DEXA Body Composition: How Nutrition and Fasting Change Your Results

What DEXA Body Composition Actually Measures

DEXA quantifies three compartments: fat mass, lean soft tissue (which approximates muscle and organ mass), and bone mineral content. The scanner uses two X-ray energies at 40 keV and 70 to 80 keV to differentiate tissue types based on differential attenuation. Unlike bioelectrical impedance, DEXA is not confounded by skin temperature or electrode placement, but it remains sensitive to the fluid environment of soft tissue [1].

The Three-Compartment Model in Practice

Fat mass is the most clinically actionable output for most patients. Lean soft tissue tracks closely with skeletal muscle mass but includes organ mass, connective tissue, and total body water. Bone mineral content is the most stable of the three compartments and changes slowly over months. For metabolic monitoring during GLP-1 or TRT therapy, clinicians primarily track the ratio of fat mass to lean mass and the regional distribution of fat, specifically the android-to-gynoid ratio and the visceral fat area estimate [2].

Android and Gynoid Regions Defined

The android region covers the torso from the pelvis to the mid-chest, roughly corresponding to abdominal and visceral fat depots. The gynoid region covers the hips and upper thighs. A high android-to-gynoid fat ratio is associated with insulin resistance, dyslipidemia, and elevated cardiovascular risk independently of total body fat percentage. The European Association for the Study of Obesity published guidance in 2020 identifying android fat distribution as a primary driver of metabolic syndrome beyond overall adiposity [3].

How Nutrition in the 24 Hours Before a Scan Alters Results

A single high-carbohydrate meal the night before a DEXA scan can shift lean soft tissue estimates by 0.5 to 1.5 kg. This is not muscle. It is water bound to glycogen stored in skeletal muscle and the liver.

Glycogen and Water Binding

Each gram of glycogen stored in muscle binds approximately 2.7 to 3 grams of water [4]. A moderate carbohydrate meal of 150 to 200 g net carbohydrate can increase muscle glycogen by 20 to 30 mmol/kg wet weight. That water shows up in the lean soft tissue compartment. If a patient eats a large pasta dinner before one scan and fasts before the next, the lean mass difference between sessions could exceed 1 kg purely from glycogen status, with no actual change in muscle fiber mass.

Sodium and Fluid Retention

High sodium intake (above 3,500 mg in the preceding 24 hours) increases extracellular water retention via aldosterone suppression and osmotic fluid shifts. A study published in the American Journal of Clinical Nutrition found that a 2,300 mg increase in daily sodium intake over four days raised total body water by approximately 0.5 L as measured by deuterium dilution [5]. Because DEXA cannot distinguish intracellular from extracellular water, this retained fluid registers as lean soft tissue.

Alcohol and Dehydration

Alcohol consumed the night before a scan suppresses antidiuretic hormone, promoting diuresis and net fluid loss. The resulting mild dehydration reduces total body water and can artificially lower lean mass estimates by 0.3 to 0.8 kg depending on the amount consumed and individual variation. Patients on GLP-1 agonists should note that nausea-related reduced intake on the day before a scan creates a similar dehydration artifact.

Fasting State and DEXA Accuracy

The standard clinical recommendation from the International Society for Clinical Densitometry (ISCD) is to avoid eating for at least 2 to 3 hours before a DEXA scan and to maintain consistent fasting conditions across serial scans [6]. For research-grade comparisons or clinical trials, an overnight fast of 8 to 12 hours is preferred.

Why Fasting Matters More for Serial Comparisons Than Single Scans

A single DEXA result tells you a snapshot value that is useful for clinical staging. Serial DEXA scans every 12 to 16 weeks during active treatment are where inconsistent fasting creates the most confusion. If a patient fasts for scan one and eats normally before scan two, the apparent lean mass loss could be entirely explained by glycogen depletion rather than true sarcopenia. The ISCD 2019 Official Positions state: "Patients should be instructed to fast (no eating or drinking other than water) for a minimum of 3 hours prior to the exam and to follow the same protocol at all follow-up scans" [6].

Exercise Before Scanning

Resistance training within 24 hours before a DEXA scan transiently increases muscle water content due to exercise-induced inflammation and glycogen resynthesis. A 2017 crossover study in the Journal of Strength and Conditioning Research (N=17) found that performing a leg resistance session 24 hours before a scan increased leg lean mass estimates by a mean of 0.41 kg compared to a no-exercise control condition [7]. Patients tracking lean mass during testosterone therapy or progressive resistance programs should schedule scans at least 48 hours after their last training session.

