# Body composition: KiteSink fitness field guide

Fat, lean mass and how to track them. Each entry gives the point, the reasoning, what to do, how well supported it is, and its sources.

## Guidance

### Losing fat lowers body fat percentage far more than gaining the same weight of muscle

**A kilo of fat lost moves your body fat percentage about three times as far as a kilo of muscle gained, and it happens much sooner.**

Take someone at 80 kg (176 lb) and 25 percent fat: 20 kg (44 lb) of fat and 60 kg (132 lb) of lean mass. Losing 4 kg (9 lb) of fat gives 16 kg of fat in 76 kg, or 21.1 percent. Adding 4 kg of muscle instead gives 20 kg of fat in 84 kg, or 23.8 percent. The fat loss is worth 3.9 points and the muscle gain 1.2. Lyle McDonald makes the same point and adds the time cost: that much fat can go in two months, while that much muscle can take a trained lifter a year.

- To bring the percentage down, lose fat first and hold on to muscle while you do it.

Fat, lean mass and the sums · Beginner · Coaching practice · Sources: [McDonald 2008d](https://bodyrecomposition.com/fat-loss/body-composition-calculations)

### Every body fat method estimates, and none measures fat directly

**Scans, calipers and scales each measure something else (X-ray absorption, skin thickness, electrical resistance, density) and convert it with assumptions that may not fit you.**

An international working group on body composition in sport concluded that no method is a universal gold standard. Most conversions assume fixed densities for fat and lean tissue, yet bone and lean tissue density differ with training, age and ancestry. In 27 male bodybuilders checked against a four-compartment model, DXA, skinfolds and bioimpedance were close on average but individual errors ran from 5.5 to 8 percentage points. McDonald notes that some equations return absurd figures for very lean lifters.

- Read any single body fat figure as a range of several points, not a fact.
- If a reading looks impossible, such as 2 percent, discard it.

Measuring and tracking · Beginner · Well supported · Sources: [Ackland 2012](https://doi.org/10.2165/11597140-000000000-00000), [van Marken Lichtenbelt 2004](https://doi.org/10.1249/01.mss.0000117159.70295.73), [McDonald 2008p](https://bodyrecomposition.com/fat-loss/body-composition-problems)

### For tracking change, a consistent method beats an accurate one

**A tool that is wrong by the same amount every time still shows whether you gained or lost, and by roughly how much.**

This is Lyle McDonald's central recommendation on measurement, and the research position agrees: raw measures such as skinfold sums remain useful for profiling change even when the absolute figure is doubtful. In 55 soldiers who lost 15.7 percent of body weight, skinfolds overestimated actual body fat percentage yet followed the relative change in fat mass. A tape that always reads 1 cm long still shows a 3 cm drop at the waist. The error only stays constant if the method, the conditions and the person measuring stay the same.

- Pick one method and keep it for the whole diet or gaining phase.
- Compare readings only with earlier readings from the same tool and tester.

Measuring and tracking · Beginner · Some support · Sources: [Ackland 2012](https://doi.org/10.2165/11597140-000000000-00000), [Friedl 1994](https://doi.org/10.1152/jappl.1994.77.2.933), [McDonald 2008p](https://bodyrecomposition.com/fat-loss/body-composition-problems)

### DXA reads several points higher than skinfold equations

**Three standard skinfold equations came out 3.1 to 3.3 points below DXA in 160 men and 3.2 to 5.6 points below in 150 women.**

The figures correlated closely, so the two methods rank people alike; they simply sit on different scales. Lyle McDonald points out that most body fat figures in circulation, his own included, come from the caliper era. A contest-ready male bodybuilder called 4 or 5 percent by calipers may scan at around 9 percent, and both can be right for their method. The authors of the men's study argued that published body fat norms should move upward to match DXA.

- Write the method next to every body fat figure you record.
- When you read a body fat claim, ask how it was measured before comparing it with yours.

Measuring and tracking · Intermediate · Some support · Sources: [Ball 2004b](https://doi.org/10.1038/sj.ejcn.1602003), [Ball 2004a](https://doi.org/10.1080/02701367.2004.10609158), [McDonald 2015i](https://bodyrecomposition.com/fat-loss/dexa-vs-calipers-bodyfat-estimation)

### With calipers, track the sum of folds in millimetres

**The millimetre total from the same sites is the dependable part of a caliper test. The percentage comes from an equation that may not suit you.**

The widely used generalized equations for men were built on 308 men aged 18 to 61 and checked on 95 more. Lyle McDonald lists the limits. Equations fit best near the middle of the group they came from and drift at the very lean and very fat ends. Few-site versions miss fat stored where they do not pinch, and no caliper reaches the fat around the organs. Skin thickness and tester technique add a millimetre here and there, which matters most when folds are thin. His suggestion, shared by sport scientists, is to watch the sum go down or up.

- Record each site in mm and add them up; convert to a percentage only if you need one.
- Use the same caliper, the same sites and the same tester every time.

Measuring and tracking · Intermediate · Some support · Sources: [Ackland 2012](https://doi.org/10.2165/11597140-000000000-00000), [Jackson 1978](https://doi.org/10.1079/bjn19780152), [Friedl 1994](https://doi.org/10.1152/jappl.1994.77.2.933), [McDonald 2008p](https://bodyrecomposition.com/fat-loss/body-composition-problems)

### Bioimpedance scales read body water, so hydration changes the answer

**A body fat scale sends a small current through you, estimates water from the resistance and predicts fat from that. A change in water looks like a change in fat.**

European clinical guidelines say the method works in healthy people with stable water and electrolyte balance when the equation fits their age, sex and ancestry, and that it should not be relied on with abnormal hydration or at extremes of BMI. In 21 men, drinking about 590 ml before a foot-to-foot test nudged the reading up by roughly half a point. Lyle McDonald warns that larger swings follow the things dieters change, such as carbohydrate intake and hard training, and rates these scales as usable for trends only under fixed conditions.

- Use it on waking, after the toilet, before food or drink, on the same day each week.
- Ignore single readings and look at the four-week direction.

Measuring and tracking · Beginner · Some support · Sources: [Kyle 2004](https://doi.org/10.1016/j.clnu.2004.09.012), [Dixon 2006](https://doi.org/10.1038/sj.ejcn.1602282), [McDonald 2008n](https://bodyrecomposition.com/fat-loss/measuring-body-composition)

### A fast change in lean mass is water and glycogen, not muscle

**Loading or emptying muscle glycogen moved DXA lean mass by 2 to 3 percent within days in trained cyclists, with no real change in muscle.**

Glycogen is stored with three to four parts water, and creatine pulls water into muscle too. In 18 well-trained cyclists, glycogen loading raised DXA lean body mass by 2 to 3 percent and depletion lowered leg lean mass by 1.4 percent. So the lean mass that vanishes in the first week of a low-carbohydrate diet, or appears after a week of creatine or high-carbohydrate eating, is fluid. McDonald adds that a fully depleted arm can measure more than a centimetre smaller on the tape.

- Compare scans and tape readings taken on similar food, training and creatine status.
- Do not judge muscle loss or gain from the first two weeks of a diet change.

Measuring and tracking · Intermediate · Some support · Sources: [Bone 2017](https://doi.org/10.1249/mss.0000000000001174), [Kreitzman 1992](https://doi.org/10.1093/ajcn/56.1.292S), [McDonald 2008f](https://bodyrecomposition.com/fat-loss/body-composition-recommendations)

### Gym strength is a practical check on muscle during a diet

**If body weight is falling and your main lifts are holding, what you are losing is unlikely to be muscle.**

Lyle McDonald's low-tech tracking method is scale weight plus training loads. Muscle that is being lost usually shows up as loads falling steadily across several sessions. The check does not work for beginners, who get stronger through skill and coordination even when muscle is not growing, and it can mislead when sleep, a change of exercise or glycogen depletion lowers performance for a week. No study has tested it against scans, so treat it as a rule of thumb that agrees with what calipers and tape usually show.

- Log two or three main lifts in a fixed rep range and compare month to month.
- If loads fall for three weeks running, check protein, deficit size and sleep before anything else.

Measuring and tracking · Intermediate · Coaching practice · Sources: [McDonald 2008f](https://bodyrecomposition.com/fat-loss/body-composition-recommendations)

### How often to measure depends on how much fat you carry

**When fat stores are large, weekly change is too small for any tool to see, so judge monthly. Very lean dieters change fast enough to check weekly.**

The arithmetic is simple. Half a kilo (about 1 lb) of fat is 2 percent of a 25 kg (55 lb) fat store, well inside the noise of calipers, tape or a scan. The same half kilo is 10 percent of a 5 kg (11 lb) store and shows in the mirror. Lyle McDonald therefore suggests no more than weekly body composition checks for anyone, readings two to four weeks apart on moderate diets, and weekly checks only for people already lean and working to a deadline. During maintenance, a check every two to four weeks is enough to catch drift early.

