Neurocognitive and Psychiatric Outcomes Following Mild to Moderate COVID-19: A Longitudinal Comparison of Alpha/Delta and Omicron Variants

Introduction

Since the onset of the COVID-19 pandemic in the United States, over 100 million Americans have been infected, and over one million have died from causes related to SARS-CoV-2 infection.¹ In addition to symptoms from acute COVID illness, many individuals have developed short- and long-term neuropsychiatric and medical symptoms, frequently referred to as post-acute sequelae of COVID-19 (PASC), which may persist for many months after recovery from acute infection.^2–4 For many, this includes cognitive complaints, including problems with processing speed, concentration, memory, and executive function,^2,4–6 as well as symptoms of depression, anxiety, and PTSD.^7,8

COVID-19 continues to present a global public health problem. However, recent predominant variants in the Omicron family compared to earlier strains of the virus (ie., Alpha and Delta), have been considerably less virulent, leading to milder symptoms, and reduced hospitalizations and mortality.⁹ Similarly, the prevalence of post-acute sequelae of COVID-19 (PASC, or Long COVID) may be lower after Omicron compared to Alpha and Delta variants, potentially attributed to reduced virulence and increased COVID-19 vaccination.¹⁰ Nonetheless, estimates vary widely, where rates of PASC from Omicron range from 0.1%-4.8% compared to 0.5%-10.8% from earlier variants.^11,12

While much has been published on PASC related to the early strains of COVID-19, little research has examined long-term effects of later variants, such as the Omicron variant. Omicron was the dominant variant of COVID-19 from December 2021 through June 2022.¹³ Research has suggested that Omicron has high transmissibility and immune evasion, but also potentially reduced virulence.^14,15

Few studies have examined cognitive effects post-recovery from the Omicron strain of COVID-19. One study of 215 participants from Shanghai found no difference between those recovered from Omicron and controls, though did find impairment in some domains in participants 50 and over.¹⁶ However, this study only included the Montreal Cognitive Assessment (MoCA) and the Mini-Mental Status Examination (MMSE) as measures, rather than more thorough neuropsychological testing, and did not directly compare Omicron to prior COVID-19 variants. These measures are more frequently used in older adults, lack age-specific controls in younger adults, and do not account for other sociodemographic factors.^17–19 Additionally, the authors have previously found that the MoCA is relatively insensitive and non-specific for COVID-related cognitive impairment.²⁰

Other studies relied on predominantly computerized surveys/assessments, including one specifically looking at college students across COVID variants, which found that the original variant had the most severe long-term effects while Omicron had the mildest effects on physical health and cognition.²¹ A computerized survey in England similarly found that the largest deficits in global cognitive scores were during the early strains compared to Omicron.²² Both studies lacked in-person neuropsychological assessments, and neither followed participants longitudinally.

In parallel with cognitive symptoms, individuals with PASC may experience significant psychiatric distress, including depression, anxiety, post-traumatic stress, and fatigue. We know less about the relative prevalence and severity of these neuropsychiatric symptoms following Omicron infection compared to earlier COVID-19 strains. The few studies which have examined this question have reported heterogeneous findings. In a comprehensive meta-analysis that included pooled data across all COVID strains, Seighali et al reported that 23% of individuals suffer from depression and anxiety in the context of PASC.²³ A survey study of individuals with mild Omicron in Shanghai, China, conducted during acute quarantine, found that 13.7% of participants met criteria for depression on the PHQ-9 and 8.6% met criteria for anxiety on the GAD-7.²⁴ Another survey-based study found that Omicron infection increased the risk for depressive, anxiety, and PTSD symptoms,²⁵ while another found that Omicron infection did not correlate with depression or anxiety symptoms.²⁶

The authors of this report have previously published findings of a longitudinal assessment of a cohort of individuals infected during the original waves of the pandemic.^4,6 Findings suggested that a significant proportion of individuals may have persistent symptoms over a year after infection, and that individuals presenting for PASC-specific care (ie. “clinical” patients) may have greater impairment.^4,6

To our knowledge, no studies have directly compared both neurocognitive performance and standardized measures of psychiatric distress and fatigue in the weeks-to-months following Omicron infection versus Alpha/Delta strain infection within the same longitudinal framework. The current study aimed to characterize PASC associated with Omicron infection and to compare neurocognitive outcomes, psychiatric distress, and fatigue between individuals infected during the Alpha/Delta period and those infected during the Omicron period. Additionally, this study aimed to follow the Omicron cohort longitudinally to characterize persistence or improvement of symptoms over time.

