Body Composition Part 1: What Is It, and How Does It Help Researchers?

This is the first post in a three-part series on body composition. Here, we lay out what body composition is and why it matters more than weight or BMI alone. Part two covers how body composition gets measured, from DEXA to smartphone imaging. Part three shows how to bring that measurement directly into an Alethios study.

When people talk about wanting to lose weight, they usually mean losing body fat, not muscle—but both look the same on the scale. Weight is the number on the scale, but it’s rarely what people are actually chasing. Changes in body composition are.

Two people can weigh exactly the same and be in completely different metabolic shape. The scale can't tell them apart. Their bodies can.

What Is Body Composition?

Body composition is what the body is actually made of: a measurement of its separate parts rather than just their sum. At the simplest level, it splits total body weight into two categories: fat mass (the body's stored fat) and lean mass (muscle, bone, organs, and water).

Where weight gives you one number, body composition gives you the breakdown behind it. A person can lose 10 pounds and have that be almost entirely fat, almost entirely muscle, or some mix of both. Those three outcomes mean very different things for their health, whether or not the scale can see the difference.

Weight vs. BMI vs. Body Composition

The three most common ways to describe someone's size answer three different questions, and it's easy to mistake one for another.

Weight answers "how much total mass does this body have?" It's fast and cheap to measure, but it can't distinguish a pound of muscle from a pound of fat—they weigh the same, even though one is metabolically active tissue and the other is stored energy.

BMI (weight in kilograms divided by height in meters squared) answers "how does this person's weight compare to a population average for their height?" It was never designed to measure an individual's health directly—it was created in the 1830s by the Belgian statistician Adolphe Quetelet to describe averages across populations, not to diagnose any one person. It's slightly more sophisticated than weight, but it still can't tell fat from muscle, determine if a person is healthy, or recognize the unique makeup of an individual.

Body composition answers the question people actually care about: "how much of this body is fat, and how much is everything else?" It's the only one of the three that can tell two people who weigh the same apart.

Why Weight and BMI Fail

The gap between what BMI measures and what people think it measures has become hard to ignore. In June 2023, the American Medical Association (AMA) adopted a policy stating that BMI has "significant limitations" as a clinical measure, noting that it "does not directly assess body fat" despite being correlated with it, and that it "loses predictability when applied on the individual level" (American Medical Association, 2023). The AMA also pointed out a more structural problem: BMI's original cutoffs were based largely on data from earlier generations of non-Hispanic white populations, so the same score doesn't carry the same health risk across genders and racial and ethnic groups. The AMA's recommendation was to use BMI, if at all, alongside other measures—visceral fat, waist circumference, and body composition analysis among them—not as a stand-alone verdict.

The clearest real-world example of this failure is playing out right now in weight-management research. GLP-1 drugs like semaglutide and tirzepatide have made "pounds lost" a headline number, but the scale doesn't say what kind of weight came off. Clinical trial data tell a more complicated story: in the STEP 1 and SUSTAIN 8 trials, an estimated 39–40% of the weight participants lost on semaglutide was lean mass, not fat (Drug Discovery Trends). Tirzepatide performed somewhat better in SURMOUNT-1, with lean mass accounting for roughly a quarter of total weight lost. Two participants could lose the same 20 pounds on the same drug and be left with very different amounts of muscle, a difference that matters for long-term metabolic health and that “pounds lost” simply cannot show.

Two participants could lose the identical 20 pounds on the identical drug and come out with very different amounts of muscle left—a difference that matters enormously for long-term metabolic health, and one that "pounds lost" simply cannot show.

Neither weight nor BMI can catch that difference. Body composition can.

When Body Composition Helps

Body composition data earns its keep anywhere a study needs to know what changed in the body's makeup, not just how much the scale moved. That question shows up across several distinct types of research, each asking a slightly different version of it.

Drug development is where body composition is moving fastest from a nice-to-have to a required outcome, especially in metabolic and weight-management trials. Beyond semaglutide and tirzepatide, newer GLP-1/GIP/glucagon triple agonists are being tested with body composition built in as a dedicated outcome rather than an afterthought: a phase 2 substudy of retatrutide in people with type 2 diabetes measured fat mass and lean mass directly, instead of relying on weight change alone (Lancet Diabetes and Endocrinology). The same logic shows up at the opposite end of the weight spectrum, too: in cancer cachexia trials, lean body mass is often the primary endpoint rather than total weight, since cachexia is fundamentally a muscle-wasting problem that weight alone can fail to capture until it's severe. Drugs like anamorelin have been evaluated using appendicular lean body mass as the key outcome (Anamorelin phase 2 trials)

Supplement and nutrition research uses body composition to answer a more everyday version of the same question: is a product or strategy helping someone lose fat while preserving muscle, or does it just help them lose weight, muscle included? That distinction is already showing up as a companion to GLP-1 research rather than a replacement for it: work presented at the Endocrine Society's 2025 annual meeting found that patients on anti-obesity medications who consumed more protein were better able to protect their lean mass during treatment Endocrine Society), which means a nutrition intervention can be tested for whether it actually protects muscle, not just assumed to (Endocrine Society).

Exercise science uses body composition to separate "this training program builds muscle" from "this training program burns calories"—claims that sound similar but require entirely different evidence to support.

Bariatric and surgical weight-loss research deals with a lean-mass cost that's easy to overlook because the headline number (pounds lost) looks like unambiguous success. A systematic review and meta-analysis found that bariatric surgery causes significant, measurable loss of lean body mass and skeletal muscle mass, separate from the intended fat loss (Obesity Reviews), which matters for how post-surgical recovery and long-term outcomes get evaluated.

Longevity and aging research increasingly treats lean mass as a biomarker worth tracking on its own—declining muscle mass, or sarcopenia, is a well-established marker of unhealthy aging, independent of what's happening to overall weight. The same logic applies to specific life stages, too: during the menopause transition, women tend to lose lean mass and gain visceral fat even when total weight barely changes, a shift that's easy to miss if a study is only tracking pounds on a scale (JCI Insight).

Curious what body composition data could add to your next study? Book a demo.