Liver Health

What a Month Without Alcohol Changes in Your Bloodwork

A BMJ Open study followed 94 adults through one alcohol-free month. Insulin resistance, blood pressure, and gamma-glutamyl transferase all fell. What those numbers can and cannot tell you.

Published

September 29, 2026

Read time

12 min read

Medically reviewed by

Rhonda Collins, FNP-C

Share

Copy link

A month without alcohol changes measurable things in blood, and the clearest evidence for which ones comes from a 2018 study in BMJ Open. Among 94 adults who stopped drinking for one month, insulin resistance fell by a median of 25.9%, systolic blood pressure by 6.6%, and gamma-glutamyl transferase, the liver enzyme that tracks alcohol exposure most closely, by 28.6%. A comparison group of 47 adults who kept drinking moved on none of those measures.

Most people judge a dry month by sleep, skin, and mood. Those are real, and they are also the parts nobody can put a number on. Bloodwork can be counted. Sober October, a 31-day alcohol-free challenge run each October as a fundraiser for Alcohol Change UK, asks for exactly the commitment that study measured: one month, and then a decision about what happens next.

What one month actually changed

One alcohol-free month improved insulin resistance, blood pressure, weight, two growth factors, and liver enzymes in the people who stopped, and changed almost nothing in the people who did not.

This was a single-center, prospective, observational study at the Royal Free London NHS Foundation Trust, published in BMJ Open in 2018. The abstinence group was 94 adults (mean age 45.5 years, SD 1.2) and the control group 47 adults (mean age 48.7 years, SD 1.8). To enter, participants had to be drinking above 64 g of alcohol per week for men or above 48 g for women. Anyone with known liver disease or alcohol dependence was excluded, as was anyone who had already been abstinent for more than three days.

Baseline intake in the abstinence group averaged 258.2 g of alcohol per week (SD 9.4). At the NIAAA standard drink of 14 g of pure alcohol, that is roughly 18 drinks a week. The authors describe the cohort as moderate-heavy drinkers, and that is the right label to keep in mind: this was not a study of people having a glass of wine on Fridays.

The primary outcome was insulin resistance, measured as a homeostatic model assessment (HOMA) score. It fell by a median of 25.9% (IQR -48.6% to +0.3%, p<0.001). The declared secondary outcomes all moved in the same direction, each at p<0.001:

• Systolic blood pressure: -6.6% (IQR -11.8% to 0.0%)
• Diastolic blood pressure: -6.3% (IQR -14.1% to +1.3%)
• Weight: -1.5% (IQR -2.9% to -0.4%)
• Vascular endothelial growth factor (VEGF): -41.8% (IQR -64.9% to -17.9%)
• Epidermal growth factor (EGF): -73.9% (IQR -86.1% to -36.4%)

Liver enzymes improved too. Gamma-glutamyl transferase fell 28.6% (IQR -43.5% to -14.4%, p<0.001) and alanine aminotransferase (ALT) fell 14.5% (IQR -28.9% to +6.7%, p<0.001). Aspartate aminotransferase (AST) fell 5.4% (IQR -16.2% to +9.5%) at p=0.03, which the authors report as a trend rather than a significant reduction, because the analysis took p<0.01 as its threshold.

In the control group, none of the primary or secondary outcomes changed significantly. The one exception was a small rise in AST of 4.5% (IQR -5.6% to +23.1%, p<0.01).

Two details in those numbers are easy to lose. First, look at the upper edge of the HOMA range: +0.3%. The interquartile range is the middle half of the individual results, so roughly a quarter of the abstinence group finished the month with insulin resistance no better than where they started, or very slightly worse. A median is not a promise.

Second, the percentages above are medians of each person’s own percent change, which is a different quantity from the change in the group median. For insulin resistance, the group median went from 1.4 to 1.0, a drop of about 29%, while the median individual change was 25.9%. Both figures are correct and they are not interchangeable.

What the study does not show

The design supports a claim about what changed over one month in a self-selected group. It does not support a claim about cause, durability, or cancer, and it does not describe anyone who is dependent on alcohol.

Participants were not randomized. They were allocated by their own intention, either to stay alcohol-free for a month or to carry on drinking. That is the central limitation the authors name, and they explain that assigning people to a drinking regimen was not considered ethically appropriate.

