WASHINGTON (WASHINGTON POST) - Something mysterious and disturbing has happened to US State Department personnel, first in Cuba and now in China.

Strange high-pitched sounds - "buzzing," "piercing squeals," "grinding metal," as the Cuba staffers later told doctors - preceded an eruption of health problems, including headaches, dizziness, confusion, ear pain, hearing loss, insomnia and fatigue.

Last year (2017), then-Secretary of State Rex Tillerson referred to what happened in Havana as "attacks."

Numerous news stories and opinion columns speculated about "sonic attacks" using some kind of unknown acoustic weapon.

The mystery spread this spring to China: A staffer in Guangzhou experienced "subtle and vague, but abnormal sensations of sound and pressure," in the words of the US State Department.

Days ago the department issued a health alert to staff in that country, telling anyone who experiences "any unusual acute auditory or sensory phenomena accompanied by unusual sounds or piercing noises, do not attempt to locate their source."

What the phenomenon lacks so far is any solid information about cause and effect. It's not even clear that the ailments suffered have an external origin.

Officials have produced no evidence that anyone has intentionally attacked the Americans, nor is there any obvious environmental cause. There is no sign of a rogue virus or other pathogen.

No one disputes that American staffers have been suffering from a long list of symptoms.

But a much-publicised research report about the Cuba staffers, published on Feb 15 as "Preliminary Communication" in the prestigious Journal of the American Medical Association (JAMA), has triggered scepticism among outside researchers.

The report came from the University of Pennsylvania Medical School's Center for Brain Injury and Repair, where 21 State Department personnel from Cuba have been examined and treated.

Most had persistent symptoms, such as cognitive problems, sleep impairment and hearing dysfunction. All but three had reported hearing unusual sounds in their homes or hotel rooms before the onset of their symptoms.

In Philadelphia, doctors concluded that the patients "appeared to have sustained injury to widespread brain networks without an associated history of head trauma."

Their report offered no explanation for the injuries, however.

The doctors detected no clear physical origin in the brain: Eighteen of the 21 patients showed nothing unusual on a brain scan, and the other three had "mild" or "moderate" damage to white matter that the investigators acknowledged could be due to pre-existing disease processes.

Under the heading of "Meaning," the paper notes, "The unique circumstances of these patients and the consistency of the clinical manifestations raised concern for a novel mechanism of a possible acquired brain injury from a directional exposure of undetermined etiology."

The statement's many caveats, as well as a separate editorial, made clear that these cases are rife with unknowns, according to Howard Bauchner, JAMA's editor in chief.

"I think the public can understand that there's just uncertainty, that it was important for there to be documentation of what is known and what isn't known," he said.

But the paper left many outside experts perplexed.

Two neuroscientists at the University of Edinburgh, Robert McIntosh and Sergio Della Sala, published a criticism in May, writing that the evidence of pathological mental impairment in the patients was "almost unbelievably flimsy."

In e-mails to The Washington Post this week, Della Sala wrote: "Whatever the cause of the symptoms, the JAMA paper presented evidence far too thin to support the existence of brain damage."

He also said the Penn doctors defined impairment too liberally. "The Penn researchers used a threshold of 40 per cent to define pathological performance," he noted.

"This means that any four out of 10 normal people would be classed pathological. They used 37 measures with this incredibly lax criterion. Hence, the chances that anyone would be 'normal' with their measures is zero."

Another University of Edinburgh neurologist, Jon Stone, said via e-mail that the data published by JAMA "is compatible with individuals who have a genuine functional disorder consisting of dizziness, headache and cognitive symptoms, which we see routinely in neurological practice. These disorders . . . are so common that it would not be hard to find a cluster of cases in a larger group of at-risk individuals."

He added, "Another possibly more important phenomenon is the way that doctors, sensing an 'outbreak,' can suggest or over-diagnose symptoms in patients in order to fit an a priori hypothesis."

A spokeswoman for Penn Medicine said the paper's authors would not be available for interviews.

"We are continuing to work with the Department of State to evaluate and treat personnel who have reported audible phenomena experiences," she added. "We are not able to provide specifics about different patient groups at this time."

The cases in Cuba began in November 2016 and were first reported to embassy officials there at the end of that year. They continued until last August, then reportedly stopped. Those 10 months were a period of political turmoil and a chilling of US-Cuba diplomatic relations after the election of President Donald Trump.

Cuban scientists looked into the cases after conferring with US officials.

The Cuban conclusion: The issue was a "collective psychogenic disorder."

According to a story in the journal Science, the US government gave Cuban investigators recordings of irritating sounds around the residences of embassy personnel - which turned out to most closely match the buzzing of Jamaican field crickets.

Cuban scientists' scepticism has now been echoed in the United States and the United Kingdom.

"We're baffled by this (JAMA) report," said Alberto Espay, a professor of neurology at the University of Cincinnati. "The symptoms are certainly very real and in no way fabricated. This is not a case of a bunch of people colluding to try to come up with some symptoms all together."

But medical literature, he noted, has long documented cases in which clusters of people in stressful situations will exhibit similar symptoms of health problems.

Last year, Trump told reporters that "some very bad things happened in Cuba. They did some bad things."

But although Tillerson described whatever happened there as "attacks," the department has not used that term in recent weeks.

US Secretary of State Mike Pompeo, testifying on Capitol Hill in late May, mentioned the health problem of the employee in China and said, "The medical indications are very similar and entirely consistent with the medical indications that have taken place to Americans working in Cuba."