How to lose fat without losing muscle, what studies found

Key takeaways

• In the studies below, most of the weight people lost was fat, and part of it was usually lean tissue. In trials, that lean part was smaller in groups that did resistance training, and in some trials in groups that ate more protein.³⁴²⁶
• Lean mass isn't all muscle. In reference figures cited by a 2024 review, a man has 56.5 kilograms of fat-free mass and 28 of skeletal muscle, and a woman has 42.0 and 17.0.⁶
• When whole-body MRI measured muscle itself in two diet studies with no structured exercise program, skeletal muscle averaged 0.22 of the weight lost in 22 men and 0.11 in 78 women.⁴
• In Villareal's 2017 trial of 160 adults aged 65 or older with obesity, lean mass fell 2.7 kilograms over 26 weeks in a dieting group that did aerobic training, 1.0 with resistance training and 1.7 with both.³
• These are small or short studies of particular groups. Longland's trial lasted 4 weeks in 40 young men, Weiss's analysis covered 34 people aged 50 to 60, and none of the three trials ran longer than 12 months.²¹³

Do you lose muscle in a calorie deficit?

On average, yes, in the diet studies that measured muscle directly. Heymsfield's 2024 paper used whole-body MRI, which measures skeletal muscle itself, in two groups who lost weight by cutting calories with no structured exercise program. One group of 74 was followed for 12 to 14 weeks and the other, 26 people, for 12 months.⁴

In the shorter study, the 14 men lost an average of 13.00 kilograms, 2.94 of it skeletal muscle, and the 60 women lost 8.34 kilograms, 0.79 of it muscle. Across the two studies, the average share of weight loss that was skeletal muscle was 0.23 and 0.20 in men and 0.10 and 0.17 in women.⁴

The paper's prediction models were built from one-time MRI scans of 897 healthy adults who weren't in exercise training, and the two weight loss studies were used to check them. The authors write that the weight loss wasn't very large and that muscle loss varied considerably between individuals.⁴

What does lean mass include in each paper?

Lean mass and muscle aren't the same thing. Tinsley and Heymsfield's 2024 review sets out the terms. Fat-free mass is every molecule in the body that isn't fat. Lean soft tissue, the figure a DXA scan estimates, is fat-free mass without bone mineral. Skeletal muscle mass is the muscles themselves, and the review lists imaging such as MRI and CT as the way to measure it. It says the plain term lean mass has no single clear definition.⁶

In the reference figures the review cites, fat-free mass is more than double skeletal muscle mass. Heymsfield's MRI paper says skeletal muscle is about one-half of fat-free mass in adults. A kilogram of lean mass lost on a scan isn't a kilogram of muscle lost.⁶⁴

Does resistance training reduce lean mass loss during weight loss?

The most direct test here is Villareal's 2017 trial. It enrolled 160 adults aged 65 or older with obesity, 40 to a group, for 26 weeks, and 141 finished. One group was a control. The other three were prescribed the same diet, set 500 to 750 calories a day below their needs with about 1 gram of protein per kilogram of body weight a day, and trained three times a week with aerobic exercise, resistance exercise or both.³

Body weight fell 9.0 kilograms in the aerobic group and 8.5 in each of the other two. Lean mass, measured across the whole body by DXA, fell 2.7 kilograms with aerobic training, 1.0 with resistance training and 1.7 with both. Thigh muscle volume by MRI moved the same way, down 77 cubic centimeters, 23 and 40 in the same order. In the control group, lean mass didn't change.³

Resistance training made the lean loss smaller. It didn't remove it. The authors describe the reductions in lean mass as attenuated but not prevented, and they note that most participants were women, white and well educated.³

Murphy and Koehler's 2022 meta-analysis asks a narrower question, whether lean mass still grows when people do resistance training in an energy deficit. It pooled randomized trials lasting 3 weeks or more. In its set of trials matched to separate studies with no deficit, the lean mass effect size was 0.11 below zero with a deficit and 0.20 above zero without one. The abstract doesn't say how many trials were pooled or how many people were in them.⁵

Does it matter whether weight comes off by diet or by exercise?

