Printed on: July 11th, 2025

Please note, once printed or downloaded, articles cannot be updated. By bookmarking the article and continuing to access it online, you can be sure you are reading the most up-to-date information and that you are not in breach of copyright laws.

Alcohol

Alcohol

About one-quarter of the UK adult population drinks alcohol in a way that is potentially or actually harmful, but few of them access treatment.

The fifth report of the Lancet commission on liver disease was published in 2018 (Lancet 2018;392:2398). The report focused on the public health aspects of liver disease in the UK. There were some stark facts regarding the increasing burden of disease in the UK:

• Hospital admissions with a primary alcohol-related diagnosis have increased by 17% between 2006/07 and 2016/17.
• Alcohol-related deaths are increasing, including an increase in deaths associated with drink driving.
• 31% of men and 16% of women drink more than the recommended weekly limit of alcohol.
• Between 2013/14 and 2017/18, funding for addiction services has been cut by 18%.

Deaths from alcohol-related causes have increased by 60% from before the pandemic (BMJ 2025;359:r681).

In the USA, mortality due to cirrhosis has been increasing since 2009. People aged 25–34y have had the greatest relative increase in mortality, driven entirely by alcohol-related liver disease (BMJ 2018;362:k2817).

This article was updated in May 2025.

You may also be interested in our separate article, Liver disease: alcohol-related liver disease.

How much alcohol is in that drink?

• One unit = 8g alcohol. That’s not very helpful, is it?
• The number of units in a drink = (volume (ml) x alcohol (% ABV)) / 1000. That doesn’t help much either!

Here’s our easy ‘ready reckoner’: because the % ABV is the number of units in 1 litre, the number of units in 250ml is a quarter of this – that’s roughly equivalent to half a pint. Based on this:

• A large (250ml) glass of wine (ABV 14%) = approximately 3.5 units.
• Half a pint of beer (ABV 4%) – approximately 1 unit.
• A single shot (25ml) of ABV 40% spirits = 1 unit.

The guidelines from the UK Chief Medical Officer based on a review of all evidence published since 1995 were updated in 2016 (DOH August 2016).

The key message is that there is no ‘safe’ level of drinking.

The risk of developing a range of illnesses increases with any amount of alcohol consumed.

Key changes in the 2016 guidelines:

• Guidance is now the same for men and women.
• A 14 units/week limit for regular drinkers was thought to be simpler to understand and more useful than daily limits.
• If drinking 14 units/w, it is best to spread this evenly over 3d or more. Having 1 or 2 heavy drinking sessions increases the risk of death from long-term illness, accidents and injuries.
• A good way to cut down on alcohol is to have several drink-free days each week.

The guidelines should help people make informed decisions about their alcohol consumption to keep risks at what they think is an appropriate level. The guidelines give specific guidance for 3 situations:

• It was previously thought that low levels of alcohol consumption had cardioprotective benefits.
• However, the latest evidence shows that net cardioprotective effects from drinking alcohol are limited to women >55y drinking around 5 units/w, and these effects are offset if they have other risk factors, e.g. being overweight.
• Any cardioprotective effects are cancelled out by heavy drinking sessions (>7.5 units at least monthly).

A BMJ paper quantified the risks:

• Non-drinking (compared with drinking within UK guidance) slightly increased the risk of CVD (by roughly 1.3x).
• Heavy drinking (exceeding UK guidance) increased the risk of most CVD (also by around 1.3x) but reduced the risk of MI and stable angina (BMJ 2017;356:j909).

NICE recommends we should carry out alcohol screening as part of routine practice, identify drinkers at increased risk of harm and provide brief advice (NICE 2010, PH24).

To improve detection, screen groups which are potentially at risk, including those with:

• With related physical conditions, e.g. GI or liver disorders, hypertension.
• With mental health problems, e.g. anxiety or depression.
• With a history of being assaulted.
• Regularly having unintentional injuries or minor trauma.
• Regularly attending GUM clinics.
• With repeat requests for emergency contraception.

