Does Everyday Worry Impair Memory? Barely, a Large Study Finds

What the research found

Five hundred and three adults sat ten tests of working memory in a single online session that ran, on average, an hour and thirty-four minutes. Working memory is what a person uses to keep a few items in mind and operate on them while they are there. The tests sampled it widely, from repeating digit sequences forwards and backwards to holding items while solving unrelated problems, tracking a list whose end was never announced, and judging whether the item on screen matched the one two steps back, each in a numerical-verbal and a visuospatial version.

Before any of the tests, every participant rated how stressed the previous month had been and how anxious they felt at that moment. The question was whether those ratings predicted the scores.

The analysis belongs to Karolina Lukasik, Otto Waris, Anna Soveri, Minna Lehtonen and Matti Laine, working at Åbo Akademi University in Turku with colleagues at the University of Helsinki, the University of Oslo and the Turku Brain and Mind Center, and it appeared in Frontiers in Psychology in 2019. The authors describe the work as exploratory. Its central number is one that a less careful write-up would have rounded in its own favour.

For anxiety, the test returned a p value of 0.06. The conventional threshold for calling a result real is 0.05, and 0.06 falls on the wrong side of it, so the authors reported a trend and declined to claim an effect. Stress fared worse again, at 0.12.

The plain correlations were small. Anxiety and verbal working memory correlated at -0.08, anxiety and the two-back updating measure at -0.06, and neither could be distinguished from zero. Only the visuospatial scores produced reliable correlations, around -0.13 for anxiety and -0.135 for stress, a size that registers as a faint tendency across a crowd and disappears at the level of one person. Whatever relationship existed was the same across all three kinds of working memory the researchers measured.

Age predicted performance far more strongly, particularly on the visuospatial and updating tests, and education mattered as well, while the socioeconomic circumstances of a participant’s childhood made no measurable difference at all. The full model, anxiety and stress included, accounted for 4.5 percent of the differences between people. Age, education and childhood circumstances on their own accounted for 2.8 percent. Adding a month of stress and a moment of anxiety to the picture bought less than two percentage points.

What it means for you and your child

This blog has reported that anxiety carries a specific cost to working memory, and that finding holds under conditions the Finnish study did not create. The earlier work drew on people with diagnosed anxiety disorders, and on experiments that manufactured threat in a laboratory and measured performance while the threat was live. Lukasik and her colleagues measured something quieter, which was ordinary worry and ordinary stress in adults going about their lives. Anyone scoring above 10 on a depression screen was excluded before analysis, which removed 159 of the 711 people who finished. The group that remained was, by the study’s own comparison against published norms, slightly less stressed than average.

The participants were adults, recruited online in the United States, younger and more educated and more often female than the country as a whole. Correlational data of this kind carries no information about direction, so the study says nothing about whether worry shapes memory or memory shapes worry. The two feelings were not cleanly separable either; the anxiety and stress measures correlated with one another at 0.499, and the authors concede that their six-item anxiety scale may have captured stress along with it. What the design does deliver is scale, and 503 people is a large place to look.

What survives is a boundary. In those 503 adults, everyday worry was a weak guide to how much a person could hold in mind and rearrange. Nothing here reaches children, and the authors would be the first to say so, but the evidence offers thin support to anyone inclined to explain a low working memory score by pointing at a worried temperament.

That leaves the worry to be addressed on its own account, and the memory difficulty to be assessed on its own terms, without each being read as the hidden cause of the other. A child who is anxious and who also loses hold of instructions halfway through presents two matters, and neither has been shown here to explain the other.