GLP-1 Race Weight: Power-to-Weight Math for Athletes

Thomas Prommer Hybrid Athlete & Technology Executive Follow on Strava

Published: March 27, 2026

Updated: April 9, 2026

Why Power-to-Weight Matters

Endurance sport is governed by physics. On a bike, you're fighting gravity on climbs and air resistance on flats. On a run, you're accelerating and decelerating your body mass with every stride. In both cases, the ratio of your power output to your body weight determines your speed more than either variable alone.

This is why a 65 kg climber with a 260W FTP (4.0 W/kg) will drop an 85 kg rider with a 300W FTP (3.53 W/kg) on every hill, despite producing 40 fewer absolute watts. And it's why a 60 kg runner at 55 ml/kg/min VO2max outpaces an 80 kg runner at the same VO2max on a flat course — because the lighter runner requires less energy to move each kilogram of body mass.

GLP-1 medications like semaglutide (Ozempic, Wegovy) and tirzepatide (Mounjaro) are fundamentally weight loss tools. The question for athletes isn't whether they work — the clinical data is unambiguous — but whether the type of weight lost translates to performance. Losing 8 kg of fat while preserving muscle is a pure performance gain. Losing 8 kg split between fat and muscle is a mixed bag. This page is about doing the math.

The Math: How Weight Loss Translates to Speed

Cycling: watts per kilogram

The relationship between weight and climbing speed is nearly linear:

• Formula: W/kg = FTP (watts) / body weight (kg)
• On a 7% gradient: Every 1 kg lost (at constant power) gains approximately 1.5-2% in climbing speed
• On flat terrain: Weight matters less — aerodynamics dominate. A 5 kg weight loss gains only ~0.5% on flat ground.

Example: An athlete with 260W FTP dropping from 82 kg to 76 kg (pure fat loss):

• Before: 260W / 82 kg = 3.17 W/kg
• After: 260W / 76 kg = 3.42 W/kg
• Improvement: +7.9% on climbs

On a 40-minute climb, that's roughly 3 minutes faster. On a hilly century, it accumulates to 10-15 minutes. This is not marginal.

Running: pace per kilogram

The weight-speed relationship in running is more complex because the body's efficiency (running economy) changes with weight. But the simplified model:

• Rule of thumb: ~3-4 seconds per km per kg lost for recreational runners
• For trained runners: ~1.5-2.5 seconds per km per kg (smaller effect because you're already more efficient)
• Marathon impact: A 3 kg fat loss at ~3 sec/km/kg = ~9 sec/km = ~6:20 faster over 42.2 km

The more useful metric for runners is pace at a fixed heart rate. If you're running 5:10/km at 150 bpm before weight loss and 4:55/km at 150 bpm after, you've captured the net effect of weight change — including any muscle loss or efficiency change. This is more reliable than pace-per-kilo estimates.

My numbers: the actual W/kg math

Here is the calculation with my real data, assuming current FTP of 281W holds constant as weight continues to drop:

• At 90.2 kg (current, Mar 29): 281 / 90.2 = 3.12 W/kg
• At 90 kg (target race weight): 281 / 90 = 3.12 W/kg
• At 88 kg (aggressive target): 281 / 88 = 3.19 W/kg

Every 1 kg lost at a maintained 281W FTP adds approximately 0.03 W/kg. That sounds small in isolation, but the cumulative effect matters. Going from 94.5 kg (baseline) to 90.2 kg (current) at constant power is a jump from 2.97 to 3.12 W/kg — a 5% improvement from weight loss alone, before counting the FTP gains from training. The protocol ended March 27 and current weight has continued to settle.

For context: Cat 4 amateur cyclists typically sit at 2.5-3.5 W/kg. Cat 3 is 3.5-4.0 W/kg. Moving from 2.76 to 3.04 in five weeks shifted me from the lower-middle to the middle of the amateur range. Reaching 3.12 at target race weight would put me comfortably in the upper half. These are not elite numbers, but for an Ironman-distance triathlon, they represent meaningful time savings on the bike leg.