Since the publication of the NICE guideline in 2010, there is emerging concern about alcohol use disorder in older adults (BMJ 2025;359:r681). In 2012, people aged >65y made up 12% of ‘heavy drinkers’ (defined as >50 units per week for men and >35 units per week for women). In 2022, this had risen to 21%. Older people are under-represented in alcohol treatment pathways, and are at higher risk of alcohol-related harm due to accidents and other comorbidities. A BMJ analysis called for a national approach targeting drinking in older people, and for primary care clinicians to be aware of higher risks in this age group.

Blood tests such as LFTs are not normally used to screen for alcohol use disorders, but an incidental finding of abnormal LFTs should prompt assessment of alcohol use. Don’t be reassured by normal LFTs in those drinking at increasing levels of risk. You may find the GEMS on Liver disease of interest.

The WHO Alcohol Use Disorder Identification Tool (AUDIT) is considered the most accurate questionnaire for assessing the risk of a person’s alcohol consumption. However, it takes times to complete, so two shorter pre-screening questionnaires (AUDIT-C and FAST) can be used to identify people who may be drinking alcohol at hazardous or harmful levels:

• FAST asks an initial screening question which may then be followed by 3 further questions if needed. Scoring is complicated so I have not included it here.
• AUDIT-C uses the first 3 questions from the full AUDIT tool and is simpler to use:
  • A positive result is score ≥5 (or ≥3 if <18y or >65y).
  • A negative result is score <5 (or <3 if <18y or >65y).

A positive pre-screening questionnaire result should prompt completion of the full AUDIT questionnaire. A negative result should be fed back favourably.

NICE says we can use AUDIT in children as young as 10y (CG115). It says the threshold for referral and intervention “should be lowered” compared with adults, but doesn’t give specifics. The American National Institute on Alcohol Abuse and Alcoholism youth screening tool is an alternative screening option in children (JAMA 2024;331:1215). It can be used in patients aged 9–18y, questions are age specific, and risk is determined by levels of drinking in both the child and their friends. The tool also gives a guide for providing brief intervention (link at end of article).

Specialist addiction services don’t typically accept under-18s, and referral should usually be to CAMHS.

Don’t forget about safeguarding if a child is using alcohol.

This is suggested by an AUDIT score ≥20 but can be present in people with lower scores.

The ICD-10 diagnostic criteria are shown below:

• There needs to be a cluster of behavioural, cognitive and physiological features that develop after repeated use of alcohol, typically including:
  • Strong desire or sense of compulsion to drink alcohol.
  • Difficulties controlling drinking behaviour in terms of starting drinking, stopping drinking or amount consumed.
  • A physiological withdrawal state when drinking stopped or reduced (as evidenced by classical alcohol withdrawal syndrome).
  • Evidence of increased alcohol tolerance.
  • Higher priority given to drinking alcohol than other activities and obligations.
  • Persisting with drinking alcohol despite clear evidence of harmful consequences, e.g. liver damage, depression or impaired cognitive functioning.

Brief advice should be given to people identified as drinking alcohol at increasing risk levels. This should take 10 minutes or less and be accompanied by self-help material. Follow-up may be offered.

Studies show that it does work: 1 in 8 people reduce their alcohol to low-risk levels after brief advice (compared with 1:20 who quit after smoking cessation advice).

eLearning for healthcare has useful modules on delivering brief advice (see link below). A suggested strategy for delivering brief advice:

• Person identified as increasing risk or higher-risk drinker by AUDIT tool.
• Make transitional statement from the AUDIT score into brief advice, e.g. “Your questionnaire results show you are drinking at an increasing risk level; how do you feel about that?”. You may assess their readiness to change their drinking behaviour, e.g. “Rate how important it is for you to change your drinking, with one being not important at all and ten being very important”.
• Offer brief advice. This has 6 essential elements summarised by the acronym FRAMES:

You may find our information on Motivating behaviour change useful.

