ORIGINAL RESEARCH published: 08 December 2021 doi: 10.3389/fnagi.2021.718949 Neuropsychiatric Profile as a Predictor of Cognitive Decline in Mild Cognitive Impairment Natalia Roberto 1,2 , Maria J. Portella 2,3,4 , Marta Marquié 1,5 , Montserrat Alegret 1,5 , Isabel Hernández 1,5 , Ana Mauleón 1 , Maitee Rosende-Roca 1 , Carla Abdelnour 1,5 , Ester Esteban de Antonio 1 , Juan P. Tartari 1 , Liliana Vargas 1 , Rogelio López-Cuevas 1 , Urszula Bojaryn 1 , Ana Espinosa 1,5 , Gemma Ortega 1,5 , Alba Pérez-Cordón 1 , Ángela Sanabria 1,5 , Adelina Orellana 1,2 , Itziar de Rojas 1 , Sonia Moreno-Grau 1 , Laura Montrreal 1 , Emilio Alarcón-Martín 1 , Agustín Ruíz 1,5 , Lluís Tárraga 1,5 , Mercè Boada 1,5 and Sergi Valero 1,5* 1 Ace Alzheimer Center Barcelona, Universitat Internacional de Catalunya (UIC), Barcelona, Spain, 2 Department of Psychiatry and Forensic Medicine, Universitat Autònoma de Barcelona (UAB), Barcelona, Spain, 3 Department of Psychiatry, Institut d’Investigació Biomèdica Sant Pau (IIB Sant Pau), Hospital de la Santa Creu i Sant Pau, Barcelona, Spain, 4 Networking Research Center on Mental Health (CIBERSAM), Madrid, Spain, 5 Networking Research Center on Neurodegenerative Diseases (CIBERNED), Madrid, Spain Edited by: Dag Aarsland, Introduction: Mild cognitive impairment is often associated with affective and other King’s College London, United Kingdom neuropsychiatric symptoms (NPS). This co-occurrence might have a relevant impact on Reviewed by: disease progression, from MCI to dementia. Jose Felix Moruno-Manchon, University of Texas Health Science Objective: The aim of this study was to explore the trajectories of cognitive decline in Center at Houston, United States an MCI sample from a memory clinic, taking into consideration a perspective of isolated Onésimo Juncos-Rabadán, cognitive functions and based on NPS clusters, accounting for the different comorbid University of Santiago de Compostela, Spain symptoms collected at their baseline visit. *Correspondence: Methods: A total of 2,137 MCI patients were monitored over a 2.4-year period. Sergi Valero svalero@fundacioace.org Four clusters of NPS (i.e., Irritability, Apathy, Anxiety/Depression and Asymptomatic) were used to run linear mixed models to explore the interaction of cluster with time Received: 01 June 2021 on cognitive trajectories using a comprehensive neuropsychological battery (NBACE) Accepted: 28 October 2021 Published: 08 December 2021 administered at baseline and at the three subsequent follow-ups. Citation: Results: A significant interaction between cluster and time in cognitive decline was Roberto N, Portella MJ, Marquié M, Alegret M, Hernández I, found when verbal learning and cued-recall were explored (p = 0.002 for both memory Mauleón A, Rosende-Roca M, functions). For verbal learning, the Irritability cluster had the largest effect size (0.69), Abdelnour C, Esteban de Antonio E, whereas the Asymptomatic cluster showed the smallest effect size (0.22). For cued- Tartari JP, Vargas L, López-Cuevas R, Bojaryn U, Espinosa A, Ortega G, recall, the Irritability cluster had the largest effect size among groups (0.64), and Pérez-Cordón A, Sanabria Á, Anxiety/Depression had the smallest effect size (0.21). Orellana A, de Rojas I, Moreno-Grau S, Montrreal L, Conclusions: In MCI patients, the Irritability and Apathy NPS clusters shared similar Alarcón-Martín E, Ruíz A, Tárraga L, patterns of worsening in memory functioning, which could point to these NPS as risk Boada M and Valero S (2021) Neuropsychiatric Profile as a Predictor factors of a faster cognitive decline, acting as early prognostic markers and helping in of Cognitive Decline in Mild Cognitive the diagnostic process. Impairment. Front. Aging Neurosci. 13:718949. Keywords: mild cognitive impairment, cognitive decline, neuropsychiatric symptoms, irritability, apathy, anxiety, doi: 10.3389/fnagi.2021.718949 depression Frontiers in Aging Neuroscience | www.frontiersin.org 1 December 2021 | Volume 13 | Article 718949 Roberto et al. Neuropsychiatric Profile Predicting Cognitive Decline INTRODUCTION AD in older adults with MCI (Jang et al., 2020), and those have also even shown synergic effects with the APOE ε4- Biological changes bonded to impairment of cognitive functions allele (Valero et al., 2020). Recently, some attempts have been are shown as humans age (Glisky, 2007). In the elderly, some made to investigate grouped NPS as possible predictors of cognitive skills such as attention, memory, executive functions or cognitive decline along the