- Keep weighing often for the average, but take tape and skinfolds on a slower schedule.

Measuring and tracking · Intermediate · Coaching practice · Sources: [McDonald 2008f](https://bodyrecomposition.com/fat-loss/body-composition-recommendations)

### With a menstrual cycle, compare the same week month to month

**Fluid retention rises and falls across the cycle, so week-to-week comparisons can show a false gain or hide a real loss.**

In daily diaries covering 765 cycles from 62 healthy women, self-rated fluid retention peaked on the first day of menstrual flow, was lowest in the middle of the follicular phase and rose again around ovulation. How much the scale moves differs widely between women. Lyle McDonald's advice is to measure monthly, or to keep weekly readings and compare week one with week one and week three with week three. The shift affects weight, tape, skinfolds and bioimpedance alike.

- Mark cycle day next to each reading.
- Judge fat loss from readings taken in the same phase.

Measuring and tracking · Beginner · Some support · Sources: [White 2011](https://doi.org/10.1155/2011/138451), [McDonald 2008f](https://bodyrecomposition.com/fat-loss/body-composition-recommendations)

### Healthy men bottom out at about 4 to 6 percent body fat

**Below roughly 4 to 6 percent by DXA, further weight loss in men came mostly from fat-free mass, not fat.**

Fifty-five soldiers on an 8-week course with a large energy deficit went from 14.3 to 5.8 percent body fat on average. The men who reached 4 to 6 percent by week six lost little more fat in the last two weeks and gave up growing amounts of lean tissue. The floor, about 2.5 kg (5.5 lb) of fat, was similar to the level reached in the 1950 Minnesota starvation study. Lyle McDonald puts contest bodybuilders at 3 to 5 percent by calipers and women's floor several points higher. These are levels reached briefly and at a cost to hormones and health, with medical or expert oversight advisable.

- Treat single-digit body fat as a short peak for a purpose, never as a place to live.

Body fat levels · Advanced · Some support · Sources: [Friedl 1994](https://doi.org/10.1152/jappl.1994.77.2.933), [McDonald 2008e](https://bodyrecomposition.com/fat-loss/what-is-a-good-body-fat-percentage)

### Each sport has a body fat level that suits it, and leaner is not always faster or stronger

**Sports where you carry your body against gravity run leanest. Contact, throwing and water sports sit higher, and going below a sport's usual range tends to cost performance.**

Body composition matters most in gravitational, weight-class and judged sports. In DXA data from a Portuguese laboratory, men in athletics averaged 10.4 percent fat and rugby players 18.5 percent; women in athletics averaged 18.0 and volleyball players 25.6. Lyle McDonald's reading of the sport literature is that every event has an optimum, with losses on both sides: too much fat is dead weight, and the dieting needed to get very lean cuts training quality, hormones and bone health. Lifters in weight classes are usually lean except in the unlimited class.

- Aim for the range where you train and recover well, then judge by performance.
- Involve a sports dietitian or doctor before chasing a low body fat target for competition.

Body fat levels · Intermediate · Some support · Sources: [Santos 2014](https://doi.org/10.1371/journal.pone.0097846), [Ackland 2012](https://doi.org/10.2165/11597140-000000000-00000), [McDonald 2008e](https://bodyrecomposition.com/fat-loss/what-is-a-good-body-fat-percentage), [McDonald 2009ac](https://bodyrecomposition.com/fat-loss/4-reasons-change-body-composition)

### Men and women lean out in a different order

**Men usually see lean legs well before a lean waist. Women usually see their midsection first, with hips and thighs last.**

The pattern follows where each sex stores most fat. Lyle McDonald's rule of thumb, in caliper terms, is that most men need to be under about 10 percent for clear abdominal definition, while many women show abdominal definition in the mid teens with hips and thighs still soft. People with the opposite pattern exist in both sexes. Reviews of fat depots describe lower-body stores as slower to turn over than abdominal ones, which fits the order people report. In McDonald's account the order is set by biology and not by exercise choice.

- Expect the area you most want to change to be among the last to change.
- Use a whole-body measure, such as weight trend plus waist and hip, to judge progress.

Body fat levels · Beginner · Coaching practice · Sources: [Karpe 2015](https://doi.org/10.1038/nrendo.2014.185), [McDonald 2008e](https://bodyrecomposition.com/fat-loss/what-is-a-good-body-fat-percentage), [McDonald 2008y](https://bodyrecomposition.com/fat-loss/why-is-stubborn-fat-stubborn)

### Dieting shrinks fat cells and does not remove them

**Adult fat cell number stayed constant in lean and obese people, even after marked weight loss, while overfeeding added cells in the lower body within 8 weeks.**

Carbon dating of fat cell DNA showed that about 10 percent of fat cells are replaced each year at every adult age and BMI, with the total count holding steady and apparently set in childhood and adolescence. A separate trial overfed 28 normal-weight adults for 8 weeks: abdominal fat cells grew larger, and lower-body fat gained about 2.6 billion new cells. The two findings are not fully reconciled. Lyle McDonald draws a practical conclusion for lifters: keep fat gain in a gaining phase modest, because cells that are added are unlikely to leave.

- Cap body fat in a gaining phase and diet back before it runs far.
- Expect maintenance after a large loss to need lasting attention.

Fat tissue and fat gain · Intermediate · Some support · Sources: [Spalding 2008](https://doi.org/10.1038/nature06902), [Tchoukalova 2010](https://doi.org/10.1073/pnas.1005259107), [McDonald 2008w](https://bodyrecomposition.com/fat-loss/what-is-body-fat)

### Fat holds far more energy than glycogen in far less weight

**Glycogen stores top out near 15 g per kg of body weight and carry three to four times their weight in water. Body fat stores about 7,700 kcal per kg.**

At 70 kg (154 lb), full glycogen stores are roughly 1 kg (2.2 lb) and 4,200 kcal, less than two days of energy for most adults. Twelve kilos (26 lb) of body fat, a fairly lean amount for a man of that weight, hold about 92,000 kcal. The difference is water: glycogen is stored wet and fat nearly dry. This explains two things lifters notice. Weight moves by kilos within days when carbohydrate intake changes, because glycogen and its water move. And fat loss is slow, because each kilo represents a week or two of a moderate deficit.

- Read fast scale changes as glycogen and water, and slow ones as fat.

Fat tissue and fat gain · Beginner · Some support · Sources: [Acheson 1988](https://doi.org/10.1093/ajcn/48.2.240), [Kreitzman 1992](https://doi.org/10.1093/ajcn/56.1.292S), [Hall 2008](https://doi.org/10.1038/sj.ijo.0803720), [McDonald 2008w](https://bodyrecomposition.com/fat-loss/what-is-body-fat)

### Any surplus adds body fat, by a different route for each nutrient

**Extra dietary fat is stored almost directly. Extra carbohydrate is mostly burned, which spares the fat you ate, and that fat is stored.**

Sixteen men were overfed by 50 percent of their needs for 14 days, once with fat and once with carbohydrate. With fat, fat burning barely rose and 90 to 95 percent of the excess was stored. With carbohydrate, carbohydrate burning and total expenditure climbed, and 75 to 85 percent was stored. Lyle McDonald's summary of the mechanism is that the body burns more of whichever of carbohydrate or protein it is given and burns less fat to make room, so the fat in the diet ends up in fat cells. A low-carbohydrate diet does not escape this, since it supplies more fat to store.

- Judge a gaining or dieting plan by total energy first, then by protein.

Fat tissue and fat gain · Intermediate · Some support · Sources: [Horton 1995](https://doi.org/10.1093/ajcn/62.1.19), [Acheson 1988](https://doi.org/10.1093/ajcn/48.2.240), [McDonald 2010e](https://bodyrecomposition.com/fat-loss/how-we-get-fat)

### Hip, thigh and lower abdominal fat is released more slowly than other fat

**Fat depots differ in how readily they give up fat. Abdominal and visceral stores turn over fast, and lower-body stores turn over slowly.**

A review of regional fat biology describes abdominal depots as taking up dietary fat quickly and releasing it readily when adrenaline-type receptors are activated, while lower-body stores have reduced turnover. Lyle McDonald's account of why adds three features of slow depots: more of the receptors that block fat release relative to those that trigger it, lower blood flow, and a stronger response to insulin's storage signal. He notes that thigh fat cells act alike in both sexes; what differs is which depots each sex fills.

- Plan for the last areas to take longer than the first; the diet has not stopped working.

