Metacognition in Community-Dwelling Older Black and African American Adults During the COVID-19 Pandemic

Abstract

Background:

Cognitive assessment of older adults typically includes symptom reports and objective evaluations. However, there is often poor agreement between these measures. Cultural norms, stress, and anxiety may also influence cognitive self-appraisal and performance. Little research describes how other factors affect the self-report/objective test discrepancies noted in the literature.

Objective:

This study investigated whether the disparity between subjective cognitive concerns and objective cognitive performance is related to measures of anxiety and stress in older Black and African American adults.

Methods:

Telephone screenings were administered to 206 older adults (ages 64–94) during the first year of the pandemic. Demographic data, objective memory (Telephone Interview for Cognitive Status [TICS-m]), an adaptation of the subjective memory measure, the Cognitive Change Questionnaire, emphasizing executive functioning in everyday life [CCQ-e]), Generalized Anxiety Disorder-7 (GAD-7), and Perceived Stress Scale-4 (PSS4) were measured. Metacognition Discrepancy Index (MDI) was calculated from the standardized residual after regressing TICS-m on CCQ-e scores to quantify the discrepancy between cognitive self-appraisal and objective cognitive functioning.

Results:

Neither GAD-7 nor PSS-4 moderated the relationship between TICS-m and CCQ-e, and TICS-m scores weakly predicted subjective CCQ-e scores (F(1, 197)=4.37, p = 0.038, R² = 0.022). The MDI correlated with stress and anxiety (rs = 0.294, 0.396, ps < 0.001).

Conclusion:

Discrepancies exist between objectively measured and self-evaluated cognition. Elevations in stress and anxiety are associated with greater overestimation of cognitive difficulties relative to objective performance. Pandemic-related stressors may have worsened anxiety and diminished self-appraisal of cognitive abilities for some individuals, while others may remain reluctant to acknowledge impairments. Social and emotional factors are meaningful considerations in assessing cognitive difficulties.

Keywords

affect Alzheimer’s disease anxiety cognitive decline metacognition psychological stress

INTRODUCTION

Metacognition, the ability to accurately assess and reflect on one’s internal mental processes, is essential in evaluating cognitive resources and performance. A perceived change in cognitive ability is often the first symptom recognized by older adults who later develop mild cognitive impairment (MCI) or, eventually, dementia [1]. However, other etiologies of cognitive symptoms are possible [2]. Prompt identification of perceived cognitive decline is critical for clinicians to treat reversible conditions and for researchers to target cognitive health-promoting interventions during the earliest possible stage of the disease.

Clinicians frequently administer subjective and objective assessments of cognition [3] to differentiate older adults with bona fide cognitive impairment (e.g., Alzheimer’s disease (AD) or MCI) from persons expressing a subjective cognitive decline (SCD) [4]. There is often a poor agreement between subjective and objective measures, with small and inconsistent associations found in meta-analyses [5, 6]. Previous research suggests that subjective cognitive appraisal, mainly when supported by informant reports, may be more sensitive to MCI than objective cognitive performance measures [7]. However, cognitive self-appraisal may be influenced by anxiety, depression, stress, and memory loss [7 –11] (for meta-analysis and review, see [11]). Cognitive, psychological, and affective processes dynamically interact with the metacognitive process to influence the accuracy of self-appraisal ofcognition.

Metacognition

Metacognition is often conceptualized as a state of real-time appraisal of confidence in cognitive performance [12] or as “stable knowledge or beliefs about one’s own cognitive system” ([13], p. 302). It can be measured using questionnaires, confidence ratings, time judgments, or discrepancies between subjective ratings and objective proxies [14, 15]. Disparities in metacognition can be measured in magnitude and direction (over- or under-estimation of cognitive function relative to objective performance).

Inaccurate metacognition may contribute to psychological dysfunction [12]. For example, negative beliefs about worry concerning uncontrollability and danger may contribute to persistent and negative interpretations of experiences, such that perceived stress leads to more intense adverse emotional outcomes. Wells [16] reported that pathological worry maintained by maladaptive metacognitions may underlie aspects of the cognitive model of generalized anxiety disorder. Further, psychotic symptoms have been associated with failures in metacognitive processes due to overestimation, overconfidence, and suspending plausibility assessment [17]. Mood and affective state are associated with distortions in cognitive self-appraisal without objective cognitive impairments [8], even though affective symptoms do not independently increase the risk for cognitive decline [2].

Symptoms of cognitive dysfunction in the absence of objective confirmation are consistent with SCD [4], whereas a lack of awareness of objective deficit would indicate a different metacognition discrepancy, anosognosia [18]. SCD is associated with an increased risk for MCI and dementia [2], with more substantial predictive accuracy when corroborated by informant reports [1 , 20].

Little research has examined the correlates of metacognition discrepancies in older adults at risk for AD and even less in older Black and African American (B/AA) adults, who have reported a high incidence of race-related stressors [21]. In two studies focused on older B/AA adults, subjective cognitive complaints were associated with mood [22, 23]. Compared to non-Hispanic White participants, B/AA older adults report subjective cognitive complaints less frequently [24, 25], although the opposite trend has been noted for middle-aged adults [26]. Subjective cognitive complaints are not consistently associated with objective cognitive performance in older B/AA adults [22 –24], yet contrasting findings have also been reported [25]. A concomitant high incidence of mood disorder and stress [21] and greater risk for dementia [27] for B/AA older adults than non-Hispanic White adults is also observed. Given this confluence of risk factors, characterizing the interaction of reported mood symptoms with SCD among older B/AA adults is needed to interpret potential applications of metacognition assessments in community settings more effectively.

