GLP-1 Maintenance: Step-Down Dose Protocol — The Evidence

Two published withdrawal trials — STEP-4 for semaglutide (Rubino 2021 JAMA^[1]) and SURMOUNT-4 for tirzepatide (Aronne 2024 JAMA^[2]) — tell the same story: stopping a GLP-1 entirely after hitting goal weight produces large, predictable regain over 48 to 52 weeks. What neither trial tested directly is whether a lower maintenance dose could preserve most of the loss at a fraction of the cost and side-effect burden. This article walks through what the published evidence actually shows, the pharmacokinetic basis for step-down protocols that telehealth providers now offer off-label, and a practical de-escalation framework anchored to the trial data.

The honest summary

• Full discontinuation produces large regain. STEP-4^[1] randomized week-20 responders to continue semaglutide 2.4 mg or switch to placebo; the continued arm went on to −7.9% additional weight loss while the placebo arm regained +6.9%, a net 14.8 percentage-point gap at week 68. SURMOUNT-4^[2]replicated the pattern for tirzepatide: continuation −5.5% vs withdrawal +14.0% over 52 weeks of maintenance.
• One-year post-stop regain is roughly two-thirds. The STEP-1 extension (Wilding 2022, Diabetes Obesity & Metabolism^[4]) followed participants 52 weeks after semaglutide withdrawal at week 68 and found about two-thirds of the lost weight was regained — the most-cited real-world anchor for “what happens if you just stop.”
• Step-down protocols are not formally tested but are pharmacologically rational. No randomized trial has compared full-dose maintenance to a structured 2.4 → 1.7 → 1.0 → 0.5 mg taper for semaglutide, nor to a 15 → 10 → 7.5 → 5 mg taper for tirzepatide. The mechanistic case — ongoing GLP-1 receptor occupancy to sustain appetite suppression — supports the approach.
• Long-term exposure looks safe in SELECT. The SELECT trial (Lincoff 2023 NEJM^[8]) exposed 17,604 adults to semaglutide 2.4 mg for a mean of roughly three years with no novel safety signal and no evidence of weight-loss tachyphylaxis at maintenance doses.

STEP-4: the semaglutide withdrawal benchmark

STEP-4 (Rubino 2021 JAMA^[1]) ran semaglutide 2.4 mg open-label for the first 20 weeks in 902 adults with overweight or obesity; those who reached the full 2.4 mg maintenance dose (n = 803) were then randomized 2:1 to continue semaglutide or switch to placebo for an additional 48 weeks while continuing lifestyle counselling. Headline results at week 68:

• Continued semaglutide arm: additional −7.9% body weight from week 20 to week 68 (further loss).
• Placebo (switched off) arm: +6.9% body weight from week 20 to week 68 (regain).
• Between-group difference: 14.8 percentage points, corresponding to a mean net regain of roughly 11.6 kg of body weight in the placebo arm relative to the continuation arm.
• Trajectory: regain was steady and did not plateau by week 68 — the trial ended before the placebo curve flattened.

STEP-4 is the canonical evidence that semaglutide’s weight-loss effect requires ongoing dosing. It is not evidence about step-down because the comparator was true placebo, not a reduced dose. That is the trial we are still missing.

SURMOUNT-4: tirzepatide replicates the pattern

SURMOUNT-4 (Aronne 2024 JAMA^[2]) ran tirzepatide open-label for 36 weeks of titration plus maintenance in 670 adults; participants who tolerated the maximum tolerated dose (5, 10, or 15 mg) were then randomized to continue tirzepatide or switch to placebo for an additional 52 weeks. Headline results:

• Mean weight loss at week 36 (end of open-label lead-in): −20.9%.
• Continued tirzepatide arm (week 36 to week 88): additional −5.5%.
• Placebo arm (week 36 to week 88): +14.0%.
• Net difference: 19.5 percentage points, favouring continuation.

SURMOUNT-4 is the larger withdrawal signal of the two trials, which fits the larger baseline efficacy of tirzepatide. The SURMOUNT-4 Reply letter (a 2024 JAMA correspondence by the same investigators) reinforced that the regain curve was still rising at the end of the study, with no plateau.

The STEP-1 extension and the one-year reality

Wilding 2022^[4] followed 327 STEP-1^[5]participants for an additional 52 weeks after semaglutide 2.4 mg was stopped at week 68. Participants regained roughly two-thirds of the weight they had lost — the most widely-cited real-world figure for what happens after a clean stop. The cardiometabolic improvements (lipids, glycaemia, blood pressure) also largely reverted toward baseline. This is the figure that should anchor any patient-facing conversation about “just stopping.”

What about step-down? The pharmacokinetic case

Semaglutide has a half-life of about 7 days; tirzepatide’s is roughly 5 days. Both achieve steady state in 4 to 5 weeks of weekly dosing. The dose-response in the SURMOUNT-1 program (Jastreboff 2022^[6]) showed graded efficacy from 5 mg through 15 mg tirzepatide: −15.0%, −19.5%, and −20.9% total body weight loss at week 72. STEP-2 and earlier dose-finding work for semaglutide showed similar graded behaviour from 0.5 mg through 2.4 mg.

The pharmacological logic of step-down: maintenance dosing does not require the same receptor occupancy as the weight-loss phase. Once a patient has reached goal, the physiological problem changes from negative energy balance to defending the new set point. A lower dose that maintains tonic GLP-1 receptor agonism may be sufficient for the defence function even though it would not have produced the same active loss. This hypothesis is biologically plausible but has not been tested head-to-head in a randomized trial.