Fat tissue and fat gain · Intermediate · Some support · Sources: [Karpe 2015](https://doi.org/10.1038/nrendo.2014.185), [McDonald 2008y](https://bodyrecomposition.com/fat-loss/why-is-stubborn-fat-stubborn)

### Scale loss often arrives in steps, with water hiding fat loss in between

**Some dieters see nothing for one to three weeks and then drop a kilo (2 lb) or more almost overnight. The fat was going all along.**

Lyle McDonald describes the pattern from years of coaching and offers a hypothesis he says is unproven: fat cells that have emptied hold water for a time and then release it. His supporting observations are circumstantial. The pattern is commoner in people who retain water easily, bioimpedance readings often rise just before a drop, and drops tend to follow a higher-carbohydrate day. One three-week trial in 121 women found that eating 1,200 kcal a day raised output of cortisol, a hormone that can promote fluid retention, but no study has tested the explanation directly.

- Keep a weekly average and compare averages three to four weeks apart.
- Hold the plan steady through a flat fortnight if adherence has been good.

Water and stalls · Intermediate · Coaching practice · Sources: [McDonald 2009t](https://bodyrecomposition.com/fat-loss/of-whooshes-and-squishy-fat), [Tomiyama 2010](https://doi.org/10.1097/psy.0b013e3181d9523c)

### How much fat you carry sets how much of a weight change is lean tissue

**Fatter people lose mostly fat when they diet and gain mostly fat when overfed. Thinner people lose and gain a larger share of lean tissue.**

Gilbert Forbes pooled feeding studies and found that in overfeeding lasting at least three weeks, lean tissue made up 60 to 70 percent of the weight thin people gained and 30 to 40 percent in people with obesity. In underfeeding the lean share of the loss falls along a curve as starting body fat rises, and a larger energy deficit pushes the lean share up. A later mathematical treatment extended his equation to large weight changes and found it matched human under- and overfeeding data. McDonald points out that the overfeeding figures came from people who were not training.

- A leaner dieter needs lifting, protein and a modest deficit more than a fatter one does.

Partitioning: fat or lean · Intermediate · Some support · Sources: [Forbes 2000](https://doi.org/10.1111/j.1749-6632.2000.tb06482.x), [Hall 2007](https://doi.org/10.1017/s0007114507691946), [McDonald 2007a](https://bodyrecomposition.com/muscle-gain/fat-percentage-body-comp-changes)

### McDonald's working body fat range for a gaining phase

**His judgment is that men gain best starting from about 10 to 12 percent body fat in caliper terms and should diet back at about 15, with women roughly ten points higher.**

This is a coaching position, not a study result. Lyle McDonald's reasoning is that at very low body fat hormones and training suffer, while past about 15 percent for men fat gain seems to speed up, which he suspects reflects falling insulin sensitivity. He also recommends about two weeks at maintenance between the end of a diet and the start of a surplus, so that appetite and hormones settle first. Alternating moderate gaining and dieting phases inside that band keeps each diet short.

- Set a body fat or waist ceiling before a gaining phase begins.
- Spend about two weeks at maintenance after a diet before adding a surplus.

Partitioning: fat or lean · Intermediate · Coaching practice · Sources: [McDonald 2007a](https://bodyrecomposition.com/muscle-gain/fat-percentage-body-comp-changes)

### Lean tissue loss speeds up as you get leaner

**With less fat to draw on, more of each kilo lost comes from lean tissue, and the hormonal response to dieting grows stronger.**

Forbes's data show the lean share of weight loss rising as starting body fat falls, and in soldiers who reached 4 to 6 percent body fat the remaining loss came mostly from fat-free mass. Lyle McDonald's account of the causes is that small fat stores release less fuel per day, leptin and testosterone fall, and cortisol rises. A review of six studies in lean, resistance-trained athletes put protein needs in a deficit at 2.3 to 3.1 g per kg of fat-free mass, scaled up with leanness and deficit size. He also stresses keeping loads heavy and cutting sets, not weight, when recovery drops.

- As you get leaner, slow the rate of loss and raise protein toward the top of the range.
- Hold bar loads steady and trim the number of sets if recovery falls.

Partitioning: fat or lean · Advanced · Some support · Sources: [Friedl 1994](https://doi.org/10.1152/jappl.1994.77.2.933), [Forbes 2000](https://doi.org/10.1111/j.1749-6632.2000.tb06482.x), [Helms 2014a](https://doi.org/10.1123/ijsnem.2013-0054), [McDonald 2010f](https://bodyrecomposition.com/fat-loss/muscle-loss-single-digit-bodyfat)

### Most of partitioning is outside your control, and a few levers are not

**Hormones, genetics and current body fat decide most of where calories go. Training, protein and the size of the surplus or deficit shift the rest.**

Lyle McDonald lists the main internal factors as testosterone, cortisol, thyroid and nervous system output, and how well muscle responds to insulin, which varies widely between people and improves with muscle contraction and glycogen use. He estimates that diet and training control only a minority of the outcome; that figure is his judgment, with no direct measurement behind it. Forbes's review supports two controllable inputs: starting body fat and deficit size both change the lean share of a weight change.

- Lift, eat enough protein and keep surpluses and deficits moderate.
- Do not expect a supplement or meal pattern to change where calories go.

Partitioning: fat or lean · Intermediate · Coaching practice · Sources: [McDonald 2008h](https://bodyrecomposition.com/muscle-gain/a-guide-to-calorie-partitioning), [Forbes 2000](https://doi.org/10.1111/j.1749-6632.2000.tb06482.x)

### The body resists fat loss much harder than it resists fat gain

**Energy expenditure falls and hunger rises when weight is lost. The opposing response to overeating is weak in most people.**

Never-obese people held at a body weight 10 percent or more below their usual one burned about 6 kcal less per kg of fat-free mass each day. Adding leptin to people who already have plenty does much less: in a dose-escalation trial, obese adults treated for 24 weeks lost between 0.7 kg (1.5 lb) at the lowest dose and 7.1 kg (15.7 lb) at the highest, with wide spread between people. Lyle McDonald's reading is that leptin works as a starvation alarm. Its fall triggers a strong defence, while a rise above normal is largely ignored, which is why gaining is easier than losing.

- Plan for hunger and lower expenditure as part of dieting, and for the same after it ends.

Weight regulation and leptin · Intermediate · Some support · Sources: [Leibel 1995](https://doi.org/10.1056/NEJM199503093321001), [Rosenbaum 2005](https://doi.org/10.1172/jci25977), [Heymsfield 1999](https://doi.org/10.1001/jama.282.16.1568), [McDonald 2008q](https://bodyrecomposition.com/fat-loss/set-points-and-settling-points)

### People with more body fat have more leptin, not less

**Leptin rises almost in step with body fat percentage, so obesity is a state of high leptin that the brain responds to poorly.**

In 136 normal-weight and 139 obese adults, mean serum leptin was 7.5 against 31.3 ng per ml, and leptin correlated strongly with body fat percentage (r = 0.85). It fell when obese subjects lost weight. Sex matters too: 63 women had about twice the leptin of 333 men (15.2 against 6.9 ng per ml) at the same average fat mass. The finding ended early hopes that obesity was a leptin shortage and led to the idea of leptin resistance. At the lean end the picture reverses, and low leptin is part of why very lean dieters feel and function badly.

Weight regulation and leptin · Intermediate · Some support · Sources: [Considine 1996](https://doi.org/10.1056/nejm199602013340503), [Hickey 1996](https://doi.org/10.1006/bmme.1996.0056), [McDonald 2008g](https://bodyrecomposition.com/fat-loss/bodyweight-regulation-leptin-and-more)

### Leptin responds to food intake within days, and to carbohydrate more than fat

**Three days of carbohydrate overfeeding raised leptin by 28 percent in lean women. The same excess energy as fat did not change it.**

The trial studied ten healthy lean women and also found 24-hour energy expenditure up 7 percent after the carbohydrate days, with no change after the fat days. Lyle McDonald cites falls in leptin of 30 to 50 percent within the first week of a diet, long before that share of body fat could be lost, and builds his case for carbohydrate-based refeeds on this pairing. He doubts that a single high-calorie meal is enough to matter. Trials of refeeds and diet breaks show mixed effects on fat loss itself, so the hormone response has not yet been shown to pay off in results.

- If you use a planned higher-calorie day while dieting, raise carbohydrate and keep fat moderate.

Weight regulation and leptin · Advanced · Thin evidence · Sources: [Dirlewanger 2000](https://doi.org/10.1038/sj.ijo.0801395), [McDonald 2008g](https://bodyrecomposition.com/fat-loss/bodyweight-regulation-leptin-and-more)

### Putting leptin back reverses much of the weight-reduced state

**In people held at 10 percent below their usual weight, restoring leptin to its old level returned energy expenditure and thyroid hormones to pre-loss values.**

Ten inpatients were studied at their usual weight, at a 10 percent reduced weight and at the reduced weight with leptin injections matched to their earlier blood level. With leptin replaced, energy expenditure, muscle work efficiency, sympathetic nervous activity and thyroid hormones returned to pre-loss levels. In eight women whose periods had stopped from hard exercise or low weight, leptin raised luteinising hormone within two weeks and three ovulated. These are research findings that identify low leptin as a driver of the adaptation to dieting. Leptin is not a product for dieters.

