Reduced awareness of cognitive deﬁcits in idiopathic normal pressure hydrocephalus and its change following lumbar puncture

Objective: Awareness of cognitive deﬁcits is related to executive functions and may, therefore, be sensitive to the effects of lumbar puncture (LP) in idiopathic normal pressure hydrocephalus (iNPH). Although a reduction in awareness of cognitive deﬁcits (RACD) has been previously described in iNPH, there is a lack of systematic, psychometrically validated reports. In this study, we investigated RACD and its LP-related changes in iNPH patients and compared them with those in clinical and healthy control groups. Methods: RACD was assessed before and after lumbar puncture (LP) in 24 patients (14 iNPH, 10 other age-associated cognitive syndromes; AACS) and compared with 23 healthy controls (HC), employing two RACD measures alongside cognitive examination. Local meta-cognition was measured using a visual percentile-based rating system and operationalized as the t-scaled distance between the participants ' task-speciﬁc performance estimations and their objective test performance ( D TSPE). Global metacognition, targeting broader estimates of cognitive functioning (ECF), was quantiﬁed by subtracting self-from informant-obtained sum scores on a questionnaire evaluating participants ' dysexecutive problems (DEX-DS). Within-group and


Introduction
Reduced awareness of cognitive deficits (RACD) affects patient self-reports and has the potential to impede the medical diagnostic process, possibly resulting in inadequate or insufficient provision of medical care (Committee on Diagnostic Error in Health et al., 2015;Vuilleumier, 2004).While inaccurate symptom reporting is frequently attributed to motivational and emotional factors (e.g., the denial of deficits to protect self-worth (Gabbard & Hobday, 2012)), neurological patients might also exhibit neuronal pathologies that impair their cognitive ability to properly recognize relevant health issues (Saj et al., 2013).To address the risk of diagnostic errors caused by imprecise or incomplete symptom reporting, instruments to assess a patient's awareness of deficits have been developed and recommended for implementation in comprehensive neuropsychological diagnostic routines (Lezak et al., 2012).Historically, deficit awareness was not systematically measured but was inferred from clinical exploration, as noticeable bodily malfunctions and the consequent reduced awareness of a disorder were reported to be easily identifiable by clinicians (Bisiach et al., 1986).Currently, various instruments, such as interviews (Cutting, 1978), clinicianassessed (Prigatano & Klonoff, 1998), informant-based (Jorm & Jacomb, 1989), and patient-reported questionnaires (Sherer et al., 1998), as well as the subsequent analysis of questionnaire discrepancies (Prigatano, 1996) and the disparity between self-estimation and objective test results (Rosen, 2011) have been established to complement clinical examination.The development of these more sophisticated assessment techniques has occurred alongside the unraveling understanding that a decrease in awareness can affect multiple and less apparent aspects of human functioning to varying degrees.Concomitant cognitive models trying to grasp the multifaceted phenomenology of awareness deficits have been both founded on these measurements' results and inspired their further advancements: Early, often hierarchical organized model structures featured "top-down" and "bottom-up" signaling, feedback-loops and domain-specificity, and revolved around a "homunculus" concept that generates awareness based on the integrity of the model's trajectories and components (Crosson et al., 1989;Schacter, 1990).In one successive approach, a metacognitive model has been proposed to describe the emergence of conscious knowledge of mental processes (e.g., awareness) from the interplay of distinct neuronal systems.The regularly updated Cognitive Awareness Model (CAM; Agnew & Morris, 1998;Hannesdottir & Morris, 2007) proposes the individual evaluation of abilities based on past successes and failures, using a "Personal Database" (PDB) for reference.Within a "Central Executive System" (CES), the ongoing performance is monitored and aligned with information stored in the PDB by a comparator mechanism, ultimately leading to conscious awareness within the "Metacognitive Awareness System" (MAS).Within the CAM, exposed awareness deficits can be classified into "primary", "executive" and "mnemonic anosognosia", depending on the malfunctioning model component: "Primary anosognosia" arises from disrupted neural connectivity and/or damage to the MAS, leading to difficulties in bottom-up integration and top-down modulation."Executive anosognosia" involves the distortion of higher-level supervisory processes, such as the CES; and finally, "mnemonic anosognosia" is characterized by a conscious awareness deficit caused by specific memory impairments affecting the PDB (Morris & Mograbi, 2013), and separate component measurement has been recommended (Piras et al., 2016).In recent literature, awareness is viewed as a continuous process, spanning from a detailed analysis of current task performance (local metacognition) to overarching global self-performance estimates (global metacognition).The gradual development of global self-performance estimates is thought to be driven by the accumulation of local confidence judgments, incorporating metacognitive sensitivity (certainty about response accuracy) and bias (discrepancy between estimated and correct responses), all without the need for external feedback (Rouault & Fleming, 2020;Seow et al., 2021).
Reduced awareness, including RACD, has already been part of Hakim and Adams' first description of normal pressure hydrocephalus, a neurological condition that predominantly occurs in the elderly population and typically comprises gait disturbances, urinary incontinence, and dementia.Adams portrayed his patients to be "bereft of thought" and "vaguely aware of their […] trouble in thinking" (Adams, 1966), and subsequent studies, employing systematic clinician-based rating and literature review (Collignon et al., 1976;De Mol, 1978), confirmed "anosognosia" to be a prevalent symptom in the examined condition.RACD also presents in the idiopathic form of normal pressure hydrocephalus (iNPH) and is conceptually associated with a cognitive deficiency within the realm of executive functions (EF; Roebers, 2017), a cognitive domain prone to inclination and responsive to treatment in iNPH patients (e.g., da Rocha et al., 2021).Nonetheless, the importance of RACD in iNPH is not solely based on its prevalence as a symptom of cognitive decline.Precise reporting of health issues plays a vital role in distinguishing iNPH from other age-related syndromes that share similar characteristics, as the elderly population often experiences motor, urinary, and cognitive impairments due to a range of neurological conditions and comorbidities (Isaacs et al., 2019).Thus, RACD may hinder the timely initiation of effective treatment that is most beneficial in the early iNPH stages (Vakili et al., 2016) by facilitating under-reporting of critical cognitive symptoms.
In our prospective case-control study, we investigated the type and change in RACD in an iNPH population during clinical examination including lumbar puncture (LP).We measured RACD using local and global metacognitive selfestimation methods of cognitive performance in reference to objective test scores and informant reports and compared the results of iNPH patients with those of a clinical group and a healthy control group.We hypothesize that RACD will be most pronounced in patient populations and that RACD in iNPH will decrease under medical care.

