DEXA scan radiation dose: how it compares with a CT scan

A whole-body DEXA scan radiation dose came out at 8.47 to 17.68 microsieverts per scan in a computer simulation of patients aged 5 to 18 [2]. A microsievert is a millionth of a sievert, the unit used for how much ionising radiation a body has taken in, and DEXA - dual-energy X-ray absorptiometry, a scan that measures bone density and body composition by passing two weak X-ray beams through you - does use X-rays. The dose is not zero.

The same simulation added up what happens to someone scanned every year from age 5 to 18: 180 microsieverts in total for girls and 168 for boys [2]. It also modelled the extra lifetime cancer risk those fourteen scans would carry and described it as minimal, while noting that girls came out at a higher risk than boys under the same protocol [2].

For scale, a CT coronary angiogram - a computed tomography scan that photographs the heart's arteries - was measured at 10.7 to 14.2 millisieverts depending on how the scan was timed to the heartbeat, and 18.8 millisieverts when that timing was not used to trim the dose [5]. One millisievert is 1,000 microsieverts, so a single bone density scan sits about a thousandfold below one cardiac CT.

Radiation has been tied to cancer in people, but at CT-scale doses. Among 3.7 million children followed in six American health systems and in Ontario, the chance of a blood cancer climbed with the dose absorbed by the bone marrow: about 40% more likely at 1 to 5 milligray, and about three and a half times as likely at 50 to 100 milligray, where a milligray measures the energy one organ absorbs and a single head CT came to 13.7 milligray [13]. In plain counts that worked out to about 26 extra blood cancers for every 10,000 children exposed to at least 30 milligray, by the age of 21 [13].

Two earlier cohorts point the same way. In 178,604 British children, cumulative doses of about 50 milligray to the bone marrow might almost triple leukaemia risk and about 60 milligray to the brain might triple brain cancer risk, which in children under 10 still came to roughly one extra case of each per 10,000 head CT scans over the following decade [14]. A nine-country European cohort of 658,752 people found brain cancer risk rising with dose, at an average cumulative brain dose of 47.4 milligray [15].

X-rays are ionising, meaning they carry enough energy to knock electrons off the atoms they pass through, and the wreckage of that collision damages DNA. The damage that matters most is a double-strand break, where both rails of the DNA ladder snap in the same place [9].

Cells mend those breaks two ways: a quick route that glues the loose ends together and makes mistakes, and a slower route that copies the missing stretch from a matching strand and is usually accurate [9]. A network of signalling proteins decides which route runs, and whether the cell pauses to repair or dies [11].

Cancer risk comes from mis-repair. A break stitched back wrongly leaves a permanent change in the genome, and enough of those changes in the wrong genes can start a tumour [8]. Radiation works less directly too: it can shift DNA methylation, the chemical tags that switch genes on and off without altering the sequence [12], and it can produce mutations in neighbouring cells the beam never hit, through signals passed between cells [8].

None of this biology identifies a dose low enough to be harmless. Radiation protection assumes every exposure counts for something, which is why bone measurement in patients who need it repeated for years - young people, and women who are pregnant or nursing - is the setting where radiation-free alternatives are being studied [6].

The dose figure is a simulation, not a measurement. The researchers fed scanner settings into software that models how X-ray energy spreads through a virtual body and calculated organ doses from that. Nobody put a detector on a patient [2].

It also covers the wrong people for most readers of this page. The evidence does not apply to:

• adults, because the simulated patients were aged 5 to 18 and children absorb radiation and respond to it differently
• anyone scanned on a different machine or a different protocol, since the dose depends on the settings used
• pregnant and nursing women, for whom X-ray-based bone measurement carries separate considerations [6]

No study has ever followed people who had DEXA scans and counted their cancers. The risk evidence here is borrowed from CT cohorts [13] [14] [15], all of them in children and young adults, whose tissues are damaged by radiation more easily than an adult's, at doses far above a bone density scan.

The units are not interchangeable either. DEXA doses are quoted as microsieverts spread across the whole body, while the CT cancer studies measured milligray absorbed by one organ, the bone marrow or the brain, so those two figures cannot be divided into each other. The comparison shows a difference in scale, not an exact ratio.

Longevity and preventative-medicine clinics offer DEXA scans, usually sold as a body composition measurement rather than a bone density test. A first visit is short: you take off anything metal, lie fully clothed on a flat table, and a scanning arm passes over you for a few minutes while the two X-ray beams separate bone, fat and lean tissue. Nothing is injected and there is no enclosed tube.

The question worth asking at the desk is how much this machine and these settings deliver, and how often the clinic expects to repeat the scan. A yearly body composition check is a different cumulative exposure from a quarterly one - the simulation above puts fourteen annual whole-body scans at 180 microsieverts in total [2].