For older B/AA adults, COVID-19 pandemic-related challenges amplified existing disparities [28]. These disparate burdens may be attributable to reduced access to healthcare, increased risk for close personal contact through employment as an essential worker or living in high-density housing, and pre-existing comorbid health conditions [29 –31]. Beyond the physical risks, psychosocial stress increased for people from racially and ethnically diverse backgrounds during the pandemic, especially for females [32]. Because physical, psychological, and social context can influence cognitive performance and perceptions, our team examined the specific dynamics of these relationships in the context of the COVID-19 pandemic in a historically disadvantaged population.

This study examined how anxiety and stress perceptions moderate the relationship between perceived cognitive changes and objectively measured cognition in older B/AA adults. We evaluated associations between subjective and objective mental functioning with the expectation that these measures would be highly correlated. Next, we examined the impact of anxiety on measures of cognition. We hypothesized that anxiety levels would moderate the relationship between subjective and objective cognitive measures. Higher anxiety would be associated with more substantial discrepancies between elevated subjective burdens relative to objective cognitive measures. Lastly, we measured the relationship between perceived stress and cognitive measures. We expected higher perceived stress to be associated with an elevated level of perceived cognitive difficulties relative to actual cognitiveperformance.

METHODS

Participants and procedure

We administered telephone screenings to 206 older B/AA adults (aged 64–94 years) during the first year of the COVID-19 pandemic (from July 28, 2020, through the reopening of our laboratory on July 16, 2021). Our recruitment source was the Healthier Black Elders Center (HBEC) research registry, a community-based participatory research initiative from the Michigan Center for Urban African American Aging Research (a Resource Center for Minority Aging Research) located at Wayne State University’s Institute of Gerontology and sponsored by the National Institutes of Health [33]. Participants were eligible for participation if they were fluent English-language speakers, self-identified as Black or African American, and were at least 64 years old. We excluded participants with unstable telephone connections or profound hearing impairment. We assessed hearing status by asking participants to repeat a phrase accurately. Verbal assent for survey participation was obtained, and this study was conducted with the approval of the Wayne State University Institutional Review Board.

We used telephone surveys as a screening step to determine eligibility for a more extensive study. We recruited potential participants through a research registry and referrals from existing participants. Each survey took approximately 45-60 min to complete. Of the 482 participants contacted, 230 agreed to complete the survey, and 206 had complete data for the variables included in the primary analysis.

Materials and instruments

Subjective Assessment of Executive Cognitive Functioning –Change Questionnaire (CCQ-e)

An international team of researchers in China, Slovenia, and the United States designed and implemented survey questions to improve the understanding of the variety of pandemic responses and experiences of participants in different environments [34]. The final survey included several core measures which were common to all sites, as well as additional site-specific items. The instrument battery captured many aspects of the COVID-19 pandemic, including perceptions of stress, cognitive performance, personal experiences with COVID-19, general health status, emotions, anxiety, and vaccination beliefs. The analyses reported here include only subjective response data collected in the United States from a sample of B/AA participants from Wayne State University’s HBEC. We present only the subset of survey instruments relevant to our research questions.

The Cognitive Change Questionnaire component from the overall international measure includes nine items designed to measure perceived changes in cognitive functioning during the pandemic compared to pre-pandemic ability [34]. There were two components. The first five items are scored on a 7-point scale (1 –much less often to 7 –more often) and relate to fatigue and effects on completing activities of daily living. The last four questions are scored on a 7-point scale (1 - much easier to 7 –much harder) and assess subjective performance of daily tasks dependent on effective executive functioning (e.g., multi-tasking, planning, remembering details of multiple tasks, task switching). Higher scores reflect increased subjective cognitive complaints.

Prior to the study, we evaluated the factor structure of the original cognitive scale based on this US sample through latent variable analysis. Because of the ordinal nature of the original item response options, we computed polychoric correlations using principal components analysis to evaluate the component structure of the instrument [35]. The principal component analysis of the nine items yielded an uninterpretable solution. Thus, we focused on the four easier-harder questions because these questions’ content focused on subjective difficulty or ease with cognitive activities that are primarily dependent on executive functioning. The first principal component explained 68.8% of the variance, including all four questions, with an eigenvalue of 2.75. Using the factor loadings of the four items on this first component, a new score, labeled the CCQ-e, was computed for each participant. This score reflects participants’ subjective appraisal of their ability to perform cognitively demanding tasks.

Objective Cognitive Performance - Telephone Interview for Cognitive Status (TICS-m)

To assess objective cognitive performance, we used the TICS-m [36]. This 22-item instrument measures global cognition with questions about orientation, immediate and delayed recall, and executive function. TICS-m is a modified version of the more extended original TICS instrument [37]. Using a score below 32 as a cut point, TICS-m scores are both sensitive (82.4%) and specific (87.0%) to the identification of MCI [38]. Possible scores ranged from 0 to 50, with higher scores indicating better cognitive performance. A score below 32 was indicative of clinically significant cognitive impairment.

Anxiety - Generalized Anxiety Disorder - 7 (GAD-7)

We measured anxiety using the GAD-7 [39]. The GAD-7 is widely used as a screening instrument to detect self-reported anxiety. The GAD-7 includes seven items measured with responses on a four-point scale (ranging from never to nearly every day). Scores from the GAD-7 have demonstrated high internal consistency (Cronbach’s α = 88–89; [34, 40]. Higher scores indicate that participants experience anxiety more frequently. In clinical settings, scores of 5, 10, and 15 indicate mild, moderate, and severe anxiety, respectively, with scores of 10 or higher serving as a widely used screening cutpoint [41].

Stress Perceptions - Perceived Stress Scale 4 (PSS-4)

We evaluated perceived stress with the four-item PSS-4 [42]. The PSS-4 is a brief adaptation of the original self-report instrument [43]. Although it contains few items, PSS-4 scores show adequate reliability in previous studies (Cronbach’s α=0.60–0.75; [34 , 45]. The PSS-4 measures generalized stress and a sense of control over self-appraised important features in life. Participants endorse each item with a rating between 1 –never, and 5 –very often. Higher scores indicate elevated perceived stress or that perceived demands exceed perceived coping ability [42].