My W/kg Progression

Here is my actual cycling power-to-weight data since starting GLP-1 (tirzepatide/Mounjaro) in February 2026. FTP estimates are from Intervals.icu based on structured cycling workouts. Weight is the measurement-day scale value — it fluctuates daily, so these are snapshots, not smoothed averages.

Two things are happening simultaneously: FTP is increasing from structured training (TrainingPeaks cycling plan), and body weight is decreasing from the GLP-1 protocol. The W/kg improvement is the compound effect of both. I cannot cleanly separate how much of the 10.1% gain comes from each factor, but both are moving in the right direction — power up, weight down.

The key pattern to watch is whether FTP holds steady or increases as weight drops. If W/kg is improving because weight is dropping AND power is holding, you're losing the right kind of weight. If W/kg improves only because weight dropped faster than power, your absolute performance may actually be declining on flat terrain. In my case, FTP went from 261W to 281W — a 7.7% absolute power gain — so the signal is clearly positive on both axes.

Pace-Per-Kilo: Running Performance

I do not run with a power meter or chest strap heart rate monitor, so I cannot produce clean pace-at-HR tables from training data. What I do have is marathon race data at different body weights, which gives directional evidence of the weight-to-pace relationship.

Marathon pace at different weights

• 92 kg (Metropolis Marathon, Feb 2025): 3:23:41 finish -- 4:49/km average pace
• 95 kg (Shanghai Marathon, Nov 2024): 3:49:47 finish -- 5:26/km average pace
• 95 kg (Valencia Marathon, Dec 2025): 3:51:01 finish -- 5:29/km average pace

The rough pattern: approximately 40 seconds per km slower at 3 kg heavier. That is higher than the textbook estimate of 3-4 sec/km/kg for trained runners, which tells me that weight was not the only variable — fitness, course profile, conditions, and taper quality all differed between these races.

I am being deliberately conservative about this data. Comparing across different race conditions is not the same as a controlled test on a treadmill at fixed heart rate. The marathon results give a directional signal, not a precise coefficient. For future races during the GLP-1 protocol, I plan to use a chest strap for cleaner heart rate data that would enable proper pace-at-HR comparisons.

The Diminishing Returns Point

There's a point where losing more weight stops making you faster and starts making you slower. Understanding where this threshold sits is critical for athletes using GLP-1, because the medication will happily take you past it.

Why lighter isn't always faster

• Muscle loss accelerates: Below a certain body fat percentage, the body preferentially catabolizes muscle for energy. Every gram of muscle lost is a gram of power-producing tissue gone.
• Hormonal disruption: Very low body fat suppresses testosterone, thyroid function, and growth hormone — all of which affect recovery, adaptation, and power output.
• Recovery suffers: Caloric restriction at low body fat percentages impairs glycogen replenishment and tissue repair. You can still train, but you stop adapting.
• Injury risk increases: Low energy availability (LEA) weakens bones, tendons, and immune function. A stress fracture undoes months of W/kg gains.
• Immune suppression: Chronically low body fat combined with high training load creates a window of vulnerability. Illness costs you more training time than a few extra kilos.

Where the line sits

These are approximate DEXA-measured body fat ranges where endurance performance typically peaks for most athletes:

• Male endurance athletes: 8-12% body fat (DEXA). Below 8%, most see declining performance.
• Female endurance athletes: 15-20% body fat (DEXA). Below 15%, RED-S risk increases substantially.

These are ranges, not targets. Your personal optimal depends on genetics, training history, and how your body responds. The only way to find it is to track both body composition and performance simultaneously, looking for where the curve of improving W/kg stops translating to faster times.