Participants and group assignment
Patients referred from in-house consultation and ambulant care to our neurological clinic with a suspected diagnosis of  1).Patients who were unable or unwilling to consent to participate in the study were excluded from the recruitment process.Data from participants were retrospectively excluded under the following conditions: a) presence of a developmental disorder (1), b) occurrence of a cerebral infarction within the past year (2), c) withdrawal of less than 30 ml of cerebrospinal fluid (CSF) during the diagnostic process (4), d) incomplete data (4), and e) unavailability or unreliability of an informant report due to infrequent personal contact (i.e., less than once a week) with the patient (3).
In our prospective control group design, clinical data were retrospectively extracted from medical patient records after discharge to ensure that neither researchers nor patients were aware of group allocation during data acquisition.The specifics of the diagnoses relevant for hospital admission can be found in the Results section.
Patients were categorized into groups based on medical records, including brain imaging (computed tomography or magnetic resonance imaging), medical history, neurological examination data (including gait assessment), cognitive assessment data, and blood/CSF laboratory results.Classification of every case into probable iNPH, possible iNPH, or unlikely iNPH was performed following the international diagnostic criteria (Relkin et al., 2005).Those meeting the criteria for possible and probable iNPH according to international guidelines were assigned to the iNPH group (n ¼ 14, 3 female, mean age 76.3 ± 4.3, 1 probable iNPH).Patients receiving alternative diagnoses explaining their condition were classified as unlikely iNPH and placed in the AACS group (AACS: other age-associated cognitive syndromes; n ¼ 10, 3 female, mean age 77.3 ± 5.5).Elderly individuals living at home or in assisted living and care facilities were recruited as healthy controls (HC) via online advertisements, pamphlets, or a direct approach.Attainment of the age of 65, normal cognition indicated by a Mini-Mental State Examination (Folstein et al., 1975) score above 27, self-reported absence of previous neurological or psychiatric illness as well as normal or corrected-to-normal vision and hearing were required for inclusion.Participants lacking available or reliable informant reports (no or infrequent contact) were not further examined, and those with !18 years of education were retrospectively excluded to match the patient sample.Detailed feedback  Please cite this article as: L€ obig, N., et al., Reduced awareness of cognitive deficits in idiopathic normal pressure hydrocephalus and its change following lumbar puncture, Cortex, https://doi.org/10.1016/j.cortex.2024.07.010 about their individual cognitive performance after completion of the examination was offered to the HC participants as an incentive for their participation.
All individual anonymized data (at raw and summary level) that are necessary and sufficient to reproduce all analyses and data presentations reported in this paper are publicly archived (URL to appendices: https://osf.io/b948p/).No part of the study procedures or its analyses were pre-registered in a timestamped, institutional registry.The study protocol was approved by the Ethics Committee of the Department of Psychology, Humboldt-University in Berlin (Reg.-Nr.2021-04), and the executive board of the recruiting site (application name "KoSSNo").All patients gave their informed consent according to the Declaration of Helsinki (2013).

