Coronavirus disease (COVID-19) is an infectious disease caused by the SARS-CoV-2 virus. Most common COVID-19 symptoms comprise fever, cough, tiredness, loss of taste or smell whereas less common symptoms comprise sore throat, headache, aches and pains, diarrhea, skin rash, discoloration of fingers or toes, and red or irritated eyes. Few people develop more serious COVID-19 symptoms which comprise difficulty breathing or shortness of breath, loss of speech or mobility, confusion, and chest pain (1). According to the World Health Organization, by end of December 2022, there have been 664,618,938 confirmed cases of COVID-19 worldwide, including 6,722,949 deaths. By end of December 2022 a total of 13,073,712,554 vaccine doses have been administered (2).

Estimations suggest that 20% of individuals who contracted the infection Covid-19 develop post-acute sequelae of COVID-19 (PASC), often referred to as long covid-19 (3–5). PASC refers to the perdurance of COVID-19 symptoms for 12 weeks or longer (6). Milder sequelae of COVID-19 can however continue after having recovered from the viral infection and last up to 18 months (3, 7, 8). Common symptoms of PASC that can persist after 12 weeks are fatigue and headache (9). Another frequent component of PASC, and common long-term consequence of COVID-19 is cognitive impairment, referred to as Cognitive Covid (10). The occurrence of Cognitive Covid is uncertain. Identified cases of Cognitive Covid vary greatly among existing research, ranging between 15% and 80% in COVID-19 patients (11). Nonetheless, numerous studies documented self-reported cognitive difficulties following also laboratory-confirmed COVID-19 infection irrespective of clinical symptoms, particularly in middle-aged and older individuals (12, 13). Subjective cognitive difficulties in COVID-19 patients appear to accurately reflect poorer performance on cognitive tasks (14, 15). Existing studies have also documented how those who contracted COVID-19 scored more poorly on objective cognitive assessments compared to those who did not contract the virus (16). Given that prevention of cognitive decline and Dementia is a global priority (17) and that estimations suggest that 55 million people worldwide are living with dementia (18), further understanding of the link between COVID-19 infection and cognitive functioning is essential.

The COVID-19 infection may cause damage to brain structure and function leading to multi-domain cognitive impairment including memory, executive functions, verbal fluency, processing speed, and visuospatial processing (11, 14, 19, 20). One possible reason for the impact of the COVID-19 infection on cognition could be neurotropism (21–29). SARS-CoV-2 can enter the central nervous system and infect neuronal cells overcoming the blood–brain barrier (21, 30). Moreover, studies using brain magnetic resonance imaging revealed brain microbleeds (30) and cortical hypertension (24) in individuals who contracted COVID-19. Neural damage caused by Covid-19 infection is further confirmed in post mortem, neuropathological studies (28). However, it is possible that neurotropism does not occur in all individuals infected with COVID-19, and that the host immune response, including anti-SARS-CoV-2 antibodies may play a modulating role (29). Moreover, substantial evidence in support of neurotropism has not yet been collected and there are few studies even arguing against neurotropism (31–33). Alternative possible explanations for the impact of Covid-19 on cognitive functioning are systemic cerebrovascular changes due to viral-induced inflammation (34).

Flogosis and/or the acquired immune response to infection might underpin cognitive covid. Therefore, irrespective of COVID-19 symptoms, hospitalization, and course of the disease it is important to investigate the association of serologically confirmed COVID-19 infections with cognitive functioning in non-clinical samples. However, while evidence on Cognitive Covid is rapidly expanding in COVID-19 symptomatic patients, epidemiological evidence on the impact on cognitive functioning in individuals with infection-induced SARS-CoV-2 antibodies irrespective of symptoms and severity is lacking. Moreover, observational evidence is very sparse on the impact of the Covid-19 infection on cognitive functioning at older ages, when cognitive decline and dementia sharply increase (34, 35), and most studies have been conducted in samples of hospitalized Covid-19 patients or often with severe COVID-19 symptoms (19, 36). Because the majority of people infected have little or no symptoms and are not hospitalized (12), clinical samples may provide a biased and incomplete picture of the impact of the virus on cognitive function. Spectrum bias may be relevant not only for COVID-19, but also for the sequelae of SARS-CoV-2 infection(s), including on the central nervous system, and thus on brain function, particularly cognition. Epidemiological studies can contribute to complete and unmuddle the clinical picture of Cognitive Covid.

