DEXA Body Composition: Sex and Menstrual Cycle Differences Explained

What DEXA Actually Measures, and Why Sex Hormones Matter

DEXA does not measure a single tissue. It uses differential X-ray absorption to partition body mass into three compartments: fat mass, lean soft tissue (mostly muscle plus water), and bone mineral content. It is distinct from bioelectrical impedance scales, calipers, and CT or MRI based body composition, each of which estimates these compartments differently and is not directly interchangeable with DEXA.

Sex hormones, principally estrogen, testosterone, and progesterone, regulate where fat is stored, how much lean tissue is built and retained, and how much water the body holds at any given moment. That last point matters for DEXA specifically, because the scan cannot distinguish a liter of retained fluid from a liter of new fat or muscle. Ignoring hormonal context is one of the more common ways serial DEXA scans get misread, particularly for people using GLP-1 medications or hormone therapy where the clinical question is whether lean mass is holding steady.

The core, quotable summary: DEXA partitions the body into fat, lean, and bone compartments with a reported test-retest precision on the order of a few percentage points, but menstrual cycle phase, hydration, and switching scanner models can each shift a fat-mass or lean-mass reading by an amount close to or exceeding that margin of error. In practice, a premenopausal woman scanned in the luteal phase and again in the follicular phase, or a patient rescanned on a different manufacturer's device, may see an apparent change in fat or lean mass that reflects timing and equipment rather than real tissue change. Sex hormones also set different population baselines, which is why clinical body-fat reference ranges are sex-specific rather than a single universal target.

Estrogen and where fat goes

Estrogen favors fat storage in subcutaneous depots, particularly the hips and thighs, and appears to restrain visceral fat accumulation during the reproductive years. This is consistent with the well-established observation that premenopausal women carry more total body fat than age-matched men but proportionally less of it viscerally, which shows up on DEXA as a lower android-to-gynoid fat ratio.

After menopause, estrogen withdrawal is associated with a shift of fat toward the central, visceral compartment. Multiple longitudinal studies of the menopausal transition report meaningful increases in visceral fat area over the years surrounding the final menstrual period, independent of total weight change, though the exact magnitude reported varies by study population and measurement method and should be checked against the specific paper before quoting a precise percentage. On a DEXA report, this shift appears as a rising android-to-gynoid ratio and rising visceral fat estimate even when the number on the scale is unchanged.

Testosterone and lean mass

Testosterone supports skeletal muscle protein synthesis, and men typically carry substantially more appendicular lean mass than women of similar height and age. Trials of testosterone therapy in hypogonadal men, including the Testosterone Trials (TTrials) published in the New England Journal of Medicine, reported increases in lean mass and decreases in fat mass on DEXA over roughly a year of treatment. Readers who want the exact effect sizes should pull the original TTrials publication rather than rely on a secondhand figure, since precise numbers are easy to garble in summary.

In women, even the modest amount of testosterone present in normal physiology appears to help maintain lean mass. Women who become androgen-deficient after surgical removal of both ovaries, without testosterone replacement, have been reported to lose lean mass faster than women with intact ovarian androgen production, though this is based on smaller observational studies and should be treated as suggestive rather than definitive.

Progesterone, fluid, and why cycle timing matters for scan comparability

Progesterone rises through the luteal phase of the menstrual cycle (roughly the two weeks before menstruation) and promotes sodium and water retention. Because DEXA lean soft tissue includes body water, not just muscle, a cycle-driven fluid shift can move a DEXA fat-mass and lean-mass reading without any real change in fat or muscle tissue. Studies that scan the same women repeatedly across cycle phases have generally found the largest apparent fat-mass readings around the mid-luteal phase and the smallest in the early follicular phase, a pattern that is easy to mistake for real fat gain or loss between two visits scheduled at different cycle points.

Practical implication: serial DEXA scans in premenopausal women should be scheduled at a consistent cycle phase, ideally days 1 to 7 (early follicular phase), and the cycle day should be recorded on the scan order so anyone reviewing the trend later can account for it.

Normal and "optimal" body fat ranges are not the same target

Normal ranges describe how a population is distributed. Optimal ranges describe the body composition associated with the lowest risk of metabolic disease and functional decline, which is a narrower and more debated concept. The American College of Sports Medicine (ACSM) publishes age- and sex-stratified fitness-category body fat ranges that are widely used in clinical and fitness settings; the exact percentage cutoffs below are commonly cited in secondary sources, but readers relying on them for clinical decisions should confirm the current figures directly from ACSM's published guidelines, since editions are periodically updated.