Neuropsychological assessment
The neuropsychological assessment of participants was conducted twice (pre-and post) with an average interval of 1.9 days (±1.3 days), and corresponding informant reports were obtained once in timely proximity.Patients underwent a neurological diagnostic process, including lumbar puncture (LP), between the pre-and post-measurement intervals, while the healthy control (HC) group only participated in the neuropsychological pre/post assessment.Before the preassessment, all participants were screened for depressiveness, using the PHQ-9 questionnaire (Kroenke et al., 2001).
For RACD measurement, global and local metacognitive assessments of cognitive performance were employed.Global metacognition was evaluated before pre-and after postmeasurement of objective cognitive performance by the patients' administration of the Dysexecutive Questionnaire e Revised (DEX-R, German translation) (Simblett et al., 2016), targeting estimates of cognitive functioning (ECF).The DEX-R is a 37-item questionnaire focusing on dysexecutive problems and emerged from a Rasch analysis of its predecessor version.
Utilizing a 5-point Likert scale with response choices ranging from "Never" to "Very often," (with higher scores indicating increased concerns) ensures ecologically valid reporting of cognitive impairments, as substantiated by empirical evidence (Azouvi et al., 2015).The questionnaire contained four subscales reflecting metacognitive, executive-cognitive, behavioraleemotional self-regulatory, and activation regulatory functions, according to Stuss' model of frontal lobe function (Wakely et al., 2022).
To assess local metacognition, parallelized subtests from the neuropsychological test battery (NAB) ( Q1 Petermann et al., 2016;Stern, 2003) were initially administered.These subtests included attention (subtests "Numbers & Letters" parts aec and "Digits Backward"), memory (subtest "Story Learning" for direct and delayed recall), and the spatial module (subtest "Design Construction").Assessments were conducted twice, both before and after the LP (for a detailed description, see Appendix A).Following each subtest, participants engaged in task-specific self-evaluation of cognitive performance, estimating the percentile score they would receive if compared to a demographically matched peer group.To aid their local confidence estimation, participants were presented with a visual representation of a normal distribution (similar to the graphic that was used in Rothlind et al., 2016;Fig. 2), which indicated that most individuals achieve average scores, while fewer achieve scores much higher or lower than the average.Participants were encouraged to guess even when they were unsure about their performance compared with their peers.All participants received neutral feedback after each subtest to avoid interference with their self-evaluation.A detailed table of all relevant data (on individual level) can be found in Appendix C. The application of the entire test battery took approximately 60 min, and functional hearing, vision, speech, and upper limb motor skills were required for the use of the tests' necessary utensils and the perception of relevant auditory and visual stimuli.

