An account of another alleged “sonic attack” has surfaced, this time from a U.S. government employee in China. The employee reported “subtle and vague, but abnormal, sensations of sound and pressure,”according to a U.S. Embassy health alert. The episode mirrors reports from American diplomats in Cuba in late 2016, and fuels the debate among scientists about what, if anything, is actually happening.

Last year, 24 of the diplomats who reported sonic attacks in Cuba were tested to gauge whether lasting harm had occurred. In March, researchers from the University of Pennsylvania Perelman School of Medicine in Philadelphia reported in JAMA that the people had balance and thinking problems, sleep disturbances and headaches, and that some had widespread injury to brain networks.

But some scientists and engineers have been questioning whether such attacks are possible, and if the diplomats’ symptoms could have been caused by a sonic attack.

The attacks were supposedly committed with sounds outside the range of human hearing. But generating enough acoustical energy to cause hearing loss and brain damage from those types of sound waves would be no easy feat, says Andrew Oxenham, a hearing researcher at the University of Minnesota in Minneapolis. The intensity of very low frequency infrasound or very high frequency ultrasound drops rapidly over distance, so attackers would need enormous loud speakers to have enough intensity to do neurological harm.

“Even to get across the street and into a building, you’d have to have a loud speaker the size of a building,”

Oxenham says.

It might be possible to focus ultrasound into a tight beam to stage a high-intensity ultrasound attack. But even with such a beam it would be difficult to make a device small enough to be used as a handheld weapon, says Tyrone Porter, a biomedical engineer at Boston University. And that device would be more likely to lead to disorientation than brain damage, he says.

Very little data exist on whether and how ultrasound in the air affects human health. One of the few people to tackle the question is Timothy Leighton, a professor of ultrasonics and underwater acoustics at the University of Southampton in England. He has investigated previous claims of people who complained that they had been victims of sonic attacks.

Some reported incidents were false alarms. But in other cases, Leighton recorded evidence of ultrasound in air at railway stations, museums and swimming pools where people had reported attacks, although the exposure was shown to be accidental, not an attack. He doesn’t know for sure how ultrasound causes symptoms such as the headaches and nausea described by the diplomats. But he suspects subaudible noise makes people anxious, which leads to the reported symptoms. The U.S. government employees in Cuba and China may be experiencing similar anxiety if exposed to ultrasound, he says.

Detected damage?

Leighton and other scientists have questioned whether the JAMA paper actually measured harm caused by a sonic attack. One symptom investigated in the study, white matter changes in the brain, made headlines. White matter is composed of axons, the long extensions of nerve cells that connect different parts of the brain.

“As a result, people got the impression this was some sort of ultrasonic death rifle,” Leighton says. But only three people in the study had white matter abnormalities, and the researchers couldn’t attribute those changes to a sonic attack. They may just have been physical differences that those people’s brains had all along.

What’s more, in the JAMA study, scores that classified diplomats as having a deficit in brain function fall into humans’ normal variation, says Sergio Della Sala, a cognitive neuroscientist at the University of Edinburgh.

The University of Pennsylvania researchers gave diplomats a failing grade on the brain tests if their score on at least one test was below the 40th percentile (meaning that 40 percent of people who take the tests have scores that fall at the low end of the scale), an impairment threshold that Della Sala argues is too high. That’s because, statistically speaking, people would get failing marks on at least one of these tests 40 percent of the time, even without an attack.

Only six of the 24 diplomats took all 37 tests, for 222 tests total. At the 40th percentile cutoff, 89 of the 222 tests would be false positives. That means a test-taker would flunk, but the result would be mistakenly chalked up to a sonic attack when it was really just a natural variation in the way people’s brains work.

In an experiment, Della Sala and University of Edinburgh colleague Robert McIntosh substituted random numbers for diplomats’ test scores and ran a simulation of possible outcomes, using the standards from the JAMA study. The result? “Everybody tested would result affected, everybody. To make sure, we repeated the simulation 1,000 times,” Della Sala wrote in an email.

He doesn’t dispute that some of the diplomats may have experienced symptoms from the incident. But the JAMA paper’s methods would make it impossible for anyone to test normal, he says. “The tests as they have been used and presented are spurious,” he wrote. (Della Sala, along with Roberto Cubelli of the University of Trento in Italy, also published a scathing review of the JAMA study in Cortex on April 5.)

One of the JAMA paper’s coauthors, Douglas Smith, says he and his colleagues have more data than were included in the study. “We note that interpretation of neuropsychological test results is somewhat more nuanced than a simple counting of scores that are lower than a conventional percentile cutoff point,” Smith wrote in an email. Instead, the researchers considered how much each person’s performance on a particular test differed from what is normal for the individual. In some cases, test scores in one aspect of brain function fell far below that person’s normal — down to the bottom 10 percent of the person’s average brain function. That low level of function counts as impairment, says Smith, who directs the Center for Brain Injury and Repair at the University of Pennsylvania’s medical school.

The researchers are currently trying to determine if the people felled by the attacks have changes in the structure of their brains that could account for the symptoms, Smith says.

Reverse engineering a ‘sonic weapon’

The sonic attacks may not have been attacks at all, but eavesdropping gone awry, says Kevin Fu, an electrical engineer and computer scientist at the University of Michigan in Ann Arbor. Fu, who studies how malicious sounds might be used to attack computers, has some of the only experimental evidence to suggest what might have happened in Cuba.

Fu’s attention was drawn to the attacks when the Associated Press released an audio clip of the sound some diplomats in Cuba heard during the incidents. He and colleagues Chen Yan and Wenyuan Xu, both of Zhejiang University in Hangzhou, China, tried to re-create the sound and surmised that an ultrasonic listening device could have developed interference that caused it to produce the unusual noise. “This seems like bad engineering rather than a deliberate attack,” Fu says.

Fu and colleagues described their experiment in a technical paper published online March 1. The researchers did not test whether such a device could have produced health and hearing problems for the diplomats.

For now, what actually happened in Cuba and China to produce the diplomats’ symptoms remains a mystery. And it’s possible we may never know. After all, Fu says, it’s unlikely that if foreign governments did have sonic weapons that they’d allow U.S. scientists to run experiments with the devices.