We aimed to study the association between serologically confirmed Covid-19 infection and cognitive function in a large sample of community dwelling individuals aged ≥65 years. We hypothesized that, compared to immunonaïve individuals (i.e., without infection-induced SARS-CoV-2 antibodies) those with a serologically confirmed Covid-19 infection have poorer cognitive function. Because more educated individuals are more favorable toward COVID-19 vaccination (37–39), and education can be a proxy of cognitive functioning in older age (40), we also hypothesized that, compared to those who are not vaccinated, those who are vaccinated against COVID-19 have better cognitive functioning. A previous study has indeed found that individuals aged 16–95 with poorer cognitive functioning report higher hesitancy toward uptake of COVID-19 vaccine (41).

We quantified the impact of COVID-19 infection on cognitive function in community dwelling older people using previously validated serological tests and in-person neuropsychological assessments, respectively. Study results showed that compared to those with no SARS-CoV-2 antibodies, those with SARS-CoV-2 antibodies due to infection had slightly poorer general cognitive functioning whereas those with SARS-CoV-2 antibodies due to vaccination had slightly better global cognitive functioning. Moreover, compared to those without SARS-CoV-2 antibodies, those with SARS-CoV-2 antibodies due to infection scored more poorly on language comprehension and temporal orientation, and those with SARS-CoV-2 antibodies due to vaccination scored better on memory. Though statistically significant, the small size of effects suggests that COVID-19 infection may have a relatively small negative impact on the cognitive functioning of community-dwelling older individuals without dementia and especially on their language comprehension and temporal orientation abilities. Our findings are aligned with, and of comparable magnitude of existing evidence linking Covid-19 infection with poorer objectively assessed cognitive functioning in clinical samples (16). Our results extend previous evidence in middle-aged individuals to older age, a critical period for cognitive impairment and decline.

Previous research linking COVID-19 infection with cognitive functioning is limited to clinical samples, and focused on hospitalized COVID-19 patients with severe symptoms. Our study suggests that the association of COVID-19 infection with poorer overall cognitive functioning may exist also in a more heterogeneous sample of community-dwelling middle aged and older adults without dementia. In line with and support of our results, another study also found a link between COVID-19 infection and poorer cognitive functioning among non-hospitalized individuals with COVID-19 (12).

One of the possible explanations for our results is that the virus neurotropism and ensuing brain damage, which in turn may lead to cognitive difficulties, may not be related to systemic COVID-19 symptoms (21, 30). However, there are also studies suggesting this possible explanation is unlikely (31–33), and that drivers of cognitive impairment in individuals recovering from COVID-19 remain largely unknown (51, 52). Alternative possible explanations for the impact of COVID-19 on cognitive functioning are systemic cerebrovascular changes due to viral-induced inflammation (53). Further evidence is warranted but our findings may have considerable public health implications because COVID-19 infections greatly outnumber symptomatic, severe, and hospitalized COVID-19 cases (54). Prevention of SARS-CoV-2 infections may be important to reduce the overall dementia attributable fraction that might be due to COVID-19, which is a function of infection-induced seroprevalence (55).

It is worth noting that different from previous studies that reported scores below the cut-off for normal cognitive functioning among those infected with COVID-19, our sample comprises older adults without self-reported diagnosis of dementia and who scored high on the CSI’D′. In our study sample the group with SARS-CoV-2 antibodies scored within normal ranges of overall cognitive scores (8), but these scores were statistically significantly lower than those of the group of participants with no SARS-CoV-2 antibodies. The mean difference in the CSI’D′ (46), which total score can range between 0 and 35, was −0.60. So, while the impact of COVID-19 infection over the cognitive functioning of older adults without dementia may be lower than that of other risks factors (e.g., smoking, hypertension) for cognitive decline (34, 56), the size of the impact on cognitive performance equated to the size of the effect of Aducanumab (i.e., the only approved drug against Alzheimer’s disease) on cognitive functioning in the ADAS-cog 70 point scale (57).

Because Cognitive Covid is a new construct, and evidence is sparse not only on its occurrence but also on its causal mechanisms, prognosis and impact in the mid-and long-term are unknown. Cognitive Covid may represent an “acute” symptom caused by the virus. Nevertheless, there is some evidence documenting how cognitive difficulties in some of the first cases of COVID-19 endured up to 1 year, and with null viral load (12). In fact, SARS-CoV-2 may cause an irreversible decline in cognitive functioning which may or not further progress after the complete clearance of the virus. Cognitive Covid could be the beginning of a trajectory of cognitive decline potentially due to the chronicity of the inflammation caused by the virus (51, 58).

