Browse the corpus

To the Editor: Since the start of the coronavirus disease 2019 (Covid-19) pandemic, an estimated 400 million persons worldwide have reported symptoms of what has been termed “long Covid-19 (long Covid).”1 Some studies have shown an association between Covid-19 and persistent memory problems.2,3 To investigate this potential association, we used a validated questionnaire about participant-reported memory problems,4 with assessments before and after tests for the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the virus that causes Covid-19. Among 188,137 participants in the nationwide Norwegian Covid-19 Cohort Study from March 27, 2020, to April 26, 2023, a total of 134,373 participants (71%) completed at least one Everyday Memory Questionnaire (EMQ), and 111,992 (60%) had documented tests showing positive or negative status for SARS-CoV-2 infection as determined from the Norwegian Surveillance System for Communicable Diseases or on the basis of participant report (Fig. S1 in the Supplementary Appendix, available with the full text of this letter at NEJM.org). The EMQ has 13 items, with a maximum total range of 0 to 52 points.4 We report here scores that were averaged over all 13 items (total score divided by 13, with each item scored on a scale from 0 to 4 [signified as A to E in the source publication4] and with higher scores indicating worse memory problems).

t NEJM.org). The EMQ has 13 items, with a maximum total range of 0 to 52 points.4 We report here scores that were averaged over all 13 items (total score divided by 13, with each item scored on a scale from 0 to 4 [signified as A to E in the source publication4] and with higher scores indicating worse memory problems). We used a linear mixed-effects model to compare mean EMQ scores among participants before testing and after testing positive or negative for SARS-CoV-2 infection. Analyses were adjusted for participant age, sex, body-mass index, vaccination status, smoking status, income level, questionnaire number, and the presence or absence of underlying medical conditions. Further details on the EMQ and statistical analyses are provided in the Supplementary Appendix.

or SARS-CoV-2 infection. Analyses were adjusted for participant age, sex, body-mass index, vaccination status, smoking status, income level, questionnaire number, and the presence or absence of underlying medical conditions. Further details on the EMQ and statistical analyses are provided in the Supplementary Appendix. Among the 111,992 participants with test results, 57,319 tested positive for SARS-CoV-2 and 54,673 tested negative. Participants who tested positive for SARS-CoV-2 were younger (mean [±SD] age, 48±13 years) than those who tested negative (50±14 years), and most of the participants in the two groups were women (Table S1). The adjusted mean EMQ scores were numerically higher (indicating worse memory problems) after a positive test than after a negative test at all time points (at 0 to 1 month after a test, 0.66 vs. 0.60; at >1 to 3 months, 0.74 vs. 0.62; at >3 to 6 months, 0.72 vs. 0.62; at >6 to 9 months, 0.71 vs. 0.62; at >9 to 12 months, 0.75 vs. 0.63; at >12 to 18 months, 0.82 vs. 0.62; and at >18 to 36 months, 0.82 vs. 0.62) (Figure 1). (The 95% confidence intervals for these scores at each time point are provided in Table S3.) In contrast, scores in the two groups were essentially the same before SARS-CoV-2 testing, with a mean score at more than 1 month before testing of 0.61 (95% confidence interval [CI], 0.60 to 0.62) among participants who tested positive and 0.60 (95% CI, 0.58 to 0.62) among those who tested negative. In the context of a score range of 0 to 4, the scores at all time points were toward the milder end of severity of memory issues.

more than 1 month before testing of 0.61 (95% confidence interval [CI], 0.60 to 0.62) among participants who tested positive and 0.60 (95% CI, 0.58 to 0.62) among those who tested negative. In the context of a score range of 0 to 4, the scores at all time points were toward the milder end of severity of memory issues. Among participants who reported memory problems (i.e., answered “yes” to a yes–no question about having memory problems) after a SARS-CoV-2 test, those who tested positive had numerically higher EMQ scores than those who tested negative (Fig. S2). However, regardless of SARS-CoV-2 infection status, participants who reported memory problems after the test (but not before the test) had higher EMQ scores before the test than participants who reported no memory problems. The mean EMQ score among SARS-CoV-2–positive participants generally increased according to the number of days that they had been bedridden (1 to 6 days, 0.79; and >13 days, 1.29 points). (The 95% confidence intervals for these scores at each time point are provided in Table S4.)

cipants who reported no memory problems. The mean EMQ score among SARS-CoV-2–positive participants generally increased according to the number of days that they had been bedridden (1 to 6 days, 0.79; and >13 days, 1.29 points). (The 95% confidence intervals for these scores at each time point are provided in Table S4.) Participant age appeared to be inversely associated with EMQ scores (Fig. S3), and body-mass index appeared to be positively associated with EMQ scores (Fig. S4). Overall, women had higher EMQ scores than men (Fig. S5), a finding that is consistent with those from several studies that have investigated psychiatric and neurologic complications of long Covid.5 Limitations of the current study include possible response bias between participants who responded to the questionnaire and those who did not and recall bias due to participant self-report on the EMQ (see Methods Section 5 in the Supplementary Appendix). In a group of Norwegian participants assessed between March 2020 and April 2023, participant-reported memory function, as scored on the EMQ, was numerically worse at several time points up to 36 months after a positive SARS-CoV-2 test than after a negative test.