Moderate drinking: Katie Witkiewitz on cutting down vs quitting

Moderate drinking means keeping to a limit you choose and can hold, at or below the lower-risk line your country sets. You can choose it instead of quitting, and research on people who cut down supports that choice. But "moderate" doesn't mean healthy or safe: less alcohol means less risk, never none. Stopping suits better when cutting down keeps failing or your health rules alcohol out.

Witkiewitz came to this question by telephone. As a graduate student at the University of Montana, her job was to call people discharged from a residential alcohol treatment center, and their partners and relatives, and record how much they had been drinking. The answers didn't fit a simple picture. Some people who weren't drinking "were really struggling", while some who had been drinking "a little" were "doing really well" (The Alcohol 'Problem' Podcast, 4:08 to 5:08).

Today she is a Distinguished Professor of Psychology with over 300 published papers, many on how people change their drinking. She also belongs to the Alcohol Clinical Trials Initiative (ACTIVE), a research group funded by pharmaceutical companies, and has reported personal fees from Eli Lilly outside that work; two of the studies below come from that group, and both declare it (Witkiewitz et al., 2025).

Her approach comes in three parts: find where you are, decide where you want to be, and know when "less" is the wrong goal.

Decide what "moderate" means for you

1. Ask whether it's serving you

Witkiewitz starts with a question, before any number: "what's your relationship with alcohol? Is it serving you in a way in your life?" (9:28). She sees drinking as a range with not drinking at one end: "I really like to think of abstinence and the choice to not drink as a spectrum" (8:16). Choosing not to drink is a valid decision, she says, and so is drinking less. The question needs no label: you don't have to decide whether you're "a problem drinker" to decide that Thursday's bottle isn't doing what you want.

2. Find your rung on the risk ladder

The next question her patients ask is "how much?" Her answer is a ladder the World Health Organization uses in research, which she has studied for years. "Humans like to think categorically," she says, and the ladder has four rungs, each with an amount of alcohol (18:57). She shows people the ladder and asks: "what level of risk are you currently drinking at and what level of risk do you want?"

Her 2025 review prints the rungs in grams of pure alcohol a day, averaged across the week (Witkiewitz et al., 2025):

• Men: low risk 1 to 40 g, medium 41 to 60, high 61 to 100, very high 101 or more.
• Women: low risk 1 to 20 g, medium 21 to 40, high 41 to 60, very high 61 or more.

A UK unit is 8 g (NHS), so 40 g is five units. Read the ladder as steps, and don't take its bottom rung as a target: the WHO's "low" band runs up to 280 g a week for men and 140 g for women, well above the UK's lower-risk guideline of 14 units a week (112 g), spread over three or more days (UK Chief Medical Officers, 2016). Our guide to the standard drink shows how to count what's really in your glass.

Here's what that looks like in practice. Say a woman drinks nothing on weekdays, then a bottle of 12% wine on Friday and another on Saturday, about 70 g each. Averaged over seven days, that's 20 g a day, the top of the low band. Counted per drinking day, both evenings sit on the very-high rung. A weekly average hides heavy evenings, so look at both numbers.

Each step down counts. In a US survey that followed 22,005 drinkers from 2001 to 2005, 37% of very-high-risk drinkers whose drinking didn't change met criteria for alcohol dependence three years later. Among those who had moved down one level, it was 14%, and lower still for two or three levels (Hasin et al., 2017). It's a link in survey data that can't show which came first, and Witkiewitz co-wrote the study.

3. Put a number on it

"Less" on its own is hard to keep. In a review of 22 treatment studies, the authors found that people more often reached low-risk drinking when their goal sat within defined limits than when they set their own vague reduction, though the comparison was indirect (Henssler et al., 2021). Pick a weekly number, a number for each drinking day and your drink-free days; the UK guidance suggests several drink-free days a week. Our guide on how to drink less covers the plan around the number.

Any step down the ladder counts, and the bottom of the ladder is not drinking at all.

Is moderate drinking healthy?

4. Drop the glass-of-red idea

Witkiewitz accepts the data "showing that there is increase in all-cause mortality and certainly increased cancer risk", and says "a glass of wine a day is not necessarily good for you" (13:41 to 14:47). She adds that researchers still argue about the heart.