Hydration Volume on the Day of the Scan

Drinking 500 mL of water in the 60 minutes before a scan adds measurable lean mass. A 2011 study in Obesity (N=31) quantified this effect at approximately 0.36 kg increase in lean soft tissue per 500 mL water ingested [8]. The practical instruction is to drink a normal amount of water, roughly 250 to 400 mL, in the two hours before the scan and to keep that volume consistent across follow-up scans.

Optimal DEXA Body Composition Ranges

"Normal range" is not the right clinical frame for body composition. The better question is: what range is associated with the lowest risk of metabolic disease, cardiovascular events, functional decline, and all-cause mortality?

Body Fat Percentage Targets by Sex and Age

The American College of Sports Medicine stratifies body fat percentage into performance, fitness, acceptable, and obese categories [9]. For metabolic health monitoring at HealthRX, clinicians use the following evidence-based targets:

Men:

• Athletic: 6 to 13%
• Fitness: 14 to 17%
• Metabolic health zone: 18 to 24%
• Elevated cardiometabolic risk: 25% and above

Women:

• Athletic: 14 to 20%
• Fitness: 21 to 24%
• Metabolic health zone: 25 to 31%
• Elevated cardiometabolic risk: 32% and above

A prospective analysis of 13,601 adults from the NHANES dataset found that a body fat percentage above 25% in men and above 35% in women was independently associated with a 1.8-fold increase in metabolic syndrome prevalence after adjustment for BMI [10].

Appendicular Lean Mass and Sarcopenia Thresholds

The European Working Group on Sarcopenia in Older People 2 (EWGSOP2) defines low appendicular skeletal muscle mass index (ASMI) as below 7.0 kg/m² in men and below 5.5 kg/m² in women when measured by DXA [11]. These thresholds identify presarcopenia, the stage at which lean mass deficit is present but functional impairment may not yet be clinically apparent. During GLP-1 therapy, ASMI tracking is the single most important metric to confirm that weight loss is coming from fat rather than muscle.

Visceral Fat Area and Cardiometabolic Risk

DEXA-derived visceral fat area (VFA) above 100 cm² in the android region is the threshold most commonly cited in Japanese metabolic syndrome criteria and validated in multiple Western cohorts [12]. A meta-analysis in Obesity Reviews (18 studies, N=9,842) found that VFA above 100 cm² was associated with a 2.3-fold higher odds of type 2 diabetes compared to VFA below 100 cm², independent of total body fat percentage [12].

DEXA Body Composition During GLP-1 Therapy

Semaglutide 2.4 mg (Wegovy) and tirzepatide 15 mg (Zepbound) produce substantial total body weight loss, but the composition of that loss matters clinically. The STEP-1 trial (N=1,961) showed a mean weight loss of 14.9% at 68 weeks with semaglutide 2.4 mg versus 2.4% with placebo [13]. A DEXA sub-study from STEP-1 published in Obesity found that approximately 39% of the total weight lost was lean mass, with the remainder being fat mass [14].

Why Lean Mass Loss During GLP-1 Therapy Raises a Clinical Flag

Losing roughly 40% of total weight loss as lean mass is substantially higher than what is seen with caloric restriction plus resistance training in exercise-based weight loss programs, where lean mass loss typically accounts for 15 to 25% of total weight lost [14]. This difference drives the current HealthRX protocol of combining GLP-1 therapy with progressive resistance training and protein intake targets of 1.6 to 2.2 g per kilogram of body weight per day, in line with the position stand of the International Society of Sports Nutrition [15].

Serial DEXA Scanning Protocol for GLP-1 Patients

The HealthRX DEXA monitoring framework for patients on GLP-1 agonists includes:

1. Baseline scan before initiating therapy (or within the first 4 weeks).
2. 12-week scan after reaching the maintenance dose to assess early lean mass trajectory.
3. 24-week scan to confirm the fat-to-lean loss ratio over the first six months.
4. Annually thereafter once weight is stable.