These use motivational interviewing techniques to explore the pros and cons of change and formulate an action plan. They involve several meetings. This would be appropriate for people still drinking at increasing risk levels after brief advice, or those drinking at harmful levels.

No!

This representative cross-sectional survey of >15 000 people aged ≥16y in England looked at whether GPs were giving brief advice to patients who smoked or drank alcohol excessively (BJGP 2016; DOI: 10.3399/bjgp16X683149).

The results suggest that brief advice for excessive alcohol consumption is rarely being given by GPs, and patients are more likely to receive advice about stopping smoking than cutting down alcohol.

The authors suggest several reasons for lower rates of alcohol advice compared with smoking:

• It’s not in QoF!
• Alcohol screening (AUDIT-C) takes longer than asking about smoking (which is just a yes/no answer).
• In England, motivation to change is lower among drinkers compared with smokers.
• It is more complex than smoking as the goal is often to reduce alcohol consumption rather than stop drinking completely.

A Cochrane review looked at whether brief alcohol interventions work (DTB 2018;56(5):51).

It defined ‘brief intervention’ as ≤5 sessions of brief advice lasting <60minutes in total – different from what we can offer!

The review found that brief interventions:

• Resulted in reduction in alcohol consumption of 2.5 units/w.
• Did not impact frequency of binges, drinking days or drinking intensity.

Extended interventions were no more effective than brief interventions.

A BMJ editorial reviewed the evidence behind smartphone apps and digital tools to reduce alcohol use (BMJ 2023;382:p1665). It cites a 2017 Cochrane review which found moderate-quality evidence that digital interventions reduced alcohol consumption (Cochrane 2017 CD011479).

A more recent study found a consistent and significant decrease in alcohol use in a group of students provided with a smartphone app designed to address unhealthy alcohol use when compared to a control group of their peers (BMJ 2023;382:e073713). In its 2020 guidance on digital and mobile health interventions, NICE recommends that we ‘consider’ digital interventions as an adjunct to usual care, but comments that effectiveness can be variable (NICE 2020, NG183).

All patients who are dependent on alcohol must be advised that they are required by law to notify the DVLA, and will have to surrender their licence. See DVLA – assessing fitness to drive (a guide for medical professionals).

People with harmful use of alcohol may be deficient in thiamine due to poor diet, malabsorption (from gastritis) or increased demands (thiamine is a co-factor in alcohol metabolism).

Thiamine deficiency can lead to Wernicke’s encephalopathy (see box below).

NICE recommends that we offer prophylactic thiamine to harmful or dependent drinkers if:

• They are malnourished or at risk of malnourishment and/or
• They have decompensated liver disease and/or
• Medically-assisted withdrawal is planned.

The dose is 50–100mg a day in mild deficiency, or 200–300mg a day in severe deficiency (see BNF).

People who are malnourished or at risk of malnourishment, or who have decompensated liver disease, should be offered parenteral thiamine if undergoing assisted withdrawal in specialist inpatient alcohol services.

Vitamin B complex preparations (such as vitamin B compound strong tablets) are no longer used (lack of evidence of effectiveness/safety) for the:

• Prevention of deficiency.
• Maintenance treatment following treatment for deficiency.
• Prevention of Wernicke’s encephalopathy.

In rare cases of medically-diagnosed deficiency or chronic malabsorption, they may be used.

(RMOC Oral vitamin B supplementation in alcoholism, 2019)

Suspect alcohol withdrawal in anyone who is alcohol dependent and has stopped/reduced their drinking within hours or days of presentation (BMJ 2023;381:951).

Common symptoms include:

• Anxiety.
• Nausea and vomiting.
• Autonomic dysfunction (sweating, tremor, tachycardia).
• Insomnia.

If untreated, this can progress to severe withdrawal, with seizures or delirium tremens (see box below for a reminder of the features of delirium tremens), both of which are life threatening.

