Nutrition 04: How fast to lose, how fast to gain

Short answer. In the reference trial, athletes losing 0.7% of bodyweight a week ended with more lean mass than those losing 1.4%, and the difference was gone at twelve months (Garthe et al., 2011a). Our position is to aim for a rate that keeps the deficit moderate, near 0.5 to 1% a week for most people. Our figures for gaining are judgment, not a finding.

---

What it is

Not how much you eat, but how quickly you change. The previous entry covered setting an intake and correcting it from your weight trend. (→ Nutrition 03: BMR, TDEE and the error around them) This one covers what trend to aim at.

The two things you want in a deficit are opposed: a larger deficit removes fat faster and removes more of everything else. Somewhere between them is a rate that gets the job done without costing the tissue you trained for.

The proposed mechanism

In a deficit, the shortfall is met from stored fat and from lean tissue, and the split is not fixed. Resistance training and adequate protein bias it toward fat. A larger shortfall is proposed to overwhelm that bias.

In a surplus, the constraint is the opposite. Muscle accrues at a limited rate and energy beyond that rate is stored as fat. A larger surplus does not raise the ceiling; it fills the excess elsewhere.

Both predict a rate beyond which the extra speed is bought with the wrong tissue. Only one has been tested properly.

What has actually been tested

The reference trial randomized athletes to two rates. Twenty-four elite athletes were assigned to lose 0.7% of bodyweight per week or 1.4% per week, all performing four resistance training sessions per week. The slow group reduced energy intake by 19 ± 2%, the fast group by 30 ± 4% (p = 0.003, well under the conventional 0.05 threshold for significance). They reached the same total weight loss over 8.5 ± 2.2 and 5.3 ± 0.9 weeks respectively (Garthe et al., 2011a).

Bodyweight fell by a comparable amount in both groups, 5.6 ± 0.8% and 5.5 ± 0.7%. What differed was composition. Lean body mass increased by 2.1 ± 0.4% in the slow group (p < 0.001) and was unchanged in the fast group (-0.2 ± 0.7%), with a significant difference between groups (p < 0.01). Fat mass fell by 31 ± 3% in the slow group against 21 ± 4% in the fast group.

The slow group lost more fat and gained lean mass while losing weight. The fast group lost the same weight and gained nothing.

And then the same researchers followed them for a year. Twenty-three of the athletes were retested at six and twelve months. There were no significant differences between groups after twelve months, which the authors read as indicating that weight loss rate is not the most important factor in maintaining body composition and performance after a weight loss period (Garthe et al., 2011b).

The deficit size that preserves lean mass has been examined separately. A meta-analysis of resistance training trials run in an energy deficit found that lean mass gains were impaired against training without a deficit, while strength gains were comparable, and its meta-regression found that a deficit of about 500 kcal per day prevented lean mass gains. The authors advise people training to preserve lean mass during weight loss to avoid deficits larger than 500 kcal per day (Murphy and Koehler, 2022). Whether that threshold holds in highly resistance-trained people has not been established.

Not every comparison finds an advantage for slow. A trial comparing progressive against severe restriction in resistance-trained women reported similar effects on body composition and strength.

The surplus side has almost nothing behind it. No randomized trial compares rates of weight gain in trained lifters against body composition outcomes at a standard approaching the deficit work above.

Our reading

Slow wins in the short term for the outcome most people care about, and the effect did not persist.

Those two sentences describe the same body of work and they usually do not appear together. The 0.7% per week figure travels everywhere; the twelve-month null result travels nowhere. We are reporting both because the second one bounds what the first one means.

The most defensible reading is that rate matters during the deficit and matters less afterward. With a date, a weight class or a photoshoot, the composition at the end is the outcome and the slow group ended with more muscle. As one phase of a long training life, the twelve-month data suggests the difference washes out.

What we would take from the trial is less about the number and more about the mechanism it exposes. The slow group ate 19% below maintenance and the fast group 30% below (Garthe et al., 2011a). The lean mass difference appeared at the point where the deficit stopped being moderate, and a 30% cut is larger than most people realize they are taking when they choose an aggressive target.

So: aim for a rate slow enough that the deficit stays moderate, and expect that to land somewhere near 0.5 to 1% of bodyweight per week for most people. Then read your trend and adjust, because the deficit that produces that rate is a moving target. (→ Nutrition 03: BMR, TDEE and the error around them)

On gaining, the trial evidence is not there and we will give you our position anyway.

The reasoning runs like this. Muscle can only be accrued at a limited rate, that rate is lower the more trained you already are, and a surplus larger than the rate has nowhere to go but fat. Nothing about that changes with a bigger surplus, so the extra energy buys weight rather than muscle. The deficit trial points the same way from the other side: the composition penalty appeared when the energy gap stopped being moderate.

So we would aim for roughly 0.25 to 0.5% of bodyweight per week for someone with a few years of training behind them, and up to about 1% for someone in their first year. Fast enough that the trend is readable within a month, slow enough that most of the added weight has a plausible chance of being what you wanted.

This is judgment, not a finding. The randomized comparisons that would settle it do not exist, and the figures in circulation are convention, not result.

And the general case is worth stating plainly. Most people want the fast version of both. Faster costs more of the tissue you were trying to keep on the way down, and adds more of the tissue you were trying to avoid on the way up. Unless you have a date on the calendar, the slower version is very likely the better trade for the ordinary case, because the thing being optimized is a training career rather than a twelve-week block. (→ Adherence 01: The first month)

Where the goals diverge

Dieting with a deadline. The slow rate has the trial behind it and the composition at the end is what you are being measured on. Sleep is the other variable with direct evidence on this outcome. (→ Sleep 01: Sleep and muscle)

Dieting as a phase. The twelve-month null result applies. Choose a rate you can hold rather than the one that optimized a short-term endpoint in athletes. (→ Nutrition sub-hub)

Trained for many years. The literature specifically flags that retaining muscle in a deficit may be harder for you, and that the trials were not run in your population.

Gaining. Our figures above are reasoned rather than trial-derived, so treat them as a starting rate and let the trend and the mirror correct them over months, not weeks. A scale alone cannot tell you what the added weight is made of.

Everything here assumes adequate protein and resistance training. Both groups lifted four times a week. Without that, the composition question is a different one. (→ Nutrition 01: Protein intake)

The case against

We are leaning on one randomized trial of 24 elite athletes. It is well conducted and rated positively in evidence appraisals, and it is 24 people in a specific population.

The two groups were not equated for duration. Five weeks against eight and a half, so the comparison confounds rate with time under deficit and with training accumulated during it.

The headline finding did not survive twelve months, and we have written the article around a number that the same authors later described as not the most important factor.

Not all comparisons agree. At least one trial in resistance-trained women found similar outcomes between progressive and severe restriction.

Our gain figures have no trial behind them. They are reasoning applied to a mechanism, and a reader who skims will carry them away as though they were measured.

What is not known

The rate question in highly resistance-trained people, which the literature itself flags as untested.

Whether rate or absolute deficit size is the operative variable. The trial varied both together.

Anything approaching a trial-based answer on the surplus side.

Whether the twelve-month convergence reflects real equivalence or the participants' subsequent training and eating overwhelming the original difference.

Safety

Aggressive restriction carries real risks including menstrual dysfunction, bone loss and disordered eating, and the athlete literature explicitly discourages weight loss interventions in under-18s and in anyone with a history of an eating disorder. If you are considering a large deficit, speak to a doctor or registered dietitian first.