How to Train in a Calorie Deficit (Without Losing What You Built)

TL;DR: People spend a cut fiddling with their training split when the variable that actually decides the outcome sits on the nutrition side. Murphy and Koehler's meta-regression put the threshold at roughly 500 calories a day: below it, lifters still added lean mass while losing fat; above it, lean mass gains stopped. Strength was fine either way. So the job in the gym is simple. Keep the sets, let the weight on the bar drop when it has to, and stop reading a slower session as muscle loss.

---

Almost every "training on a cut" article gets the causality backwards. It treats the program as the thing that saves your muscle and the diet as background. The evidence runs the other direction. Your training keeps the signal alive. How aggressively you eat decides whether that signal has anything to work with.

That reframe changes what you should actually adjust.

The deficit size is the variable that matters

Murphy and Koehler pooled the randomized trials that put lifters through at least three weeks of resistance training under energy restriction (Murphy and Koehler, 2022). Two findings are worth building a cut around.

First, lean mass gains were clearly impaired in a deficit compared to the same training at maintenance. No surprise. But the meta-regression found a threshold: an energy deficit of roughly 500 kcal per day was the point where gains in lean mass disappeared entirely. Under that, trained subjects were still adding tissue while losing fat.

Second, strength gains were statistically comparable between the deficit and control groups. The effect was negative but not significant. Your ability to produce force is far more resistant to underfeeding than your ability to build new tissue.

The practical version: a 300 to 500 calorie daily deficit is not a "slow cut," it's the setting where the training you're doing can still pay you. Push to 800 or 1,000 and you're paying for speed with the exact tissue you came to protect. Run your numbers in the cut calculator and check what your target actually implies as a percentage of bodyweight per week.

Garthe's elite-athlete trial makes the same point in the field. Athletes losing 0.7% of bodyweight per week gained 2.1% lean mass over the intervention. Athletes on the same training who lost 1.4% per week gained nothing (Garthe et al., 2011). Same lifting, different rate of loss, opposite body-composition result.

Keep the sets. Let the load fall.

The gym-side rule follows from that. Volume is the retention signal, so protect it first.

If you can't hit your normal reps at your normal weight, take 5 to 10% off the bar and complete the sets. Cutting from three sets to one so you can keep a number you're attached to is the wrong trade. You've kept the load and thrown away the stimulus.

You also have more room than you think. Muscle maintenance takes far less volume than muscle growth, and the minimum effective dose is low enough that a busy or depleted week doesn't have to cost you anything. That threshold is worked through in detail in our piece on how many sets it actually takes to keep muscle. Knowing your floor is what lets you drop volume deliberately during a rough stretch instead of panicking and either overreaching or quitting.

What not to do: swap your compound work for circuits, higher reps, and "fat-burning" formats. That trades the retention signal for a trivial calorie burn you could have gotten by eating 150 fewer calories. The lifting isn't the deficit tool. It's the reason the deficit comes out of fat.

Protein is doing more work than your program

Longland and colleagues ran a brutal test of this. Four weeks, a 40% energy deficit, resistance and interval training six days a week, one group at 1.2 g/kg of protein and one at 2.4 g/kg. The low-protein group essentially held their lean mass (+0.1 kg). The high-protein group gained 1.2 kg of lean mass and lost more fat, 4.8 kg versus 3.5 kg (Longland et al., 2016).

Identical training. Identical deficit. The only variable was protein, and it moved both sides of the body-composition ledger.

That study used an aggressive deficit and a young trained population, so don't read 2.4 g/kg as a universal requirement. But it does establish the direction, and 2.0 to 2.4 g/kg (roughly 0.9 to 1.1 g per pound) is a sensible target while cutting rather than a bodybuilding superstition. The protein calculator will size it for your bodyweight.

Protein also does a second job here, which is keeping you from abandoning the cut. High-protein intakes reliably reduce spontaneous eating, and appetite is the real failure mode of most diets. We covered the full set of levers in managing hunger on a cut.

Sleep is not a soft recommendation

Nedeltcheva's crossover study is still the cleanest demonstration of what sleep restriction does to a diet. Same 14-day calorie restriction, run twice, once with 8.5 hours in bed and once with 5.5. Short sleep cut fat loss by 55% and increased fat-free mass loss by 60% (Nedeltcheva et al., 2010).

Read that carefully, because it's not a weight-loss finding. Total weight lost was similar. The composition of that loss inverted. You do the diet, you see the scale move, and you end up smaller and softer instead of leaner.

There is no training adjustment that compensates for this. Seven hours is the floor, and it belongs in the same tier as your protein target, not in the "nice to have" pile. The mechanisms are in sleep, fat loss, and muscle gain.

Cardio has a compensation tax

Cardio is a legitimate way to widen a deficit. It's just less efficient than people assume, for three reasons.

Non-exercise activity drops to partly offset it. Add three hours a week of training and your body quietly takes some of it back through less fidgeting, fewer steps, and more sitting. Appetite typically rises with longer aerobic work, which claws back more. And high-intensity conditioning competes directly with lifting for the recovery capacity you already reduced by underfeeding.

Start low, one or two sessions a week, and add only if your rate of loss stalls. Cardio should be the adjustment you reach for after the food is already dialed in, not a penance for a bad weekend.

What to expect, so you don't misread it

In a deficit you will see some combination of: slightly lower top-end strength, noticeably worse work capacity on high-rep sets, flatter-looking muscles from reduced glycogen and intramuscular water, and higher perceived effort at the same load.

None of that is muscle loss. Glycogen depletion alone explains most of it, and it reverses within days of eating at maintenance. If you want a definitive read on whether you're actually keeping tissue, judge it on measurements and strength on your main lifts across a month, not on how a Tuesday session felt. Our guide to losing fat without losing muscle covers the markers worth watching.

A workable week

| Day | Session | |-----|---------| | Mon | Upper body, push emphasis | | Tue | Lower body | | Wed | Walk or easy cardio, 30 min | | Thu | Upper body, pull emphasis | | Fri | Lower body | | Sat | Optional moderate cardio, 30 to 45 min | | Sun | Rest |

Four lifting sessions, one or two cardio sessions, full rest day. Nothing exotic, because the exotic part isn't where the result comes from.

Bottom line

Set the deficit at 300 to 500 calories, hold your protein near 2 g/kg, sleep seven hours, keep every set you were doing before and let the load float down when it needs to. Everything else is noise. If the cut has run long enough that adherence is slipping and performance is genuinely falling off, that's a signal to eat at maintenance for a while, which is the case laid out in how to structure a diet break.

Protokl runs both sides of this at once: it sets your deficit from your actual intake and weight trend rather than a formula, and it adjusts your programming for the phase you're in, so your training targets aren't quietly written for a surplus you're not eating. Download it on the App Store.