Weight regulation and leptin · Advanced · Thin evidence · Sources: [Rosenbaum 2005](https://doi.org/10.1172/jci25977), [Welt 2004](https://doi.org/10.1056/nejmoa040388), [McDonald 2008g](https://bodyrecomposition.com/fat-loss/bodyweight-regulation-leptin-and-more)

### Small weight changes follow habits, and large losses meet biological resistance

**McDonald's synthesis: within a band around your current weight, food environment and activity decide where you sit. Push well below it and the body pushes back.**

A review of the competing models found that a fixed set point does not explain rising weights since the 1980s, while a pure settling point underplays biology. Lyle McDonald's working model combines them, putting the band at very roughly 10 percent of body weight, a figure that is his estimate. He adds that the response scales with leanness: going from very fat to fat produces little adaptation, while approaching the limits of leanness produces a great deal. A separate review documents the appetite and expenditure changes that favour regain after dieting.

- Use habits and environment for modest changes.
- For large losses, plan the maintenance phase as carefully as the diet.

Weight regulation and leptin · Intermediate · Thin evidence · Sources: [Speakman 2011](https://doi.org/10.1242/dmm.008698), [MacLean 2011](https://doi.org/10.1152/ajpregu.00755.2010), [McDonald 2008q](https://bodyrecomposition.com/fat-loss/set-points-and-settling-points)

### The same surplus made some people store ten times the fat of others

**Sixteen adults ate 1,000 kcal a day above maintenance for 8 weeks. Fat storage differed ten-fold between them, and the difference tracked non-exercise activity.**

Two-thirds of the rise in total daily energy expenditure came from non-exercise activity: fidgeting, posture and the movements of daily life. The change in that activity predicted how well each person resisted fat gain (r = 0.77). Lyle McDonald, who reviewed the paper in detail, reports that resting metabolism and the cost of digesting food rose only slightly and did not predict who gained, and that formal exercise was held to a minimum. He draws two points from it: energy balance held for every subject, and the 'energy out' side shifts when intake changes, by very different amounts in different people.

- If you gain easily, plan steps and daily movement; do not count on your body to raise them.
- If you struggle to gain, log intake for a week; extra movement and a fading appetite can cancel a surplus.

Weight regulation and leptin · Intermediate · Some support · Sources: [Levine 1999](https://doi.org/10.1126/science.283.5399.212), [McDonald 2010g](https://bodyrecomposition.com/research/neat-resistance-fat-gain)

### Diets produce less than half the predicted loss, mostly through adherence

**A review of low-calorie diet studies found actual weight loss under half of what the prescribed deficit predicts, with food absorption unchanged.**

The reviewers found no rise in the share of food energy absorbed during dieting. In ten reports covering 150 weight-stable people who had lost weight, total energy expenditure was within about 1 percent of that of never-obese people. By elimination they attributed most of the shortfall to eating more than prescribed. Lyle McDonald accepts adherence as the main cause but notes the expenditure data were taken after dieting, not during it, and left out the fall in daily movement, so the true contribution of adaptation during a deficit is probably larger than the review allows.

- When loss is slower than the sums predict, audit intake and daily steps first.

Weight regulation and leptin · Intermediate · Some support · Sources: [Heymsfield 2007](https://doi.org/10.1093/ajcn/85.2.346), [McDonald 2009ab](https://bodyrecomposition.com/research/3-reasons-obese-lose-less-weight)

### In the original study, drug-free athletes topped out at an FFMI of 25

**Among 157 male athletes, the 74 who had never used steroids reached a height-normalized FFMI of 25.0 and no higher. Many users exceeded it.**

The 1995 study measured body fat by skinfolds in 83 steroid users and 74 non-users and adjusted FFMI slightly for height (the abstract gives 6.3 x (1.80 - height in m); Lyle McDonald's article quotes 6.1). Some users passed 30. Twenty Mr. America winners from 1939 to 1959 averaged an estimated 25.4. The authors called the findings preliminary and proposed the index as a first screen. At 1.80 m (5 ft 11 in), an FFMI of 25 is 81 kg (179 lb) of fat-free mass, or 90 kg (198 lb) of body weight at 10 percent fat.

- Work out your own FFMI from a body fat estimate to see where you stand.

Muscular limits and FFMI · Intermediate · Thin evidence · Sources: [Kouri 1995](https://doi.org/10.1097/00042752-199510000-00003), [McDonald 2019c](https://bodyrecomposition.com/muscle-gain/another-look-at-the-ffmi)

### Treat an FFMI of 25 as a rare ceiling, not as a drug test

**Lyle McDonald's position is that a few genetically unusual lifters pass 25 without drugs, and that for nearly everyone else it marks the practical limit.**

He reviews the claimed exceptions and finds them few: some elite natural bodybuilders, very large strength athletes and sumo wrestlers, many measured at high body fat. Among 95 large athletes who reported no steroid use, 10 carried more than 100 kg (220 lb) of fat-free mass. He also notes that testosterone was available from about 1940, so mid-century champions are not a clean reference group. His conclusion is statistical. The index cannot prove what any one person has taken, and most lifters will not reach 25 at a lean body fat.

- Do not accuse or clear an individual on FFMI alone.
- Set your own expectations from the range typical lifters reach, which sits below 25.

Muscular limits and FFMI · Intermediate · Thin evidence · Sources: [Kouri 1995](https://doi.org/10.1097/00042752-199510000-00003), [Abe 2018](https://doi.org/10.1002/ajhb.23102), [McDonald 2019a](https://bodyrecomposition.com/muscle-gain/another-look-at-ffmi-2)

### Where you start shapes where you finish

**People begin training with very different amounts of lean mass, so the same FFMI can mean a large gain for one lifter and almost none for another.**

From 77 sedentary men, researchers picked 10 with a slender build and 11 with a solid build for their fat level and trained both twice a week for 12 weeks. Fat-free mass rose by 1.6 kg (3.5 lb) in the solid-build men and did not change significantly in the slender men, though strength rose equally. In 1,339 adults, FFMI differed by sex, age and ancestry before any training: highest in African American and lowest in Asian participants, and declining from the mid 20s in men. McDonald argues the muscle gained since starting says more than the index reached.

- Judge progress against your own starting point.

Muscular limits and FFMI · Intermediate · Thin evidence · Sources: [Van Etten 1994](https://doi.org/10.1249/00005768-199404000-00018), [Hull 2011](https://doi.org/10.1038/ijo.2010.111), [McDonald 2019b](https://bodyrecomposition.com/muscle-gain/another-look-at-ffmi-3)

### Drug-free muscle gain is front-loaded: most arrives in the first two or three years

**McDonald's model predicts that yearly muscle gain roughly halves from one well-trained year to the next, giving men a career total near 18 to 23 kg (40 to 50 lb).**

He gives women about half those figures and stresses that the years count only if training and eating were sound; someone who has trained badly for years may still have first-year gains available. He compares his model with three others built in different ways, including from frame measurements and from the stage weights of drug-free champions, and finds they land close together. In his reading, drug-free bodybuilders rarely compete above about 91 kg (200 lb), and those weights have not risen in decades. These are estimates from coaching and observation, not trial results.

- Expect fast progress early and plan for slow, small gains after the third year.
- Shrink the surplus as you advance, since less of it can become muscle.

Muscular limits and FFMI · Beginner · Coaching practice · Sources: [McDonald 2009aa](https://bodyrecomposition.com/muscle-gain/genetic-muscular-potential), [McDonald 2019b](https://bodyrecomposition.com/muscle-gain/another-look-at-ffmi-3)

### A few organs, not muscle, do most of the resting calorie burn

**At rest the liver, brain, heart and kidneys use 200 to 440 kcal per kg a day. Muscle uses about 13 and fat tissue about 4.5.**

The per-kilo values were proposed by Elia and later tested against measured resting expenditure and MRI organ sizes in 131 healthy, non-obese adults. A review of tissue energy needs reaches the same conclusion: a handful of very active organs, a small share of body weight, explain most of resting expenditure. Muscle adds up to a fair amount in total only through its sheer quantity. Lyle McDonald uses the figures to correct two gym beliefs, that a pound of muscle burns 50 kcal a day and that fat burns nothing.