Demographics

Participants reported their sex, age (in years), years of education completed, status of current employment (employed, retired), number of household residents (1, 2, 3, or more), and general perceived health status (poor, fair, good excellent) in Table 1.

Table 1
MDI differences by demographic group

Variable n (%) M (SD) Tests for group differences Group MDI M (SD) Effect size

Sex t (197) =0.50, p=0.31 d = 0.12

Male 18 (9) 0.11 (0.94)

Female 188 (91) -0.01 (1.00)

Age (y) 200 75.4 (7.2) r = 0.03, p=0.71 η2 = 0.03

Education (y) 114 15.5 (2.2) r = 0.13, p=0.18 η2 = 0.04

Occupation t (192) =0.58, p=0.34

Employed 29 (14.1) 0.08 (0.87) d = 0.09

Retired 166 (80.6) -0.01 (1.00)

Members of household F (2, 196) =2.05, p=0.13, MSE=0.071 η2 = 0.02

1 124 (60.2) -0.10 (1.00)

2 53 (25.7) 0.09 (0.86)

3 or more 23 (11.2) 0.33 (1.18)

General Health F (3, 195)=3.89, p = 0.01, MSE = 0.07 η2 = 0.06

Poor 3 (1.5) -0.22 (0.14)

Fair 66 (32.0) 0.32 (1.11)

Good 116 (56.3) -0.13 (0.95)

Excellent 15 (7.3) -0.39 (0.41)

Variable	n (%)	M (SD)	Tests for group differences	Group MDI M (SD)	Effect size
Sex			t (197) =0.50, p=0.31		d = 0.12
Male	18 (9)			0.11 (0.94)
Female	188 (91)			-0.01 (1.00)
Age (y)	200	75.4 (7.2)	r = 0.03, p=0.71		η2 = 0.03
Education (y)	114	15.5 (2.2)	r = 0.13, p=0.18		η2 = 0.04
Occupation			t (192) =0.58, p=0.34
Employed	29 (14.1)			0.08 (0.87)	d = 0.09
Retired	166 (80.6)			-0.01 (1.00)
Members of household			F (2, 196) =2.05, p=0.13, MSE=0.071		η2 = 0.02
1	124 (60.2)			-0.10 (1.00)
2	53 (25.7)			0.09 (0.86)
3 or more	23 (11.2)			0.33 (1.18)
General Health			F (3, 195)=3.89, p = 0.01, MSE = 0.07		η2 = 0.06
Poor	3 (1.5)			-0.22 (0.14)
Fair	66 (32.0)			0.32 (1.11)
Good	116 (56.3)			-0.13 (0.95)
Excellent	15 (7.3)			-0.39 (0.41)

MDI, Metacognition Discrepancy Index.

Data analysis

We examined each variable’s central tendency and distribution measures (Table 1). We computed Welch’s t tests and ANOVAs (Table 1) to compare metacognition discrepancy indices across subsamples based on self-reported health status and demographic groups (sex, age, education, occupation, household members). We also calculated scale reliabilities and descriptive statistics for each study variable (Table 2). We tested hypotheses using linear and multiple regression and correlation with an alpha level of 0.05. To quantify the magnitude and direction of metacognition discrepancies, we measured the mismatch between subjective and objective cognitive measures as the standardized residual variance resulting from the regression of the objective TICS-m on the subjective CCQ-e. We, therefore, renamed this standardized residual a Metacognition Discrepancy Index (MDI). Negative MDI values indicated more observed impairment than expected from the level of subjective complaints; positive MDI values reflected less objective impairment than expected from subjective complaints. We analyzed data with IBM SPSS Statistics, version 28, and Stata, version 16.1 [46].

Table 2

Descriptive statistics of other measures

Variable	Number of items	Possible range	Observed range	n	M	SD	Skewness	Kurtosis	α
TICS-M	22	0–50	18–49	206	36.5	5.36	-0.097	0.297	0.78
CCQ-e	9	9–63	15–43	198	28.3	4.66	0.335	0.920	0.75
GAD-7	7	0–21	0–21	200	6.2	0.52	0.718	–0.269	0.90
PSS-4	4	4–20	4–17	199	8.0	2.80	0.474	–0.600	0.57

TICS-M, Telephone Interview for Cognitive Status; CCQ-e, Change Questionnaire; GAD-7, Generalized Anxiety Disorder - 7; PSS-4, Perceived Stress Scale 4; α, Cronbach’s α coefficient of internal consistency.

RESULTS

Demographics

Table 1 displays complete demographic data for the final sample. Participants were primarily female (91%). Participants’ ages ranged from 64 to 94 years (M = 75.4, SD = 7.2). Most had received at least some university education (years of education ranged from 12-20). Participants were generally retired from formal employment (80.6%), but 14.5% maintained an actively employed job status. Household composition and residence type varied. Most participants lived alone (60.2%) and in single-family, detached dwellings (63%), although some lived in households of up to 9 individuals. Generally, participants rated their health status as “good” (56%).

3.2 Anxiety and cognition

To evaluate the role of anxiety in subjective cognitive complaints, we tested GAD-7 as a moderator of the TICS-m and CCQ-e relationship. The interaction between GAD-7 and TICS-m did not demonstrate a significant moderating effect on the linear relationship between TICS-m and CCQ-e (see Table 3A, b = 0.005, SE = 0.003, p = 0.10). However, GAD-7 scores correlated moderately with the MDI (r(197)=0.40, p < 0.001; Fig. 1A).

Fig. 1

Correlates of Metacognition Discrepancy Index. Correlates of the Metacognition Discrepancy Index (MDI) are depicted. Above and to the right of the scatterplots, shaded areas represent histogram data distributions. A) MDI and General Anxiety (GAD-7) correlations are visualized with a scatter plot (r(197)=0.40, p < 0.001; shaded area shows 95% CI). B) MDI and Perceived Stress (PSS-4) are visualized with a scatter plot (r(197)=0.33, p < 0.001; shaded area shows 95% CI).

