Maintenance-Phase Dosing Research: Evidence Behind Lower-Dose Continuation

Lower-dose maintenance is one of the most widely discussed strategies in clinical practice and one of the least supported by randomized evidence. The gap between how often it is done and how well it has been studied is worth stating plainly.

What the withdrawal trials did and did not test

STEP 4 and SURMOUNT-4 are the two most relevant trials, and both used a binary structure: after an open-label lead-in, participants were randomized to continue at their established dose or to switch to placebo. There was no intermediate arm.

These trials therefore establish that continuing at full dose maintains and extends loss, and that stopping entirely produces regain. They say nothing about the space between those two conditions, which is precisely where the maintenance-dose question lives.

The mechanistic argument, stated fairly

The reasoning offered for reduced-dose maintenance runs roughly as follows. Physiologic adaptation to weight loss — increased appetite signalling, reduced energy expenditure relative to predicted — is the mechanism opposing maintained weight loss. A GLP-1 receptor agonist counteracts part of that adaptation. Counteracting an ongoing defence of a lower weight may require less pharmacologic pressure than driving weight down against that defence in the first place.

The argument is coherent. It is also untested. Two competing possibilities have equal a priori plausibility: that the adaptive drive is strong enough to require the same dose indefinitely, or that the required maintenance dose varies enough between individuals that no general rule applies.

The dose-response literature is sometimes recruited in support, on the grounds that efficacy increments shrink at higher doses. But those data describe the induction phase. Whether the same curve shape applies to maintenance is an assumption, not a finding.

Why the trial has not been run

A properly designed maintenance-dose trial would require a lead-in to achieve weight loss, randomization to full dose versus one or more reduced doses, a non-inferiority margin agreed in advance, and follow-up long enough for divergence to appear. That is a large, long and expensive study with a commercial rationale that points the wrong way for the sponsor most able to run it.

Independent and publicly funded research could fill the gap. Until it does, the evidence base for a widely used clinical strategy will remain mechanistic reasoning plus accumulated practice experience.

Reading claims about maintenance dosing

Three things to check whenever you encounter a maintenance-dosing claim:

Is the cited trial a withdrawal trial? If the citation is STEP 4 or SURMOUNT-4 in support of reduced-dose maintenance, the citation does not support the claim — those trials had no reduced-dose arm.

Is the evidence observational? Retrospective series of patients maintained on lower doses suffer from severe selection effects: patients who do well on less are the patients who remain on less. This does not establish that reducing the dose is safe for someone doing well on more.

Is a cost argument being presented as an efficacy argument? Reduced-dose maintenance may be entirely reasonable on cost or tolerability grounds. That is a different claim from equivalent efficacy, and the two are frequently blurred.