Is there a risk to having a DEXA scan

Hi, I had a bone density scan a year ago, I suppose it is the same thing as a DEXA scan. I started treatment for Osteoporosis in the Spine and have a follow up appontment coming up soon.
I'm wondering how often we are supposed to have a repeat scan. Is there a risk from the radiation or is it really just a low dose that they use?
Another question; are there any other ways to scan for this for eg does it show on an MRI?

Report post

30 replies. Join the discussion

Very low dose.

Report post

There are three accepted methods for measuring or estimating bone density. DXA is by far the most accepted, has the best accuracy (unless you are taking large doses of strontium) and precision. IT also has established strong relationships with fracture risk, and the ability to monitor tiny changes in density over time (if the scans at both time points are done on the same machine, by the same skilled operator and the patient positioned in exactly the same manner). It does use a tiny bit of x-radiation. Put into perspective, the average radiation dose from natural radiation we receive at sea-level is about 3000 microSieverts per year. (A Sievert is a very large dose, and a micro-sievert is million times less).
The average DXA scan will give you between 1-10 microSieverts, or the equivalent extra radiation you get in a leap-year compared to a regular year. A 4 hour airplane trip at 30,000 feet is about 40 microSieverts. A mammogram is about 600 micro-Sieverts and a CT of the abdomen with contrast is about 1000. TO know how much risk from cancer there is from common radiation exams see:

It will vary with age as far as cancer risk as the older you are, the less risk of cancer because your risk of dying from something else first competes with it. The calculator needs you to enter the risks at time of exam.

QCT is the other radation based measure of density, and it is about 60 times higher than DXA at the spine, and 300 times higher at the hip because of the radiation to the gonads.

The only non-radiation method is ultrasound. It can measure something related to density. It measures SOS (Speed Of Sound) in meters/second, and BUA (Broadband Ultrasound Attenuation) in dB/MHz( decibels per megaHertz.) It is limited to the heel, fingers, calf and forearm.

Only DXA T-scores can be used with the -2.5 and -1.0 cutpoints.

No, MRI cannot measure bone density, but can measure structural changes and bone marrow edema that occurs at the sites of new fractures, and is often used to age compression fractures in the spine to separate something new from a football injury 30 years ago.

Report post


The risk of having a scan is hardly measureable by any technique. The risk of shortening your lifespan or reducing your mobility via an osteoporotic fracture is very measureable and far out weighs the risk of a BMD scan, via QCT or DXA.

dxauguru; QCT T-Scores of the hip ARE recognised by the World Health Organisation for diagnosis of normal bone density, osteopenia and osteoporosis (using the -1 and -2.5 thresholds). You may have noticed it as a selection on the WHO FRAX tool. QCT also has the ability to detect changes in the spine earlier than DXA as it only measures the more metabolically active trabecular bone, which is the first to deteriorate and the leading cause of vertebral crush fractures.

Another risk of having a DXA scan that is often swept under the carpet is the incidence of false negatives (ie the DXA scan can incorrectly give a normal result in a patient with low bone density due to other common conditions such as osteoarthritis, obesity, and scoliosis to name a few). There are members here with relatively normal DXA results who have suffered multiple fractures. That is not to suggest that DXA is inaccurate across the board, but if you suffer from one of the contraindications your results may be misleading.

Good luck with your BMD results Dormouse.

Report post

Thank you dxaguru and MWA, I feel much better knowing that the risk is hardly measurable and seeing the comparisons is most helpful. I have just had a closer look at my last test and always assumed it was a Bone Density scan but in fact it was an
OSTEODENSITOMETRIE BIPHOTONIQUE which is apparently not quite the same but in any case it uses 5 à 20 microsieverts per exam and so very small comparing it to your figures of other xrays.
That said I am now not happy about having my next mammogram!.

Report post

One risk in having a DXA scan is falling off the table and breaking your hip, wrist or femoral neck.

Report post

Ewcollins, that is the funniest thing that you have ever said! Thanks for the laugh. :)

Report post

Good one Ewcollins!!!