We hypothesized that individuals infected during the earlier phase of the pandemic (Alpha/Delta period) would demonstrate greater neurocognitive impairment, psychiatric distress, and fatigue compared to those infected during the Omicron period, reflecting both increased psychosocial stressors early in the pandemic and viral factors, including greater virulence, lower vaccination rates, and limited availability of antiviral treatments.

Methods

Study Design and Participants

This study was conducted by the Department of Psychiatry at New York Medical College (NYMC), a medical school affiliated with Westchester Medical Center Health Network (WMC Health), and was approved by the Institutional Review Board of NYMC (Protocol #14400) and the WMC Health Clinical Research Institute. The study was conducted according to the principles of the Declaration of Helsinki, and all patients gave written informed consent.

Data for this manuscript included the assessments of participants enrolled in a longitudinal investigation of the neuropsychiatric sequelae of COVID-19. This included an initial wave of participants who were infected with the Alpha and Delta strains of SARS-CoV-2 (hereafter referred to as Alpha/Delta), evaluated between December 2020 and November 2021, for which results have been published previously, as well as a second wave of participants who had been infected with the omicron strain.^4,6

Recruitment and Eligibility

Participants were recruited through social media, Email communication, flyers, word of mouth, and the WMC Health Post-COVID Recovery Program, a specialized clinical program for individuals with PASC. Prospective participants were screened via telephone to determine eligibility based on the following criteria: (1) age 20 years or older; (2) documentation of a positive COVID-19 nasopharyngeal test or positive antibody test prior to vaccination; (3) recovery from acute COVID-19 infection as defined by Centers for Disease Control and Prevention criteria (10–20 days after symptom onset and >24 hours without fever); (4) completion of at least an eighth-grade education; (5) fluency in English; and (6) ability to provide informed consent.

Participants were excluded if they had a prior history of a major neurocognitive disorder, traumatic brain injury with loss of consciousness, uncorrected visual or hearing impairment, intellectual disability, or an acutely unstable psychiatric condition (such as acute psychosis or suicidality).

Clinical vs. Community Recruitment

During recruitment for the Alpha/Delta cohort, 47 of 75 participants (63%) were referred from the WMC Health Post-COVID Recovery Program.^4,6 In contrast, during the Omicron recruitment period, only one participant was seeking care for PASC, consistent with an observed >80% decline in referrals to this clinical program during the Omicron era (Gary Rogg, M.D., personal communication).

Because prior analyses demonstrated that individuals seeking treatment for PASC exhibited significantly poorer psychiatric and neuropsychological functioning compared to non-treatment-seeking participants,^4,6 clinical participants were excluded from the present analyses (47 from Alpha/Delta and 1 from Omicron). This approach was taken to minimize referral bias and to allow for a more balanced comparison between community-recruited participants across viral strains.

Study Procedures

After the risks and benefits of participation were explained and written informed consent was obtained, participants completed a comprehensive in-person assessment during their study visit. Participants met with trained study assessors (S.L., S.S., R.M., S.A., A.L)., who were supervised by the principal investigators (S.F. and R.D.) and assessments were conducted using standardized administration procedures. Participants were compensated $40.00 for each study visit.

Participants completed an initial baseline assessment and were invited to return approximately six months later for a follow-up assessment. To minimize practice effects at follow-up, alternate test forms were used when available; the same examiner did not necessarily conduct both visits.

Measures

A complete list of measurements obtained at each visit can be found in Table 1. Data collected at the assessment visits included sociodemographic information, medical history, COVID-19 illness characteristics, psychiatric symptom measures, and neuropsychological testing. Time since acute COVID-19 infection was defined as the interval between the participant’s infection and the study visit. When available, this was calculated using the date of the first positive COVID-19 test. For participants identified by antibody testing prior to vaccination, the date of symptom onset as reported by the participant was used. Prior psychiatric history was assessed as lifetime history of psychiatric diagnosis or treatment, based on participant self-report.

Table 1 COVID-19 and Neurocognitive Assessments

Neuropsychological measures included an estimate of premorbid functioning and assessments across multiple cognitive domains, including processing speed, attention/concentration, memory, and executive function. Raw scores were converted to standardized T-scores or scaled scores according to test manuals and, when available, compared to age- and education-adjusted normative values.