• Offer acute medical admission for urgent benzodiazepine treatment and monitoring if:
  • Seizures or delirium tremens are present, or there is a past history of these.
  • Signs of autonomic dysfunction are present (sweating, tremor, tachycardia).
  • Age <16y.
• Other risk factors for severe withdrawal which should make us consider admission include:
  • Alcohol intake >30 units/d.
  • Fever.
  • High anxiety.
  • Tachycardia.
  • Hypoglycaemia, hypocalcaemia, hypokalaemia.
  • Poor background physical health.
  • Other psychiatric disorders and/or use of psychotropic drugs.
  • Multiple previous medically-assisted withdrawals.
• Admit urgently for parenteral thiamine if there are features of Wernicke’s encephalopathy (reminder of these in the box below).

As well as being high risk in itself, this presentation in primary care is a chance to make a holistic assessment of a vulnerable group who may present to healthcare rarely or chaotically. Things to think about (particularly if you’re not planning admission) include (BMJ 2023;381:951):

• Other substance misuse +/- screening for HIV and hepatitis B/C.
• Why did they stop drinking? Physical symptoms, e.g. abdo pain? Social reasons, e.g. lack of money?
• Is this definitely alcohol withdrawal or could other diagnoses actually be causing their symptoms? Don’t miss:
  • Sepsis.
  • Head injury (keep in mind that this group are more likely to have a significant bleed after a head injury due to deranged clotting and low platelets).
  • Decompensated liver disease.
  • Gastritis.
  • Pancreatitis.
  • Metabolic derangements.
  • Other drug intoxication or withdrawal.

Unnecessary inpatient management of alcohol withdrawal isn’t just a poor use of resources – it also has very poor outcomes in terms of long-term abstinence. Usually, patients not requiring medical admission will need referral to the community addiction service for medically-assisted withdrawal (more on this in the next section). There is a group with mild symptoms and lower baseline drinking levels who may only need monitoring, not drug treatment, but they are still likely to benefit from the psychological support offered by addiction services.

We should advise people awaiting a medically-assisted withdrawal that they must NOT suddenly stop or suddenly significantly reduce their alcohol intake because they may:

• Experience unpleasant withdrawal symptoms.
• Develop seizures or delirium tremens.

A commonly-quoted rule of thumb is to not reduce alcohol intake by more than 25% every 2 weeks (BMJ 2023;381:951).

Emergency syndromes of alcohol dependence and tolerance:

NICE recommends that we offer assessment for assisted withdrawal for patients who:

• Typically drink >15 units/d or
• Score ≥20 on the AUDIT score.

This may be offered in the community (for those with mild to moderate dependence) or as an inpatient, but it is a specialist service. It is not just a case of providing chlordiazepoxide/diazepam; appropriate psychological support, usually including motivational interviewing, group therapy and family/carer support, is required.

This means, in general, unless we are specialists in drug and alcohol dependence, we will not be offering medically-assisted withdrawal in primary care. If we are asked to prescribe the medication under a shared care protocol, this will usually be instalment prescribing of no more than 2 days at a time.

Alcoholics Anonymous and UK SMART Recovery are alcohol dependency mutual aid groups that provide a source of ongoing support for patients, family and friends. Long-term cohort studies suggest that these are effective.

NICE recommends offering those with harmful drinking or alcohol dependence an ‘alcohol-focused psychological intervention’ (usually via addiction services); this may include CBT, behavioural therapy, social-network-based therapy or behavioural couples therapy (NICE 2011, CG115).

Is my patient likely to be able to access medication to help prevent relapse?

If they want it, the answer should be yes! Either acamprosate or oral naltrexone can be used first line. According to NICE, these drugs should be considered alongside psychological interventions in:

• Anyone recovering from harmful or dependent drinking who specifically requests drug therapy, as well as...
• Those who have severe alcohol dependence.
• Those who have mild to moderate alcohol dependence which hasn’t responded to psychological intervention alone.

Are these medications effective?

