Can a DEXA Scan Be Wrong? Causes & Next Steps

DEXA scans can absolutely produce misleading results, and the ways this happens are more varied than most people realize. The technology is considered the gold standard for measuring bone mineral density, but “gold standard” does not mean infallible. Everything from how your leg was positioned on the table to which brand of scanner the clinic uses to whether you drank a bottle of water in the waiting room can shift the numbers enough to change a diagnosis. Understanding the specific sources of error gives you the leverage to question a result that does not match the rest of your clinical picture, and to take practical steps if something seems off.

How Precise Is a DEXA Scan to Begin With

Before getting into outright errors, it helps to know that DEXA has a built-in margin of wobble even under ideal conditions. Researchers quantify this with something called precision error, which is essentially how much the reading would vary if you scanned the same person twice in a row without anything changing. One study reported short-term precision errors of about 1.6% at the lumbar spine, 2.4% at the femoral neck, and 1.1% at the total hip.¹ Those percentages might sound small, but they matter when you are trying to detect genuine bone loss over time, because annual bone density changes can be in a similar range. This is why clinicians use a threshold called the least significant change: the smallest shift that can be trusted as real rather than noise. At the femoral neck, for instance, that threshold can be around 6.7%, meaning a change smaller than that might just be measurement wobble.

The practical upshot is that a single DEXA scan is a reasonable snapshot, but small differences between two scans do not necessarily mean your bones have gotten better or worse. The scan’s inherent imprecision sets a floor under the kinds of errors we are about to discuss, and the other sources of error stack on top of it.

Positioning on the Table Makes a Bigger Difference Than You Would Think

The most common and most preventable source of DEXA error is how you are positioned during the scan. Your hip, in particular, is sensitive to rotation. Research has shown that the femoral neck gives its lowest reading at a neutral position and reads higher when the leg is rotated inward or outward.² That means if a technologist does not secure your foot properly, you could get a falsely elevated hip reading one visit and a falsely low one the next, creating the illusion of bone loss or gain where none occurred. Using a proper positioning device can cut this precision error roughly in half.

A large descriptive study found that hip rotation errors affected about one in five scans, and spinal mispositioning showed up in roughly 4% of cases. Older adults fared worse on both counts: hip rotation errors appeared in nearly a quarter of scans in the older group compared to about 16% in younger adults, and spinal mispositioning rates were higher as well.³ This is partly because older bodies are harder to position consistently due to stiffness, joint replacements, and postural changes. If you have had two scans at different facilities and the results seem contradictory, positioning differences are one of the first things worth investigating.

Switching Machines Can Change Your Diagnosis

Not all DEXA scanners read bone the same way. The two dominant manufacturers produce machines that use different calibration methods, and the numbers they output are not directly interchangeable. In a head-to-head comparison of healthy adults scanned on both systems, one brand consistently measured bone density 14% to 20% lower at the lumbar spine and 8% to 17% lower at the hip compared to the other.⁴ A 14-20% difference in spine BMD is enormous in clinical terms; it could easily shift someone from a “normal” classification on one machine to an “osteopenia” classification on another.

The same pattern shows up in body composition measurements. Lean mass readings differed by 4% to 10% between the two systems, and the gap widened at the extremes of body fat percentage.⁵ Cross-calibration equations exist that allow clinicians to translate results between systems, but they are not always applied in routine clinical practice. The straightforward advice: try to get follow-up scans on the same machine, ideally at the same facility. If you have to switch, make sure whoever interprets your results knows about the change and applies the appropriate conversion.

What Your Body Brings to the Scanner

DEXA measures density by passing two X-ray beams of different energies through your body and using the absorption difference to distinguish bone from soft tissue. Anything that changes the soft tissue the beams pass through can throw off the calculation.

Body fat is the most studied confounder. As body fat increases, the precision error of DEXA measurements rises with it. Research has found that the least significant change threshold climbs in step with BMI class, meaning that for people with higher degrees of obesity, a larger shift in measured BMD is needed before you can be confident the change is real.⁶ Simulation work has also shown that adding fat around the trunk can directly introduce errors in spinal BMD readings, because the extra tissue alters how the X-ray beams are absorbed.¹

Even something as mundane as hydration matters, at least for body composition analysis. Drinking about 500 mL of water (roughly a standard bottle) before a scan significantly inflated lean mass readings and slightly lowered body fat percentage in a study of 100 participants.⁷ The changes were small in absolute terms, but they illustrate how DEXA treats all non-bone tissue as a single “soft tissue” bucket and can be fooled by shifts within that bucket. If you are using DEXA to track body composition over time, consistency in hydration and meal timing before each scan matters more than most clinics mention.