Data processing and statistical analysis
For global metacognitive assessment, all DEX-R responses were quantified by aggregating their item responses relative to the 5-point Likert scale, separately for the participants' preand post-measurement assessment, as well as for the informant reports.Given the superior accuracy of informantreports over self-reports in identifying cognitive impairment in the elderly (Carr et al., 2000), the pre-and postmeasurement DEX-R self-reports were subtracted from the corresponding DEX-R informant-reports, and the two resulting discrepancy scores (DEX-DS) represent the participants' global metacognitive outcomes in terms of ECF.Regarding local metacognition, absolute differences between the participants' objective, age-adjusted test scores, and distributionequivalent (i.e., T-scale) self-evaluations were calculated to quantify task-specific performance estimations (DTSPE).DTSPE scores were obtained for all employed cognitive subtests and subsequently averaged to produce a DTSPE general score.Regarding the objective assessment of cognitive performance, test results were transformed into z-scores, as outlined in the NAB manual and subsequently averaged across all subtests to facilitate further analysis.
Subsequently, the normal distribution of the assessed and generated data was scrutinized with the ShapiroeWilk test before using the pre-and post-measurement scores for conducting between-group and within-group comparisons.Demographic and clinical between-group differences (sex, age, education, CSF withdrawal, days between measurements, days between LP and post-measurement) and neuropsychological data (depressiveness, objective cognitive test performance, DTSPE, DEX-DS) were analyzed using chi-square, KruskaleWallis, and ManneWhitney U tests according to the data scale level and number of groups compared.The main neuropsychological outcomes (averaged objective test performance z-scores and DTSPE scores across subtests, and preand post-measurement DEX-DS) were additionally averaged across pre-and post-measurement scores to allow a global between-group comparison.Furthermore, pre-/post-measurement changes in the main neuropsychological outcomes were compared between groups by first subtracting the variables' post-measurement values from their pre-measurement values.The resulting DTSPE and DEX-DS score differences were subsequently z-standardized so that variable changes could be compared on the same scale.In addition, Wilcoxon signed-rank tests (two-tailed) were conducted to detect pre-/ post-measurement within-group differences.Regarding multiple within-group comparisons, Bonferroni-corrected pvalues below .05were considered significant (Table 1).Effect sizes were obtained in form of Pearson's r according to Fritz et al. (2012).The data were analyzed using SPSS software 27 (IBM, 2020) and plotted using JASP (JASP, 2022).

Results
Analysis of skewness and kurtosis of the test variables revealed multiple distributional abnormalities.The ShapiroeWilk test revealed that a majority (16/19) of the relevant variables (partial/averaged cognition and DTSPE scores, DEX-R scores) did not follow a normal distribution in at least one group (Appendix B).Thus, non-parametric methods were used for data analysis.Missing data were found in 12 items distributed across four DEX-R informant reports and imputed by the rounded average item value of the individual report, representing 1% of all DEX-R informant data.
Regarding the demographic and clinical group characteristics, depressiveness measured by the PHQ-9 score [H(2) ¼ 13.476, p < .001]and the percentage of female participants differed significantly between the groups [X 2 (2, N ¼ 47) ¼ 6.357, p ¼ .042].The variables age, years of education, and days between measurements were equally distributed across the three groups (p > .05for both, Table 1).In instances where LP was performed, the post-measurement was conducted after an average of 1.3 days (±.9 days) and 34.8 ml (±4.4 ml) of CSF was withdrawn, and LP-specific variables did not differ between the patient groups (p > .05for withdrawn CSF and days between LP and post-measurement).A betweengroup comparison of the main neuropsychological outcomes averaged across pre-and post-measurement showed significant group differences for the averaged objective test performance [H( 2 Here, a more detailed examination of single-group differences showed that HC differed significantly from both patient groups regarding the averaged objective test performance and averaged DTSPE score (p .005for all) in the way of better objective test performance and more accurate task-specific self-estimation of cognitive performance, but no differences in these variables could be found between the groups iNPH and AACS (p > .05for all).Accompanying the higher DTSPE scores, a significant impairment of objective cognitive performance was apparent in both patient groups, with average z-values falling below one standard deviation.
Pre-/post-measurement changes in the main neuropsychological outcomes showed significant between-group dif-  1.

