Is Addiction a Disease or a Choice
Why effective recovery needs compassion for what happened and agency over what happens next
By Chris Howarth • Approximately 10 minute read
Arguments about addiction often begin with a choice between two explanations. Either addiction is a disease that takes control away, or it is a behaviour that a person can decide to change.
Both positions contain something important. Neither is sufficient on its own.
The disease model explains why repeated alcohol use can alter motivation, reward, stress and decision-making. It supports treatment rather than condemnation. The language of choice protects something equally important: the possibility that a person can take action, accept help and build a different life.
The difficulty begins when either idea is pushed too far. Disease can be misheard as destiny. Choice can be turned into blame. One risks telling people that they are powerless to change; the other suggests that they would change if only they wanted it enough.
My own view comes from lived experience of alcohol addiction and recovery, as well as years spent talking with and supporting other people in recovery. The most useful position I have found is neither moral judgement nor biological fatalism. Addiction can profoundly constrain choice without erasing it. Recovery works by strengthening the choices that remain and by building the medical, psychological and social conditions in which better choices become possible.
The false choice between illness and agency
Modern clinical definitions are already more nuanced than the public argument. The American Society of Addiction Medicine describes addiction as a treatable chronic medical disease involving interactions among brain circuits, genetics, the environment and a person’s life experiences. That is a much broader claim than saying that the brain is the only thing that matters.
Alcohol dependence is not simply a series of ordinary decisions made under ordinary conditions. Repeated use can change the relative value attached to immediate relief and longer-term consequences. Craving, withdrawal, stress, habit, mental health, trauma and a person’s environment can all narrow the realistic options available in a particular moment.
Severe dependence also has medical consequences that cannot be overcome by determination alone. The current UK clinical guidelines warn that suddenly stopping alcohol can cause seizures, delirium tremens and Wernicke’s encephalopathy. Some people need a planned, medically assisted withdrawal. Any account of addiction that reduces this to weak character is medically unsafe as well as unfair.
But constraint is not the same as the complete disappearance of agency. A contemporary review of the brain disease model makes this distinction clearly: addiction can shift the probability of harmful choices and compromise a person’s capacity to choose advantageously without annihilating that capacity. The same brain in which addiction develops is also the brain through which learning, treatment and change occur.
This is why disease and choice are not opposites. They describe different parts of the same problem.
What the disease model gets right
The disease model has done important work. It challenges the idea that addiction is a moral defect and makes a straightforward case for healthcare. A person with alcohol dependence may need clinical assessment, medication, psychological treatment and continuing support. They should not have to prove that they are virtuous before receiving any of them.
The model can also relieve destructive shame. There is a difference between recognising the harm caused during addiction and believing that addiction proves someone is fundamentally bad. Shame tends to collapse that distinction. It turns a serious, treatable condition into an identity.
For some people, learning that alcohol has affected their brain and behaviour provides an explanation for experiences that previously felt incomprehensible. It can replace the question ‘What is wrong with me?’ with more practical questions about risk, treatment and recovery.
The medical lens is therefore valuable. It becomes less useful when it is presented as the complete explanation for every person, every stage and every severity of alcohol problem. Not everyone follows a chronic relapsing course. People differ in vulnerability, circumstances, impairment and the help they need. A model that earns access to treatment should not become a label that fixes a person’s future in advance.
Where disease language can go wrong
People do not hear clinical language in a laboratory. They interpret it through fear, hope, stigma and their previous attempts to change.
When ‘brain disease’ is translated into ‘my brain makes me do it’, the language can weaken self-efficacy. A lapse may feel inevitable. Treatment may become something professionals do to a passive patient rather than work undertaken with an active participant. Family members may lower their expectations so far that support becomes resignation.
This is not a necessary consequence of the disease model, but it is a possible one. Good practice has to address it directly. A diagnosis should explain why change is difficult and why support matters. It should never imply that effort, learning or personal action are irrelevant.
The opposite message is equally damaging. Telling someone that alcohol use is always freely chosen ignores how craving, withdrawal and altered reward learning affect behaviour. It also ignores poverty, trauma, isolation, mental illness, availability and the social environment. Calling every act a choice does not tell us how free, informed or realistically available that choice was.
The task is to preserve agency without pretending that everyone starts from the same position.
Responsibility without blame
The word responsibility causes confusion because it is used to mean several different things. It can refer to causing a problem, deserving blame for it, accepting the consequences of an action or taking part in changing what happens next. Those meanings should not be treated as interchangeable.
A person may not be morally responsible for the combination of vulnerability, experience and exposure through which dependence developed. They are still the person whose participation is required in recovery. No clinician, sponsor, partner or service can practise a coping skill, take medication, make a telephone call or decline the next drink on someone else’s behalf.
That does not mean leaving the person alone with an instruction to choose better. Agency is affected by conditions. Safe withdrawal, stable housing, medication, therapy, supportive relationships, meaningful activity and distance from high-risk environments can all expand a person’s capacity to act. Exhaustion, fear, homelessness, untreated trauma and easy access to alcohol can shrink it.
Supported responsibility is therefore a better aim than either blame or passivity. It says that the person matters as an active participant and that the surrounding system has duties too.