progression of MCI toward dementia processing speed suffer from subtle changes associated with the (Palmer et al., 2007; Edwards et al., 2009). In a recent 2-year normal aging process (Park et al., 2002; Park and Reuter-Lorenz, prospective study, and according to the three classes found 2009), whereas others suffer a greater cognitive decline beyond in terms of NPS trajectories (stable, improved and worsened) expected, but not all decrement in cognitive functioning in this in MCI patients, it was found that the NPS worsened class population is a precursor of disease. Therefore it is important to suffered the greatest cognitive and functional decline, as well distinguish between normal and pathological cognitive decline, as the highest conversion rate in comparison with the stable mainly because it could affect the patient’s daily functioning class and the improved class (David et al., 2016). Other clinical (Ginsberg et al., 2019) worsening their quality of life. The studies exploring associations of NPS by using factor analysis accurate measurement of cognitive decline over time is of utmost in MCI and mild AD dementia were focused on conversion importance as it could help in the diagnosis and posterior to dementia and/or its relation to the severity of cognitive prognosis of different neurodegenerative diseases and other decline, but not specific cognitive domains (Siafarikas et al., syndromes (Grober et al., 2008; Wise et al., 2019). 2018; Liew, 2019). There are two studies in the same line Cognitive impairment is often associated with affective exploring NPS clusters and conversion to dementia in cognitively symptoms, such as anxiety or depression (Geda et al., 2008; healthy volunteers (Leoutsakos et al., 2015; Forrester et al., Hermida et al., 2012; Singh-Manoux et al., 2017), which have been 2016). However, there is still no consensus in the findings, widely reported in different populations. While more is known probably due to dissimilarities in the design and methodology about the former in relation to cognitive decline (Gonzales of these studies (different diagnostic criteria, sample selection et al., 2017), there is still not much agreement about other or neuropsychological assessment applied) (Ma, 2020). Likewise, neuropsychiatric symptoms (NPS) that could interfere with or there is a conceptual void when exploring the most common NPS relate somehow to a worsening in neuropsychological measures in patients with MCI and their implications in cognitive decline over time in early stages of different diseases. Some studies in a long-term follow-up to analyze patients’ progression in demonstrate the co-existence of both factors, with NPS being specific domains. Only a few studies in neurological patients, such the predecessors of cognitive decline, often for many years (Wise as those with Parkinson’s and Huntington’s diseases (Pirogovsky- et al., 2019; Tsunoda et al., 2020). There is no consensus on the Turk et al., 2017), are going in this direction of assessing NPS in order of appearance of both neural insults; previously, it was the MCI population and their implications for cognitive decline thought that cognitive deficits were the main reason for medical (Weintraub et al., 2015; Donaghy et al., 2018). consultation, while studies increasingly claimed that NPS were Therefore, this longitudinal study aims at investigating the the precursors initially detected before any cognitive decline is existence of different trajectories of specific cognitive-domain shown (Mortby and Anstey, 2015; Ismail et al., 2018). In any decline over time in an MCI sample from a memory clinic, event, it is important to delve into early cognitive decline and try considering baseline NPS clustering. to elucidate the factors favoring it. At this early stage, another important feature to keep in mind is that comorbid NPS are often found in the clinical practice, and this co-occurrence of NPS MATERIALS AND METHODS and cognitive decline might have a cumulative effect on disease progression (Geda et al., 2013). Many attempts have been made Participants to identify specific profiles of NPS associated with Alzheimer’s The study was conducted at the Memory Clinic of ACE disease (AD). Some studies have explored the existence of Alzheimer Center Barcelona (Spain), a private non-profit neuropsychiatric subsyndromes or the genetics of NPS that could institution focused on the diagnosis, care and research of be the basis of AD, but no clear conclusions have been raised so cognitive disorders and providing