signs cannot distinguish between them or explain why a pattern exists, which is one reason a label from an article is no substitute for a conversation with a professional.
Codependency vs Anxious Attachment
The concept most often confused with codependency is anxious attachment. The two overlap: both involve fear of abandonment, a strong need for closeness, and difficulty tolerating distance in a relationship. But they are not the same thing, and the difference matters.
Anxious attachment is one of the attachment styles identified in research on adult relationships. In a landmark study, Hazan and Shaver (1987) showed that the attachment patterns first observed in children — secure, avoidant, and anxious — also appear in adult romantic relationships, and that anxiously attached adults tend to worry about abandonment, want more closeness than their partner offers, and feel distressed by separation. Attachment styles are a well-studied framework: they have been measured with validated questionnaires, replicated across many samples, and linked to specific behaviours in relationships. The site’s guide to attachment styles compares all four styles, and the guide to anxious attachment covers this style in depth.
Codependency, by contrast, is a label for a caretaking dynamic, not a style of attachment. An anxiously attached person may or may not be codependent, and a codependent person may or may not be anxiously attached. The overlap is real — both can involve clinging to a relationship out of fear — but the concepts describe different things: attachment describes how a person bonds and seeks closeness, while codependency describes compulsive caretaking at one’s own expense. Attachment theory also has something codependency lacks: decades of research with agreed measures and consistent findings. That does not make attachment theory “true” and codependency “false” — it makes attachment the more solidly grounded framework for understanding why someone fears abandonment.
Codependency vs Healthy Interdependence
The other confusion worth clearing up is between codependency and interdependence — the healthy mutual reliance that good relationships are built on. The difference is not whether partners depend on each other. They do, and they should: needing each other is part of what a relationship is. The difference is how the dependence works.
In a healthy interdependent relationship, support flows both ways. Both partners ask for help and offer it. Both can say no without fearing the relationship will end. Each person keeps a life of their own — friends, interests, opinions — alongside the shared life. Caring for a partner through illness or difficulty is part of the deal, not a sign of dysfunction. The site’s guide to the psychology of relationships describes this balance as one of the things that makes relationships work.
In the pattern codependency describes, the flow is one-sided and compulsive. One person gives and the other takes, and the giver cannot stop — not because they have decided to be generous, but because stopping feels dangerous. Boundaries are absent rather than flexible. The caretaker’s identity shrinks to the role of helper. The test is not “do you care for your partner?” but “can you stop caring at your own expense when it is harming you — and do you still have a self left when you do?” If the answer is no, that is the pattern the term is trying to name, whether or not the term itself is scientifically solid.
What Researchers Criticize About the Concept
The criticisms of codependency are serious, and an honest article should give them full weight. The most thorough early critique came from Anderson (1994), who argued that the concept was poorly defined, applied so broadly that it could describe almost any caring relationship, and used in ways that pathologised ordinary caregiving — particularly the caregiving traditionally expected of women. When a concept covers everything from the spouse of a person with addiction to anyone who is “too nice” in a relationship, it stops being a useful explanation and becomes a label that can be stuck on almost anyone.
Stafford (2001) asked the question directly — “Is codependency a meaningful concept?” — and found the evidence wanting. Her review pointed to the same core problems: no agreed definition, measures that overlap heavily with other well-established traits such as low self-esteem and anxiety, and little evidence that codependency is a distinct condition rather than a new name for familiar difficulties. If a measure of codependency mostly captures low self-esteem, then calling the result “codependency” adds a label without adding understanding.
The measurement problem is the practical heart of the criticism. Dear and Roberts (2000) built a serious instrument, validated in 2005, but one questionnaire cannot create a consensus. Different measures of codependency define it differently, so studies using different measures may not be studying the same thing at all. Without agreed criteria, researchers cannot reliably estimate how common codependency is, what causes it, or what helps — which is why there are no trustworthy statistics on its prevalence and no established treatment for it. A concept that cannot be measured consistently cannot be tested properly, and one that cannot be tested properly should be used with caution.
There is also a moral criticism, raised in the critical literature, that the label can blame the wrong person. When the partner of someone with an addiction is called “codependent”, the focus shifts from the person causing the harm to the person coping with it — as if the partner’s caretaking were the problem, rather than a response to a difficult situation. That does not mean caretaking patterns cannot be unhealthy. It means the label carries a judgement, and that judgement deserves to be examined before it is applied to anyone.
None of this means the experiences people describe with the word are unreal. Many people recognise themselves in the pattern — the compulsive giving, the lost boundaries, the fear of leaving — and find the word useful for naming what is happening to them. The criticism is not that these experiences do not exist. It is that the word is a blunt tool for describing them: popular, evocative, and scientifically shaky. Readers deserve to know both sides.
What Helps, According to the Evidence
Honesty requires saying this plainly: there is no established, evidence-based treatment specifically for codependency, because there is no agreed definition of the condition to treat and no clinical trials of a codependency-specific therapy that meet normal research standards. Anyone promising a proven cure for codependency is overstating what the evidence supports.