In practice, responsibility may mean attending an assessment, being honest about consumption, agreeing a safe withdrawal plan, taking medication as prescribed, noticing triggers, repairing relationships, returning after a lapse and asking for help earlier next time. These are actions, not verdicts on character.
Recovery is active and shared
The first UK-wide clinical guidelines for alcohol treatment, published in 2025 and updated in 2026, reflect this combined approach. They describe personalised, non-judgemental care in which the practitioner and the person reach a shared understanding of needs, strengths and goals. They include medical treatment, psychological interventions, harm reduction, peer support and longer-term recovery support.
This matters because evidence-based practice is sometimes mistaken for a contest in which one intervention must defeat all the others. Alcohol problems do not occur in a vacuum, and treatment is rarely one event. A person may need medication for withdrawal or relapse prevention, cognitive behavioural strategies, help with housing or mental health, family work, peer connection and a plan for rebuilding daily life.
The guidelines also emphasise self-efficacy: confidence in the ability to make and sustain change. Confidence is not created by telling someone that recovery should be easy. It develops when a person carries out achievable actions, observes change and receives help to understand setbacks without turning them into proof of failure.
This is the practical middle ground. The condition is taken seriously enough to warrant care, while the person is treated as capable of learning and acting within that care.
Where Alcoholics Anonymous fits
Alcoholics Anonymous is often pulled into the disease-versus-choice debate as evidence for one side or the other. That misses how the programme operates for many of the people who use it.
AA asks members to accept that controlling alcohol has become unreliable, but it does not ask them to become inactive. The Twelve Steps require sustained action: seeking support, examining behaviour, acknowledging harm, making amends where appropriate, practising reflection and helping others. The idea of powerlessness over alcohol sits alongside responsibility for the work of recovery.
The evidence is stronger than it was when many familiar arguments about AA were formed. A 2020 Cochrane review included 27 studies with 10,565 participants. Manualised Twelve-Step Facilitation and AA-related interventions performed at least as well as established treatments such as cognitive behavioural therapy on most drinking outcomes and often performed better for continuous abstinence. They also produced healthcare cost savings.
That finding needs careful interpretation. Structured Twelve-Step Facilitation is not the same thing as attending one meeting and deciding whether AA ‘worked’. Nor does evidence of average benefit mean that AA is right for every person. Some people object to its language, culture or abstinence focus. Others prefer SMART Recovery, professional therapy, medication, another peer community or a combination of approaches.
Current UK guidance reflects that plurality. It recommends discussing mutual aid in relation to the person’s own goals and helping them engage with an appropriate group, including AA or SMART Recovery. If one group is not a good fit, the answer is not to declare the person resistant to recovery. It is to help them find support they can use.
AA should therefore be neither dismissed nor imposed. It is one evidence-supported route within a wider recovery system, and for some people it provides the social connection, practical structure and example of long-term change that other services struggle to reproduce.
What evidence based practice really requires
Evidence-based practice is broader than selecting the intervention with the strongest headline. It brings together the best available research, professional judgement and the preferences, circumstances and experience of the person seeking help.
Research can tell us what tends to help groups of people. It cannot decide, on its own, what a particular person is ready to attempt, which setting is safe, what barriers they face or which source of support they will trust. Lived experience cannot replace controlled research, but research that ignores lived experience can misunderstand why an intervention is accepted, rejected or sustained.
The original disease-versus-choice argument is attractive because it promises a single answer. Real practice requires a formulation: what is maintaining this person’s alcohol use, what risks are present, what strengths can be mobilised and what combination of help is most likely to work now.
The answer may change over time. Early recovery can require more external structure and clinical support. As stability grows, the person may take on more responsibility, test new skills and build a life that makes alcohol less central. A lapse may require the plan to change rather than the person to be written off.
A better question for recovery
Asking whether addiction is a disease or a choice encourages us to decide who is right before deciding what is useful. A better question is this: what combination of medical care, psychological skill, social support and personal action will give this person the best chance of change?
Compassion and agency are not competing values. Compassion recognises the forces that restrict choice and removes shame from the search for help. Agency recognises that the person is more than those forces and has a meaningful part in what happens next.
The clearest principle I have taken from recovery is that accepting a limit does not require surrendering the whole of one’s power. A person may be unable to make alcohol behave predictably. They can still choose to seek treatment, enter a room, call someone, follow a plan, learn from a lapse and try again.
Good treatment protects that capacity. It neither blames people for being ill nor asks them to wait passively to be cured.
A safety note
Alcohol withdrawal can be dangerous. A person who may be dependent on alcohol should seek medical advice before stopping suddenly. Severe withdrawal symptoms, seizures, hallucinations or confusion require urgent medical help.
Sources and further reading
UK Clinical Guidelines for Alcohol Treatment
UK guidance on assessment and treatment and recovery planning
UK guidance on psychosocial interventions mutual aid and self efficacy
UK guidance on recovery support services and lived experience
UK guidance on medication and medically assisted withdrawal
American Society of Addiction Medicine definition of addiction
Addiction as a brain disease revised
Cochrane review of Alcoholics Anonymous and Twelve Step Facilitation
NHS alcohol support and treatment information