All scans are scheduled in the morning after an overnight fast of 8 hours, no exercise in the prior 48 hours, and consistent hydration of 300 mL of water consumed 1 hour before the appointment. This standardization makes changes as small as 0.5 kg of lean mass statistically detectable above the typical DEXA precision error of 1 to 2% for lean mass.

DEXA Body Composition During TRT

Testosterone replacement therapy in hypogonadal men reliably shifts body composition toward more lean mass and less fat mass. A meta-analysis published in the Journal of Clinical Endocrinology and Metabolism (29 RCTs, N=1,083) found that testosterone therapy increased lean body mass by a mean of 1.6 kg and reduced fat mass by a mean of 1.6 kg over 3 to 12 months [16]. DEXA is the preferred tracking tool because it captures regional changes, including visceral fat reduction, that the scale cannot detect.

Pre-Scan Protocol Consistency Matters Even More During TRT

Patients on TRT often report increased training volume and carbohydrate intake as energy improves in the first 8 to 12 weeks. Without standardized pre-scan fasting and exercise abstinence, the glycogen-and-water effect described earlier can inflate apparent lean mass gains by 0.5 to 1.5 kg, making the biological testosterone effect appear larger than it is. Serial scans should follow the same overnight fast and 48-hour exercise rest protocol described above.

How Diet Quality Affects Long-Term DEXA Trends

Acute nutrition effects on scan-day measurements are reversible within 24 to 48 hours. Chronic dietary patterns over months shape actual tissue composition and produce real, sustained changes in DEXA outputs.

Protein Intake and Lean Mass Preservation

A 12-week RCT published in the American Journal of Clinical Nutrition (N=130) found that doubling protein intake from 0.8 g/kg/day to 1.6 g/kg/day during a 500 kcal/day caloric deficit preserved 2.1 kg more lean mass as measured by DXA, compared with the lower protein group [17]. The practical target for patients in active weight loss is 1.6 g/kg/day minimum; for patients over 60 or those with low baseline ASMI, 2.0 to 2.2 g/kg/day is appropriate.

Very Low-Carbohydrate Diets and Apparent Lean Mass Loss

Patients who switch to a ketogenic diet often report alarm at their first follow-up DEXA scan, which may show a 1 to 2 kg apparent lean mass decline. This is mostly glycogen depletion. Muscle glycogen stores can hold 300 to 500 g of glycogen plus 900 to 1,500 mL of associated water, all of which is lost within the first 1 to 2 weeks of carbohydrate restriction [18]. Clinicians should interpret a lean mass drop on DEXA in the context of dietary change and not assume sarcopenia without functional testing (grip strength, gait speed) and a second scan after dietary stabilization.

Caloric Restriction Rate and Lean Mass Loss

Aggressive caloric restriction below 1,000 kcal/day accelerates lean mass loss regardless of protein intake. A Cochrane systematic review of 20 RCTs found that very low energy diets (<800 kcal/day) produced greater total lean mass loss on DXA than moderate caloric restriction (<1,200 kcal/day), even when protein was equated [19]. The clinical recommendation is to target a caloric deficit of 500 to 750 kcal/day, sufficient to produce 0.5 to 0.75 kg per week of total weight loss while limiting lean mass erosion.

Precision, Reproducibility, and What a Real Change Looks Like

DEXA precision error for body fat percentage is typically 1 to 2% in certified facilities using the same scanner and operator. For absolute lean mass, the least significant change (LSC) at 95% confidence is approximately 0.8 to 1.2 kg for total lean mass and 0.3 to 0.5 kg for appendicular lean mass [20]. Changes smaller than the LSC should not be treated as clinically meaningful without corroborating data.

Scanner and Software Variability

Different DEXA platforms (Hologic, GE Lunar, Norland) use different algorithms and generate systematically different absolute values for body fat percentage, sometimes differing by 3 to 5 percentage points for the same patient scanned on the same day [20]. Patients should use the same scanner model and ideally the same machine at the same facility for all serial comparisons. Switching platforms mid-protocol renders the data difficult to interpret.

Positioning and Scan Reproducibility

Patient positioning, particularly arm placement, affects android and gynoid region boundaries. Most facilities use a body composition analysis template where the arms rest at the sides within the scan field. A 2019 paper in the Journal of Clinical Densitometry found that repositioning artifacts contributed up to 0.7% variability in body fat percentage in the same session [20]. Requesting the same technologist or using facilities with written positioning protocols reduces this source of error.