Table 3

A) Moderator Analysis: Anxiety and Cognition predicting CCQ-e

Effect	Estimate	SE	95% CI		P
			LL	UL
Intercept	8.23	0.88	6.50	9.96	< 0.001^*
TICS-m	–0.06	0.02	–0.11	–0.02	0.01^*
GAD-7	–0.07	0.10	–0.28	0.13	0.50
TICS-m^*GAD-7	0.005	0.003	–0.001	0.01	0.10

Perceived stress and cognition

We examined perceived stress as a potential moderator of the relation between subjective and objective cognitive functioning. There was no interaction effect between the PSS-4 total score and the TICS-m (see Table 3B; b = 0.01, SE = 0.01, p = 0.16), indicating that PSS-4 scores did not moderate the linear relationship between TICS-m and CCQ-e. In contrast, the PSS-4 and the MDI were moderately correlated (r(197) = 0.33, p < 0.001; Fig. 1B).

Table 3

B) Moderator Analysis: Perceived Stress and Cognition predicting CCQ-e

Effect	Estimate	SE	95% CI		p
			LL	UL
Intercept	8.39	1.74	4.95	11.83	<0.001^*
TICS-m	–0.08	0.05	–0.17	0.01	0.08
PSS-4	–0.12	0.20	–0.51	0.27	0.55
TICS-m^*PSS-4	0.01	0.01	–0.003	0.02	0.16

CI, confidence interval; LL, lower limit; UL, upper limit. ^*p<0.05.

Group differences

Table 1 illustrates comparisons in average discrepancy scores and tests for group differences. We found no significant group differences in average MDI between sexes or employment status using t tests and no significant differences in average MDI between individuals living alone, with a partner, or with others using ANOVA. Years of age, education, and MDI were uncorrelated. Individuals reporting different perceived health statuses reflected the only group differences in mean MDI (F(3, 195)=3.89, p = 0.01; Fig. 2). Those who acknowledged “Fair” health had the highest average MDI (M = 0.32, SD = 1.11), and those endorsing “Excellent” health had the most negative MDI (M = -0.39, SD = 0.41).

Fig. 2

MDI by General Health Group. Metacognition Discrepancy Index (MDI) mean and dispersion plotted by self-reported perceived health groups.

DISCUSSION

No significant stress or anxiety moderation effect was observed when examining the relationship between subjective and objective cognitive functioning. Our analyses suggest that, although not wholly explanatory, subjective and objective measures of cognition are related. The discrepancy between subjectively and objectively measured cognitive performance is related to anxiety and perceived stress, such that increases in these affective experiences occur alongside increased perceived cognitive difficulty beyond expectation based on objective cognitive performance. There were no observable differences in MDI between any of the examined subgroups except for individuals with differences in perceived health status. Participants who reported “Fair” health demonstrated the highest MDI, and those that reported “Excellent” health status showed the most negativeMDI.

Metacognition: Subjective complaints and objective measures

The direction of the metacognitive discrepancy may provide clinically meaningful information beyond what can be independently ascertained through symptom reports or cognitive tests. An MDI of zero indicates a perfect agreement between objective and subjective cognitive measures. A positive discrepancy corresponds to an excess of subjective cognitive complaints relative to what would be expected based on objective criteria, not unlike SCD. In contrast, a negative MDI describes a state similar to anosognosia in which an individual’s subjective report does not acknowledge measured cognitive impairments.

Previous studies of subjective and objective cognitive measures in older adults report inconsistent findings. Diverse populations have often been excluded from cognitive aging research and studies of metacognition [24, 27]. In a study of a diverse sample, Jackson and colleagues [24] observed different patterns of inaccurate metacognition across racial groups. Although all older adult participants reported similar levels of cognitive concerns, non-Hispanic white participants’ self-reported cognitive concerns were significantly related to poorer measurements of objective performance. In contrast, no similar associations were found among B/AA participants. These differential findings were attributed to differences in culture or lifestyle factors. Some evidence also suggests that the reliability of subjective reports in predicting objective performance may be compromised in older adults with MCI compared to normal cognition [47], which further complicates assessment efforts in this subset of the population with significant needs for accurate results. Across all samples, statistically detectable associations between subjective ratings and cognitive performance are weak and explain only a small portion of variance by this relationship alone [5 , 48]. The remaining variance must be attributable to other physical, psychological, or social factors. This study further explored two possible psychological factors, anxiety, andstress.

Metacognition and psychological factors

Anxiety represents a psychological burden that may interrupt the process of cognitive self-evaluation and reporting. Even though the MDI, as a composite measure of the discrepancy between the subjective and objective evaluation of cognitive ability, did not detect the moderating effect of anxiety, our findings align with previous studies cited above that found anxiety to be associated with metacognition. Our study shows that in older Black/African American participants, higher self-reported anxiety was associated with more challenging self-perceived cognitive difficulties, primarily executive functioning. Previous studies of the relationships between anxiety and the closely related concepts of worry and negative affect have consistently demonstrated the detrimental effect of these emotional processes on subjective, self-reported cognitive status [1 , 20]. However, previous studies have not explained the influences of psychological burdens on metacognition to contextualize the nature of cognitive complaints within objective cognitive performance. Other samples have failed to support a relationship between anxiety and subjective complaints when studied across all age groups[34].

Similarly, stress impacts both subjective and objective functions. Our finding that elevated self-perceived stress is associated with more self-perceived executive functioning difficulties aligns with other studies that found that acute and chronic stress is associated with self-perceived cognitive problems. In its acute and chronic stages, stress affects cognitive functions such as planning, working memory, attention, and cognitive flexibility [49, 50]. The combination of daily and pandemic-related stressors may also promote subjective cognitive impairment in samples of older and younger adults [34, 51].