services to the Catalan Public far (Canevelli et al., 2013; Huang et al., 2020). A high prevalence Health Service (Xarxa Hospitalària d’Utilització Pública, XHUP) of NPS in AD has commonly been associated with a worsening in (Boada et al., 2014). the patient’s functionality (Karttunen et al., 2011). A total of 2,137 patients diagnosed with MCI were selected It is well known that NPS seem to play a critical role in from a pool of patients evaluated at the Memory Clinic, see early clinical stages of the dementia continuum (Karttunen et al., Roberto et al. (2021) for more information; MCI subtypes 2011; Burhanullah et al., 2020), such as in Mild Cognitive diagnoses were based on modified Petersen’s criteria and Lopez Impairment (MCI) (Lyketsos et al., 2002; Geda et al., 2008; and colleagues’ classification, defined as amnestic (aMCI) or Peters et al., 2012). In a search of profiles of clustered symptoms non-amnestic (naMCI), and possible or probable MCI due that could serve as markers of disease progression in early to AD, respectively (Petersen et al., 1999; Lyketsos et al., stages, NPS would act as early clinical manifestations of an 2002; Petersen, 2004). All patients had to fulfill the following emergent process of neurodegeneration (Gallagher et al., 2017). inclusion criteria: (i) more than 44 years old; (ii) a Mini- In particular, affective NPS (depression, apathy, anxiety and Mental State Examination (MMSE) total score of 24 or above; irritability) were associated with a more rapid progression to (iii) a Clinical Dementia Rating (CDR) score of 0.5; (iv) a Frontiers in Aging Neuroscience | www.frontiersin.org 2 December 2021 | Volume 13 | Article 718949 Roberto et al. Neuropsychiatric Profile Predicting Cognitive Decline Global Deterioration Scale (GDS) score of three or below; (v) obtaining three scores derived from the number of words at least six total years of formal education; (vi) absence of recalled; and the abbreviated similarities subtest of the WAIS-III severe visual or auditory disturbances that could hinder the for abstract reasoning. neuropsychological examination; (vii) presence of an informant or relative to complete the baseline administration of the NPI- Neuropsychiatric Symptoms Measures Q; and, (viii) a baseline neuropsychological visit completed along Neuropsychiatric symptoms were evaluated at the with at least one follow-up. All clinical data were collected from baseline clinical assessment using the Neuropsychiatric January 2006 to June 2017. In all cases, the date of the MCI Inventory-Questionnaire (NPI-Q) (Boada et al., 2002). diagnosis was taken as the starting point or inclusion date for The NPI-Q is a simplified and widely-used scale that this study. Patients were followed up approximately annually assesses 12 behavioral disturbances including delusions, with a clinical assessment that included a neurology and a hallucinations, agitation/aggression, depression/dysphoria, neuropsychological visit. anxiety, elation/euphoria, apathy/indifference, disinhibition, irritability/lability, aberrant motor behavior, sleep and night- Cognitive Measures time behaviors, and appetite and eating disorders, in the Cognitive data were collected at baseline and at every follow- dementia-related population. The NPI-Q was completed up visit, using The Neuropsychological Battery of Fundació through information provided by a patient’s reliable informant ACE (NBACE). The NBACE is a 50-min battery designed to (family member or caregiver). A change during the previous assess cognitive domains especially affected in the elderly when month in each one of the 12 behavioral domains was recorded as dementia due to AD or other neurodegenerative processes is a dichotomized measure (present or absent). For more details on suspected (Alegret et al., 2012). The NBACE was proposed this measure and procedures see Roberto et al. (2021). as a brief, easy-to-administer and goal-directed compilation of globally-used neuropsychological tests in our target population, provided that it is focused on verbal memory and learning, Analytical Approach visual perception, and executive functions, which are affected The present study is based on the results of a previous early in the course of the disease. However, these are not the Latent Class Analysis (LCA) for clustering participants by only explored domains. In our study we included tests sensitive means of a dichotomized NPI-Q measure (Roberto et al., to the following cognitive domains: attention, working memory, 2021). Each participant was