Weiss's 2007 trial compared the two over 12 months in healthy men and women aged 50 to 60. One group cut calories and kept its activity the same. The other added exercise about 6 days a week and kept its food intake the same. The exercise was mostly walking, elliptical machines, cycling and running, and nobody used strength training. The analysis covers 18 people in the diet group and 16 in the exercise group.¹

The groups lost similar amounts of weight, 8.1 and 7.7 kilograms. Lean mass by DXA fell in both, by 3.5 percent in the diet group and 2.2 percent in the exercise group, and the difference wasn't significant. Thigh muscle volume by MRI fell 6.9 percent in the diet group and didn't fall in the exercise group.¹

So in the same people, whole-body lean mass and a muscle scan of one region gave different answers about the exercise group. The authors didn't measure total muscle mass, and the thigh scan covered a 10-centimeter stretch just above the kneecap. The trial had no group that combined diet and exercise.¹

Does eating more protein reduce muscle loss?

One trial among the six tested protein, and it paired protein with hard training. For 4 weeks, Longland's 2016 trial fed 40 young men a diet that supplied about 40 percent less energy than they needed. Half got 1.2 grams of protein per kilogram a day and half got 2.4. Every man did resistance exercise and high-intensity interval training 6 days a week.²

Neither group lost lean body mass on average, by a 4-compartment model. It was close to unchanged on the lower protein diet, up 0.1 kilograms, and rose 1.2 kilograms on the higher protein diet. Both groups lost fat. The abstract calls it a proof-of-principle trial, and because both groups trained, it can't show what the added protein would have done without the training.²

Tinsley and Heymsfield's review cites a meta-analysis of weight loss studies in adults with average ages of 50 or more. It found that higher protein diets kept more fat-free mass, and the review calls that benefit modest, at 0.45 to 0.83 kilograms. Those results are fat-free mass, not muscle measured directly.⁶

Where do these papers differ?

On some questions the papers' results point in different directions.¹²³⁴⁶

What don't these papers show?

They don't show long-term results. The longest of the three trials ran 12 months, and Longland's lasted 4 weeks.¹²³

They don't cover everyone. Villareal's participants were 65 or older with obesity and mild to moderate frailty, Longland's were young men, and Weiss's were aged 50 to 60. Heymsfield's authors say their estimates fit middle-aged adults most closely.³²¹⁴

No trial here tested protein without training, or resistance training at more than one protein level. Among the papers that took their own measurements, only Heymsfield's measured whole-body muscle.²³⁴¹

A personal plan for losing weight is a question for a doctor or a registered dietitian.

Worked example

Take three made-up people, each matching one group's average in Villareal's trial. One does aerobic training and loses 9.0 kilograms, 2.7 of it lean mass. Dividing 2.7 by 9.0 gives 0.30, so 30 percent of that loss is lean mass. One does resistance training and loses 8.5 kilograms, 1.0 of it lean mass, which is about 0.12, or 12 percent. One does both and loses 8.5 kilograms, 1.7 of it lean mass, which is 0.20, or 20 percent. The shares are arithmetic on the trial's kilograms, and they are shares of lean mass by DXA, not of muscle. Each made-up person stands for the average of 40 older adults, so the numbers describe that trial, not what happens to any one person.³

A closer look

Tinsley and Heymsfield's review describes an MRI study by Bosy-Westphal's team that traced where lost fat-free mass came from. Women with overweight and obesity on a low-energy diet lost 9.5 kilograms of body mass, 1.5 of it fat-free mass. Of that 1.5 kilograms, an estimated 0.9 came from skeletal muscle, 0.1 from the kidney, heart and liver, and 0.5 from the fat-free part of fat tissue, the gut, the skin and parts that weren't measured. Of the 0.9 kilograms from muscle, 0.7 was water and 0.15 was protein. The review doesn't give the number of women or the length of the diet.⁶

Questions and answers

Related guides

• Building muscle in a calorie deficit
• How much protein to lose weight
• Protein per meal
• Strength training per week
• How big a calorie deficit is safe
• What a DEXA scan measures
• Lean body mass calculator
• Body fat calculator

Sources

1. Journal of Applied Physiology. Weiss, Racette, Villareal, Fontana, Steger-May, Schechtman, Klein, Ehsani, Holloszy. 2007.
2. American Journal of Clinical Nutrition. Longland, Oikawa, Mitchell, Devries, Phillips. 2016.
3. New England Journal of Medicine. Villareal, Aguirre, Gurney, Waters, Sinacore, Colombo, Armamento-Villareal, Qualls. 2017.
4. Obesity. Heymsfield, Yang, McCarthy, Brown, Martin, Redman, Ravussin, Shen, Müller, Bosy-Westphal. 2024.
5. Scandinavian Journal of Medicine and Science in Sports. Murphy, Koehler. 2022.
6. Journal of the Endocrine Society. Tinsley, Heymsfield. 2024.