The ranges are sex-specific because normal female reproductive function requires a minimum essential body fat fraction that is higher than the male equivalent. Values well below the low end of the female range, especially in athletes, can be associated with menstrual irregularity and reduced bone density and are not a target for most patients.

Visceral fat area

Visceral adipose tissue (VAT) is more strongly associated with cardiometabolic risk than total body fat percentage in most of the epidemiological literature. A VAT area threshold around 100 cm² is commonly cited as a lower-risk cut point, with substantially higher insulin resistance prevalence reported at larger VAT areas in some population studies. These specific thresholds originate in cross-sectional cohort data and should be treated as population-level risk markers, not individual diagnostic cutoffs; an individual patient's VAT number needs interpretation alongside waist circumference, metabolic labs, and family history.

Women generally show lower VAT than men at a similar BMI during the reproductive years, and that difference narrows after menopause as estrogen declines.

Appendicular lean mass index (ALMI) and sarcopenia

Sarcopenia screening commonly uses DEXA-derived appendicular lean mass index, calculated as appendicular lean mass in kilograms divided by height in meters squared. The Foundation for the National Institutes of Health (FNIH) Sarcopenia Project proposed sex-specific low-lean-mass thresholds based on pooled cohort data, using absolute cut points rather than a percentage of body weight, specifically because percentage-based thresholds make taller people look falsely lean. The exact numeric thresholds are widely reprinted on DEXA reports; readers making a clinical determination based on a borderline ALMI value should confirm the current FNIH-recommended cutoff with their clinician rather than relying on a number pulled from a secondary source.

GLP-1 medications and sex-specific body composition tracking

GLP-1 receptor agonists (such as semaglutide and liraglutide, prescribed for weight management and type 2 diabetes) are now one of the most common reasons a clinician orders serial DEXA scans, because the clinical question is rarely just "how much weight was lost" but "how much of that weight was fat versus muscle."

What the major trials showed, with the honest caveat

DEXA substudies of the semaglutide STEP program and related GLP-1 trials have reported that the large majority of weight lost is fat mass rather than lean mass, which is reassuring, but a meaningful minority of the total weight lost is lean tissue. The exact percentage split, and whether it differs by sex in a clinically important way, should be verified against the specific trial publication before being quoted as a precise figure, since summaries of this kind are one of the more commonly garbled numbers in body composition writing. What is well established, independent of the exact percentage, is that some lean mass loss accompanies GLP-1-driven weight loss, and that this loss is large enough in absolute terms (kilograms, not grams) to matter for strength and function in older adults.

Why sex hormones are relevant here

Estrogen appears to play a protective role in muscle protein turnover, which is the mechanistic reason some clinicians are more cautious about lean mass loss in postmenopausal women on GLP-1 therapy who are not also on estrogen therapy, compared with premenopausal women. This is a plausible, biologically grounded concern rather than a directly proven clinical outcome from a dedicated trial, and it should be described to patients as such. It is a common enough reason that some longevity-oriented clinicians pair GLP-1 therapy with a resistance training plan and a protein intake target, though the exact optimal protein target for this specific population has not been established by a dedicated trial and current recommendations are extrapolated from general sarcopenia-prevention literature.

A reasonable monitoring interval

Because the minimum change that clinical DEXA machines can reliably detect for lean mass is on the order of roughly a kilogram, scanning more often than every few months is likely to produce noise rather than a meaningful signal. A baseline scan before starting GLP-1 therapy, followed by a repeat scan at 6 months and again at 12 months during active dose titration, is a reasonable and commonly used interval, though this is a matter of clinical judgment rather than a formal guideline recommendation specific to GLP-1 monitoring.

Is this DEXA change real, or is it artifact? A working decision framework

Use this when comparing two DEXA scans and deciding whether a fat-mass or lean-mass change reflects real tissue change or a confound.

These DEXA results should serve as a basis for discussing bone health with your healthcare provider rather than standing in for that discussion, and they cannot take the place of personalized assessment, bone density interpretation, or treatment planning.

