What Quitting or Cutting Does, Measured

The case against drinking rests on what it does at higher doses. This page runs the experiment in reverse: when the dose drops, which numbers move, by how much, and on what timeline. The answer is uneven: blood pressure and liver fat respond fast, the glucose curve barely notices, and weight moves slower than promised.

Three Instruments, One Question

Asking what quitting does runs straight into the association-versus-causation trap that undid the J-curve: people who quit differ from people who don't. So this page leans on three instruments, kept separate: randomized trials of reduction (cleanest but short); Mendelian randomization, comparing people whose genes nudge them to drink less — the method the parent topic, Alcohol & Metabolic Health: The Honest Read, introduced; and biomarker studies showing what moves and when.

Blood Pressure: The Fastest Measurable Win

Cutting delivers its least arguable benefit here, established the old-fashioned way: randomized trials. Xin and colleagues pooled 15 trials in which reduced drinking separated the groups — 2,234 participants — and found systolic pressure fell 3.31 mmHg and diastolic 2.04 mmHg on average, with a dose-response pattern (Hypertension, 2001). The 2017 update, pooling 36 trials and 2,865 participants, sharpened who benefits: at two or fewer drinks a day, reduction barely moved pressure; above that, it fell dose-dependently, topping out at 5.50 mmHg systolic among people drinking six or more a day who halved their intake (Lancet Public Health, 2017).

• 🩸 The reduction is trial-derived, not observational: effects were larger in people starting higher, and they showed up over weeks of sustained reduction.
• 🧬 The genetics agree: East Asian men carrying the heavy-drinking ALDH2 genotype carried 7.44 mmHg more systolic pressure than the rarely-drinking genotype (PLoS Medicine, 2008); in the UK Biobank, one more drink a day tracked with 2.65 mmHg (PLOS ONE, 2021) — though that estimate did not replicate in the external cohorts the same paper tested.

Sizing that honestly: three millimeters of mercury is meaningful at a population level — comparable to a modest sodium cut — but not a medication substitute for established hypertension. The non-drug levers, ranked, live in the blood-pressure topic's lever-ranking page.

Weight and Body Fat: What the Genetic Dice Say

The weight story is less crisp, and the strongest new evidence is genetic. Lu and colleagues ran Mendelian randomization in 408,540 UK Biobank participants: below about seven drinks a week, genetically predicted intake showed no association with body size. Above 14 drinks a week, each additional weekly drink predicted 0.36 kg more fat mass, an 8-percent higher odds of obesity, and a 10-percent higher odds of type 2 diabetes (Journal of Clinical Endocrinology & Metabolism, 2023). The observational J-shape — moderate drinkers looking slimmer — did not survive it.

Two honest footnotes. An earlier UK Biobank analysis found the BMI signal fragile across methods (PLOS ONE, 2021), so treat the fat-mass numbers as a converging picture, not a settled constant. And the arithmetic is modest but real: ethanol carries about seven calories per gram, roughly 100 per standard drink, before mixers and late-night food. Drop two nightly drinks and the deficit runs near 6,000 calories a month — under two pounds of fat, and only if nothing else fills the gap. The mechanism page explains why the body stops burning fat while alcohol is on board; cutting lifts that blockade.

The Glucose Curve: What CGM Shows, and What It Doesn't

Here is the surprise for anyone expecting quitting to transform their continuous glucose monitor chart. In the cross-sectional AEGIS study, 581 Spanish adults wore CGM for up to six days; among those without diabetes, heavy drinkers — above 140 grams a week — had lower day-to-day glucose variability than abstainers, not higher, and light drinkers were indistinguishable from abstainers (Journal of Diabetes Science and Technology, 2017). Alcohol suppresses the liver's glucose output and blunts counter-regulatory hormones, flattening curves — sometimes into delayed hypoglycemia, a real risk in type 1 diabetes after evening drinking (Diabetes Care, 2001).

• 📊 CGM is not where alcohol's damage lands: pooled short-term interventions even found moderate intake slightly lowered fasting insulin (Diabetes Care, 2015). Alcohol's bill arrives via blood pressure, liver, cancer risk, and sleep — the REM-suppressing nightcap — not a jagged glucose line.
• 📉 What quitting does fix: morning-after dips and beer-carb spikes fade within days — which matters on glucose-lowering medication, where insulin and sulfonylureas make alcohol's delayed lows hazardous.
• 🔬 The honest read: if you quit and your CGM curve looks identical, the monitor isn't broken and neither are you. The wins sit in the cuff reading and the liver.

The Liver: The Earliest Structural Win

The liver produces quitting's most dramatic anatomy. Roughly 90 percent of people drinking more than 60 grams a day — four to five US drinks — accumulate fat in liver cells, and simple alcohol-related steatosis is usually fully reversible with abstinence within about 4–6 weeks (Hepatology, 2010). That is AASLD guidance: the fat is a dynamic pool, drained when the ethanol inflow stops and the fat-oxidation blockade lifts.

• 🧪 Enzymes move first: GGT and the transaminases drift down over the following weeks; GGT is slowest — a half-life of roughly two to four weeks — and its poor specificity means one value proves little.
• ⚠️ The stage matters: the 4–6 week reversal applies to simple fatty change. Once fibrosis sets in, abstinence still improves survival at every stage, but reversibility is partial — 5–15 percent of patients with alcohol-related liver disease still progress to fibrosis or cirrhosis despite abstinence.
• 🧭 Where the threshold sits: grams per day that start the fat accumulating are covered in the dose-thresholds page; the broader toolkit in the fatty-liver topic's what actually reverses it.

What Changes on What Timeline

Watch-Items and Honest Limits

• 🚨 Withdrawal risk trumps everything: tremor, sweating, anxiety, or seizure history on stopping means supervised tapering — the callout above is this page's most important sentence.
• 🩺 Elevated enzymes route to a clinician: GGT or ALT above range warrants evaluation before self-interpretation — the fatty-liver topic's ALT problem shows why "normal" can mislead.
• ⚖️ Association still haunts the edges: trial data cover heavy reducers, not casual drinkers; for the light end, genetics is the evidence, and it prices lifetime exposure, not your next month.
• 🎗️ Cancer risk falls slowly, not to zero: stopping is linked to lower oral-cavity and oesophageal cancer risk, and it can take years for risk to approach never-drinker levels (NCI).
• 🧠 Benefits beyond metabolism arrive too: sleep architecture recovers over weeks to months, and often worsens first — rebound insomnia outlasts the first nights — plus the brain risks covered in Alcohol and the Brain.

Questions, Answered Briefly

• 🗓️ "How fast will I see anything?" Blood pressure: several weeks. Liver enzymes: two to six. Weight: one to three months, only if the removed calories aren't replaced. Sleep: often worse for a week or two, then better.
• 🍷 "Does cutting halfway help?" The trials were reductions, not abstinence, and still lowered pressure dose-dependently above two drinks a day. For the liver, abstinence is the standard; for pressure and calories, less is better.
• 📈 "My CGM looks unchanged — did quitting fail?" No. People without diabetes rarely show glucose-curve changes; the measurable effects sit in pressure, liver fat, and sleep, which a sensor cannot see. Gene assignment happens at conception, so those studies carry no sick-quitter bias — and this page flags where they and the trials diverge.

Related Topics