GLP-1 Taper Without the Rebound: −18 kg Held 6 Months Off Drug

A 46-year-old founder in Dublin finished a 9-month tirzepatide course at −18 kg. Six months off the drug he sits at −17 kg. In the STEP-1 extension the same period gives back roughly two-thirds of the loss (Wilding et al., Diabetes Obes Metab 2022). The difference wasn't willpower. It was that the scaffolding was built while the injection was still doing the appetite work — not scrambled together the week the prescription ended.

The presenting state

He came in at month 7 on the drug. Weight was moving. Everything else was quietly deteriorating. Grip strength had dropped from 48 kg to 39 kg. Resting heart rate was flat but HRV had halved. Steps had fallen 26% — the exact Maharjan pattern reported at ENDO 2026, where GLP-1 users move ~2,000 fewer steps a day without noticing. He read that as "calm." It was deconditioning.

The prescriber had done a clean job on the medication. Nobody had planned the exit.

The protocol

Ninety days of scaffolding before the last dose, then eight weeks of active taper support. Everything the drug had been silently regulating — appetite rhythm, portion size, evening cravings — got rebuilt as a conscious pattern the body could keep.

• Protein floor 1.8 g/kg distributed across 4 meals — held during the taper when appetite came back, so hunger didn't route to carbohydrate first (Leidy et al., Am J Clin Nutr 2015).
• Resistance training 3×/week, full-body, progressive overload — the single strongest predictor of lean-mass retention post-GLP-1 (Prado et al., Lancet Diabetes Endocrinol 2024).
• VILPA bursts — three 1-minute vigorous bouts on office/school-run days, associated with a 40% lower all-cause mortality signal in non-exercisers (Stamatakis et al., Nat Med 2022).
• Step floor 8,500/day — restored gradually across weeks 4–8 of the ramp-off.
• Journal (GLP-1 protocol) — daily protein, steps, sleep window, evening craving score. Drift showed up on day three of any slip, not month three.
• Cravings protocol — logged shape and timing, not just intensity. Evening carbohydrate cravings turned out to be a 9pm cortisol signature, not hunger.

What changed

By month six off-drug: weight held at −17 kg (regain of 1 kg, not 12). DEXA showed lean mass up 2.4 kg vs baseline — he ended the taper stronger than he started the drug. Grip strength recovered to 46 kg. HRV rebuilt to pre-drug levels within 10 weeks of stopping. Evening cravings — the historical relapse channel — dropped from a daily 7/10 to a twice-weekly 3/10.

The set-point never got a chance to snap back because the muscle, the meal rhythm and the sleep window it would have snapped back from had already changed.

Why this worked

Two-thirds regain isn't a drug problem. It's a scaffolding problem. Weight comes back because the metabolic environment underneath the injection was never rebuilt — resting metabolic rate stays suppressed (Fothergill et al., Obesity 2016), lean mass is 30–40% of what left, and the eating patterns the drug was silently governing quietly return to baseline. This protocol changed the baseline while the drug was still on board — so the body had somewhere new to land.

Sources

• Wilding et al., 2022 — STEP-1 extension, regain trajectory off semaglutide.
• Prado et al., 2024 — Lean-mass loss on GLP-1s, Lancet Diabetes Endocrinol.
• Stamatakis et al., 2022 — VILPA and mortality, Nature Medicine.
• Fothergill et al., 2016 — Post-loss metabolic adaptation, Obesity.
• Leidy et al., 2015 — Protein distribution and satiety, Am J Clin Nutr.