A total of 11 neuropsychological test scores were analyzed at each assessment. The five Repeatable Battery for the Assessment of Neuropsychological Status (RBANS) index scores were included, while the RBANS total score was excluded to avoid redundancy. Participants were categorized according to widely accepted clinical conventions: performance ≥1 standard deviation below normative values on two or more tests was classified as “low,” and ≥2 standard deviations below normative values on two or more tests was classified as “extremely low”.^38,39 Alternate RBANS forms (Form A at baseline and Form B at follow-up) were used to reduce practice effects.

Vaccination for COVID-19 became available during the Alpha/Delta recruitment period and was therefore present in a relatively small proportion of those participants.

Statistical Analysis

Statistical analyses were performed using SPSS software version 30.⁴⁰ Descriptive statistics were used to characterize the study population (frequency, mean, standard deviation); Chi-square tests were used to compare categorical variables; independent t-tests were used to compare continuous variables between groups; and one-way t-tests were used to compare the groups to population-based normative values for neuropsychological testing. Effect sizes were calculated for results which were statistically significant, using Cohen’s d statistic for t-tests (d) and Cramer’s V statistic for variables compared with chi square (V). General linear models were developed to examine the interaction between the two COVID-19 variants and repeated NP testing measures (appointment one to appointment two). General linear models were also developed to examine interaction effects between COVID-19 variant and time, as well as with severity of acute infection.

Results

Sample Characteristics at Initial Assessment

Participants overall (N=55) were on average 32 years old (SD 11.2), had approximately a bachelor’s degree level of education, and were predominately White (70.9%) and female (67.3%). They had on average 0.6 medical comorbidities (SD 1.0) (Table 2). Participants rated their acute COVID-19 symptom severity as moderate (mean scores of 10.6–13.7) and reported an average length of acute illness of about 10 days (SD 10.7). Participants were on average 122 days out from their acute infection (SD 100.0) (Table 3).

Table 2 Sociodemographic and Medical Factors

Table 3 NP Test Scores of Alpha/Delta Group Compared to Omicron Group at Appointment 1

When comparing the Alpha/Delta and Omicron groups, participants who were infected with the Alpha/Delta strain had a greater number of medical comorbidities (1.0 compared to 0.2, p<0.001, d=0.967), and reported a more severe acute COVID-19 infection (13.7 vs 10.6, p=0.041, d=0.565). The Alpha group was also further removed from their acute illness date (171.8 days vs 70.3 days, p<0.001, d=1.173) (Table 3). There was no other significant difference (age, gender, education, length of acute illness) between the Alpha and Omicron groups in terms of baseline characteristics.

Initial Neuropsychological Testing

At the initial evaluation, 46.4% of the Alpha/Delta group demonstrated some degree of NP test impairment, with about 7% scoring in the “extremely low” range. When comparing average test scores to normative values, the Alpha/Delta group performed significantly worse than expected on RBANS Immediate Memory (t=−2.636, df=27, p=0.014, d=−0.498), while performing significantly better than expected on RBANS Visuospatial/Constructional (t=4.510, df=27, p<0.001, d=0.852). No other individual test average was significantly different than expected based on population norms (Table 3).

Just over half of the Omicron group demonstrated some degree of NP test impairment at the initial evaluation, with about 7% scoring in the “extremely low range”. When comparing average test scores to normative values, the Omicron group performed significantly worse than expected on RBANS Immediate Memory (t=−2.594, df=26, p=0.015, d=−0.499) and on TMT-A (t=−3.499, df=26, p=0.002, d=−0.673). No other individual test average was significantly different than expected (Table 3).

General linear models were developed to examine the interaction between COVID-19 variant with time since acute COVID-19 infection, and also with the severity of acute infection. No interactions were found that impacted any NP test score (data not shown).