Dormouse, the risk/benefit for mammograms has been studied time and again, mammograms offer a clear benefit to higher risk segments of the population.

There are people out there who protest against immunisation as there is a minuscule risk associated, yet rampant disease is a far bigger problem and some refuse to acknowledge the sense in accepting the lesser of two evils. 1000 years ago there were no X-rays, drugs etc and people certainly did not live longer. Is today's technology perfect? No, but we are experiencing benefits from it.


Report post

MWA - Thank you for the clarification. Some QCT scanners do have hip software (many do not)and normative data that can be converted into DXA like grams/square centimeter from gram/cubic centemeter, and thus give equivalent T-scores. But this is true only for FRAX, and ony for the hip. Ths spine T-scores would over-diagnose osteoporosis and give too many false positives, the opposite of DXA false negatives in osteoarthritic spines. But hip DXA T-scores for FRAX use much less radiation to get the same thing and can be done to measure a single hip without the extra gonadal dose of QCT. Further, the extra skills required to properly reconstruct a 2D hip from 3D data, and the even more unlikely possibility that the QCT center would have established its own LSC's, makes finding a good QCT center even harder than finding a qualified DXA facility.
QCt in the right hands, has value. I dind;t mean to besmirch the technique, but only offer real world advice, and I see far more poorly reconstructed QCT hip scans than I do DXA scans.

And you are also correct in that there are some people with normal DXA scores that get osteoporotic fractures. But that is also true of QCT.
And while you can monitor changes with QCT, if DXA is also available and for most people with an intact hip, I don't think it is ethical to expose them to the higher radiation measurements serially to obtain something just as valid with 100 times less radiation dose, as amsall as it is.

Report post

For your information: OSTEODENSITOMETRIE BIPHOTONIQUE is French for Dual Energy X-ray Absorptiometry (DXA).

Report post

Well, how about that, I did have a DXA after all, thank you Dxaguru for your wealth of information.

Report post

Another risk is that the scan will be done incorrectly, and your insurance company will refuse to pay for it because it is "experimental".

Report post

I can't image why a QCT scan would be classified as "experimental", QCT was developed before DXA and has been an approved medical device in the USA since the 1980's. Medicare reimburses it for BMD measurement.

Report post

dxaguru, the difference in T-Scores measured at the spine via different methods are well documented. I can convey some of the details for you.

First of all, since trabecular bone loses density faster than cortical bone, it makes sense that QCT will detect a change sooner than DXA since QCT measures only trabecular bone - not a hybrid of cortical and trabecular bone like DXA does. Since all the models I have seen only demonstrate a decline in BMD after a peak in the late teens, I think QCT offers the patient an early warning and therefore an opportunity to make less drastic changes to help maintain or build BMD. A patient may be able to employ dietary changes and exercise program rather than using drugs which may have unappreciated side effects and a significant cost.

I see a lot of negative feedback here regarding the side effects of certain drugs, and a demand for a more natural approach. Late detection requires extreme measures - early detection allows more management options.

Now we move into the grey area of T-Scores and fracture predictions. The -2.5 and -1 thresholds were originally decided upon for risk of fracture at the hip but actually measured at the forearm, and later applied to the spine in a loose manner. When I say 'decided on,’ we have to remember that the osteoporosis and osteopenia thresholds are NOT a specific finding. For example, we can decide to label the boiling point of water at sea level 100 degrees Celsius and apply this as a standard as one will get consistent results globally boiling water at sea level, and so we can use that fact as a standard reference.

Supposing that every patient with a T-Score of -2.5 is osteoporotic was a subjective decision and one that is not a scientific standard. Someone had to draw a line in the sand and lead the way, and so it was done, right or wrong. They may well have picked -2.6 or -2.4 and I doubt much would have changed in the world of fracture reduction.

