In Generalised Anxiety, the Difference Appears in Milliseconds

What the electrodes recorded

Fifteen medication-free adults with generalized anxiety disorder completed a laboratory matching task with an accuracy close to perfect, and so did the fifteen healthy adults matched to them for age and education. Correct responses reached about 99 percent in both groups on the simplest trials, and neither accuracy nor reaction time separated the two to a degree the statistics would credit. On any record a school or clinic keeps, the groups performed the same.

The task was plain. Two shapes appeared in sequence, each for half a second with a fifth of a second between them, and the participant judged whether the second matched the first in colour while ignoring any change of shape. Four kinds of trial occurred equally often, varying whether the colour changed, the shape changed, both, or neither. Nothing in the material carried emotional content.

While the participants worked, electrodes on the scalp recorded the electrical activity beneath. The measure of interest was the N270. It appears roughly 270 milliseconds after a stimulus conflicts with what was expected, and it indexes how that conflict is handled. The patients were recruited from an outpatient ward in Beijing and screened to exclude other psychiatric and neurological conditions. The results reached Neuropsychiatric Disease and Treatment in 2015, from Yingxue Yang and colleagues in the neurology department at Xuanwu Hospital, Capital Medical University, working with the Suzhou Institute of Biomedical Engineering and Technology.

Both groups produced a larger N270 to a mismatch than to a match, and both slowed by about fifty milliseconds when the colour changed. The groups parted over the left front of the head. When the colour changed, the response at the left frontal electrode averaged minus 0.46 microvolts in the controls and plus 1.33 in the patients, and because the N270 is a negative-going wave the higher figure is the smaller response; when colour and shape both changed, the same gap appeared at the left frontal and the left central site. A change of shape alone, which participants had been instructed to ignore, produced no group difference, and neither did any electrode over the right hemisphere.

The timing of the response was identical in the two groups. So was the P300, a later and far more commonly measured signal, in size and in speed alike. The authors conclude that the N270 is more sensitive than the P300 in detecting minor cognitive impairment.

What it means for you and your child

A correct-response rate records whether the answer was right, and a reaction time records how long the answer took. The interval between them, which neither measure captures, was the only place these two groups differed at all.

Why performance held up is where the paper stops reporting and begins proposing. Following Etkin and Schatzberg, the authors suggest that intact behaviour in generalized anxiety disorder may rest on compensatory processes, meaning additional work that keeps the output normal. No measure of effort was taken and nobody was asked how hard the task had felt, so that account stands as a hypothesis offered to explain an absence.

The limits are substantial. Fifteen people to a group is a small sample, and the design leaves open which came first, the anxiety or the diminished response. Patients with severe anxiety were excluded because they would have struggled to sit still through the task, so what was studied is a moderate middle. The correlations the authors report between symptom scores and amplitude were computed with patients and controls pooled together, which means they largely restate the group difference and establish no graded relation among anxious people. Every participant was a Chinese adult of around thirty with roughly fourteen years of education, working on coloured shapes rather than schoolwork. Nothing here describes a child.

What the study establishes is narrow and unusually clean. Two groups can be indistinguishable on the outcome measure and distinguishable on a record of the processing, and the second fact had no visible consequence for the first. Carrying that to a child at a desk goes beyond what fifteen anxious adults in Beijing can show, and it should be held as the inference it is.

The correct answer, delivered on time, is a thin guide to how the answer was produced. If your child’s work comes back right and the evening that produced it did not look easy, the marks will not tell you which of those two things to believe, and there is no hurry to decide.