Initial Psychiatric Measures

Approximately 40% of Alpha/Delta and Omicron groups reported a history of prior psychiatric treatment for any mental health problem, and approximately 1 in 5 were taking psychiatric medication at time of assessment (Table 3). These percentages are similar to USA national estimates of lifetime psychiatric treatment and current psychiatric medication use.^39,40–42

There were no significant differences between Alpha/Delta and Omicron groups in terms of mean symptom measures for depression, anxiety, PTSD and fatigue. Generally, the mean scores were in the non-clinical range. The number and percentage of participants meeting criteria for clinically significant symptoms or disorder did not differ between groups. However, individuals with Alpha/Delta infection numerically reported higher rates of clinically significant fatigue compared to Omicron (29% vs. 11%), although this difference did not reach statistical significance (χ²=2.85, df=1, p=0.09). This trend should be interpreted cautiously given limited power. Although between-group differences were not observed, both groups demonstrated higher rates of depressive symptoms compared to population norms, affecting 21% of Alpha/Delta and 19% of Omicron, as compared to 8% of the general population.⁴³ No other disorder appeared to diverge from national prevalence estimates.

Given the absence of between-group differences and the relatively low symptom severity on psychiatric measures, multivariable interaction models were not pursued for psychiatric outcomes in order to limit overfitting.

Follow-Up Neuropsychological Assessment

Twenty-seven participants from the Alpha/Delta group returned for a second evaluation, on average 381 days after their acute COVID-19 infection. One person remained in the “extremely low” NP Test group at this appointment, while 10 (37.0%) remained in the “low” group. The Alpha/Delta group showed some improvements, in that at the follow-up appointment had only test which was significantly different from established norms, Stroop Color-Word, which was above average (t=3.015, df=26, p=0.006, d=0.580) (Table 4).

Table 4 NP Test Scores of Alpha/Delta Group Compared to Omicron Group at Appointment 2

Twenty-one participants from the Omicron group returned for a second evaluation, which was on average 275 days after their acute COVID-19 infection. No participant remained in the “extremely low” NP test group at this appointment, while one third “low” NP test group. The Omicron group showed improvements in that there were no persistent significant differences from normative values (Table 4).

General linear models were developed to examine the interaction between COVID-19 variant with time since acute COVID-19 infection, and also with the severity of acute infection. The only test for which any interaction was found was RBANS Delayed Memory, for which number of days between acute infection and follow-up appointment was found to impact participants’ scores (F(1,42)=4.370, p=0.043, η_p²=0.094), indicating improvement for both groups over time; however, n interaction was also noted between variant and time since acute infection (F(1,42)=5.145, p=0.029, η_p²=0.109), which suggested that the Alpha/Delta group showed a greater improvement over time. No other interactions were found for any NP test variable.

Follow-Up Psychiatric Measures

At time of reassessment, there were no significant differences between the Alpha/Delta and Omicron groups in terms of mean score on any psychiatric measure, nor as far as proportion meeting criteria for clinically significant symptoms (Table 4). Mean scores remained primarily in the non-disordered range, and the proportion meeting criteria remained similar to that at Appointment 1.

Discussion

This longitudinal study compared neurocognitive performance, psychiatric distress, and fatigue among individuals who experienced mild to moderate acute COVID-19 infection during the Alpha/Delta and Omicron eras. Participants were not hospitalized and, following exclusion of clinical referrals, were not seeking care for PASC. Overall, relatively few differences were observed between variants across neuropsychological or psychiatric outcomes.

Across both groups, cognitive performance improved between the initial and follow-up assessments, and notably, no neuropsychological test scores in either group remained significantly below population-based norms at follow-up. No participant demonstrated worsening neuropsychological performance over time. These findings suggest that recovery trajectories following mild to moderate COVID-19 infection may be favorable across variants, including Omicron.

At the initial assessment, both groups had similar levels of subjective cognitive impairment. The Alpha/Delta group reported a more severe acute illness as well as a greater number of medical comorbidities, which is consistent with prior reports.^14,15,44,45 The Alpha/Delta group was evaluated substantially farther from the time of acute infection compared to the Omicron group, which is reflective of the timeline of this study’s development. Despite this difference in time, early test performance was largely similar between variants.

At the initial assessment, both groups performed significantly below normative expectations on RBANS Immediate Memory. This is consistent with prior reports which have suggested impaired immediate memory is a frequent feature of PASC. Given normal scores on RBANS Attention, these findings likely reflect inefficiencies in learning or cognitive stamina, which is consistent with prior literature.^46–48 Immediate memory performance could potentially be used as a more sensitive indicator of PASC across variants.

