Juli Engel was delighted when a neurologist recommended a PET scan to determine whether amyloid — the protein clumps associated with an increased risk of Alzheimer’s disease — was accumulating in her mother’s brain.

“My internal response was, ‘Yay!’ ” said Engel, 65, who travels almost monthly from her home in Austin, Texas, to help her mother in Florida. “He’s using every tool to try to determine what’s going on.”

She checked off the warning signs: Her mother, Sue Engel, 83, has been financially exploited, suffered an insurance scam, caused an accident.

Medicare officials decided in 2013, shortly after PET (positron emission tomography) amyloid imaging became available, that they lacked evidence of its benefits. So outside research trials, Medicare doesn’t cover the scans’ costs ($5,000 to $7,000, the Alzheimer’s Association said); private insurers don’t, either.

Juli Engel thinks Medicare should reimburse for the scan, but “if necessary, we’ll pay for it out of pocket,” she said.

The Alzheimer’s Association and nuclear medicine experts said the criteria calls for PET scans to be used only in cases of unexplained or unusual symptoms and unclear diagnoses. But as evidence mounts that brain damage from Alzheimer’s begins years before people develop symptoms, people may start turning to PET scans to learn whether they have this biomarker.

They have few alternatives. Scientists at Washington University in St. Louis said they have developed a blood test for amyloid that can predict the development of plaques in the brain, but it is years away from use.

Some experts fear PET scans offer few benefits, at substantial costs. “There are lots of incentives, including financial incentives, for doing more testing and interventions,” said Dr. Kenneth Langa, a researcher at the University of Michigan. “My hope is we’ll think hard about the unintended downsides.”

Amyloid plaques occur commonly in older people’s brains, but not everyone with amyloid will develop dementia. Nor does a negative PET scan mean someone won’t develop dementia.

Biostaticians at the University of California, Los Angeles, calculated that a 75-year-old man with amyloid has a 17.2% lifetime risk of developing Alzheimer’s; for a woman that age, with a longer life expectancy, the figure is about 24%.

But older people may also be among the 30% of those with amyloid deposits who, for unexplained reasons, retain normal cognition. “If we start treating everyone with preclinical Alzheimer’s, we’ll treat a lot of people who would never have gone on to have dementia at all,” Langa said.

Moreover, what treatments would those be? Multiple trials have failed to find drugs that prevent, reverse or slow Alzheimer’s. Results from a different approach, an infusion drug being tested in older people with amyloid but without cognitive impairment, remain years away.

“As a clinician, would I like amyloid information about my symptomatic patients? Yes,” said Dr. Ronald Petersen, a neurologist and director of the Mayo Clinic Alzheimer’s Disease Research Center. “Am I going to be able to do something about it? Not at present.”