Discussion
In our prospective case-control study, we examined the ability of iNPH patients to evaluate their cognitive performance during a clinical examination process that included LP.Two RACD measures were employed and corrected based on objective task performance, respectively informant reports, and results were compared to those of clinical and healthy control participants.Regarding the main neuropsychological outcomes, lower overall cognitive test performance together with a generally elevated DTSPE score were found in both patient groups, indicating a reduction in their local metacognitive performance.Following LP, iNPH patients' improvement in local metacognition exceeded the post-measurement changes of the two other groups.Furthermore, the iNPH group's post-LP shift to a negative DEX-DS score indicated an increase in patients' reports of ECF, surpassing the magnitude of the postmeasurement changes observed in the healthy control group.As these reports correspond more closely with the group's objectively reduced cognitive capabilities (Azouvi et al., 2015), they can be interpreted as an improvement in global metacognitive performance.
Our findings align with previous research describing RACD in iNPH (Adams, 1966;Collignon et al., 1976;De Mol, 1978) and other age-associated cognitive disorders (e.g., Alzheimer's disease; Ansell & Bucks, 2006), and allow a more detailed examination of RACD form and variability due to the repeated measurement design.Within the means of the CAM, our results suggest the contribution of "executive anosognosia" to the patient groups' RACD by the overall increased distance between their task-specific self-estimation and their objective test scores (DTSPE), indicating neurologically caused malfunction of the CES component.Regarding the dynamics of the patient group's post-LP changes in RACD, the AACS group's self-estimations remained statistically unchanged, whereas the iNPH group's RACD scores became more adequate, thereby proving them incompatible with persistent "primary anosognosia".Since the complete assessment of cognitive performance was administered without any evaluating feedback before and after cognitive testing, and the short time interval between measurements should have prevented the actualization of self-estimates by extensive real-world observations, these improvements must be prompted by a different mechanism.In a similar study, Ansell and Bucks (2006) attributed improved outcomes of local and global metacognition in Alzheimer's patients to the experience of disparity between predicted and achieved results ("failure") within the testing process and the consequent adjustment in subsequent self-estimates.On the same note, the stimulating character of the repeated acquisition of performance-based self-evaluation in our paradigm (e.g., by DTSPE score; Rothlind et al., 2016) seems akin to specific elements in neuro-rehabilitation programs targeting awareness deficits (Ownsworth et al., 2010) and might therefore influence RACD outcomes during the study process.Nevertheless, the absence of changes in the AACS group's RACD measures suggests that the iNPH-specific variability in self-evaluation may be attributed to the LP conducted between pre-and post-measurements, impacting RACD in iNPH beyond the stimulating effect on metacognitive abilities by the test procedure itself.On a cognitive level, the assessment of DTSPE requires a comparative analysis between an individual's task performance and the expected performance of others, thereby necessitating updating and monitoring of working memory representations (Williamson et al., 2010).As these functions are already sensitive to deterioration in the early stages of iNPH pathogenesis (Xiao et al., 2022), impaired local metacognitive operations resulting in RACD seem unsurprising for this neurological condition.
Improvement of RACD in our iNPH sample presented with lowered DTSPE and DEX-DS scores, indicating more realistic local and global metacognitive self-estimates with regard to the participants' low cognitive test performance.This interdependence between local and global metacognitive abilities was recently addressed by Rouault and Fleming (2020), who modeled the formation of global metacognitive estimates (e.g., ECF) as a probability distribution derived from the compilation of local metacognitive operations (Rouault & Fleming, 2020).This construction of global metacognitive estimates (represented within the (ventro-)medial prefrontal cortex and precuneus) is believed to depend on the activation of the ventral striatum, reinforcing correct (positive) estimations when matched by monitoring-driven, local metacognitive operations processed within the anterior cingulate cortex and other anterior prefrontal regions in the manner of a prediction error based learning mechanism (Fleming et al., 2010).In iNPH, the integrity of these neuronal structures and circuits is affected by pathognomonic sequelae of CSF malabsorption, such as ventricular dilatation, impaired blood flow, and axonal damage (Tan et al., 2021).Improvement of iNPH symptoms is frequently described following LP (Nakajima et al., 2021), as the removal of CSF during diagnostic or therapeutic interventions initiates periventricular decompression, thereby facilitating the recovery of local white-matter regions as well as the attenuation of striatal dopamine receptor downregulation (Demura et al., 2012;Nakayama et al., 2007).As callosal white matter fibers and striatal activity are strongly correlated with local (Park et al., 2008) and global metacognitive estimations (Fleming et al., 2010;Rouault & Fleming, 2020), their regained structural integrity and functionality following CSF removal might lead to the RACD improvements we observed in iNPH patients.On a functional level, LP-induced physiological changes may manifest as an increase in psychomotor speed, attention function, and working memory, all of which contribute to the smooth functioning of other cognitive processes (Nakajima et al., 2021), including EF (e.g., da Rocha et al., 2021).As both EF and (local) metacognition share overlapping neuronal substrates distributed across prefrontal areas (Fernandez-Duque et al., 2000) and depend on feedback from lower-level cognitive functions (e.g., to enable individuals to adjust to demanding challenges (Wu & Was, 2023)), improvement in more basic cognitive functioning after LP might attenuate metacognitive dysfunction in iNPH.Alongside a metacognitively stimulating testing procedure, improved local metacognitive performance may facilitate the calibration of global metacognitive estimates (Rouault & Fleming, 2020), resulting in the changed metacognitive measurements we observed.
While our paradigm allowed the detection of change in local and global metacognition by examining of DTSPE and DEX-DS scores, no general between-group differences were found in ECF by self-and informant reports.This finding contrasts with similar awareness studies reporting associations between patient/informant-report discrepancies and (cognitive) symptom severity in neurological patients (e.g., Smith & Arnett, 2010).Beyond the reduced visibility and subsequent hindered reportability of cognitive symptoms for informants (McDade-Montez et al., 2008), the increased depressiveness in our patient groups might have contributed to equal reports of ECF across patients and their corresponding informants, leading to insignificant DEX-DS between-group comparisons.While negative emotions were found to improve local metacognitive operations by diminishing metacognitive bias (Massoni, 2014), depressiveness affects the assessment of global metacognition by increasing symptom reports (Wakely et al., 2022) and could thereby outweigh the more discrepant patient/informant questionnaire results we expected to find.As depressiveness is frequently exhibited in the iNPH population and has a tendency to increase along the course of treatment, possibly due to an increase in awareness (Israelsson et al., 2016), we estimate our iNPH sample to adequately represent a broader population.
Our study faced limitations due to a shortage of reliable instruments, highlighting the necessity for more tests that are appropriate for repeated measurement designs and capable of simultaneously detecting cognitive improvement in iNPH patients following LP.Moreover, the majority of patients included in the iNPH group fulfilled the ambiguous criteria of "possible" iNPH, compared to the certainty of "probable" or "definite" iNPH diagnoses (Relkin et al., 2005).As we identified significant changes within our medically heterogeneous group, we posit that the LP effect on RACD is robust within the broader iNPH patient population (Nakajima et al., 2021;Relkin et al., 2005).In this matter, the authors also acknowledge the small sample sizes used per group and, due to the large effect sizes, primarily recommend memory-related tasks and change of discrepancy scores between self-and informant reports in future studies targeting RACD.Additionally, to ensure disease specificity and to examine the temporal dynamics of CSF alteration on RACD, iNPH patients' awareness of cognitive performance, and of other symptoms like gait disturbance, could be further scrutinized in the context of long-term interventions, such as external lumbar CSF drainage or intracranial shunt placement.Nevertheless, we hope that our results draw attention to the importance of RACD assessment in the iNPH diagnosis process and recommend local metacognitive measurements for this purpose.