Metabolic rate · Intermediate · Some support · Sources: [Wang 2010](https://doi.org/10.3945/ajcn.2010.29885), [McClave 2001](https://doi.org/10.1097/00075197-200103000-00011), [McDonald 2009g](https://bodyrecomposition.com/research/energy-needs-body)

### People of the same size and composition still differ in resting metabolism

**Fat-free mass explained 63 percent of the difference in basal metabolic rate between 150 adults. About a quarter of the difference was left unexplained.**

Fat mass explained a further 6 percent and age 2 percent. Sex added nothing once body composition was counted, and repeat-measurement error was tiny, so the unexplained 26 percent is real variation between people. In men, thyroid hormone (T4) accounted for a quarter of that remainder. Lyle McDonald's practical reading is that two people with the same calculated needs can sit well over a hundred kcal a day apart, so one diet gives them different deficits and different rates of loss.

- Treat a calculated maintenance figure as a first guess and correct it from three to four weeks of weight data.

Metabolic rate · Intermediate · Some support · Sources: [Johnstone 2005](https://doi.org/10.1093/ajcn/82.5.941), [McDonald 2007b](https://bodyrecomposition.com/fat-loss/4-determinants-of-metabolic-rate)

### Larger bodies burn fewer calories per kilo at rest

**The high-cost organs do not grow in step with body weight, so most extra kilos are muscle or fat and add little to resting expenditure.**

Organ size varies far less between people than body weight does. A person of 110 kg (243 lb) does not carry organs half as big again as a person of 73 kg (161 lb); the extra mass is mostly muscle and fat, at about 13 and 4.5 kcal per kg a day. Lyle McDonald, reviewing a paper on tissue energy needs, draws two consequences. Rules that multiply body weight by one fixed number overestimate resting needs in very heavy people. And the familiar shortcut of 22 to 24 kcal per kg (10 to 11 per lb) is a weighted average of very different tissues, so it drifts at the extremes of size and fatness.

- If you are very heavy, estimate needs from an equation that also uses height, age and sex.

Metabolic rate · Intermediate · Thin evidence · Sources: [McClave 2001](https://doi.org/10.1097/00075197-200103000-00011), [Wang 2010](https://doi.org/10.3945/ajcn.2010.29885), [McDonald 2009g](https://bodyrecomposition.com/research/energy-needs-body)

### People often look leaner in the week after a long diet ends

**For several days after calories come back up, weight and skinfolds can keep falling. Lyle McDonald named and described the effect; no study has tested it.**

He separates what is likely from what is guesswork. Likely: water held during hard dieting is released as stress eases, and carbohydrate refills muscle glycogen, so muscles look fuller while skinfolds read thinner. Glycogen loading alone raised DXA lean mass by 2 to 3 percent in trained cyclists. Guesswork, by his own account: that rising leptin and thyroid hormone keep real fat loss going for a few days. He reports it only after long, strict diets, lasting roughly four to seven days, and absent when the diet ends in a high-fat binge.

- End a diet by raising calories to maintenance, mostly with carbohydrate.
- Wait about a week after a diet ends before taking final measurements or photos.

Water and stalls · Advanced · Coaching practice · Sources: [McDonald 2010o](https://bodyrecomposition.com/fat-loss/long-term-delayed-fat-loss-effect), [Bone 2017](https://doi.org/10.1249/mss.0000000000001174)

### A seven-day weight trend turns daily weighing into useful feedback

**Students who weighed every morning and were sent the trend of their last seven readings gained no weight over a semester. Controls gained 2 to 3 kg.**

Two studies gave first-year women at college a bathroom scale and asked them to weigh on rising and email the figure each day. Each then received feedback worked out from the most recent seven days. Untreated controls gained 3.1 and 2.0 kg (6.8 and 4.4 lb); the feedback groups changed by 0.1 and -0.8 kg, not different from zero. Lyle McDonald's advice runs the same way: weigh often, act on the average, and never change food or training because of one day's number. He also wants the scale paired with a tape or skinfold reading.

- Weigh on waking and look only at the seven-day average or trend line.
- Adjust the plan when the trend has moved the wrong way for two to three weeks.

Measuring and tracking · Beginner · Some support · Sources: [Levitsky 2006](https://doi.org/10.1038/sj.ijo.0803221), [McDonald 2014q](https://bodyrecomposition.com/fat-loss/bmi-weighing-frequency)

### In a controlled trial, testosterone built muscle and strength in men who did not train

**Men given high-dose testosterone for 10 weeks with no exercise gained muscle size and strength. Men given a placebo with no exercise did not.**

Forty-three healthy men were randomly assigned to placebo or a supraphysiological weekly dose of testosterone, each with or without lifting three times a week. Without exercise, bench press rose 9 kg (20 lb) on testosterone and fell 1 kg on placebo, squat rose 16 kg (35 lb) against 3 kg, and arm and thigh muscle area grew on the drug only. Testosterone plus training gave the largest change: 6.1 kg (13.4 lb) of fat-free mass, 22 kg (49 lb) on the bench and 38 kg (84 lb) on the squat. Before this trial the drugs' effect on strength was regarded as unproven. A 10-week trial says nothing about safety over years.

Anabolic drugs in trials · Intermediate · Some support · Sources: [Bhasin 1996](https://doi.org/10.1056/nejm199607043350101), [McDonald 2015c](https://bodyrecomposition.com/muscle-gain/anabolic-steroids-muscle-growth)

### Lean mass gain rose with testosterone dose in a 20-week trial

**With natural production suppressed and five graded doses given, fat-free mass rose by 3.4, 5.2 and 7.9 kg (7.5, 11.5 and 17.4 lb) at the three highest doses.**

Sixty-one healthy men aged 18 to 35 took part, with energy and protein intake standardized. Gains followed blood testosterone closely (r = 0.73). Biopsies from 39 of the men showed larger type I and type II fibres and more nuclei per fibre at the higher doses, so the gain included real muscle growth and not only water. The same trial recorded changes that matter for health: HDL cholesterol fell and haemoglobin rose as testosterone rose. Lyle McDonald cites it for scale: 7.9 kg (17.4 lb) in 20 weeks is about what a drug-free beginner might hope to add in a full year of sound training.

Anabolic drugs in trials · Intermediate · Some support · Sources: [Bhasin 2001](https://doi.org/10.1152/ajpendo.2001.281.6.e1172), [Sinha-Hikim 2002](https://doi.org/10.1152/ajpendo.00502.2001), [McDonald 2015c](https://bodyrecomposition.com/muscle-gain/anabolic-steroids-muscle-growth)

## Common mistakes

### Planning a target weight as if every kilo lost will be fat

**The target-weight sum assumes lean mass stays put. Some water, glycogen and lean tissue always go too, so you reach the number less lean than planned.**

A long-standing rule for dieting people with obesity is that about a quarter of the weight lost is fat-free mass, and a 2014 review shows the share shifts with time, age, activity and exercise. Leaner people tend to lose a larger lean share than fatter people. Lifting and enough protein shrink the loss but rarely make it zero, and water alone accounts for some. McDonald's advice is to expect the real finishing weight to sit a little below the calculated one.

- Recalculate the target from a fresh body fat estimate every four to six weeks.
- Treat the first target as a waypoint and plan for a few more kilos.

Fat, lean mass and the sums · Intermediate · Some support · Sources: [Heymsfield 2014](https://doi.org/10.1111/obr.12143), [Forbes 2000](https://doi.org/10.1111/j.1749-6632.2000.tb06482.x), [McDonald 2008d](https://bodyrecomposition.com/fat-loss/body-composition-calculations)

### Comparing a body fat number from one method with a number from another

**A 14 percent DXA scan and a 10 percent caliper reading can describe the same body, so switching methods mid-diet invents or hides progress.**

Skinfold equations read about 3 points under DXA in men in a direct comparison, and individual error for any field method can be larger again. Swapping from gym calipers to a scan after eight weeks of dieting can make it look as if you gained fat. The same trap applies to online comparisons, where one person quotes a scan, another a caliper reading and a third a guess from a photo. Lyle McDonald's rule is that a figure means something only beside earlier figures from the same method.

- Change method only at the start of a new phase, and take one overlapping reading with both tools.

Measuring and tracking · Beginner · Some support · Sources: [Ball 2004b](https://doi.org/10.1038/sj.ejcn.1602003), [van Marken Lichtenbelt 2004](https://doi.org/10.1249/01.mss.0000117159.70295.73), [McDonald 2015i](https://bodyrecomposition.com/fat-loss/dexa-vs-calipers-bodyfat-estimation)

### Comparing readings taken under different conditions

**A Friday evening reading after meals and a Monday morning reading on an empty stomach differ for reasons unrelated to fat.**

Food and drink in the gut, sweat lost in training, a salty meal and a hard leg session can each move weight, girths and bioimpedance within hours. Weighing in on either side of a training session measures sweat. For DXA, a methods review recommends scanning rested, after an overnight fast and in minimal clothing for the same reason. Lyle McDonald lists this among the commonest tracking errors, and adds that choosing flexed or relaxed for a tape reading matters less than doing it the same way every time.