The cohort was narrow. Recruitment ran through email advertising at University College London, Queen Mary University of London, and New Scientist Magazine. The authors note this group probably had higher educational attainment and health-related motivation than the general population.

Alcohol dependence was screened out at entry. Anyone with known liver disease or alcohol dependence was excluded, so none of these numbers describe a dry month for someone who is dependent. That exclusion is a safety line as well as a statistical one. NIAAA warns that when someone who has been drinking heavily for a prolonged period suddenly stops, the body can go into a painful or even potentially life-threatening withdrawal, with symptoms that can include nausea, rapid heart rate, and seizures, and it advises seeking medical help to plan a safe recovery. If that could be you, plan the month with your doctor rather than starting it alone.

Lifestyle is not the loose end it looks like. The primary and secondary outcomes were adjusted for changes in diet, exercise, and cigarette smoking, and the reported changes were not associated with them. Exercise and smoking scores did not move and diet changed slightly. The honest limitation here is self-selection and the absence of randomization, not a suspicion that the abstinence group quietly fixed everything else too.

VEGF and EGF are laboratory measures, not a risk result. The researchers followed these two growth factors because other work reports a link between them and cancer biology. A fall in either one is a change in a blood concentration, not a measured change in anyone’s cancer risk. This study did not measure cancer outcomes.

One month says nothing about the following year. The authors address this directly, warning against reading their results as evidence that a short detox period is all that is needed to refresh the liver. The durability of the changes they measured remains to be established.

The marker that tracks drinking most directly

If you measure one thing, measure gamma-glutamyl transferase. It fell further than either ALT or AST in the study above, and it is the liver-associated enzyme most responsive to alcohol.

The reason is where the enzyme sits. Gamma-glutamyl transferase is concentrated on the outer membranes of liver and bile-duct cells and is central to glutathione metabolism, so stress on those cells releases it into the blood. Our gamma-glutamyl transferase biomarker page has the full measurement detail, including the reference ranges most laboratories use: 8 to 61 U/L for men and 5 to 36 U/L for women. The enzyme has a half-life of roughly 14 to 26 days, and after complete cessation in alcohol-related elevation, levels typically normalize within four to eight weeks.

Here is where the one-month number gets interesting. The usual rule of thumb is that gamma-glutamyl transferase falls by about half every two to four weeks once drinking stops. The study measured a median fall of 28.6% over a month, noticeably less than half.

Both can be true, because the rule of thumb is drawn largely from people who started well above the reference range, often with dependence, where there is a large elevation to clear. This cohort did not start there. In the table of paired values, the abstinence group’s gamma-glutamyl transferase (n=92) had a baseline median of 22.5 IU/L (IQR 14.0 to 34.8) and a one-month median of 15.0 IU/L (IQR 10.3 to 24.0), p<0.001. That is inside the normal range for both men and women. Less elevation means less to fall, and a percentage drop from an already normal value looks smaller than the same biology would look in someone starting at 150.

That is worth stating plainly: drinking at a level a clinician would flag does not mean your liver enzymes are abnormal today. In this cohort, the typical result before the dry month was a normal one.

One more reconciliation, because Mito publishes a longer figure elsewhere. Our article on how alcohol ages the body places significant liver enzyme improvement at six months to a year, scoped specifically to fatty liver. That is a different question. How quickly an enzyme clears once the exposure stops, and how long it takes structural fatty change in the liver to resolve enough to show up in enzymes, are separate timelines. Neither figure corrects the other.

Finally, a single gamma-glutamyl transferase result is nonspecific. Fatty liver disease, enzyme-inducing medications, metabolic syndrome, and bile duct disease all raise it, and it cannot prove that anyone drank or stopped drinking. Alcohol is one explanation for a high result, not the explanation.

Why the baseline has to come first

A month covers only about one to two half-lives of gamma-glutamyl transferase, so a single result drawn at the end of a dry month shows a partial fall that carries no meaning on its own.

Work through it with a real number. Someone finishes their alcohol-free month, tests, and gets 15 IU/L. That is a normal result. It is also exactly what the study’s median participant looked like after falling from 22.5, and exactly what someone who has always sat at 15 looks like. The two are indistinguishable from one draw. The information is in the pair, not in either value.