The Muscle Loss Trap

This is the central risk of using GLP-1 for performance. The medication reduces appetite indiscriminately — it doesn't know whether you need those calories for muscle maintenance or whether they're excess. Without active intervention, clinical trials show 25-40% of weight lost on GLP-1 is lean mass.

Let's run the math on what muscle loss costs an athlete:

• Scenario A (fat loss only): 82 kg athlete, 270W FTP. Loses 6 kg of pure fat to 76 kg. New W/kg: 270/76 = 3.55 W/kg (+12% from 3.29).
• Scenario B (60/40 fat/lean): Same athlete. Loses 6 kg: 3.6 kg fat, 2.4 kg lean. FTP drops proportionally to ~255W. New W/kg: 255/76 = 3.36 W/kg (+2% from 3.29).
• Scenario C (worst case, 50/50): Same athlete. Loses 6 kg: 3 kg fat, 3 kg lean. FTP drops to ~245W. New W/kg: 245/76 = 3.22 W/kg (-2% from 3.29 — net SLOWER).

Scenario C is an athlete who lost weight, looks leaner, and is measurably slower. This is not hypothetical — it's what happens when athletes use GLP-1 without a structured muscle preservation protocol.

The rule is simple: monitor power output and body composition, not just scale weight. If your FTP or threshold pace is declining while weight drops, you're losing muscle and need to adjust your protein intake, resistance training, or dose.

Optimal Race Weight: A Framework

There is no universal formula for optimal race weight. The "ideal weight" calculators you find online are based on population averages and don't account for individual muscle mass, bone density, or genetic variation. What works instead is a framework for finding your personal optimum.

Step 1: Establish your baseline

• Get a DEXA scan before starting GLP-1 (or as soon as possible)
• Record your current FTP or threshold pace
• Calculate your starting W/kg or pace-at-heart-rate
• Note your current body fat percentage

Step 2: Track both variables monthly

• Monthly DEXA scans (or at minimum every 6-8 weeks)
• Monthly FTP test or threshold pace test under consistent conditions
• Plot W/kg (or pace at HR) against body fat percentage over time

Step 3: Find the inflection point

• Early in the process, W/kg should improve as body fat drops
• At some point, the rate of improvement slows or reverses
• That inflection point is approximately your optimal race composition
• It's a body fat range, not a specific number on the scale

Step 4: Maintain, don't minimize

• Once you've found the inflection point, the goal shifts from losing weight to maintaining composition
• Discuss maintenance dosing with your physician — you may not need the same dose for maintenance as for weight loss
• Continue resistance training and protein targets — the work doesn't stop when the scale stops moving

Where I am now

The 5-week protocol ended March 27. As of March 29, I am at 90.2 kg — down 4.3 kg from the 94.5 kg baseline — while FTP increased from 261W to 281W. That puts me at 3.12 W/kg, up from 2.76 — solidly in the "clear improvement, no power loss" zone. In fact, power went up, which means I am losing the right kind of weight.

My target race weight for Ironman is the 88-90 kg range. At 281W FTP:

• At 90.2 kg (current): 281 / 90.2 = 3.12 W/kg
• At 90 kg: 281 / 90 = 3.12 W/kg
• At 88 kg: 281 / 88 = 3.19 W/kg

I am essentially at race weight target already. Below 88 kg I would need to monitor carefully for power loss. The 5-week GLP-1 protocol was designed to get from 94.5 kg to a launchable Ironman training weight, not to reach minimum race weight. The remaining work is maintaining this composition through training volume and nutrition management during the build phase.

I do not have DEXA body fat data to plot the inflection point precisely. For now, FTP trajectory is my proxy: as long as power is holding or increasing while weight drops, I have not hit the diminishing returns zone. If FTP starts declining, that is the signal to stop cutting and switch to maintenance.

The temptation on GLP-1 is to keep going because the medication makes continued weight loss relatively easy. The discipline is knowing when to stop losing and start maintaining. Your performance data will tell you — if you're tracking it.