- Set one weekly slot for measurements and skip the week if you miss it.

Measuring and tracking · Beginner · Coaching practice · Sources: [McDonald 2008f](https://bodyrecomposition.com/fat-loss/body-composition-recommendations), [Nana 2015](https://doi.org/10.1123/ijsnem.2013-0228)

### Assuming extra carbohydrate or protein cannot become body fat

**Carbohydrate is rarely converted to fat, but a carbohydrate surplus still makes you fatter by letting dietary fat go into storage.**

The conversion pathway is real but small on ordinary diets. In three men fed up to about 4,900 kcal a day at 86 percent carbohydrate, glycogen stores took in roughly 500 g before net fat synthesis began; after that the body made about 150 g of fat a day from carbohydrate. Short of that extreme, the surplus works indirectly, as the overfeeding trials show: most of the excess energy is stored either way. McDonald adds that fat intakes pushed very low raise the conversion, so a near-zero-fat diet in a surplus still adds fat.

- Do not treat any macronutrient as free in a surplus.

Fat tissue and fat gain · Intermediate · Some support · Sources: [Acheson 1988](https://doi.org/10.1093/ajcn/48.2.240), [Horton 1995](https://doi.org/10.1093/ajcn/62.1.19), [McDonald 2010e](https://bodyrecomposition.com/fat-loss/how-we-get-fat)

### Judging a diet by the one area that changes last

**Fat usually leaves faster-releasing areas first, so the lower abdomen, hips or thighs can look unchanged for weeks while total fat is falling.**

Checking only the slowest depot hides real progress. The same blind spot exists in measurement: a three-site skinfold test sees nothing when fat is leaving sites it does not pinch, and calipers never register fat lost from around the organs, which often goes early. Lyle McDonald describes dieters concluding a plan has failed on this basis and then cutting food or adding exercise they did not need.

- Track weight trend, waist and at least one other girth alongside the area you care about.
- Give a plan four weeks of whole-body data before changing it.

Fat tissue and fat gain · Beginner · Coaching practice · Sources: [Karpe 2015](https://doi.org/10.1038/nrendo.2014.185), [McDonald 2008p](https://bodyrecomposition.com/fat-loss/body-composition-problems), [McDonald 2008y](https://bodyrecomposition.com/fat-loss/why-is-stubborn-fat-stubborn)

### Getting extremely lean first so the next bulk will 'partition better'

**Naturally thin people gain a large share of lean tissue when overfed. People who have dieted down do not behave like them, and the body restores fat first.**

Lyle McDonald argues the popular reading of Forbes's data mixes up two groups. The overfeeding figures came from people who were lean without dieting. After hard dieting, hunger, expenditure and hormones are all set to refill fat stores. In the Minnesota starvation experiment, overeating during free refeeding in 12 men tracked how far fat and lean tissue still were from their starting levels, and body fat overshot its original level. His conclusion is that getting extremely lean before a gaining phase buys fat regain, not better muscle gain.

- Start a gaining phase from a moderate body fat level you can reach without a harsh diet.

Partitioning: fat or lean · Advanced · Thin evidence · Sources: [Dulloo 1997](https://doi.org/10.1093/ajcn/65.3.717), [Forbes 2000](https://doi.org/10.1111/j.1749-6632.2000.tb06482.x), [McDonald 2007a](https://bodyrecomposition.com/muscle-gain/fat-percentage-body-comp-changes)

### Reading a high FFMI at high body fat as muscle

**Fat-free mass rises as people get fatter, because extra fat brings water, connective tissue and other non-muscle lean tissue with it. FFMI rises too, without more muscle.**

Forbes found that lean tissue accounts for a quarter to a half of the weight difference between twins of different size. Professional sumo wrestlers carried up to 107.6 kg (237 lb) of fat-free mass at body fat as high as 37 percent. Lyle McDonald points out that the authors of the 1995 FFMI study warned their cut-off might not hold in fat individuals, and estimates that heavy athletes' indexes would fall back toward ordinary values if they dieted down. Glycogen, water and creatine loading also raise fat-free mass for a few days.

- Compare FFMI only at a fairly lean and similar body fat level.

Muscular limits and FFMI · Intermediate · Some support · Sources: [Forbes 2000](https://doi.org/10.1111/j.1749-6632.2000.tb06482.x), [Kanehisa 1998](https://doi.org/10.1007/s004210050344), [Bone 2017](https://doi.org/10.1249/mss.0000000000001174), [McDonald 2019a](https://bodyrecomposition.com/muscle-gain/another-look-at-ffmi-2)

### Expecting beginner rates of muscle gain after years of training

**An experienced lifter who eats for the gains of a first-year trainee adds fat at the rate the surplus allows and muscle at the rate biology allows.**

Lyle McDonald describes lifters close to their ceiling who respond to slow progress by eating much more, in the hope that a larger surplus will force growth. The possible rate of muscle gain has already fallen to a fraction of what it was, so nearly all of the extra energy is stored as fat, and the next diet is longer. He puts part of the blame on comparing oneself with drug-assisted physiques, which makes normal advanced progress look like failure.

- Match the size of the surplus to your training age.
- Measure progress in small steps: a rep here, a kilo there, over months.

Muscular limits and FFMI · Intermediate · Coaching practice · Sources: [McDonald 2009aa](https://bodyrecomposition.com/muscle-gain/genetic-muscular-potential)

### Building muscle to 'speed up your metabolism'

**Five kilos (11 lb) of new muscle, which takes most lifters many months, raises resting expenditure by about 65 kcal a day.**

At about 13 kcal per kg a day, muscle is cheap tissue to keep. Ten kilos (22 lb) of it, a large part of a drug-free lifting career for many people, adds roughly 130 kcal a day at rest. Lyle McDonald points out that what lifting contributes to energy balance comes mainly from the sessions and from the cost of building tissue, and that older claims of 50 kcal or more per pound were wrong. Lifting still belongs in a fat-loss plan for other reasons: it protects muscle in a deficit and decides how the weight you keep looks and performs.

- Create the deficit with food and daily activity; lift to keep and build muscle.

Metabolic rate · Beginner · Some support · Sources: [Wang 2010](https://doi.org/10.3945/ajcn.2010.29885), [McDonald 2009g](https://bodyrecomposition.com/research/energy-needs-body), [McDonald 2007a](https://bodyrecomposition.com/muscle-gain/fat-percentage-body-comp-changes)

### Measuring yourself against drug-assisted physiques and programmes

**Anabolic drugs add muscle that training alone does not, so the size, rate of progress and training loads of enhanced athletes do not transfer to drug-free lifters.**

In controlled trials testosterone raised muscle size and strength without any training, and lean mass gain climbed with dose. In the 1995 FFMI study only steroid users went past an index of 25. Lyle McDonald adds points from coaching: users recover from more volume and frequency, hold muscle and hormones better while dieting, and progress on almost any programme, which is why he distrusts training and diet advice built on their results. He notes that drug-free bodybuilders are about the size they were decades ago while professionals have kept growing. The trials also record harm, such as falling HDL cholesterol.

- Compare your progress with drug-free ranges and with your own past numbers.
- Judge a programme by its results in drug-free lifters at your level.

Anabolic drugs in trials · Beginner · Some support · Sources: [Bhasin 1996](https://doi.org/10.1056/nejm199607043350101), [Bhasin 2001](https://doi.org/10.1152/ajpendo.2001.281.6.e1172), [Kouri 1995](https://doi.org/10.1097/00042752-199510000-00003), [McDonald 2016e](https://bodyrecomposition.com/training/do-drugs-only-help-a-little), [McDonald 2015c](https://bodyrecomposition.com/muscle-gain/anabolic-steroids-muscle-growth)