Degenerative Disease and Structural Artifacts in the Spine

This is arguably the most clinically consequential source of error, and it disproportionately affects the people most likely to need the scan. The lumbar spine is the standard site for DEXA measurement, but it is also where arthritis, disc narrowing, compression fractures, bone spurs, and calcified aortic deposits accumulate with age. All of these add extra dense material to the area the scanner reads, which makes bone density look higher than it truly is.

A clinical review noted that a finding of high BMD on a routine scan most commonly reflects degenerative disease rather than genuinely strong bones, and that elevated BMD can paradoxically occur in conditions where fracture risk is actually increased.⁸ In other words, a reassuringly “normal” spine reading in an older adult may be masking genuine osteoporosis because calcified arthritis is propping up the number. This is one reason many clinicians give more weight to the hip measurement in older patients, since the hip is less affected by these overlying artifacts.

Researchers have documented five distinct mechanisms that cause T-score discordance between the hip and spine: physiologic (natural variation in how different skeletal sites respond to loading and aging), pathophysiologic (disease effects), anatomic (body size and shape differences), artifactual (the degenerative artifacts just described), and technical (positioning and equipment).⁹ Getting different categories at different body sites is not rare, and it does not necessarily mean one measurement is “wrong.” But when the spine reads normal while other clinical signs point toward fragility, the artifact explanation should be on the table.

Metal Hardware and Contrast Agents

If you have metal in your body near the scan site, that is an obvious concern. Phantom testing, however, has shown that a steel implant adjacent to (but not overlapping) the measurement area does not significantly affect DEXA readings.¹⁰ The key word is “adjacent.” If metal directly overlaps the region of interest, such as a spinal fusion cage sitting right in the lumbar vertebrae being measured, it will absolutely inflate the reading, and the technologist needs to exclude that area or scan a different site. The takeaway is nuanced: nearby hardware is fine, overlapping hardware is a problem, and the distinction depends on careful scan setup.

Contrast agents used in CT scans are a more clear-cut issue. A study found that DEXA measurements performed shortly after a contrast-enhanced CT showed dramatically distorted results: whole-body bone mineral content jumped to about 125% of baseline, fat mass dropped to about 75%, and lean mass rose to about 110%. These distortions disappeared by seven days later.¹¹ Nuclear medicine tracers and MRI contrast, by contrast, did not cause significant interference. If you have had a contrast CT recently, wait at least a week before getting a DEXA scan.

The Reference Database Problem

Your DEXA result is not just a raw measurement; it is compared against a reference population to produce a T-score or Z-score. Which reference population the software uses can shift the diagnosis. Research has shown that applying a Caucasian reference database to East Asian patients produces systematically lower T-scores, leading to overdiagnosis of osteoporosis. Conversely, using a female reference database for male patients leads to higher T-scores and underdiagnosis.¹²

This is not an error in the machine or the technologist; it is an error in the interpretive framework. Two patients with identical raw BMD could receive different diagnoses depending on which reference population their clinic’s software defaults to. If you belong to a population that may not be well represented in the default reference database, this is worth discussing with your provider. It is a particularly common issue for people of East Asian, South Asian, and Hispanic descent, and for men in general, since osteoporosis research has historically been built around postmenopausal white women.

Technologist Skill and Interpretation Variability

The human element is easy to underestimate. A review in a rheumatology journal put it plainly: there is a high degree of variability in the skills of technologists performing DEXA tests and the clinicians interpreting them, and poor-quality testing and reporting can lead to inappropriate clinical decisions, ranging from unnecessary treatment to treatment being withheld when it is actually needed.¹³ These are not hypothetical risks. They include selecting the wrong vertebrae for the region of interest, failing to exclude artifact-dense areas, misidentifying the femoral neck boundaries, and entering incorrect patient demographics that shift the reference comparison.