The mechanisms underlying the relationships between anxiety or stress and metacognition are not known. In situations with measurable cognitive decline, physical neurodegeneration, chronic health conditions, or physiological aging processes may explain some metacognition discrepancies arising from objective reductions in cognitive performance [52 –54]. In these scenarios, individuals’ perceptions of cognitive decline may be denied or repressed due to fears about AD [18]. This position is congruent with our finding of group differences in perceived health status: those who already perceived poorer overall health reported more significant cognitive difficulties than expected for their objective performance. In contrast, those who perceived having an optimal health status acknowledged fewer cognitive complaints than expected from their actual performance.

In other cases, psychological and social phenomena may influence metacognition by causing increased vigilance and sensitivity to expected age-related changes, motivating an increased utilization of memory care evaluations and a higher likelihood of a diagnosis of MCI or AD [10]. Both objective and subjective cognition could be at risk for decline during heightened stress or anxiety. These stressful circumstances may divert already limited cognitive resources for processing stress- and anxiety-related thoughts and emotions, reducing the ability to allocate cognitive capacity to the performance of cognitive tasks or accurate metacognitive evaluation [54].

4.3 Metacognition and contextual factors

In a COVID-19 telephone outreach initiative conducted by the HBEC, our target population expressed their experiences with issues such as grief, difficulty obtaining necessities through reduced in-person shopping outings, and confusion about COVID-19 [55]. Many participants also expressed a need for further information or specific resources [55]. Considering the likely combination of the acute stress of the pandemic and the chronic stress of structural racism within this particular population [21 , 29–32], the magnitude of perceived stress was surprisingly low in this survey. The extent to which practical stress-related coping skills support accurate metacognition remains unknown. Further, although this study did not measure COVID-19 exposure or recovery, accumulating evidence demonstrates the physical risks to subjective and objective cognition related to SARS-CoV-2 infection [56 –59].

Limitations

The cross-sectional sample of participants limits the extent to which conclusions can be drawn about the long-term clinical utility of the MDI over time. We collected data remotely with telephone interviews, which may threaten the reliability of the measures compared to better-validated, traditional, pencil-and-paper cognitive evaluation. However, when considering both the physical risks and sequelae of COVID-19 and the technological limitations of older adults with limited financial means, telephone contact was the most ethically responsible and inclusive communication during the pandemic [55].

Our study focused only on older B/AA participants from a single urban metropolitan area, which limits the ability to draw comparisons with other groups and may limit the generalizability of these findings to other populations. Further, the sample included primarily women (91%), another limitation to the generalizability of these findings. Participants in this study were volunteers, and the risk of self-selection bias is possible. In studies where recruitment is not population-based, altruism or volunteerism can further complicate bias. In a study of aging and cognition [41, 60], volunteers, in comparison to randomly selected participants, were significantly more likely to be women, more educated, demonstrate higher cognitive performance, and have fewer Instrumental Activities of Daily Living (IADL) impairments. Participants in this non-randomly selected sample may be similarly differentiated from the general population. In addition, several studies using cognitive tasks have shown gender differences in metacognition, demonstrating that relative to women, men generally underestimated their performance [61, 62]. However, females have been more precise in questioning the author’s credibility and the source of the article they read than the male participants [63]. Future research should further explore the potential effects of selection bias related to investigations of the association between subjective and objective measures of cognition.

Lastly, additional contextual covariates, such as depression [5 , 18], comorbid health conditions, or formal dementia diagnosis, were not measured in these analyses. Other unmeasured determinates of the relationship between subjective cognitive complaints and objective cognitive status may exist.

Notably, general SCD is not always indicative of disease-related neurodegenerative processes. In many cases, SCD may reflect expected age-related changes or the harmful cognitive effects of chronic health conditions rather than impending AD [52, 53]. Considering the potential etiologies of SCD and the limited reliability of self-reported cognition, it is essential to emphasize that a subjective cognitive complaint does not always indicate preclinical AD.

Conclusion

This study demonstrated that subjective and objective measures of cognitive status are related and that metacognition discrepancy has clinically relevant interpretations. The MDI, which anchors subjective cognitive symptoms in the context of objectively measured cognitive performance, may provide more comprehensive insights into the individual’s experience than either subjective or objective reports measured independently. Furthermore, metacognition discrepancies relate to affective experiences such as anxiety and perceived stress. These findings also have several practical implications. First, while subjective and objective measures are critical data sources linked to an individual’s cognitive capacity, subjective complaints are more uniformly interpretable when subjective complaints are anchored within the context of objectively measured cognitive performance. Second, during experiences of heightened anxiety or stress, an overestimation of cognitive difficulties relative to actual capability is likely. Future studies may extend these findings by investigating the potential for focused interventions to reduce anxiety and stress, which may, in turn, reduce metacognition discrepancies. Further testing of such interventions in persons with different cognitive diagnoses or psychosocial backgrounds may also help realign subjective and objective perceptions. Clinical diagnosis and future research should also consider the association between subjective and objective cognitive decline within the context of neural, psychological, and social factors.

Footnotes

ACKNOWLEDGMENTS

The authors thank the research participants for participating in this study and the research staff for their recruitment and data collection efforts.

FUNDING

We thank our funders for the funding programs that support this research, including the NIA/NIH, R01 AG 054484 (PI, V. Kavcic) and partial support for K.K. and B.G. from NIH/NIA grant to the Michigan Alzheimer’s Disease Research Center (P30AG053760; PI, H. Paulson).

CONFLICT OF INTEREST

The authors have no conflict of interest to disclose.

DATA AVAILABILITY

The data supporting the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.

References

Jessen

(2010) Prediction of dementia by subjective memory impairment effects of severity and temporal association with cognitive impairment dementia and subjective memory impairment. Arch Gen Psychiatry 67, 414.

Pike

, Cavuoto

, Li

, Wright

, Kinsella

(2022) Subjective cognitive decline: Level of risk for future dementia and mild cognitive impairment, a meta-analysis of longitudinal studies. Neuropsychol Rev 32, 703–735.

