The exit is where most GLP-1 protocols fail. The studies are blunt about it: those who eventually get off the drug quickly regain most of the weight and lose almost all of the other health benefits — blood sugar, lipids, inflammation. The reason isn't willpower. The drug suppressed the signal; it didn't rebuild the system. Stop the injection, and the hypothalamic setpoint, the gut hormones, the muscle-poor metabolic engine all snap back to the configuration that made you fat in the first place. The Path's last move is the only one that decides whether the prior nine weeks were a rental or a purchase.
What's actually happening when you stop
GLP-1 receptor agonists act at the area postrema and arcuate nucleus to suppress hunger and slow gastric emptying. The peripheral signal is loud — porous blood-brain barrier in that region lets the analogue act as a chemosensory trigger — and it overrides the body's endogenous appetite drive. It does not rewrite the defended weight. The hypothalamus still has its setpoint. Leptin sensitivity, insulin signaling, and the gut-brain hunger loop are all running underneath the suppression, waiting.
The second failure mode is body composition. Across the SURMOUNT/SUSTAIN trials, anywhere from a quarter to nearly half of the lost weight is lean mass when the drug is run aggressively without a protein-and-training scaffold. Muscle is the metabolic engine. After age 30 you lose 5% of muscle mass per decade by default; an aggressive GLP-1 run can compress a decade of sarcopenia into 10 months. Stop the drug and the now-smaller engine, with its lower resting expenditure, is what's defending against the returning appetite signal. You lose the fight before it starts.
The third failure is metabolic flexibility — the mitochondrial capacity to switch between burning fat and glucose. Chronic caloric restriction without mitochondrial work shrinks mitochondria and crashes resting energy expenditure. The fat-loss period was a controlled injury. The post-fat-loss window — practitioner consensus calls it the rehab — is where you rebuild the engine. Skip it, and the body interprets re-feeding as famine recovery and partitions the calories to fat storage.
The exit is a protocol, not a stop
Maintenance is built on three substitutions: replace pharmacological appetite suppression with endogenous metabolic competence; replace muscle-loss tolerance with active hypertrophy; and replace GLP-1's mitochondrial glucose-handling with peptide-driven mitochondrial biogenesis. You taper the drug only as those three are coming online — not before.
The taper itself is slow. Cut the maintenance GLP-1 dose by 25% every 4 weeks, not weekly. Watch the appetite signal. If hunger returns hard at a step, hold that dose for another 4 weeks before cutting again. The pituitary, the hypothalamus, and the gut hormones are renegotiating; rushing this is what triggers the rebound binge cycle.
Maintenance protocol — months 3 through 12
| Compound | Dose | Route | Frequency | Evidence Tier | Notes |
|---|---|---|---|---|---|
| Semaglutide (taper) | 0.25 mg → 0.125 mg microdose | SubQ | 1x/week, cut 25% every 4 weeks | Clinical | Hold step if hunger returns; substitute Tirzepatide at equivalent low microdose if appetite drift is severe |
| Berberine | 500 mg | Oral | 2–3x/day with meals | Expert | The bridge compound — replaces GLP-1's glycemic floor. Cycle 8 weeks on / 4 weeks off |
| MOTS-c | 5–10 mg total per week (split 2–3 doses) | SubQ, fasted | 2–3x/week | Expert | Mitochondrial biogenesis. Builds the factory that builds the factory. Run long-term |
| 5-Amino-1MQ | 50–150 mg | Oral | Daily | Experimental | Stacks with MOTS-c. NNMT inhibition — restores NAD+, fat oxidation. Long-term |
| Methylene Blue | 0.5–1 mg | Oral | Daily, 2–4 weeks before MOTS-c, then cycled | Expert | Stabilizes electron transport chain; reduces baseline ROS so MOTS-c lands cleanly. Cycle to avoid hormesis flip |
| Ipamorelin | 200–300 mcg | SubQ | Before bed, fasted, 5 nights/week | Expert | Clean GH pulse — no cortisol/prolactin spillover. Preserves lean mass through the taper |
| CJC-1295 (no DAC) | 100 mcg | SubQ | With Ipamorelin, same injection | Expert | Adds GHRH signal. Skip if fasting insulin >10 µIU/mL — GH secretagogues can worsen insulin resistance |
| Essential amino acids | 20–40 g | Oral | Daily, split around training | Expert | Non-negotiable. Anabolic resistance rises with age; muscle protein synthesis needs the leucine threshold |
| Resistance training | Progressive overload | — | 3–4x/week | Clinical | The single highest-leverage lifestyle input. Muscle is the organ of longevity |
The CJC-1295/Ipamorelin pair is the muscle insurance policy. Practitioner consensus runs the combination 5 nights on, 2 off, with periodic 4-week washouts every 4–6 months to keep the pituitary responsive — desensitization is the documented risk of continuous GHRH stimulation. If fasting insulin sits above 10 µIU/mL when you re-test at week 16, hold the GH stack and run berberine plus MOTS-c alone until insulin clears the gate. GH antagonizes insulin; stacking on a resistant background defeats the purpose.
The Berberine bridge is the metabolic understudy. It activates AMPK through the same downstream pathway GLP-1 partially recruits, holds postprandial glucose, and gives the gut-brain axis a softer landing as the GLP-1 signal fades. The 8-on/4-off cycle is mandatory — continuous use blunts the AMPK response.
What you should feel — taper weeks 1 through 16
- Weeks 1–4 (first 25% cut): appetite uptick, mostly evening. Manageable. Sleep quality often improves as gastric emptying normalizes.
- Weeks 5–8 (second cut): food noise returns at a low murmur. If it's a roar, hold the dose. Energy in training sessions climbs as glycogen storage normalizes.
- Weeks 9–12 (third cut): the body is now mostly running on its own appetite regulation. MOTS-c effects — sustained energy, better fasted-state tolerance — are detectable.
- Weeks 13–16 (final taper or microdose hold): if labs hold (fasting insulin <10, HbA1c stable, body weight ±2%), drop the GLP-1 entirely. Many practitioners hold a 0.125 mg semaglutide microdose monthly as insurance. Both approaches work.
What's NOT happening yet
- Hunger is not gone. It's regulated. The drug masked it; the system now manages it. Treating returning hunger as failure is the most common reason people re-up the dose and stall the rebuild.
- The setpoint hasn't moved permanently. Six months of stable weight at the new baseline is what convinces the hypothalamus. Less than that, and the defended weight is still the old one.
- Mitochondrial density is not built yet. MOTS-c is biogenesis, not turbocharging. The new mitochondria take months. Subjective energy lags the labs by 8–12 weeks.
- You are not on a maintenance diet — you are on a maintenance protocol. Calorie counting is a poor substitute for muscle mass, insulin sensitivity, and mitochondrial competence. The lifestyle layer (protein floor, resistance training, sleep) is the actual maintenance dose. The peptides keep the engine running; they don't replace the inputs.
- The GH stack is not optional theater. Without an anabolic signal during the taper, the lean mass that survived the cut keeps shedding. Lose another 5% of lean mass post-taper and the rebound math gets ugly.
Research describes this. Track your labs every 90 days, hold any taper step the body refuses, and trust that the engine you rebuilt is the real protocol.