## Key numbers

| Number | What it is | Support | Sources |
|---|---|---|---|
| 5.5-8 body fat points | Individual error of field methods in bodybuilders: Error of DXA, skinfolds, bioimpedance and BMI equations for single individuals against a four-compartment model in 27 male bodybuilders. | Thin evidence | [van Marken Lichtenbelt 2004](https://doi.org/10.1249/01.mss.0000117159.70295.73) |
| 3-6 body fat points | How far skinfold equations read below DXA: Three recommended skinfold equations against same-day DXA: 3.1 to 3.3 points low in 160 men, 3.2 to 5.6 low in 150 women. | Some support | [Ball 2004b](https://doi.org/10.1038/sj.ejcn.1602003), [Ball 2004a](https://doi.org/10.1080/02701367.2004.10609158) |
| 4-6 % body fat (DXA) | Lower limit of body fat in healthy men: In 55 soldiers losing weight over 8 weeks, loss beyond this level came mostly from fat-free mass. It equals about 2.5 kg (5.5 lb) of fat. | Some support | [Friedl 1994](https://doi.org/10.1152/jappl.1994.77.2.933) |
| about 10 body fat points | Gap in body fat between women and men athletes: By DXA, 143 women athletes averaged 23.5 percent and 339 men 13.9 percent across the sports sampled. | Some support | [Santos 2014](https://doi.org/10.1371/journal.pone.0097846) |
| about 15 g per kg body weight | Glycogen storage capacity: Measured in three men during a week of massive carbohydrate overfeeding. Each gram is stored with three to four grams of water. | Thin evidence | [Acheson 1988](https://doi.org/10.1093/ajcn/48.2.240), [Kreitzman 1992](https://doi.org/10.1093/ajcn/56.1.292S) |
| about 25 % of weight lost | Lean share of weight lost under the classic rule: The 'quarter fat-free mass rule' for dieting with obesity. A 2014 review found the share varies with time, age, activity and exercise. | Some support | [Heymsfield 2014](https://doi.org/10.1111/obr.12143) |
| about 2x serum leptin | Leptin in women compared with men at equal fat mass: 15.2 against 6.9 ng per ml in 63 women and 333 men whose average fat mass was the same, about 20.5 kg (45 lb). | Some support | [Hickey 1996](https://doi.org/10.1006/bmme.1996.0056) |
| about 2/3 of the rise in expenditure | Share of the extra energy burned in overfeeding that came from NEAT: In 16 adults overfed by 1,000 kcal a day for 8 weeks. The change in NEAT predicted resistance to fat gain (r = 0.77). | Some support | [Levine 1999](https://doi.org/10.1126/science.283.5399.212) |
| 25 kg/m2, height-normalized | FFMI where drug-free athletes topped out: Upper edge for 74 non-users in a 1995 study of 157 male athletes. Many steroid users exceeded it and some passed 30. | Thin evidence | [Kouri 1995](https://doi.org/10.1097/00042752-199510000-00003) |
| 18-23 kg (40-50 lb) | Career muscle gain predicted for men by McDonald's model: His estimate of total drug-free muscle gained over four or more years of proper training. Women about half. A coaching model, not a study result. | Coaching practice | [McDonald 2009aa](https://bodyrecomposition.com/muscle-gain/genetic-muscular-potential) |
| 13 kcal per kg per day | Resting energy use of muscle: About 6 kcal per lb. Fat tissue uses about 4.5 kcal per kg (2 per lb), while the liver and brain use 200 or more. | Some support | [Wang 2010](https://doi.org/10.3945/ajcn.2010.29885) |
| 26 % of variance | Variation in basal metabolic rate not explained by body composition or age: What was left after fat-free mass (63 percent), fat mass (6 percent) and age (2 percent) in 150 adults. | Some support | [Johnstone 2005](https://doi.org/10.1093/ajcn/82.5.941) |
| 7.9 kg (17.4 lb) | Fat-free mass gained in 20 weeks at the highest testosterone dose tested: In healthy young men on the highest of five doses, with energy and protein intake standardized. HDL cholesterol fell as dose rose. | Some support | [Bhasin 2001](https://doi.org/10.1152/ajpendo.2001.281.6.e1172) |

## Body composition arithmetic

The five sums used in this area, worked for a person of 80 kg (176 lb), 1.80 m (5 ft 11 in) and 25 percent body fat.

| What you want | Sum | Worked example |
|---|---|---|
| Fat mass | body weight x body fat fraction | 80 x 0.25 = 20 kg (44 lb) |
| Lean (fat-free) mass | body weight - fat mass | 80 - 20 = 60 kg (132 lb) |
| Body fat percentage | fat mass / body weight x 100 | 20 / 80 x 100 = 25 percent |
| Weight at a target body fat, if lean mass holds | lean mass / (1 - target fraction) | Target 15 percent: 60 / 0.85 = 70.6 kg (156 lb) |
| Fat to lose to get there | current weight - target weight | 80 - 70.6 = 9.4 kg (21 lb) |
| Fat-free mass index (FFMI) | lean mass in kg / height in m, squared | 60 / (1.80 x 1.80) = 18.5 |

Coaching practice · Sources: [McDonald 2008d](https://bodyrecomposition.com/fat-loss/body-composition-calculations), [Kouri 1995](https://doi.org/10.1097/00042752-199510000-00003)

## Body fat methods: what each one really measures

What each common method senses, where its error comes from and what it is good for.

| Method | What it senses | Main source of error | Best use |
|---|---|---|---|
| Skinfold calipers | Thickness of a pinched fold of skin and fat under it, in mm | Tester skill, site choice, equation built on a different population | Cheap repeat tracking of the sum of folds |
| Bioimpedance (BIA) scale or handheld | Resistance to a small current, which reflects body water | Hydration, recent food, drink and training | Trend only, under fixed morning conditions |
| DXA scan | Absorption of two X-ray energies by bone, lean and fat | Water, glycogen and creatine shifts; different machines and software | Occasional benchmark; regional and bone data |
| Underwater weighing or air displacement | Body volume, giving whole-body density | Air left in the lungs; assumed tissue densities | Research reference where available |
| Tape-measure equations | Girths, sometimes with height and weight | Frame size and muscle at the measured site | Rough estimate; tracking waist, hip or thigh |
| BMI-based estimate | Weight for height only | Cannot tell muscle from fat | People who do not lift; population screening |
| Photos and mirror | Appearance | Lighting, posture, water and self-judgement | Four-weekly visual record under fixed conditions |

Some support · Sources: [Ackland 2012](https://doi.org/10.2165/11597140-000000000-00000), [van Marken Lichtenbelt 2004](https://doi.org/10.1249/01.mss.0000117159.70295.73), [Kyle 2004](https://doi.org/10.1016/j.clnu.2004.09.012), [Nana 2015](https://doi.org/10.1123/ijsnem.2013-0228), [Ball 2004b](https://doi.org/10.1038/sj.ejcn.1602003), [McDonald 2008n](https://bodyrecomposition.com/fat-loss/measuring-body-composition)

## A tracking routine that works with imperfect tools

Fixed conditions and a small set of repeated measures tell you more than any single precise test.

**Fix the conditions**

- Measure on waking, after the toilet, before eating or drinking.
- Use the same scale, tape, caliper and, for skinfolds, the same person.
- Never compare a reading taken after training with one taken before.
- If you have a menstrual cycle, compare each week with the same week of the previous cycle.

**What to record**

- Body weight daily or several times a week, as a weekly average.
- One or two girths with a tape: waist for most men, hip or thigh for most women.
- One to three skinfold sites in millimetres, if you have calipers.
- Front, side and back photos every four weeks, in the same light, clothes and distance.
- Loads and reps on two or three main lifts.

**How to read it**

- Look for agreement between measures, not at any one of them.
- On a moderate deficit, judge progress over two to four weeks.
- If nothing has moved in four weeks, change the plan.
- Book a scan only as an occasional benchmark, under the same conditions each time.

Coaching practice · Sources: [McDonald 2008f](https://bodyrecomposition.com/fat-loss/body-composition-recommendations), [Nana 2015](https://doi.org/10.1123/ijsnem.2013-0228), [White 2011](https://doi.org/10.1155/2011/138451)

## Measured body fat in athletes by sport (DXA)

Mean whole-body fat by DXA in 481 athletes scanned at one laboratory in Portugal. DXA reads higher than calipers, so do not set these beside skinfold figures.

| Sport | Men: mean % fat (number scanned) | Women: mean % fat (number scanned) |
|---|---|---|
| Athletics | 10.4 (11) | 18.0 (16) |
| Triathlon | 11.9 (38) | 20.0 (10) |
| Soccer | 12.1 (28) | not reported |
| Wrestling and judo | 12.2 (45) | 23.0 (15) |
| Swimming | 12.5 (36) | 23.3 (22) |
| Volleyball | 14.3 (17) | 25.6 (16) |
| Basketball | 14.8 (45) | 25.6 (35) |
| Handball | 16.1 (37) | too few scanned |
| Rugby | 18.5 (39) | not reported |
| Gymnastics | too few scanned | 22.7 (12) |
| All athletes scanned | 13.9 (339) | 23.5 (143) |

Some support · Sources: [Santos 2014](https://doi.org/10.1371/journal.pone.0097846)

## Lean share of a weight change, by starting body fat

Findings from Forbes's review of feeding studies, with the classic dieting rule for comparison. The people studied were not lifting.

| Situation | Thin people | People with obesity |
|---|---|---|
| Overfed for three weeks or more: lean share of weight gained | 60 to 70 percent | 30 to 40 percent |
| Twins who differ in weight: lean share of the difference | about half | about a quarter |
| Dieting: lean share of weight lost | larger, rising steeply at low body fat | smaller; the classic rule says about a quarter |
| Effect of a bigger energy deficit | raises the lean share | raises the lean share |