The Omicron group additionally performed significantly lower than norms on TMT-A, which measures processing speed and simple attention, which is also consistent with prior literature.^45,46,48 Surprisingly, this was not seen in the Alpha/Delta group. No difference was found between either group and norms in measures of complex attention, delayed memory, or executive function. This is inconsistent with prior research, and may reflect the wide variation across studies of which testing batteries have been used, as well as heterogeneity between studies in terms of sample populations.^46,47

Follow-up data were available for over three-quarters of participants. Both groups demonstrated improvement in neuropsychological performance over time, with declining rates of classification-based impairment. At follow-up, neither group showed persistent deficits relative to normative values. These findings are consistent with earlier longitudinal work suggesting gradual cognitive recovery following COVID-19 infection.^4,44,49,50

General linear models revealed few significant interactions between NP Test scores, COVID-19 variant, illness severity and time. For RBANS Delayed Memory, time since acute infection predicted performance at follow-up, and a time-variant interaction was observed, in which the Alpha/Delta group showed greater improvement over time. This group was overall significantly further out from acute infection when compared to the Omicron group. Given the limited sample size and inconsistent broader patterns across measures, these results should be interpreted cautiously.

In parallel with neurocognitive outcomes, psychiatric distress and fatigue were largely similar between individuals infected during the Alpha/Delta and Omicron eras. Although participants in the Alpha/Delta group reported more severe acute illness and had more medical comorbidities, the groups did not differ in psychiatric treatment history, psychotropic medication use, or dimensional or categorical measures of depression, anxiety, PTSD, or fatigue in the months following infection. Taken together, these data suggest that individuals who experienced mild to moderate COVID-19 infection with Omicron are neither more nor less likely to experience psychiatric distress or fatigue compared to those infected during the Alpha/Delta period.

This finding is surprising for two reasons. First, pandemic-related psychosocial stresses were more prevalent during the Alpha/Delta-strain era, including increased morbidity and mortality, fear of contagion, the effects of quarantine, loss of employment and economic burdens.^51,52 Second, Alpha/Delta strains were more virulent and have been associated with greater acute and protracted inflammatory responses, which may increase risk for psychiatric complications and fatigue.⁵³ In contrast, vaccination and acute COVID-19 treatments were more widely available during the Omicron era, potentially mitigating downstream neuropsychiatric effects.^54,55 The absence of group differences likely reflects the generally mild to moderate severity of illness in both groups, with relatively short duration of acute symptoms and no requirement for hospitalization or specialized PASC care.

Psychiatric symptom levels observed in this study were broadly similar to pre-pandemic population estimates with respect to treatment history, medication use, anxiety symptoms, and fatigue. However, rates of depressive symptoms appeared higher than national prevalence estimates in both groups, suggesting that depression may remain a salient concern following mild to moderate COVID-19 infection regardless of variant. This observation warrants further investigation in larger samples with formal diagnostic assessments.^56,57

Several limitations should be acknowledged. The sample size was modest and demographically homogeneous, limiting generalizability. COVID-19 variant was inferred based on timing rather than virological sequencing. Psychiatric diagnoses were not established through structured clinical interviews. Vaccination status differed between groups due to real-world availability during different phases of the pandemic and may have influenced both acute illness severity and recovery trajectories. Multiple statistical comparisons were conducted, increasing risk of Type I error; however, effect sizes were reported to aid interpretation of clinical relevance. Strengths of the study include in-person, comprehensive neuropsychological assessment, longitudinal follow-up, and the ability to directly compare outcomes across COVID-19 variants using a consistent methodology.

Conclusion

This study aimed to determine whether neurocognitive outcomes differed between individuals infected with earlier COVID-19 variants (Alpha/Delta) and those infected with the Omicron variant, and to characterize longitudinal changes in cognitive performance. We hypothesized that individuals recovering from Omicron infection would demonstrate less severe or more rapidly improving neurocognitive deficits compared to earlier variants. However, our findings showed largely similar patterns of impairment and recovery across groups, with only limited variant-specific differences. Both groups demonstrated improvement in neuropsychological test performance over time, and no participant showed worsening performance.

Overall, participants recovered from both variants of COVID-19 demonstrated improvement in NP test scores over time, with no evidence of progressive cognitive decline and no persistent deficits relative to population norms at follow-up. There were no major differences in the pattern or persistence of cognitive deficits. These findings are reassuring and suggest that while mild impairment may be present for many months following infection, improvement is the predominant trajectory for individuals with mild-to-moderate COVID-19. Future studies should include larger and more diverse samples and longer follow-up periods to clarify the mechanisms, timing, and durability of cognitive recovery, as well as the neuropsychiatric impact of emerging variants.