A 2023 JAMA meta-analysis found that the NNT to prevent return to any drinking was 11 for acamprosate and 18 for oral naltrexone (JAMA 2023;330:1653). The evidence for the second-line drugs is weaker.

Here’s a closer look at these drugs. Check your local formulary; it’s likely that most of these will be specialist initiation only, and remember that they should only be used alongside structured psychological support:

Carbohydrate deficient transferrin is a test that can be used in alcohol use disorders where abstinence is the key (so it may be used by the DVLA to look for hazardous drinking in those reapplying for a driving licence after losing it for an alcohol-related offence) (DTB 2017;55(6):69).

• It is more specific than other tests such as GGT, and is less affected by liver disease, but various things can affect the test (obesity, smoking, pregnancy, sepsis, anorexia, iron overload, primary biliary cirrhosis, and those who have a genetic variant in their transferrin molecule).
• It should not be used in isolation and is not performed by all laboratories, so not one for primary care!
• It correlates well with individuals’ drinking pattern in the preceding 30d, showing a response within 1–2w of heavy drinking.

In 2018, Scotland introduced ‘minimum unit pricing’ on alcohol sales (50p per unit).

Following the introduction of this policy (Lancet 2023;401:1361):

• Alcohol sales fell by 3%.
• Deaths wholly attributable to alcohol fell by 13.4%.
• Hospitalisation wholly attributable to alcohol fell by 4.1%.

The greatest reductions were seen in the lower socioeconomic groups, suggesting that the policy is effectively addressing deprivation-based inequalities in harms of alcohol.

Consider referral for those who:

• Failed to benefit from brief advice or extended brief intervention, and want further help.
• Show signs of moderate to severe alcohol dependency.
• Have severe alcohol-related physical impairment or related comorbid conditions, e.g. liver disease or mental health problems. See our article on Liver disease: alcohol-related liver disease for more information.

	Alcohol Guidelines from the UK Chief Medical Officer to keep risks from alcohol to a low level recommend that regular alcohol consumption is limited to 14 units/w spread over 3 or more days. The limit is the same for men and women. About one-quarter of the UK adult population drinks alcohol at hazardous or harmful levels. FAST or AUDIT-C questionnaires can be used as an initial screen for alcohol use disorders. An AUDIT-C score ≥5 should prompt completion of the full AUDIT questionnaire to assess the risk of the person’s alcohol consumption. Alcohol consumption may be classified as low, increasing or high risk. In those drinking at higher-risk levels, it is important to assess for alcohol dependency. Brief advice should be given to those identified as drinking at increasing or high-risk levels. People with features of moderate or severe alcohol dependency, or those with severe alcohol-related problems, should be referred to specialist services. Offer prophylactic thiamine where there is risk of malnourishment, decompensated liver disease or a plan for medically-assisted withdrawal. Acute alcohol withdrawal can be life threatening. We should advise patients with harmful or dependent drinking not to stop suddenly, and, where we suspect acute withdrawal, assess for features requiring admission. Everyone with harmful or dependent drinking should be told about mutual aid groups and be offered alcohol-focused psychological support, e.g. CBT. Acamprosate or naltrexone (accessed via addiction services) can be considered alongside psychological therapies in most people recovering from harmful or dependent drinking. Disulfiram and nalmefene are second line, with weaker evidence for efficacy.
	Do you known what mutual aid groups (e.g. Alcoholics Anonymous) are available locally and how to access these? How do you assess alcohol use in patients newly registering at your practice, and who acts on the results?
	Useful resources: Websites (all resources are hyperlinked for ease of use in Red Whale Knowledge) Elearning for healthcare - alcohol identification and brief advice DVLA – assessing fitness to drive (a guide for health professionals) The National Institute on Alcohol Abuse and Alcoholism - youth screening tool (for children aged 9–18y) For patients: Drink Aware (information for patients on alcohol, units, tracking alcohol consumption and cutting down) Alcoholics Anonymous Smart Recovery