DEXA T-Score vs. Z-Score: What Your Bone Density Numbers Actually Mean

How a DEXA Scan Produces Two Different Scores

A dual-energy X-ray absorptiometry (DEXA) scan measures bone mineral density in grams per square centimeter (g/cm²). That raw number alone does not tell you much. A BMD of 0.85 g/cm² at the femoral neck could be perfectly normal for a 35-year-old woman or a sign of significant bone loss in a 70-year-old man. The scan software converts that raw value into two standardized scores, each comparing you to a different reference group.

The T-score expresses how many standard deviations your BMD falls above or below the mean BMD of a young-adult reference population, typically White women aged 20 to 29 from the National Health and Nutrition Examination Survey (NHANES III) database [1]. The Z-score does the same math but substitutes a reference group matched to your age, sex, and sometimes ethnicity. Both scores use the same underlying BMD measurement. The difference is entirely in which comparison group the software selects [2].

A standard deviation is roughly 10 to 12 percent of peak bone mass. So a T-score of −2.0 means your BMD sits about 20 to 24 percent below peak. That distinction matters because fracture risk approximately doubles with each standard deviation decline below the young-adult mean, according to a meta-analysis of 12 prospective cohorts published in the BMJ (N = 39,000+) [3].

T-Score: The Diagnostic Standard After Age 50

The World Health Organization established T-score thresholds in 1994, and they remain the global diagnostic standard for postmenopausal women and men aged 50 and older [4]. The cutoffs are straightforward.

A T-score of −1.0 or above is classified as normal bone density. A T-score between −1.0 and −2.5 indicates osteopenia, a zone of reduced density that carries moderate fracture risk. A T-score at or below −2.5 meets the WHO definition of osteoporosis. If a patient has both a T-score at or below −2.5 and one or more fragility fractures, the diagnosis becomes severe (or established) osteoporosis [4].

These thresholds were calibrated so that roughly 30 percent of postmenopausal White women would meet the osteopenia definition and about 16 percent would meet the osteoporosis definition, figures that aligned with the lifetime fracture prevalence observed in epidemiologic studies at the time [5]. The International Society for Clinical Densitometry (ISCD) recommends using the lowest T-score from the lumbar spine, femoral neck, or total hip for diagnosis [6]. A patient with a spine T-score of −1.8 and a femoral neck T-score of −2.6 is diagnosed with osteoporosis based on the femoral neck value.

One point that trips up many patients: the WHO thresholds apply only at the spine, hip, and femoral neck. Peripheral sites like the forearm or heel use different reference data, and a T-score of −2.5 at the calcaneus does not carry the same diagnostic weight as −2.5 at the femoral neck [6].

Z-Score: Why Age-Matched Comparison Matters for Younger Patients

For premenopausal women, men younger than 50, and children, the ISCD position is clear: do not use T-scores to diagnose osteoporosis [6]. A 32-year-old woman who has not yet reached peak bone mass may have a T-score of −1.5 simply because she is still accruing bone. Labeling her with osteopenia based on a comparison to peak values would be misleading.

The Z-score answers a different question. Instead of asking "How do you compare to a healthy 30-year-old?", it asks "How do you compare to other people your age?" A Z-score of −2.0 or lower is reported as "below the expected range for age," signaling that something beyond normal aging may be driving bone loss [6]. That finding should prompt a secondary workup for causes like vitamin D deficiency, celiac disease, hyperparathyroidism, hyperthyroidism, chronic glucocorticoid use, or hypogonadism [7].

Dr. E. Michael Lewiecki, director of the New Mexico Clinical Research & Osteoporosis Center and past president of ISCD, has stated: "A very low Z-score is a red flag. It tells you this patient is losing bone faster than their peers, and you need to find out why" [8]. That clinical instinct holds true regardless of the patient's sex. Even in postmenopausal women, a Z-score of −2.0 or below should raise suspicion for a secondary cause, even though the T-score is the value used for the formal diagnosis [7].

In pediatric patients, the situation is even more specific. Because children and adolescents are still building bone, DEXA results should be reported as Z-scores adjusted for height or body size, not raw Z-scores, to avoid overdiagnosing short or small children [6].

How the Same BMD Can Produce Very Different Scores

Consider a 45-year-old premenopausal woman with a lumbar spine BMD of 0.88 g/cm². If the young-adult reference mean is 1.05 g/cm² with a standard deviation of 0.11 g/cm², her T-score calculates to roughly −1.5. That would technically fall in the osteopenia range. But if the age-matched mean for 45-year-old women is 0.95 g/cm² with a standard deviation of 0.10 g/cm², her Z-score is −0.7, which is well within normal limits for her age.

The T-score makes her look worse than she is for her life stage. The Z-score contextualizes her density against realistic expectations. This is exactly why the ISCD mandates Z-score reporting for younger populations [6].

The reverse scenario also matters. A 72-year-old woman with a femoral neck BMD of 0.62 g/cm² might have a Z-score of −0.5, suggesting she is losing bone at about the same rate as her peers. Her T-score, however, could be −2.8, placing her squarely in the osteoporosis range. Her fracture risk is high regardless of what her peers are doing. The T-score captures that absolute risk. The Z-score simply tells her she is aging "normally" for her cohort, which is not reassuring when that cohort has a one-in-three lifetime fracture probability [9].

FRAX and How T-Scores Feed Into Fracture Risk Calculators

A T-score alone does not determine treatment. The FRAX calculator, developed by the University of Sheffield and endorsed by the WHO, integrates femoral neck T-score with clinical risk factors (age, sex, BMI, prior fracture, parental hip fracture, glucocorticoid use, rheumatoid arthritis, smoking, and alcohol intake) to generate a 10-year probability of major osteoporotic fracture and hip fracture [10].

