Are full-body MRI scans worth it? Advantages and disadvantages

Are full-body MRI scans worth it? For a healthy adult with no symptoms, no study has shown that they are, and the two largest reviews of the evidence both end by saying the consequences are unknown [1] [2].

Start with what the scan finds. Across 12 studies covering 5,373 adults with no symptoms and no known disease, about 32 in every 100 had at least one finding rated critical or unclear [1]. Split apart, about 13 in 100 had a finding rated critical and about 14 in 100 a finding rated unclear [1]. The studies disagreed with each other a great deal, so read 32 as a rough middle rather than a precise number [1].

What the scan covers drives what it turns up. Protocols that included the heart, the blood vessels or the colon flagged something in about 50 of every 100 people, against about 23 of every 100 when they left those out [1].

Then the harder question: how many of those findings mattered? A separate pooling of 32 studies covering 27,643 adults put potentially serious findings at about 4 in every 100 people given a brain-and-body scan, rising to about 13 in 100 once findings of uncertain seriousness were counted as well [2]. Around half of the potentially serious ones were suspected cancers [2].

Of the 234 potentially serious findings that could be traced to a final diagnosis, 48 turned out to be serious and 186 did not [2]. The authors concluded that little is known about what such findings do to people's health, and called for the long-term follow-up that has not been done [2].

False alarms are part of the arithmetic. Across the six studies in the whole-body review that reported them, false positives came to about 16 in every 100 [1]. No study in that review went back five years or more to check whether a clean scan had really been clean [1].

Single studies land in the same place:

• 148 healthy volunteers in London, median age 36, meaning half were younger and half older: 29% of scans were abnormal, and 12.8% of the volunteers had a finding of clinical significance [4]
• 132 doctors in Hong Kong with no symptoms, average age 56: 94% had some positive finding, 18% went on to further tests, and 1.5% were found to have a proven cancer [5]
• 298 adults averaging 50 years old: the scan produced 169 findings worth reporting, and of the 128 that could later be checked, one was a false alarm and 41 were never resolved either way [3]

That is not how the scans are sold. Across 982 Instagram and TikTok posts about five popular tests, full-body MRI among them, 87% mentioned benefits and 15% mentioned harms [7]. Only 6% mentioned overdiagnosis, which means finding something real that would never have caused trouble if left alone [7]. About 84% of the posts were promotional in tone, and 68% of the accounts had a financial interest in the test [7].

Why does a scan of a healthy person find so much? Because it looks at everything and asks nothing.

Magnetic resonance imaging, or MRI, builds pictures from how water molecules behave inside a strong magnetic field. It shows soft tissue - organs, muscle, bone marrow - in fine detail, and it uses no ionising radiation, the kind of energy in X-rays that can damage DNA [10].

Those two properties are why whole-body MRI is established in cancer care, where the scan has a specific question to answer: mapping cancers that have spread to bone, and watching people who carry genes that raise their cancer risk [10].

Sharper machines find more, not less. Pooling 16 studies of 19,559 people with no neurological symptoms, brain scans run at high resolution turned up incidental findings in 4.3% of people against 1.7% at standard resolution [6]. You would have to scan 37 people to turn up one incidental finding of any kind, and those authors concluded that this was not enough to justify screening healthy people [6].

Bodies accumulate harmless oddities, and a scanner cannot sort those from the ones that matter. In the London volunteers, findings rose with both age and body weight: people with obesity were about five times as likely to have an incidental finding, though the range around that estimate was wide [4].

The evidence has a hole in the middle of it. Nobody has run the study that would settle the question: scan one group of healthy adults, leave another unscanned, and compare how long they live and how well [1] [2]. Every number on this page describes what a scan found, not what happened to the person afterwards.

The pooled figures are unstable as well. The studies inside both reviews disagreed with each other enough that the ranges around each headline percentage are wide, and the whole-body review calls its own verification data lacking [1] [2].

Vintage and make-up matter. The 298-person study is from 2005, ran at a single centre, and 247 of its 298 participants were men [3]. Its protocol covered four organ systems rather than the whole body clinics advertise today. The Hong Kong cohort was 132 doctors who volunteered, 111 of them men [5].

Who this evidence does not describe:

• people with symptoms or an existing diagnosis, for whom whole-body MRI has a defined clinical job [10]
• people carrying an inherited gene that raises cancer risk, who have their own guidance and their own numbers [9]
• women in anything close to equal numbers, in either of the two cohorts above [3] [5]
• adults in their twenties, and adults over 70, thinly represented across all of these studies [3] [4] [5]

The inherited-risk case is worth stating plainly, because it is the one place the scan has a clear job and the numbers are still mixed. European guidance recommends a yearly whole-body MRI for adults carrying a disease-causing variant of the TP53 gene, the change behind Li-Fraumeni syndrome, an inherited condition that makes cancer far more likely [9]. In 75 such adults given 325 scans over twelve years of surveillance, 17 cancers were diagnosed and the whole-body scan picked up 9 of them, 7 of those at an early stage [8]. It did not pick up the other 8, and benign findings in 89 of the 325 scans led to 53 extra investigations [8].

A 48-year-old with no symptoms pays for a private full-body scan, and the report comes back describing a small spot on a kidney.

The evidence can supply a proportion, not a verdict on that spot. In the pooled data, a finding that looks potentially serious turns out to be serious in roughly one case in five [2]. The other four end in a follow-up scan, sometimes a biopsy, and a wait.

What the evidence cannot supply is whether finding it changed anything, because the comparison that would answer that has never been run [1] [2].

The scan itself is quick: in the Hong Kong cohort it averaged 33 minutes [5]. The report is the long part, and what to do about a line in it depends on the person's own history and on a doctor who knows it.