Now this gets interesting because different populations with identical T-Scores experience different fracture rates. So how constant and reliable are T-Scores? (What is a T-Score? - It is a statistical number, quantifying how many standard deviations a BMD is from the young mean normal). It is not a measurement of strength. It is a convenience - a comparison to average - since absolute DXA measurements vary across different populations.

SO, QCT spine T-Scores are different because they measure different bone! Should we apply a different T-Score threshold to QCT that is calibrated to DXA T-Scores? For that to be a good idea we would first need to be certain that DXA T-Scores are the best measurement standard.

Since BMD contributes to bone strength, a BMD scan is useful to determine how strong the bones are. So why not measure the BMD directly and quote a BMD figure that can be standardised like blood test results and other scientific measurements?

The American College of Radiologists and other bodies have adopted thresholds for Osteoporosis and Osteopenia for Spine QCT using absolute bone mineral values. Under 120mg/cm3 has been determined indicative of osteopenia, and under 80mg/cm3 indicative of osteoporosis. These figures are applicable across different populations and can be measured directly without errors due to scoliosis, osteoarthritis, obesity, and other geometric anomalies that impact on DXA results.

At the hip, patients of any size can be scanned and accurate BMD values obtained using QCT since the soft tissue can be stripped away in the CT image very effectively and in a very reproducible manner. This is important as different patients have a different ratio of lean tissue, fat and bone. DXA can make measurements through two tissue types very well. Throwing three in there makes problems, and the DXA method of guessing the average ratio of fat and lean tissue, then applying that average degrades the accuracy of DXA. I guess that makes DXA accurate if you are average.

Your statement that the false negative of DXA are evened out by the false positives of QCT is not a logical conclusion, since one difference is due to using a different standard (QCT), the other due to inherent and inconsistent errors in measurement (DXA). The false positive of QCT you speak of simply require correct interpretation using the standards above. The raw data received for QCT measurements is very consistent across a wide range of patient shapes, patient sizes and in patients with common medical conditions which degrade DXA scans - making DXA unreliable for the masses.

In short, QCT offers patients an opportunity for early detection, and potentially less aggressive treatment. Standard thresholds for osteoporosis and osteopenia have been researched and can be applied with confidence. QCT offers advantages in accuracy for a larger percentage of the population.

A thorough physician would assess a patient and request QCT if there are any doubts that the patient may be contraindicated for DXA. And accordingly, patients without DXA contraindications should feel confident with their DXA results. Tarring all with the same brush is not a clever approach.


Report post

MWA - ALl your points about the accuracy of QCT are well taken, and are not in dispute. But better accuracy in measuring bone density does not also mean better prediction of fracture risk. We don't have any evidence that in severe osteoarthritis and scoliosis QCT's superior accuracy provides better correlation with fracture risk because those with osteoarthritis at the spine have lower fracture rates than those without OA. Apparently the scerolis from artritis that occures in the outer cortex of the vertebral body, which is NOT meaured by QCT, does contribute to bone strength and resistance to fractures which is measured by DXA but not QCT. There are two other errors in DXA that do not exist in QCT which you have correctly pointed out. DXA is affected by fat, and it is affected by bone size and QCT is immune to these two variables.

But QCT BMD at the hip, for example, doesn't predict fracture as well as DXA at the hip in obese individuals (especially women)precisely because it isn't affected by the increase in fat and tissue thickness. Fat falsely elevated DXA bone density. That is very true. There are plenty of studies that show some of the increase in BMD is obesity and some of the "loss" seen after bariatric weight loss is simply artifact. However, that fat which I admit is an error in DXA and not in QCT of the hip, provides fracture protection as most hip fractures occur after falls, and those who are walking around with a mattress of fat wrapped around their thighs are at much lower risk of hip fracture. The same for extremely thin persons, where DXA underestimates the true QCT derived bone density. But because of the lack of padding from fat, these people have higher fracture risks despite similar bone density to average subjects, and thus again, the error in DXA actually makes it work better at fracture risk protection at the hip.