For decades, studies found that light drinkers had less heart disease than non-drinkers. Tim Stockwell, a psychologist at the University of Victoria, thought otherwise: "This is so fundamentally flawed" (Science Vs, 11:03). With the sociologist Kaye Fillmore he looked at who the non-drinkers were. Many had stopped because they were already ill, the "sick quitters", and they made moderate drinkers look healthy by comparison. Once the team kept only studies without that problem, in Stockwell's words, "the significance fell away" (14:46).

A 2023 review pooled 107 cohort studies with more than 4.8 million people. After adjusting for these biases, drinking up to 25 g a day showed no significant reduction in the risk of dying early compared with never drinking, and the risk rose at a lower daily amount for women than for men (Zhao et al., 2023).

Genetics gives a second test: gene variants common in East Asia make alcohol unpleasant, so their carriers drink less for reasons unrelated to the rest of their health. George Davey Smith, an epidemiologist at the University of Bristol, says studies built on this find "no beneficial effect on coronary heart disease of alcohol consumption, and a substantial adverse effect on overall cardiovascular disease" (Freakonomics, M.D.). In one, which followed 512,715 adults in China, the apparent protection against stroke was "largely non-causal" (Millwood et al., 2019).

Keith Humphreys, a professor of psychiatry at Stanford, put it plainly on Huberman Lab: "I like red wine, I would love to believe it is healthy. It's not" (Huberman Lab, 24:45). If you drink moderately, do it because you enjoy it, never as a health habit.

When stopping is the better goal

5. Know who should stop

Witkiewitz doesn't think the data can say in advance who will do better with less and who with none: "I don't think we know from the data who should be recommended towards abstinence, who should be recommended toward reduced drinking" (27:25). She means people in treatment, where the clearest evidence is a review of 22 studies with 4,204 patients in treatment for alcohol use disorder or problem drinking (Henssler et al., 2021). People who aimed to drink within limits reached low-risk drinking about as often as people who aimed to stop, when their support fitted their goal, and 58% in each group cut down substantially. People who aimed to stop more often ended up not drinking, 21.2% against 9.7%. The authors' verdict: "Available evidence does not support abstinence as the only approach in the treatment of alcohol use disorder."

For your own choice, the US National Institute on Alcohol Abuse and Alcoholism advises quitting if you've tried cutting down but can't stay within the limits you set (NIAAA). Stopping is also the goal in these cases:

• Pregnancy, or trying to conceive: stop for the period, and talk to your midwife or doctor.
• Liver disease diagnosed by a doctor: stop, guided by your doctor.
• Some medicines don't mix with alcohol: ask your pharmacist or doctor about yours.
• Driving: when you're driving, the right number is none.

In one case a doctor should set the pace before you change anything. If your body may have got used to alcohol, stopping or cutting down quickly can make you unwell, sometimes seriously. Talk to a doctor before you change how much you drink. A doctor can help you cut down or stop safely, usually without a hospital stay. Don't stop suddenly on your own.

If you choose a break or a stop, read this once before you start. If, after stopping or cutting down, you get shaky, sweaty, sick or very anxious in the hours after your last drink, talk to a doctor before going further. If you, or someone with you, notice a fit, confusion (not knowing where you are or what day it is), or seeing or hearing things that aren't there, get medical help straight away. Don't wait.

6. Treat a change of goal as part of the plan

A goal isn't a vow. Many people in those studies switched; in one, with people who had long drinking histories, only 51% of those who started with a drink-less goal kept it (Henssler et al., 2021). If "less" keeps breaking after the first drink, a dry month or a stop can be easier to hold than a number, and our guide to alcohol cravings covers the moment the number usually gives way. Witkiewitz is clear about that direction too: "for many people, that's abstinence. And that's lovely and wonderful and beautiful too" (43:21 to 44:23).

Build a life that gets in the way

Her closing advice works for either goal: "figuring out ways to build a life where the life gets in the way of ... alcohol use" (46:27). Her own review of US surveys found that drinking tends to fall from the mid-20s as life fills up, and that more than 70% of resolutions of drinking problems happened outside treatment (Tucker, Chandler & Witkiewitz, 2020). If you do one thing this week, find your rung, then choose the one you want. Our guides to drinking every night, being sober curious and alcohol-free drinks help with the evenings themselves.