We combined information on serological assessment with self-reported vaccination status to discern between those with SARS-CoV-2 antibodies due to infection and those with SARS-CoV-2 antibodies due to vaccination. We found that older adults who vaccinated against Covid-19 had slightly better cognitive functioning, and in particular memory, compared to those who did not vaccinate against COVID-19 (and were not infected). Contrary to our expectations and to existing evidence suggesting that less-well educated individuals are less likely to get vaccinated (37–39), educational level did not confound the association between vaccination status and overall cognitive function. However, vaccinated individuals were statistically significantly better educated than the two remaining study groups.

This study has several strengths. First, differently from all other studies on the topic that relied on self-reported COVID-19 symptoms or positive SARS-CoV-2 test (PCR test or antigen rapid test) as indicator of COVID-19 infection, we used a serological test to detect anti-SARS-CoV-2 antibodies. Spectrum bias is highly unlikely. Second, we conducted in-person cognitive assessment. Self-reported, subjective cognitive difficulties reported by COVID-19 patients have been previously used but are prone to information biased, and may not relate to objective cognitive function (59). Third, the study sample comprises randomly recruited individuals in old and advanced old age, who were underrepresented in previous studies on the topic. Fourth, different from many previous studies that selectively focused on hospitalized individuals with COVID-19, we focused on a representative sample of community dwelling older adults without self-reported dementia. Fifth, in the analyses we controlled for the potential confounding effects of health conditions and age on cognition. Sixth, data was collected early in the vaccination campaign, allowing the study of a mixed population of vaccinated and non-vaccinated individuals.

This study has also several limitations too. Even though our measure of cognitive functioning (46) is more sensitive than those used in most previous studies (19, 60), such as the Mini Mental State Examination (61) and the Montreal Cognitive Assessment (62), it is not free from bias. The CSI’D′ participant part (46) was designed as a screening tool for pathological cognitive decline. A ceiling effect may not be excluded because our participants had an overall high cognitive functioning. Our results may be an underestimate of the true effect of COVID-19 infection on cognitive functioning. Next, issues of directionality due to the cross-sectional design seem implausible but cannot be excluded also because of the lack of pre-pandemic cognitive assessments that did not allow to capture changes in cognitive functioning prior and after infection. Moreover, although we had information about self-reported diagnosis of dementia and all participants scored high on the CSI’D′ – indicating cognitive health - we cannot completely exclude that a minority of participants may have had cognitive decline.

Although only 5.6% of the study sample had SARS-CoV-2 antibodies due to infection we could detect a statistically significant effect on cognition. That the temporal span between the cognitive assessment and the blood test varied among participants is a potential additional limitation. However, considering that in person assessment of cognitive functioning is more time consuming than self-reporting used in previous studies, our study retained a good balance between internal validity and feasibility. A final limitation of our study is that we did not consider SARS-CoV-2 variants, which may have differential pathogenicity, neurotropism, and immune escape potential (63–65). However, we collected blood samples before the Omicron variants appeared, and mainly during the concomitant presence of the alpha and delta variants, which share similar infectiousness and pathogenic characteristics (64).

We found preliminary and novel epidemiological evidence in support of Cognitive Covid, by showing that, in a sample of community dwelling individuals aged 65 years and over and without dementia, serologically-confirmed COVID-19 infection was cross-sectionally associated with lower cognitive performance, especially in the domains of language comprehension and temporal orientation. We extend evidence from individuals infected with COVID-19 to the much larger population of those who got infected with the virus and were pauci-or a-symptomatic. The scale of Cognitive Covid may be larger than previously thought, and may even slightly contribute to worsen the expected increases in dementia occurrence, and require recalculating projections based on demographic changes only. Given the preliminary nature of our results, further cognitive assessment of older people with evidence of COVID-19 infection is warranted to monitor trajectory of cognitive decline, and accordingly implement secondary prevention interventions at large scale.

The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author.

The studies involving human participants were reviewed and approved by Swiss BASEC Registration No 2020–01247 University of Italian Switzerland and University of Zurich. The patients/participants provided their written informed consent to participate in this study.

SS performed the data analyses and drafted the manuscript. EA contributed to the data collection and co-edited the manuscript. DP, AF, and MP contributed to the data collection and provided comments to the manuscript. DP and GG contributed in terms of data collection. All authors contributed to the article and approved the submitted version.

This study is part of the Corona Immunitas research network, coordinated by the Swiss School of Public Health (SSPH+), and funded by fundraising of SSPH+ that includes funds of the Swiss Federal Office of Public Health and private funders (ethical guidelines for funding stated by SSPH+ were respected), by funds of the canton of Zurich, and by institutional funds of the university. SwissDEM was funded by the Swiss National Science Foundation (SNSF), 320030_184794. The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.