assigned to the best fitting processing speed, executive functions, verbal memory, language, cluster with the highest membership probability using baseline gnosis, visuospatial skills and praxis. Normative data and cut-off NPS. A 4-cluster was considered the optimal solution: Class scores of the NBACE subtests for individuals more than 44 years 1 = Irritability; Class 2 = Apathy; Class 3 = Anxiety/Depression; old can be found elsewhere (Alegret et al., 2013). Class 4 = Asymptomatic. Then, linear mixed-effects models Processing speed was measured with the Automatic Inhibition (LMMs) were executed to explore cognitive decline for specific subtest of the Syndrome Kurtz Test (SKT; Erzigkeit, 1989), domains including NPS clusters and time of assessment (four using execution time as the raw score. Attention and working time points, baseline, 1−, 2−, and 3-year follow-ups). Individual memory scores were obtained by means of the digit span LMM models were calculated, one for each neuropsychological forward and backward subtests of the (WAIS–III; Wechsler, domain explored. Interaction of NPS cluster by time of 1997a). Verbal learning and memory were measured through assessment was considered the main effect of interest in the word list learning test from the Wechsler Memory Scale– the model, also incorporating the corresponding cluster and Third Edition. Verbal learning trials, long-term retention and time main effects. For these two factors, fix and also as cued-recall were used as raw scores; the interference list was a random effect were analyzed, considering that assessment not included in the battery (WMS–III; Wechsler, 1997b). Verbal time points could vary among participants, and they also learning was composed of the sum of raw scores obtained in had different conditional probabilities of cluster belonging the four trials of the learning phase (61st + 2nd + 3rd + 4th ); (see Roberto et al., 2021), i.e., individual differences had long-term retention was the total amount free recalled words; to be modeled. Mean differences (SD) accounting for time and cued-recall was the total number of words correctly between baseline and every follow-up (times of assessments) recognized among the correct items and the same amount were as it follows: from baseline to the first follow up were of “interference” items. Language was measured with the 15- 11.26 months (6.22); from baseline to the second follow-up were item version of the Boston Naming Test (BNT; Kaplan et al., 22 months (7.48); and from baseline to the third follow-up were 1983). Gnosis, with a single score, was evaluated by means of 32.21 months (9.01). Both random intercept and slopes were the Poppelreuter test (Della Sala et al., 1995). Visuoconstructive included in the analyses. Asymptomatic class was considered praxis was evaluated with the abbreviated block design subtest the reference category. As a controlling factors, age, MMSE, of the Wechsler Adult Intelligence Scale–Third Edition WAIS– educational level, sex, conversion to dementia (yes/no), MCI type III (Wechsler, 1997a). Visuospatial skills were measured with (amnestic/non-amnestic), and MCI profile (possible/probable) Luria’s Clock Test (Golden, 1980), providing a single score. were also included in the models and were considered in the Finally, executive functioning was measured through different models as fixed factors. Only when a significant interaction tests: the Automatic Inhibition of the SKT accuracy score (cluster × time) was obtained in a specific cognitive domain, of inhibition ability; phonetic, semantic and verb fluencies, simple effects were calculated contrasting differences among Frontiers in Aging Neuroscience | www.frontiersin.org 3 December 2021 | Volume 13 | Article 718949 Roberto et al. Neuropsychiatric Profile Predicting Cognitive Decline clusters across the time points. Syntax of LMM is provided in 51 to 66%. According to the classification of possible or Supplementary Material. probable MCI profile, the Irritability and Apathy classes had a higher percentage of patients with a diagnosis of probable amnestic (47 and 48.2%, respectively). Finally, patients in the RESULTS Irritability and Apathy classes showed a higher proportion of conversion to dementia. According to previously published findings from our group Table 2 shows the LMM results for cognitive domains (Roberto et al., 2021), the whole sample of 2,137 MCI patients accounting for cluster, time, and cluster by time interaction. Only was divided into four NPS clusters. Class 1-Irritability