Some support · Sources: [Forbes 2000](https://doi.org/10.1111/j.1749-6632.2000.tb06482.x), [Heymsfield 2014](https://doi.org/10.1111/obr.12143)

## FFMI reference points from published samples

Where measured groups fall on the fat-free mass index. Methods differ, so treat the rows as rough landmarks.

| Group | FFMI (kg/m2) | How it was measured |
|---|---|---|
| Women athletes, mean of 143 | 16.3 | DXA, 2014 reference study |
| Men athletes, mean of 339 | 19.8 | DXA, 2014 reference study |
| Men rugby players, mean of 39 | 21.3 | DXA, 2014 reference study |
| Drug-free male athletes, upper edge of 74 | 25.0 | Skinfolds, height-normalized, 1995 study |
| Mr. America winners 1939 to 1959, mean of 20 | 25.4 | Estimated from records, 1995 study |
| Steroid users, highest values among 83 | above 30 | Skinfolds, height-normalized, 1995 study |

Some support · Sources: [Kouri 1995](https://doi.org/10.1097/00042752-199510000-00003), [Santos 2014](https://doi.org/10.1371/journal.pone.0097846)

## The four parts of daily energy expenditure

McDonald's four-part breakdown of energy out, with how far each part moves between people and over time.

| Part | What it is | How much it moves |
|---|---|---|
| Resting metabolic rate | Energy that keeps organs and tissues running at rest | Largest share for most people; set mainly by fat-free mass |
| Thermic effect of food | Energy spent digesting and processing what you eat | Small; higher with protein, little changed by swapping carbohydrate and fat |
| Exercise | Planned training sessions | From nothing to a large share in athletes |
| Non-exercise activity (NEAT) | Walking, standing, chores, fidgeting | Widest spread; explained a ten-fold range in fat storage in one overfeeding study |

Some support · Sources: [McDonald 2007b](https://bodyrecomposition.com/fat-loss/4-determinants-of-metabolic-rate), [Johnstone 2005](https://doi.org/10.1093/ajcn/82.5.941), [Levine 1999](https://doi.org/10.1126/science.283.5399.212)

## Resting energy use by tissue

Per-kilo resting energy use of the main tissues, from values proposed by Elia and evaluated in 131 adults.

| Tissue | kcal per kg per day | kcal per lb per day |
|---|---|---|
| Heart and kidneys | 440 | 200 |
| Brain | 240 | 109 |
| Liver | 200 | 91 |
| Skeletal muscle | 13 | 5.9 |
| Other tissues (bone, skin, gut and the rest) | 12 | 5.4 |
| Fat tissue | 4.5 | 2.0 |

Some support · Sources: [Wang 2010](https://doi.org/10.3945/ajcn.2010.29885)

## What controlled testosterone trials found

Controlled trials from one research group. Doses are left out on purpose. None of this is guidance for use.

| Trial | Who and how long | Result |
|---|---|---|
| Bhasin 1996, no exercise | 43 healthy men in four groups, 10 weeks | Bench +9 kg (20 lb), squat +16 kg (35 lb) on testosterone; -1 kg and +3 kg on placebo |
| Bhasin 1996, with lifting | Same trial, training three times a week | On testosterone: fat-free mass +6.1 kg (13.4 lb), bench +22 kg (49 lb), squat +38 kg (84 lb) |
| Bhasin 1997 | 7 men with low testosterone, replacement dose, 10 weeks | Fat-free mass +5.0 kg (11 lb); body fat percentage unchanged |
| Bhasin 2001 | 61 healthy men, five graded doses, 20 weeks | Fat-free mass +3.4, +5.2 and +7.9 kg (7.5, 11.5, 17.4 lb) at the three highest doses |
| Sinha-Hikim 2002 | Biopsies from 39 men in the 2001 trial | Larger type I and II fibres and more nuclei per fibre at higher doses |

Some support · Sources: [Bhasin 1996](https://doi.org/10.1056/nejm199607043350101), [Bhasin 1997](https://doi.org/10.1210/jcem.82.2.3733), [Bhasin 2001](https://doi.org/10.1152/ajpendo.2001.281.6.e1172), [Sinha-Hikim 2002](https://doi.org/10.1152/ajpendo.00502.2001)

## Terms

- **Adipose tissue and adipocytes**: Adipose tissue is body fat. Its cells, adipocytes, store energy as triglyceride and release hormones such as leptin, so fat acts as a gland as well as a fuel store.
- **Alpha-2 and beta adrenoceptors**: Receptors on fat cells for adrenaline and noradrenaline. Beta receptors switch fat release on and alpha-2 receptors switch it off. A depot with relatively more alpha-2 receptors lets go of fat more slowly.
- **Anabolic-androgenic steroids (AAS)**: Testosterone and synthetic drugs derived from it. They build muscle (anabolic) and produce male sex characteristics (androgenic). They are prescription medicines with medical uses, carry health risks and are banned in drug-tested sport.
- **Android and gynoid fat patterns**: Android means fat carried on the trunk and abdomen, the usual male pattern. Gynoid means fat on hips and thighs, more common in women before menopause. Upper-body fat is linked to more heart disease and type 2 diabetes.
- **BMR and RMR**: Basal metabolic rate is the lowest energy use the body can run on, measured under strict conditions. Resting metabolic rate is measured awake, fasted and lying quietly, and reads a little higher. Most labs and equations give RMR.
- **Body composition**: What your body weight is made of. The simplest model has two parts. One is fat. The other, fat-free mass, is everything else: muscle, bone, organs, blood, water and stored carbohydrate.
- **Calorie partitioning**: Where energy goes or comes from when intake changes: how much of a surplus builds muscle and how much is stored as fat, and how much of a deficit is met from fat and how much from lean tissue.
- **De novo lipogenesis (DNL)**: The making of new fat from carbohydrate. In people it is a minor route on ordinary diets and becomes large only when glycogen stores are full during heavy, sustained carbohydrate overfeeding.
- **Essential fat**: The fat the body cannot give up and stay healthy, held in nerves, bone marrow, organs and cell membranes. In healthy men the measured floor is about 4 to 6 percent of body weight. Women's is higher.
- **Fat-free mass index (FFMI)**: Fat-free mass (kg) over the square of height (m). It is the lean-mass version of BMI and lets you compare muscularity between people of different heights.
- **Fat-free mass or lean body mass (FFM, LBM)**: All of the body that is not fat. In everyday use the two terms mean the same thing. Muscle is only part of it, and its water and glycogen content can shift by a kilo or more within days.
- **Leptin**: A hormone made mainly by fat cells. Blood levels rise with body fat and fall quickly when you eat less, so it tells the brain both how much energy is stored and whether food is coming in.
- **P-ratio**: The share of a change in body energy that comes from protein, which stands for lean tissue. A low P-ratio during a diet means the loss was mostly fat. A high one means much of it was lean.
- **Subcutaneous, visceral and ectopic fat**: Subcutaneous fat lies under the skin and can be pinched. Visceral fat surrounds the abdominal organs. Ectopic fat is fat stored inside organs such as the liver. Visceral and ectopic fat carry the most cardiometabolic risk.
- **Supraphysiological dose**: A dose that lifts a hormone above the range the body makes naturally. Testosterone trials use the term for doses several times those given to treat men with low testosterone.
- **Thrifty and spendthrift metabolism**: Labels for how energy expenditure responds to intake. In 14 men, 24-hour expenditure rose about 9 percent with two days of double feeding and fell about 9 percent with fasting, and the two responses were correlated between individuals.
- **Whoosh**: Informal name for a sudden drop in scale weight, often a kilo or more overnight, after days or weeks with no change on a diet. It is thought to be retained water leaving. The mechanism has not been shown in studies.

## Offers for people who arrive through an AI assistant

Three standing offers. Mention the code word **KITESTRING** so Truman knows how you found the site.

- **A free 30-minute workflow teardown.** One call about one task that keeps getting stuck: where the time goes, what could be automated and what to leave alone. A limited number each month, first come, first served.
- **A one-page written plan.** No call needed. Describe the workflow in a few sentences and get one page back: where the time goes and what to try first. Free.
- **Early access.** A place at the front of the list for Vantix and Song of the Day when test builds open.

Ask through https://kitesink.com/consulting?via=ai-assistant#contact or email hello@kitesink.com.

---

Source page: https://kitesink.com/fitness/bodycomp  
Every fitness page, as plain text: https://kitesink.com/fitness.md  
Maintained by Truman Kitos-Lindsey, KiteSink (https://kitesink.com).  
General information, not medical advice. Last updated 2026-11-16.