Albert

, DeKosky

, Dickson

, Dubois

, Feldman

, Fox

, Gamst

, Holtzman

, Jagust

, Petersen

, Snyder

, Carrillo

, Thies

, Phelps

(2011) The diagnosis of mild cognitive impairment due to Alzheimer’s disease: Recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement 7, 270–279.

Jessen

, Amariglio

, Boxtel

, Breteler

, Ceccaldi

, Chételat

, Dubois

, Dufouil

, Ellis

, Flier

, Glodzik

, Harten

, Leon

, McHugh

, Mielke

, Molinuevo

, Mosconi

, Osorio

, Perrotin

, Petersen

, Rabin

, Rami

, Reisberg

, Rentz

, Sachdev

, Sayette

, Saykin

, Scheltens

, Shulman

, Slavin

, Sperling

, Stewart

, Uspenskaya

, Vellas

, Visser

, Wagner

, Subjective Cognitive Decline Initiative (SCD-I) Working Group (2014) A conceptual framework for research on subjective cognitive decline in preclinical Alzheimer’s disease. Alzheimers Dement 10, 844–852.

Burmester

, Leathem

, Merrick

(2016) Subjective cognitive complaints and objective cognitive function in aging: A systematic review and meta-analysis of recent cross-sectional findings. Neuropsychol Rev 26, 376–393.

Crumley

, Stetler

, Horhota

(2014) Examining the relationship between subjective and objective memory performance in older adults: A meta-analysis. Psychol Aging 29, 250–263.

Rami

, Mollica

, García-Sanchez

, Saldaña

, Sanchez

, Sala

, Valls-Pedret

, Castellví

, Olives

, Molinuevo

(2014) The Subjective Cognitive Decline Questionnaire (SCD-Q): Avalidation study. J Alzheimers Dis 41, 453–466.

Dux

, Woodard

, Calamari

, Messina

, Arora

, Chik

, Pontarelli

(2008) The moderating role of negative affect on objective verbal memory performance and subjective memory complaints in healthy older adults. J Int Neuropsychol Soc 14, 327–336.

Jenkins

, Tree

, Tales

(2021) Distinct profile differences in subjective cognitive decline in the general public are associated with metacognition, negative affective symptoms, neuroticism, stress, and poor quality of life. J Alzheimers Dis 80, 1231–1242.

10.

Norman

, Woodard

, Calamari

, Gross

, Pontarelli

, Socha

, DeJong

, Armstrong

(2020) The fear of Alzheimer’s disease: Mediating effects of anxiety on subjective memory complaints. Aging Mental Health 24, 308–314.

11.

Desai

, Whitfield

, Said

, John

, Saunders

, Marchant

, Stott

, Charlesworth

(2021) Affective symptoms and risk of progression to mild cognitive impairment or dementia in subjective cognitive decline: A systematic review and meta-analysis. Ageing Res Rev 71, 101419.

12.

Spada

, Nikčcević

, Moneta

, Wells

(2008) Metacognition, perceived stress, and negative emotion. PersIndivid Dif 44, 1172–1181.

13.

Wells

(1995) Meta-cognition and worry: A cognitive model of generalized anxiety disorder. Behav Cogn Psychother 23, 301–320.

14.

Hallam

, Chan

, Gonzalez Costafreda

, Bhome

, Huntley

(2020) What are the neural correlates of meta-cognition and anosognosia in Alzheimer’s disease? A systematic review. Neurobiol Aging 94, 250–264.

15.

Kononowicz

, Roger

, Van Wassenhove

(2019) Temporal metacognition as the decoding of self-generated brain dynamics. Cereb Cortex 29, 4366–4380.

16.

Wells

, Carter

(2001) Further tests of a cognitive model of generalized anxiety disorder: Metacognitions and worry in GAD, panic disorder, social phobia, depression, and nonpatients. Behav Ther 32, 85–102.

17.

Gohari

, Moore

, Depp

, Ackerman

, Pinkham

, Harvey

(2022) Momentary severity of psychotic symptoms predicts overestimation of competence in domains of everyday activities and work in schizophrenia: An ecological momentary assessment study. Psychiatry Res 310, 114487.

18.

Tagai

, Shinagawa

, Kada

, Inamura

, Nagata

, Nakayama

(2018) Anosognosia in mild Alzheimer’s disease is correlated with not only neural dysfunction but also compensation: Anosognosia in mild AD. Psychogeriatrics 18, 81–88.

19.

Rahman-Filipiak

, Giordani

, Heidebrink

, Bhaumik

, Hampstead

(2018) Self- and informant-reported memory complaints: Frequency and severity in cognitively intact individuals and those with mild cognitive impairment and neurodegenerative dementias. J Alzheimers Dis 65, 1011–1027.

20.

Wion

, Hill

, Reed Bell

, Mogle

, Yates

, Bhang

(2022) The role of cognitive self-report measure type in predicting cognitive decline among older adults: A systematic review. J Geriatr Psychiatry Neurol 35, 487–511.

21.

Williams

(2018) Stress and the mental health of populations of color: Advancing our understanding of race-related stressors. J Health Soc Behav 59, 466–485.

22.

Sperling

, Tsang

, Williams

, Park

, Helenius

, Manning

(2017) Subjective memory change, mood, and cerebrovascular risk factors in older African Americans. J Geriatr Psychiatry Neurol 30, 324–330.

23.

Sims

, Whitfield

, Ayotte

, Gamaldo

, Edwards

, Allaire

(2011) Subjective memory in older African Americans. Exp Aging Res 37, 220–240.

24.

Jackson

, Rentz

, Aghjayan

, Buckley

, Meneide

T-F

, Sperling

, Amariglio

(2017) Subjective cognitive concerns are associated with objective memory performance in Caucasian but not African-American persons. Age Ageing 46, 988–993.