So if a dry month is the experiment, the blood draw belongs before it starts, not after it ends. Mito sells gamma-glutamyl transferase as a standalone test, processed at the same CLIA and CAP-accredited national labs your doctor uses, with typical results in 1 day. Availability depends on your state.

On preparation, requirements differ by laboratory, so follow the instructions attached to your order rather than a general rule. Whether to avoid alcohol before the draw is a separate question from the dry month itself, and our guide to alcohol before a blood test covers it. Use the same laboratory and comparable preparation for both draws, fasting status included, because gamma-glutamyl transferase falls after a meal. Even then, a difference between two draws shows that something changed, not what changed it.

What to do with the number

Read a gamma-glutamyl transferase result next to the rest of the liver picture, not alone. The pairing with alkaline phosphatase is the most useful one: elevated together points toward a liver or bile-duct source, while a normal gamma-glutamyl transferase alongside a high alkaline phosphatase makes a non-liver source such as bone more likely.

If a result comes back elevated, the useful next step is inventory rather than alarm. Write down your recent drinking, every prescription and supplement you take, and your weight trend, because medications and metabolic factors sit alongside alcohol as ordinary causes. If you changed something modifiable, retesting after eight to twelve weeks is a reasonable interval to see whether the trend moved.

Two things belong with your doctor rather than in a self-experiment: an elevation that persists after you have addressed the obvious causes, and any result accompanied by symptoms such as jaundice, dark urine, pale stools, itching, or discomfort under the right ribs. Inflammation is a separate question: our article on whether alcohol causes inflammation covers hs-CRP and what the liver enzymes add.

Frequently asked questions

Can a GGT test prove that someone stopped drinking? No. Alcohol can raise gamma-glutamyl transferase, but the result is nonspecific and individual responses vary. It cannot establish that a person drank, or that a person abstained, and it should never be used alone for that purpose.

Which liver marker responds fastest to stopping alcohol? In the BMJ Open study, gamma-glutamyl transferase moved most over one month, falling a median of 28.6%. ALT fell 14.5%, and AST fell 5.4%, which the authors reported as a trend rather than a significant change.

Do you need to fast before a GGT test? It depends on the laboratory running the assay. Fasting requirements differ between providers, so follow the preparation instructions attached to your order rather than a general rule.

Does a GGT result inside the reference range still tell you anything? It can, as a trend. In population data, levels above roughly 25 to 30 U/L correlate with higher long-term rates of metabolic syndrome and cardiovascular events, even though those values sit within many laboratories’ normal windows. A single normal result is reassuring context, not a finished answer.

Does it matter whether the dry month is October or January? Not biologically. The study measured one month of abstinence, not a particular calendar month. What changes the usefulness of the experiment is whether you have a blood result from before it started.

Does a month without alcohol undo liver damage? That is not what the study measured, and the authors specifically cautioned against reading it that way. They describe the idea that a short detox period refreshes the liver as untrue, and state that the durability of the changes they observed remains to be established.

Related reading

• Gamma-glutamyl transferase biomarker reference
• Does alcohol cause inflammation? What blood tests can show
• Alcohol before a blood test: when to avoid it and what to do
• Beyond the buzz: how alcohol ages you faster and what to do about it
• Alkaline phosphatase biomarker reference

Sources

• Mehta G, Macdonald S, Cronberg A, et al. Short-term abstinence from alcohol and changes in cardiovascular risk factors, liver function tests and cancer-related growth factors: a prospective observational study. BMJ Open 2018;8(5):e020673
• BMJ Open record for the same study (doi:10.1136/bmjopen-2017-020673)
• NIAAA: what is a standard drink?
• Sober October: the 31-day alcohol-free challenge for Alcohol Change UK
• NIAAA Rethinking Drinking: to cut down or to quit
• MedlinePlus: gamma-glutamyl transferase (GGT) test
• Whitfield JB. Gamma glutamyl transferase. Critical Reviews in Clinical Laboratory Sciences, 2001

Medical disclaimer

This article is for informational purposes only. It does not provide medical advice, diagnosis, or treatment, and it is not a substitute for care tailored to your health history and symptoms. Gamma-glutamyl transferase is not a diagnostic test: a result does not identify a liver condition, establish a cause, or demonstrate alcohol use or abstinence. Any result should be interpreted by a qualified clinician in the context of your history, examination, medications, and other testing.