Then there is this issue of grams per square centimeter versus cubic centimeter. DXA makes larger bones of the same QCT density appear more dense and smaller bones of the same QCT density appear less dense. That is an error in the technique for measuring true density. However, smaller bones of the same true density are not as strong, and larger bones are stronger and explain most of the reason why men have higher DXA BMD than women but not higher QCT BMD than women. So you would have to establish different fracture thresholds for QCT in men and women and in blacks and perhaps even other ethnicities rather than a single value. Thus again, the size error in DXA makes it a better predictor of bone strength and fracture risk. This has been validated in studies of in vitro studies of fracture resistance to loads where DXA correlated better than QCT.

Lastly, I am not disparraging the judicious use of QCT in appropriate populations. However, I don't think you are advocating that it replace DXA as the scan of choice for screening populations are you? The topic of this thread, which we have really gotten away from is, about radiation dose and potential harm. Even with the lowest dose technique, a single QCT is still 10-60 DXA scans worth of radiation dose for the spine and over a 100 DXA scans if you include a hip meausrement. So for a mass screening of high risk populations perspective, it is not ethical to use a higher dose test when a lower dose test can provide the same information for the greatest numbers of those screened. As for monitoring therapy, rightly or wrongly, DXA in the USA at least, is the only approved method recognized by Medicare. You may argue that QCT is the same if not better at this, but again, if the same results can be had with a lower dose technique, it is unethical to use one with a higher dose. If someone already had a CT scan for some other condition, you may be able to get some density data from it, and that is fine as the potential damage of the radiation was already done, but again, we are drifing away from the topic at hand. However low the radiation risk of QCT is, it is still much higher than DXA, and the benefits in a small proportion of the general population it may have does not offset the higher radiation risk for the rest.

Report post

Oh, I almost regards to osteoarthritic changes and DXA for spinal fracture risk. THere is a new software that measures the differences in textures of the DXA image that apparently is not influenced by arthritic changes called a trabecular bone score. It was recently approved by the FDA, and TBS can be added to most DXA unitsand provides much of the trabecular data of QCT, without any additional radiation dose, and the cost of the software is about the cost of adding QCT software to a CT scanner. IT may even be incorporated into FRAX and a number of researchers are actively pursuing its incorporation into the FRAX model. DXA plus TBS may be coming to a theatre near you soon. The trailer looks very promising.

Report post

Absolutely NO RISK..........Dexa is the same as a bone density. Very LOW radiation. You could probably have one a month and it wouldn't be a problem......notice the tech does not even go behind a lead wall................No an MRI would not help (why would you want to be in that tube?) MRI's are good for soft tissue mainly................bones show much better with a CT, but a Dexa Scan measures the density (strength of your bones) I've had osteoporosis for year............took Actonel and then Fosomax and then went off of everything for about 6 years until I fractured my femur! I am DEFINITELY back on meds! I just had my first injection of every six months..............just like a flu shot. No side effects so far...........too soon to see if it is stopping the progression.

Good luck to you.

Report post

QCT Scans are just a lot more expensive than a DEXA Scan.............Medicare will pay for both.

Report post

I'm curious KAR49, what makes you think that QCT scans are "a lot more expensive than a DEXA Scan"?

Report post

I was a hospital administrator for many years. But I have been retired for several, so I'm sure technology has changed as well as machinery and equipment, and the people who are trained to run them.

Report post

This quite an eye openener for me, so now I'm wondering if either the QCT or Dexa scanner or software can differentiate between oesteoporosis and osteomalacia?

Report post

This discussion is closed to replies. We close all discussions after 90 days.

If there's something you'd like to discuss, click below to start a new discussion.

Things you can do

Support NOF

Help the National Osteoporosis Foundation reach its goals and support people like yourself by making a donation today.

Donate to the National Osteoporosis Foundation

Learn more about osteoporosis awareness and prevention

Discussion topics

Links and resources from NOF

Community leaders


The National Osteoporosis Foundation would like to remind visitors and community members that the views and opinions expressed on this site are not necessarily those of NOF. Please consult your personal healthcare provider regarding any medical information that is shared on this site.