included the memory domain showed a significant interaction in cluster by 134 patients (6.3%) with high probability of irritability (0.93), time, with verbal learning and cued-recall in particular being the together with lower probability of anxiety (0.64) and apathy only processes showing significant differences. Simple effects for (0.63). Class 2-Apathy comprised 272 patients (12.7%) and these two cognitive functions comparing baseline with the third it was strongly represented by this symptom (1). Class 3- follow-up are displayed in Table 3, revealing significant results Anxiety/Depression included 1,056 patients (49.4%) who showed in all four clusters (p < 0.001). When differences between the a high probability of depression (0.95), anxiety (0.93) and, by final follow-up and the baseline scores were calculated in terms far, apathy (0.61). Class 4-Asymptomatic included 675 patients of verbal learning, a faster decline was shown in the Irritability (31.6%) with low probabilities (<0.3) in all NPS. class, with double the difference, versus a slower decline in the Demographic and clinical characteristics of the sample Asymptomatic class, with 3.66 and 1.34, respectively. With regard stratified by NPS cluster are shown in Table 1. There were to effect size (Cohen’s D), the Irritability class had the highest significant differences in age, gender, educational level, MMSE score (0.69), whereas for Apathy and Anxiety/Depression classes total score, MCI type, MCI profile, and conversion rates to the effect sizes were medium (0.49 and 0.31, respectively). The dementia among the four NPS clusters, thus those were included Asymptomatic class showed the smallest effect size (0.22). Similar in the models and were considered as controlling factors. results were obtained for cued-recall, but this time the Irritability Differences among groups in age distribution showed that Apathy and Apathy classes had quite similar differences between follow- patients were the oldest (mean age 76.2). In relation to gender, up measures and the baseline (1.80 and 1.88, respectively), and women were more prevalent in the Anxiety/Depression and the Anxiety/Depression class had the lowest score difference Asymptomatic classes (65.2 and 60.9%, respectively). Educational from the baseline. In terms of effect sizes, the Irritability class level attained was higher in patients in the Apathy class again showed the largest effect size (0.64), followed by the (8.04 years of education). In relation to MMSE score obtained, Apathy and Asymptomatic classes (0.46 and 0.30, respectively), the Anxiety/Depression class had the highest results (MMSE mean with the Anxiety/Depression class having the smallest effect size 27.02). Regarding the MCI type (amnestic vs. non-amnestic), (0.21). Cognitive decline for memory domains (learning and our sample was quite balanced in general terms, having cued recall) were calculated for each cluster trajectory (see percentages in the four classes ranging from approximately Figures 1, 2, respectively). Slopes of the trajectories were also TABLE 1 | Demographic characteristics and clinical variables of our final sample (n = 2137) stratified by neuropsychiatric symptoms cluster (NPS cluster). Irritability class Apathy class Anxiety/Depression class Asymptomatic class F/χ2 p n = 134 n = 272 n = 1056 n = 675 Age 75.17 (7.99) 76.20 (7.32) 73.82 (8.31) 75.23 (8.15) 8.30 <0.001 (years) Gender 39 (29.1%) 112 (41.2%) 688 (65.2%) 411 (60.9%) 102.14 <0.001 (% of females) Education 7.38 (3.80) 8.04 (4.23) 7.10 (3.79) 7.25 (4.08) 4.14 0.006 (years) MMSE 27.00 (1.72) 26.65 (1.74) 27.02 (1.69) 26.97 (1.76) 3.46 0.016 (total score) MCI type 69 (51.5%) 180 (66.2%) 609 (57.7%) 417 (61.8%) 11.51 0.009 (% of amnestic) MCI profile 64 (47.0%) 131 (48.2%) 292 (27.7%) 266 (39.4%) 59.86 <0.001 (% of probable) Conversion to 72 (53.7%) 149 (54.8%) 394 (37.3%) 264 (39.1%) 37.21 <0.001 dementia Results are shown as mean (SD) for age, education and MMSE; whereas for gender, mild cognitive impairment (MCI) type, MCI profile and conversion to dementia, data are showed as n (%). MMSE, Mini-Mental State Examination. MCI type, amnestic/non-amnestic. MCI profile, probable/possible. Conversion to dementia was reported independently of the etiology. Class is related to neuropsychiatric-cluster belonging (Class 1 = Irritability; Class 2 = Apathy; Class 3 = Anxiety/Depression; 4 = Asymptomatic). Frontiers in Aging Neuroscience | www.frontiersin.org 4 December 2021 | Volume 13 | Article 718949 Roberto et al. Neuropsychiatric