25.

John

, Evans

, Hanfelt

, Loring

, Goldstein

(2020) Subjective memory complaints in white and African American participants. J Geriatr Psychiatry Neurol 33, 135–143.

26.

Gupta

(2021) Racial and ethnic disparities in subjective cognitive decline: A closer look, United States, 2015-2018. BMC Public Health 21, 1173.

27.

Weuve

, Barnes

, Mendes De Leon

, Rajan

, Beck

, Aggarwal

, Hebert

, Bennett

, Wilson

, Evans

(2018) Cognitive aging in Black and White Americans: Cognition, cognitive decline, and incidence of Alzheimer disease dementia. Epidemiology 29, 151–159.

28.

Daugherty

, Chopra

, Korzeniewski

, Levy

(2020) COVID-19 as a risk factor for Alzheimer’s disease and related dementia: A perspective from Detroit, MI. Psychiatry Res 294, 113557.

29.

Schulz

, Mehdipanah

, Chatters

, Reyes

, Neblett

, Israel

(2020) Moving health education and behavior upstream: Lessons from COVID-19 for addressing structural drivers of health inequities. Health Educ Behav 47, 519–524.

30.

Farley

, Hines

, Lee

, Brooks

, Nair

, Brown

, Doll

, Sullivan

, Chapman-Davis

(2020) Promoting health equity in the era of COVID-19. Gynecol Oncol 158, 25–31.

31.

Gray

, Anyane-Yeboa

, Balzora

, Issaka

, May

(2020) COVID-19 and the other pandemic: Populations made vulnerable by systemic inequity. Nat Rev Gastroenterol Hepatol 17, 520–522.

32.

Thomas

, Osbourne

, Appelhans

, Roisman

, Booth-LaForce

, Bleil

(2022) Disparities in COVID-19-related stressful life events in the United States: Understanding who is most impacted. Health Social Care Comm 30, 1199–1211.

33.

Mitchell

, Perry

, Rorai

, Ilardo

, Lichtenberg

, Jackson

(2020) Building and sustaining a community advisory board of African American older adults as the foundation for volunteer research recruitment and retention in health sciences. Ethn Dis 30, 755–764.

34.

Podlesek

, Komidar

, Kavcic

(2021) The relationship between perceived stress and subjective cognitive decline during the COVID-19 epidemic. Front Psychol 12, 647971.

35.

Byrne

, Byrne

(2013) Structural Equation Modeling With EQS, Routledge.

36.

van den Berg

, Ruis

, Biessels

, Kappelle

, van Zandvoort

MJE

(2012) The Telephone Interview for Cognitive Status (Modified): Relation with a comprehensive neuropsychological assessment. J Clin Exp Neuropsychol 34, 598–605.

37.

Brandt

, Spencer

, Folstein

(2022) Telephone Interview for Cognitive Status. APA PsycTests.

38.

Cook

, Marsiske

, McCoy

KJM

(2009) The use of the Modified Telephone Interview for Cognitive Status (TICS-M) in the detection of amnestic mild cognitive impairment. J Geriatr Psychiatry Neurol 22, 103–109.

39.

Spitzer

, Kroenke

, Williams

JBW

, Löwe

(2006) A brief measure for assessing generalized anxiety disorder: The GAD-7. Arch Intern Med 166, 1092.

40.

Löwe

, Decker

, Müller

, Brähler

, Schellberg

, Herzog

, Herzberg

(2008) Validation and standardization of the Generalized Anxiety Disorder Screener (GAD-7) in the general population. Med Care 46, 266–274.

41.

Williams

(2014) The GAD-7 questionnaire. Occup Med 64, 224–224.

42.

Warttig

, Forshaw

, South

, White

(2013) New, normative, English-sample data for the Short Form Perceived Stress Scale (PSS-4). J Health Psychol 18, 1617–1628.

43.

Cohen

, Kamarck

, Mermelstein

(1983) A global measure of perceived stress. J Health Soc Behav 24, 385–396.

44.

Ingram

, Clarke

, Lichtenberg

(2016) Confirmatory factor analysis of the Perceived Stress Scale-4 in a community sample: PSS-4 CFA. Stress Health 32, 173–176.

45.

Mitchell

, Crane

, Kim

(2008) Perceived stress in survivors of suicide: Psychometric properties of the Perceived Stress Scale. Res Nurs Health 31, 576–585.

46.

(2019) Stata Statistical Software.

47.

Park

, Lee

J-H

, Lee

, Cho

, Park

, Yoo

, Youn

J-H

, Ryu

S-H

, Hwang

, Kim

, Lee

J-Y

(2019) Interactions between subjective memory complaint and objective cognitive deficit on memory performances. BMC Geriatr 19, 294.

48.

Fausto

, Badana

ANS

, Arnold

, Lister

, Edwards

(2018) Comparison of subjective and objective measures of hearing, auditory processing, and cognition among older adults with and without mild cognitive impairment. J Speech Lang Hear Res 61, 945–956.

49.

Chattarji

, Tomar

, Suvrathan

, Ghosh

, Rahman

(2015) Neighborhood matters: Divergent patterns of stress-induced plasticity across the brain. Nat Neurosci 18, 1364–1375.

50.

Shields

, Sazma

, McCullough

, Yonelinas

(2017) The effects of acute stress on episodic memory: A meta-analysis and integrative review. Psychol Bull 143, 636–675.

51.

Fiorenzato

, Zabberoni

, Costa

, Cona

(2021) Cognitive and mental health changes and their vulnerability factors related to COVID-19 lockdown in Italy. PLoS One 16, e0246204.

52.

Gagnon

, Baril

A-A

, Montplaisir

, Carrier

, De Beaumont

, D’Aragon

, Chami

, Pelleieux

, Poirier

, Gauthier

, Lafond

, Gagnon

J-F

, Gosselin

(2019) Disconnection between self-reported and objective cognitive impairment in obstructive sleep apnea. J Clin Sleep Med 15, 409–415.