To work out your goal with your own week in front of you, try the 1h Protocol to Reduce Your Drinking. The first chapter is free: it asks about your goal, your usual week and your real drink, checks how your body reacts when you drink less, then finds your pattern and gives you one first step. The full program is practical education and takes about an hour. You choose to drink less, take a break or stop, set your number with your country's guidance beside it, and leave with a page holding your goal and starting changes, a card for your lock screen and a fourteen-day sheet, with days 7 and 14 in your calendar. Short check-ins by email follow on days 1, 7, 14 and 30; the day-7 review asks the body questions again, and you can change your goal at each review. Start free

Frequently asked questions

What counts as moderate drinking?

There's no single definition. Each country sets its own lower-risk line; in the UK it's no more than 14 units a week on a regular basis, spread over three or more days (UK Chief Medical Officers, 2016). Use your country's line, and count every rung down as progress.

Is moderate drinking healthy?

No. It carries less risk than heavier drinking, but no amount is safe. After correcting for older studies' flaws, a review of 107 studies found that people who drank a little didn't live longer than people who never drank (Zhao et al., 2023).

Can I cut down instead of quitting?

Yes, many people do. In treatment studies, people who aimed to drink within clear limits reached low-risk drinking about as often as those who aimed to stop, when their support fitted their goal (Henssler et al., 2021). Put a number on it, and if your body may have got used to alcohol, talk to a doctor before you change how much you drink.

When is stopping the better goal?

When you've tried cutting down and can't stay within your limits (NIAAA), when you're pregnant or trying to conceive, and when a doctor has told you that you have liver disease. If your body may have got used to alcohol, a doctor should set the pace. If a plan isn't enough, talk to a doctor or a local alcohol service.

Is a glass of red wine good for your heart?

The evidence says no. The heart benefit in older studies came largely from comparing drinkers with people who had stopped because they were ill, and genetic studies found "no beneficial effect on coronary heart disease" (Freakonomics, M.D.). Colin Angus, a senior research fellow in the Sheffield Alcohol Research Group at the University of Sheffield, put the practical point simply: nothing in the evidence suggests that a non-drinker should "take up drinking because it'll improve your health" (The Alcohol 'Problem' Podcast, 1:01:52).

Sources

This article draws on Katie Witkiewitz's interview on The Alcohol 'Problem' Podcast and her published research, interviews with Tim Stockwell (Science Vs), George Davey Smith (Freakonomics, M.D.), Keith Humphreys (Huberman Lab) and Colin Angus (The Alcohol 'Problem' Podcast), and the studies and guidance listed below.

• Non-abstinent recovery: examining drinking reductions for public health, with Prof Katie Witkiewitz (podcast episode, The Alcohol 'Problem' Podcast, 2024)
• Witkiewitz et al, Reductions in World Health Organization risk drinking levels as a primary efficacy endpoint for alcohol clinical trials: a review (narrative review, JAMA Psychiatry, 2025; authors in the pharma-supported ACTIVE group)
• Hasin et al, Change in non-abstinent WHO drinking risk levels and alcohol dependence (3-year follow-up of 22,005 US drinkers, The Lancet Psychiatry, 2017; authors in the ACTIVE group)
• Henssler et al, Controlled drinking: non-abstinent versus abstinent treatment goals in alcohol use disorder (systematic review and meta-analysis, 22 studies, 4,204 patients, Addiction, 2021)
• Tucker, Chandler & Witkiewitz, Epidemiology of recovery from alcohol use disorder (narrative review of US population surveys, 2020)
• NIAAA Rethinking Drinking, Should you cut down or quit? (public information page)
• UK Chief Medical Officers, Low risk drinking guidelines (government guideline, 2016)
• NHS, Alcohol units (health information page, 2024)
• Alcohol: time to put a cork in it? (podcast episode with Tim Stockwell, Science Vs, 2023)
• Zhao et al, Association between daily alcohol intake and risk of all-cause mortality (systematic review and meta-analysis, 107 cohort studies, JAMA Network Open, 2023)
• How do we know if alcohol is bad for us? (podcast episode with George Davey Smith, Freakonomics, M.D., 2022)
• Millwood et al, Conventional and genetic evidence on alcohol and vascular disease aetiology (prospective cohort of 512,715 adults in China with genetic analysis, The Lancet, 2019)
• Alcohol risk, guidelines and messaging, with Tom Chivers and Colin Angus (podcast episode, The Alcohol 'Problem' Podcast, 2021)
• How to overcome addiction to substances or behaviors, with Dr Keith Humphreys (podcast episode, Huberman Lab, 2026)