An off-label variant some telehealth providers use is every-other-week dosing at the same nominal strength. For tirzepatide with its 5-day half-life, Q2-week dosing produces roughly half the steady-state area-under-the-curve of weekly dosing — mathematically equivalent to halving the weekly dose, but with larger peak-to-trough swings. The clinical equivalence of those two approaches has not been formally compared.

Magnitude: weight change in the year after week-20 maintenance

The cost economics behind step-down

Cash list price for Wegovy 2.4 mg is roughly $1,349 per 28-day supply; LillyDirect self-pay for Zepbound 5 mg vials currently lists at $499 per month. A patient maintaining post-goal on a half-dose — obtained either by splitting a 2.4 mg pen across two months or by stepping down to a lower-strength pen — can roughly halve out-of-pocket cost. Many U.S. commercial plans only cover 12 months of anti-obesity medication; for those patients, a structured step-down at month 10 or 11 is the practical way to extend the prescription duration without falling off the SURMOUNT-4 / STEP-4 cliff.

The side-effect calculus also favours step-down. Most GLP-1 adverse events — nausea, constipation, eructation — are dose-dependent and tend to fade at lower maintenance doses. Patient-reported tolerability typically improves on a 1.0 mg or 0.5 mg semaglutide maintenance dose compared to 2.4 mg, even when efficacy is partially preserved.

The practical step-down protocol

1. Reach goal weight + a 90-day stabilization window on the full maintenance dose. Do not step down while the weight curve is still falling. The signal you are looking for is a 12-week period of weight stability (within ±2 kg) on the full dose.
2. Step down one increment. Semaglutide: 2.4 → 1.7 mg. Tirzepatide: 15 → 10 mg, or 10 → 7.5 mg. Hold at the new dose for 8 to 12 weeks before evaluating.
3. Weigh weekly; track the 4-week moving average. A single week-to-week fluctuation is noise; the trend line is the signal. The decision rule: if the 4-week moving average drifts upward by more than ~1 kg over the 8-week observation window, step back up to the prior dose.
4. Continue the behavioural foundation. Protein at 1.6 g/kg of current body weight, resistance training 2 to 3 times per week, and 7 to 8 hours of sleep per night. These are the same levers used during the weight-loss phase; the published evidence for muscle and metabolic-rate preservation does not disappear at maintenance.
5. If stable at the new dose: step down again. The next increments are semaglutide 1.7 → 1.0 mg or tirzepatide 10 → 7.5 → 5 mg. Most patients find a floor below which regain begins; the floor is highly individual and most often lands in the 0.5 to 1.0 mg range for semaglutide.
6. If regain > 2 kg over any 4-week interval: step back up. Do not chase. Return to the prior dose, hold for 12 weeks, and re-evaluate.
7. Discontinuation candidates. A small fraction of patients can come off the medication entirely — usually those whose BMI has normalized, who have 12+ months of stability at the lowest dose, and who have a strong behavioural foundation. The STEP-1 extension two-thirds regain figure^[4] is the realistic expectation; patients planning full discontinuation should know this in advance.

Tirzepatide vs semaglutide: does step-down differ?

SURMOUNT-4’s placebo arm regained +14.0% over 52 weeks — numerically larger than STEP-4’s +6.9% over 48 weeks, but the comparisons are not head-to-head and the starting losses were also larger for tirzepatide. There is no published evidence that tirzepatide has a lower regain rate at matched percent-loss anchors. For step-down planning, the operating assumption should be that the regain physiology is similar between the two drugs, scaled to baseline loss.

Lean mass on step-down

The SURMOUNT-1 body composition substudy (Look 2025^[7]) found that roughly 25% of total weight loss on tirzepatide was lean mass. Step-down protocols do not reverse that signal in either direction — they preserve the achieved body composition rather than restoring lost lean mass. Patients with sarcopenic obesity risk should continue the muscle-preservation protocol (protein 1.6 g/kg plus resistance training) regardless of which maintenance dose they end up on.

What we still do not know

• No randomized trial has compared full-dose maintenance to a structured step-down. The trials that exist (STEP-4, SURMOUNT-4) compared full dose to true placebo.
• The minimum effective maintenance dose has not been formally defined for either drug.
• Long-term cardiometabolic outcomes on reduced maintenance doses are not in the published record. SELECT^[8] exposed participants to full 2.4 mg dosing.
• Cagrilintide-semaglutide (CagriSema) will change this landscape if approved; the dose-response and step-down profile of dual amylin-GLP-1 agonism may differ materially from GLP-1 monotherapy.

Related research and tools

• Stopping a GLP-1 and rebound weight gain — the published regain curves and what they imply for patients planning to come off
• Life after a GLP-1: maintenance decision hub — the broader framework around long-term GLP-1 use vs structured discontinuation
• Zepbound maintenance dose after goal weight — tirzepatide-specific step-down considerations
• GLP-1 muscle loss prevention protocol — the protein and resistance-training levers that carry over to maintenance
• GLP-1 pregnancy washout protocol — a different discontinuation scenario with its own published guidance
• GLP-1 protein calculator — daily protein target for the maintenance phase

Important disclaimer. This article is educational and does not constitute medical advice. Step-down dosing of GLP-1 therapy is currently off-label; no randomized trial has compared structured step-down to full-dose maintenance. Any dose adjustment should be made in consultation with the prescribing clinician, with weight and side-effect monitoring at each step. Cost figures reflect U.S. cash list prices as of mid-2026 and may change. PMIDs were verified live against the PubMed E-utilities API on 2026-05-29.

Last verified: 2026-05-29. Next review: every 12 months, or sooner if a randomized step-down trial is published or if CagriSema phase 3 data materially changes the maintenance landscape.