Q2
The authors hereby report that there are no competing interests to declare and that this research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.The authors also state that the determination of the sample size, all data exclusions, all inclusion/ exclusion criteria, whether inclusion/exclusion criteria were established prior to data analysis, all manipulations, and all c o r t e x x x x ( x x x x ) x x x Please cite this article as: L€ obig, N., et al., Reduced awareness of cognitive deficits in idiopathic normal pressure hydrocephalus and its change following lumbar puncture, Cortex, https://doi.org/10.1016/j.cortex.2024.07.010 measures in the study were addressed in the Methods section.During the preparation of this work the authors used "Paperpal" (https://paperpal.com/) to assist with editing tasks.After using this service, the corresponding author reviewed and edited the content as needed and takes full responsibility for the content of the publication.

Open practices
The study in this article has earned Open Data and Open Materials badges for transparent practices.The data and materials are available at: https://osf.io/c5wt6/.

Fig. 1 e
Fig. 1 e Flow-chart of consecutive inclusion of study participants.Patients were first assessed for eligibility and, if they met the study requirements, included into the study by assignment to iNPH or AACS group.Parallelly, eligible elderly volunteers with reliable informant reports were matched with the patient group based on their education and included into the study.

Fig. 2
Fig.2e A visual tool was provided to the participants to help them estimate their distribution-equivalent score relative to a demographically similar peer group in selected neuropsychological tests.Task-specific performance estimation (DTSPE) was operationalized by calculation of the absolute differences between participants' age-adjusted test scores and distribution-equivalent estimations.

Fig. 3 e
Fig. 3 e Boxplots of the main neuropsychological outcomes' z-standardized pre-/post-measurement discrepancies.NAB: neuropsychological test battery; DTSPE: absolute differences between participants' task-specific performance estimations and objective, age-adjusted test scores; ECF: estimates of cognitive functioning e calculated by the measurement disparity between self-and informant rated estimates of everyday dysexecutive problems.Outliers are indicated by , significant differences are indicated by *(p < .05)and **(p < .01).iNPH, participants with suspected and possible idiopathic normal Pressure hydrocephalus; AACS, other age-associated cognitive syndromes; HC, healthy controls; iNPH and AACS patients received lumbar puncture between pre-and post-measurements.

Table 1 e
Data of the three groups assessed for objective cognitive performance and self-awareness using two different methods.