Hormone therapy and DEXA: distinguishing treatment effect from artifact

Estrogen therapy in postmenopausal women

Observational and trial data, including analyses from the Women's Health Initiative, suggest that estrogen-based hormone therapy is associated with slower lean mass loss and slower visceral fat accrual in postmenopausal women, largely by redistributing fat from visceral to subcutaneous depots rather than eliminating it. This redistribution is generally considered metabolically favorable, since subcutaneous fat is less strongly linked to insulin resistance than visceral fat. Exact effect sizes vary across studies and depend on the specific formulation, dose, and route of estrogen used, and should be confirmed against the specific paper before being cited precisely.

Testosterone therapy in men

The Testosterone Trials, a randomized placebo-controlled study in older men with confirmed low testosterone, reported increases in lean mass and decreases in fat mass on DEXA over about a year of treatment. Endocrine Society guidance recommends periodic body composition monitoring for men on testosterone therapy, though the exact recommended interval should be confirmed against the current guideline text rather than assumed.

Testosterone therapy in women

Low-dose testosterone is used off-label in some women for hypoactive sexual desire disorder or after surgical removal of both ovaries. Reported lean-mass effects in this population are modest, consistent with the intentionally low, sub-male-range doses used. This is an off-label use in the United States, and patients considering it should discuss risks, monitoring, and alternatives with a clinician experienced in this area.

Getting a comparable scan: practical standardization

Small procedural inconsistencies compound across serial scans and can create the appearance of a body composition change that never happened.

• Hydration and food timing. A meaningful difference in hydration status between two visits can shift a lean-mass reading. A common protocol is a 3-hour fast from food and caloric beverages, with normal water intake, before the scan.
• Same machine, same protocol. Different DEXA manufacturers (Hologic and GE Lunar are the two most common in the United States) use different underlying algorithms and are not directly interchangeable for absolute values. Serial tracking should use the same scanner model, ideally the same physical unit, for any comparison to be meaningful.
• Clothing, metal, and positioning. Metal objects, thick clothing, and inconsistent limb positioning introduce measurement error. A light gown, no metal, and a standardized position (arms at the sides, legs in neutral rotation) reduce this source of noise.
• Cycle-phase documentation for premenopausal women. Recording the menstrual cycle day on every scan order costs nothing and removes one of the most common sources of misread trend data.

Where DEXA fits among body composition methods

DEXA is one option among several, and its relative strengths matter for choosing the right tool for a given question.

DEXA versus bioelectrical impedance analysis (BIA). BIA devices, from consumer scales to clinical multi-electrode units, are faster and cheaper than DEXA but are generally reported to have a wider test-retest error for fat mass and are more sensitive to hydration status. Because GLP-1 therapy and hormone therapy both produce real changes in body water, BIA is a weaker tool than DEXA for tracking lean mass through these treatments specifically.

DEXA versus CT and MRI. Single-slice CT at the L4 vertebral level remains the reference standard for visceral fat quantification and cross-sectional muscle area, and DEXA's visceral fat estimate is a modeled approximation rather than a direct measurement, with reported correlation to CT that is good but not perfect. For research-grade visceral fat quantification, CT remains preferred. For routine, repeated clinical monitoring, DEXA's much lower radiation dose makes it the more practical and repeatable choice.

What is established, what is plausible, and what is not established

Established: Men and women differ systematically in fat distribution and lean mass, driven substantially by estrogen and testosterone. Menstrual cycle phase can shift DEXA fat-mass and lean-mass readings through fluid retention, independent of any change in tissue. Different DEXA scanner models are not directly interchangeable. Testosterone therapy in hypogonadal men increases lean mass and decreases fat mass on DEXA in randomized trial data. GLP-1 therapy causes weight loss that is mostly, but not entirely, fat mass.

Plausible but not firmly established by dedicated trials: That postmenopausal women without estrogen therapy lose disproportionately more lean mass than premenopausal women on GLP-1 therapy specifically. That a particular protein intake target or resistance training prescription optimally prevents GLP-1-associated lean mass loss. That a specific ALMI or VAT cutoff should trigger a specific clinical action for an individual patient rather than serving as a population risk marker.

Not established from the material available here: Precise numeric effect sizes for many of the claims above (exact kilograms, exact percentages, exact years) that appear in secondary summaries of the underlying trials. Anyone relying on an exact number for a clinical decision should pull the original trial or guideline rather than a secondary citation.