Quality control extends beyond the moment of scanning. Each step of the DEXA process, from daily phantom calibration to patient setup to edge detection by the software to final report generation, is subject to potential errors and artifacts that require attention from both the technologist and the interpreting physician.¹⁴ A well-run densitometry program with certified technologists and systematic quality assurance will produce far more reliable results than a facility that treats DEXA as a simple push-button test.

What to Do If You Suspect Your Result Is Wrong

If a DEXA result does not match your clinical situation, whether it is an unexpected diagnosis of osteoporosis in a healthy young athlete, a suspiciously normal spine in someone who has already had a fragility fracture, or a dramatic change from a prior scan, several concrete steps can help clarify what happened.

Request the images: Ask your provider to review the actual scan images, not just the summary report. Many errors become visible when someone looks at the image: vertebral artifacts, incorrect region-of-interest placement, rotation issues, and overlying hardware are all detectable by eye.
Confirm the machine: If your new result was done on a different scanner than your baseline, the numbers are not directly comparable without cross-calibration. Ask whether the systems are the same manufacturer and model.
Check the reference database: Verify that the correct sex and, if available, ethnicity-specific reference population was used for your T-score calculation.
Consider the least significant change: If you are being told your bone density has dropped, ask whether the change exceeds the facility’s calculated LSC. A small decline that falls within the precision error range is not evidence of bone loss. The commonly used threshold requires the change to exceed about 2.77 times the precision error before it can be called real with 95% confidence.¹⁵
Repeat the scan: A simple rescan at the same facility, ideally on the same machine, can resolve many questions. Whether the follow-up is done by the same or a different technologist does not appear to have a clinically significant impact on precision.¹⁶

When Complementary Tests Fill the Gaps

DEXA measures density, which is essentially how much mineral is packed into a given area of bone. But density is not the whole story of bone strength. Bone architecture, the internal scaffolding of tiny struts and plates that gives bone its resilience, can deteriorate even when density looks adequate. This is why some people fracture despite having normal T-scores, and why clinicians sometimes look beyond DEXA for a fuller picture.

The trabecular bone score (TBS) is a texture analysis derived from the same lumbar spine DEXA image you already had. It provides information about the underlying bone microarchitecture that is independent of and complementary to the BMD number.¹⁷ A low TBS can flag elevated fracture risk even when BMD looks fine, and adding TBS to standard fracture probability calculations produces a small but statistically significant improvement in risk prediction.¹⁸ TBS is especially useful in conditions where bone quality deteriorates out of proportion to density, such as in diabetes, glucocorticoid use, and other causes of secondary osteoporosis.¹⁹

Blood markers of bone turnover offer another angle. Organizations focused on bone health have identified specific markers in blood as reference tools for predicting fracture risk and monitoring how someone responds to osteoporosis treatment.²⁰ These markers can detect changes in bone metabolism within weeks of starting treatment, far faster than DEXA, which typically needs one to two years between scans to detect real change. They do not replace DEXA, but they give your clinician a faster feedback loop and an independent data point when the scan results seem questionable.

CT as an Alternative or Supplement

Standard CT scans that you might already have for other reasons can provide useful bone density information without a separate DEXA appointment. CT technology has advanced to the point where Hounsfield unit values from routine abdominal scans can serve as an opportunistic screen for osteoporosis. The correlation between these CT-derived values and DEXA T-scores is moderate, and mean Hounsfield unit values differ meaningfully between groups classified as normal, osteopenic, and osteoporotic.²¹ This kind of opportunistic screening is not as precise as a dedicated DEXA scan, but it can catch cases that would otherwise go undetected in patients who were getting a CT for unrelated reasons.

Quantitative CT, a more formal application, measures volumetric bone density rather than the areal (two-dimensional) density that DEXA reports. This distinction matters because DEXA’s areal measurement is influenced by bone size: larger bones look denser even if the mineral concentration per unit volume is the same. Volumetric CT avoids this issue and is also less affected by overlying degenerative artifacts. The trade-off is a higher radiation dose and greater cost, which limits its use in routine screening. But when a DEXA result is genuinely puzzling and clinical decisions hinge on the answer, quantitative CT can serve as an independent check that approaches the question from a fundamentally different direction than DEXA does.