What does exist is help for the patterns the word describes. Therapy, particularly therapy focused on boundaries, self-esteem, and communication, can help a person practise saying no, notice their own needs, and tolerate the discomfort of disappointing someone. Support groups, including Codependents Anonymous, offer a place to talk with people in similar situations, though their effectiveness has not been tested in the way a medical treatment would be. Learning to set boundaries — a skill the site’s guide to boundaries in relationships covers in practical detail — is often the most concrete place to start.
Two cautions belong here. First, change is usually slow: patterns built over years do not reverse in a weekend, and anyone selling a quick fix should be treated with scepticism. Second, if a relationship involves abuse, coercion, or fear, the priority is safety, not self-improvement — professional help and, where needed, support services for abuse are the right first step, and no amount of boundary-setting makes an unsafe relationship safe. The word “codependency” should never be used to suggest that a person being harmed is responsible for the harm.
Key Takeaways
- Codependency describes a pattern of caretaking at your own expense — compulsive, one-sided giving in which your own needs, boundaries, and identity shrink — not a diagnosis.
- The term began in addiction treatment in the 1970s and 1980s, was formalised in a 1986 proposal by Cermak that was never adopted, and then spread through popular self-help writing.
- Commonly described signs include always putting the other person’s needs first, difficulty saying no, self-worth tied to being needed, control through caretaking, trouble naming your own needs, and staying in one-sided relationships — but these are descriptions, not diagnostic criteria.
- Codependency is not the same as anxious attachment: attachment styles are a well-researched framework with validated measures, while codependency is a looser label for a caretaking dynamic.
- Healthy interdependence is mutual, flexible, and keeps both partners’ autonomy; codependency is one-sided and compulsive, and the giver cannot stop even when giving harms them.
- Researchers criticise the concept as poorly defined, hard to measure, overlapping with low self-esteem and anxiety, and at risk of blaming the person coping with a difficult partner rather than the difficulty itself.
References
- American Psychiatric Association. (2013). Diagnostic and Statistical Manual of Mental Disorders (5th ed.). American Psychiatric Publishing. https://doi.org/10.1176/appi.books.9780890425596
- Anderson, S. (1994). A critical analysis of the concept of codependency. Social Work, 39(6), 677–685. https://doi.org/10.1093/sw/39.6.677
- Cermak, T. L. (1986). Diagnosing and Treating Co-Dependence: A Guide for Professionals Who Work with Chemical Dependents, Their Spouses, and Children. Johnson Institute Books.
- Dear, G. E., & Roberts, C. M. (2000). The Holyoake Codependency Index: Investigation of the factor structure and psychometric properties. Psychological Reports, 87(3), 991–1002. https://doi.org/10.2466/pr0.2000.87.3.991
- Dear, G. E., & Roberts, C. M. (2005). Validation of the Holyoake Codependency Index. The Journal of Psychology, 139(4), 293–314. https://doi.org/10.3200/jrlp.139.4.293-314
- Hazan, C., & Shaver, P. (1987). Romantic love conceptualized as an attachment process. Journal of Personality and Social Psychology, 52(3), 511–524. https://doi.org/10.1037/0022-3514.52.3.511
- Stafford, L. L. (2001). Is codependency a meaningful concept? Issues in Mental Health Nursing, 22(3), 273–286. https://doi.org/10.1080/01612840152053101
Keep Exploring
- The Psychology of Relationships: What Makes Them Work
- Boundaries in Relationships: How to Set Healthy Ones
- Anxious Attachment: Signs, Causes & Growth Strategies
- Attachment Styles: The 4 Types Compared (Secure to Avoidant)
Is codependency a mental health diagnosis?
No. Codependency does not appear in the DSM-5 and is not a recognised diagnosis. It is a popular term for a pattern of caretaking at your own expense, and researchers disagree about whether it is a distinct, measurable concept at all.
What are the signs of codependency?
Commonly described signs include always putting the other person’s needs first, difficulty saying no, self-worth that depends on being needed, trying to control a partner through caretaking, trouble naming your own feelings and needs, and staying in one-sided relationships. These are descriptions from the literature, not diagnostic criteria, and several of them also appear in anxiety, depression, and other patterns.
Is codependency the same as anxious attachment?
No, though they overlap. Anxious attachment is a well-researched attachment style involving fear of abandonment and a strong need for closeness. Codependency is a looser label for a pattern of compulsive caretaking. A person can be anxiously attached without being codependent, or codependent without being anxiously attached.
Can codependency change?
The patterns it describes can change, but usually slowly — through work on boundaries, self-esteem, and communication, often with a therapist or a support group. There is no established treatment specifically for codependency and no quick fix, because the concept itself has no agreed definition to treat.
Is caring for someone the same as codependency?
No. Caring deeply for a partner, including through illness or hard times, is normal and healthy. It becomes the pattern codependency describes when the caring is compulsive and one-sided, you cannot stop even when it harms you, and your own needs and identity have disappeared. Healthy relationships are interdependent: support flows both ways and both partners keep their autonomy.