The National Osteoporosis Foundation (now the Bone Health & Osteoporosis Foundation) treatment threshold is a 10-year major osteoporotic fracture probability of ≥20% or a hip fracture probability of ≥3% [11]. A patient with a T-score of −1.8 (osteopenia, not osteoporosis) who also has a prior vertebral fracture and glucocorticoid use may cross these thresholds and warrant pharmacologic treatment. A patient with a T-score of −2.6 but no other risk factors may fall below them.

FRAX uses femoral neck T-score specifically, not spine or total hip [10]. This is a common source of confusion. If a patient's worst site is the spine, FRAX may underestimate their risk. The ISCD acknowledges this limitation and allows clinicians to use clinical judgment in such cases [6].

When to Get a DEXA Scan

The U.S. Preventive Services Task Force (USPSTF) recommends bone density screening for all women aged 65 and older and for younger postmenopausal women whose fracture risk equals or exceeds that of a 65-year-old White woman (approximately a 9.3% 10-year risk of major osteoporotic fracture by FRAX) [12]. For men, the USPSTF found insufficient evidence to recommend universal screening, though the Endocrine Society recommends screening men aged 70 and older, or at age 50 to 69 if clinical risk factors are present [13].

The ISCD also recommends DEXA for any adult with a fragility fracture, any adult with a disease or condition associated with bone loss, anyone taking a medication associated with bone loss (such as glucocorticoids, aromatase inhibitors, or androgen deprivation therapy), and anyone being considered for pharmacologic treatment of osteoporosis [6].

Younger patients should not be overlooked. A case series from the Hospital for Special Surgery documented that 22% of female athletes evaluated for stress fractures had Z-scores below −1.0, with relative energy deficiency in sport (RED-S) as the most common underlying cause [14]. Early DEXA screening in this group can catch bone loss while the window for full recovery is still open.

Tracking Changes Over Time: Least Significant Change

Serial DEXA scans are used to monitor treatment response or continued bone loss. The critical concept is the least significant change (LSC), the minimum change in BMD that exceeds measurement error. At most centers, the LSC at the lumbar spine is 3 to 5 percent and at the total hip is 4 to 6 percent [6]. A patient whose spine BMD changes by 1.5% between two scans has not shown a statistically meaningful change, regardless of whether the T-score shifted from −2.4 to −2.5.

The ISCD recommends follow-up DEXA no sooner than one year after starting therapy, and typically every one to two years thereafter [6]. For untreated patients, intervals of two years are common, though shorter intervals may be appropriate for patients on glucocorticoids or with rapidly progressive conditions. Scans should be performed on the same machine each time, as cross-calibration between different manufacturers (GE Lunar, Hologic, Norland) can introduce systematic differences of 5 to 15% in raw BMD values [15].

Common Misunderstandings About DEXA Scores

The first misconception is that a T-score of −2.5 means bones are "half gone." It does not. A T-score of −2.5 represents roughly a 25% reduction from peak bone mass, not a 50% loss. The second is that Z-scores are less important than T-scores. For younger patients, Z-scores are the only clinically appropriate metric. A low Z-score should trigger a thorough medical evaluation, not just a recommendation to drink more milk.

The third misunderstanding involves vertebral artifacts. Degenerative changes, aortic calcification, and compression fractures can falsely raise spine BMD, making the T-score appear better than it truly is [15]. This is particularly common in patients over 65, which is one reason the ISCD recommends using the hip if the spine values seem discordant with clinical findings. If the spine T-score is −0.5 but the hip T-score is −2.7, the spine result is almost certainly artifactual.

A fourth issue is precision error compounding with patient positioning. If the patient is rotated during a hip scan, the projected area changes and the BMD calculation shifts. Certified densitometry technologists are trained to minimize this error, but patients should ask whether their scan center follows ISCD positioning protocols [6].

Practical Steps After You Get Your Results

If your T-score is −1.0 or above and you are postmenopausal or over 50, you are in the normal range. Standard bone-health maintenance applies: 1,000 to 1 to 200 mg of calcium daily (preferably from food), 600 to 800 IU of vitamin D daily (more if deficient), weight-bearing exercise, and fall prevention [11].

If your T-score is between −1.0 and −2.5, your clinician should calculate your FRAX score. Depending on the result, treatment might range from lifestyle optimization alone to pharmacologic therapy with a bisphosphonate such as alendronate (70 mg weekly) or risedronate (35 mg weekly) [16].

If your T-score is −2.5 or below, or you have had a fragility fracture regardless of T-score, pharmacologic treatment is indicated. Options include oral bisphosphonates, intravenous zoledronic acid (5 mg once yearly), denosumab (60 mg subcutaneously every 6 months), and anabolic agents like teriparatide (20 mcg daily) or romosozumab (210 mg monthly for 12 months) for patients at very high fracture risk [16].

If your Z-score is −2.0 or lower at any age, expect your doctor to order labs including serum calcium, phosphorus, alkaline phosphatase, 25-hydroxyvitamin D, intact PTH, TSH, complete blood count, comprehensive metabolic panel, and possibly celiac serologies and 24-hour urine calcium [7]. The goal is to find and treat the secondary cause before the bone loss becomes irreversible.

Repeat DEXA at the same facility on the same machine, ideally with the same technologist, at intervals your clinician specifies based on your risk profile and treatment status [6].