53.

Salthouse

(2019) Trajectories of normal cognitive aging. Psychol Aging 34, 17–24.

54.

Boals

, Banks

(2020) Stress and cognitive functioning during a pandemic: Thoughts from stress researchers. Psychol Trauma 12, S255–S257.

55.

Rorai

, Perry

, Whitney

, Gianfermi

, Mitchell

, Key

, Lichtenberg

, Taylor

, Ilardo

, Knurek

, Conyers

(2021) “It takes some empathy, sympathy, and listening”: Telephone outreach to older detroiters in a pandemic as a modality to gain an understanding of challenges and resiliency. J Urban Health 98, 91–102.

56.

Liu

Y-H

, Chen

, Wang

Q-H

, Wang

L-R

, Jiang

, Yang

, Chen

, Li

, Cen

, Xu

, Zhu

, Li

, Wang

Y-R

, Zhang

L-L

, Liu

, Xu

Z-Q

, Wang

Y-J

(2022) One-year trajectory of cognitive changes in older survivors of COVID-19 in Wuhan, China: A longitudinal cohort study. JAMA Neurol 79, 509.

57.

Kas

, Soret

, Pyatigoskaya

, Habert

M-O

, Hesters

, Le Guennec

, Paccoud

, Bombois

, Delorme

, on the behalf of CoCo-Neurosciences study group and COVID SMIT PSL study group Delorme

, Corvol

J-C

, Delattre

J-Y

, Carvalho

, Sagnes

, Dubois

, Navarro

, Louapre

, Stojkovic

, Idbaih

, Rosso

, Grabli

, Gales

, Millet

, Rohaut

, Bayen

, Dupont

, Bruneteau

, Lehericy

, Seilhean

, Durr

, Kas

, Lamari

, Houot

, Brochard

, Dupont

, Lubetzki

, Seilhean

, Pradat-Diehl

, Rosso

, Hoang-Xuan

, Fontaine

, Naccache

, Fossati

, Arnulf

, Durr

, Carpentier

, Lehericy

, Edel

, Di Stefano

, Robain

, Thoumie

, Degos

, Sharshar

, Alamowitch

, Apartis-Bourdieu

, Peretti

C-S

, Ursu

, Dzierzynski

, Bourron

, Belmin

, Oquendo

, Pautas

, Verny

, Delorme

, Corvol

J-C

, Delattre

J-Y

, Samson

, Leder

, Leger

, Deltour

, Baronnet

, Gales

, Bombois

, Touat

, Idbaih

, Sanson

, Dehais

, Houillier

, Laigle-Donadey

, Psimaras

, Alenton

, Younan

, Villain

, Grabli

, Del Mar Amador

, Bruneteau

, Louapre

, Mariani

L-L

, Mezouar

, Mangone

, Meneret

, Hartmann

, Tarrano

, Bendetowicz

, Pradat

P-F

, Baulac

, Sambin

, Salachas

, Le Forestier

, Pichit

, Chochon

, Hesters

, Nguyen

BHH

, Procher

, Demoule

, Morawiec

, Mayaux

, Faure

, Ewenczyk

, Coarelli

, Heinzmann

, Charles

, Stojkovic

, Masingue

, Bassez

, Navarro

, An

, Worbe

, Lambrecq

, Debs

, Musat

, Lenglet

, Lambrecq

, Hanin

, Chougar

, Shor

, Pyatigorskaya

, Galanaud

, Leclercq

, Demeret

, Rohaut

, Cao

, Marois

, Weiss

, Gassama

, Le Guennec

, Degos

, Jacquens

, Similowski

, Morelot-Panzini

, Rotge

J-Y

, Saudreau

, Millet

, Pitron

, Sarni

, Girault

, Maatoug

, Gales

, Leu

, Bayen

, Thivard

, Mokhtari

, Plu

, Gonçcalves

, Bottin

, Yger

, Ouvrard

, Haddad

, Ketz

, Lafuente

, Oasi

, Megabarne

, Herve

, Salman

, Rametti-Lacroux

, Chalançon

, Herve

, Royer

, Beauzor

, Maheo

, Laganot

, Minelli

, Fekete

, Grine

, Biet

, Hilab

, Besnard

, Bouguerra

, Goudard

, Houairi

, Al-Youssef

, Pires

, Oukhedouma

, Siuda-Krzywicka

, Malkinson

, Agguini

, Douzane

, Said

, Houot

(2021) The cerebral network of COVID-19-related encephalopathy: A longitudinal voxel-based 18F-FDG-PET study. Eur J Nucl Med Mol Imaging 48, 2543–2557.

58.

Solomon

, Heyman

, Ko

, Condos

, Lynch

(2021) CT of post-acute lung complications of COVID-19. Radiology 301, E383–E395.

59.

Evans

, Leavy

, Richardson

, Elneima

, McAuley

HJC

, Shikotra

, Singapuri

, Sereno

, Saunders

, Harris

, et al. (2022) Clinical characteristics with inflammation profiling of long COVID and association with 1-year recovery following hospitalisation in the UK: A prospective observational study. Lancet Respir Med 10, 761–775.

60.

Ganguli

, Lytle

, Reynolds

, Dodge

(1998) Random versus volunteer selection for a community-based study. J Gerontol A Biol Sci Med Sci 53A, M39–M46.

61.

Hertzog

, Dixon

, Hultsch

(1990) Relationships between metamemory, memory predictions, and memory task performance in adults. Psychol Aging 5, 215–227.

62.

Lemieux

, Collin

, Watier

(2019) Gender differences in metacognitive judgments and performance on a goal-directed wayfinding task. J Cogn Psychol 31, 453–466.

63.

Srinivasan

, Crooks

(2005) Does gender influence critical thinking attitudes? In Association for the Advancement of Computing in Education (AACE), pp. 3376-3382.