Profile Predicting Cognitive Decline TABLE 2 | Linear mixed model results of cluster by time interaction and main effects in cognitive domains. Cluster Time Interaction cluster × time Attention Digit span forward 1.59 (0.189) 12.82 (<0.001) 1.34 (0.212) (WAIS III) Working memory Digit span backward 0.53 (0.661) 14.33 (<0.001) 1.28 (0.244) (WAIS III) Processing speed Execution time in sec 6.71 (< 0.001) 1.87 (0.134) 0.59 (0.803) (SKT) Executive Phonetic fluency 6.86 (< 0.001) 5.27 (0.001) 1.59 (0.112) Semantic fluency 9.79 (< 0.001) 55.64 (< 0.001) 0.78 (0.638) Verbal fluency 5.40 (0.001) 0.69 (0.559) 0.73 (0.682) Inhibition ability 3.66 (0.012) 3.55 (0.014) 1.48 (0.152) (SKT errors) Abstract reasoning 7.58 (< 0.001) 41.93 (< 0.001) 1.30 (0.230) (WAIS III) Verbal memory Verbal learning 5.27 (0.001) 45.86 (< 0.001) 2.93 (0.002) (WMS III) Long-term retention 3.45 (0.016) 43.70 (< 0.001) 2.51 (0.008) (WMS III) Cued-recall 3.15 (0.024) 45.46 (< 0.001) 2.92 (0.002) (WMS III) Language Naming 0.61 (0.612) 45.05 (< 0.001) 1.10 (0.363) (BNT abbreviated) Gnosis Poppelreuter 1.33 (0.262) 21.12 (< 0.001) 1.14 (0.333) Visuospatial skills Luria’s clock 3.14 (0.024) 20.32 (< 0.001) 0.82 (0.597) Praxis Block-design 3.19 (0.023) 18.34 (< 0.001) 1.30 (0.230) (WAIS III) General cognition Total score (sum) 10.34 (< 0.001) 138.86 (< 0.001) 2.60 (0.006) Results are shown as follows: F(p). After Bonferroni correction for multiple testing, an effect is significant when p < 0.003 (in bold). Asymptomatic class was considered the reference category in the LMM analysis. Age, MMSE, educational level, sex, conversion to dementia (yes/no), MCI type (amnestic/non-amnestic), and MCI profile (possible/probable) were also included in the models and were considered as fixed factors. WAIS, Wechsler Adult Intelligence Scale; SKT, Syndrom-Kurztest; WMS, Wechsler Memory Scale; BNT, Boston Naming Test. Cluster is the neuropsychiatric class (Class 1 = Irritability; Class 2 = Apathy; Class 3 = Anxiety/Depression; Class 4 = Asymptomatic). Time refers to the assessment at every follow-up for our study period. presented for each cluster. Figures 1, 2 show speed of decline, (Irritability, Apathy, Anxiety/Depression, and Asymptomatic) in with the Irritability class being the faster decliner (cognitive slope patients with MCI. In our sample of 2,137 MCI patients, the −0.98) and the Asymptomatic class the slower decliner (cognitive Irritability and Apathy NPS classes shared a similar pattern slope −0.43) in relation to verbal learning. For cued-recall, the of faster cognitive decline in two memory domains (verbal Irritability and Apathy classes had the same cognitive slope value, learning and cued-recall), compared to the Anxiety/Depression and the Anxiety/Depression class showed a slower decline in and Asymptomatic classes, which showed a slower cognitive this memory domain. worsening over the stipulated follow-up period. Even though Irritability was the least prevalent neuropsychiatric condition in this sample, it proved to be the NPS class with the worst and DISCUSSION fastest cognitive decline. Therefore, the present findings suggest that although Irritability and Apathy are less frequent NPS in The findings in our study revealed different trajectories of MCI, these symptoms should be taken into account to improve cognitive decline in memory domains depending on NPS clusters the quality and usefulness in diagnostic and prognostic evaluation Frontiers in Aging Neuroscience | www.frontiersin.org 5 December 2021 | Volume 13 | Article 718949 Roberto et al. Neuropsychiatric Profile Predicting Cognitive Decline TABLE 3 | Simple effects and effect sizes of significant cognitive domains (i.e., verbal learning and cued-recall) between baseline (X1 ) and third follow-up (X4 ) measures. Cluster 1 X1 –X4 Confidence interval for difference (95%) P Cohen’s d difference Verbal learning Irritability 3.66 2.48–4.84 <0.001 0.69 Apathy 2.54 1.59–3.48 <0.001 0.49 Anxiety/Depression 1.72 1.24–2.19 <0.001 0.31 Asymptomatic 1.35 0.76–1.92 <0.001 0.22 Cued recall Irritability 1.80 1.08–2.53 <0.001 0.64 Apathy 1.88 1.30–2.47 <0.001 0.46 Anxiety/Depression 0.70 0.40–0.99 <0.001 0.21 Asymptomatic 1.05 0.69–1.41 <0.001 0.30 Cluster reflects neuropsychiatric symptoms class: Class 1 = Irritability; Class 2 = Apathy; Class 3 = Anxiety/Depression; Class 4 = Asymptomatic. Confidence interval refers to mean differences. Asymptomatic class was considered the reference category. Age, MMSE, educational level, sex, conversion to dementia (yes/no), MCI type (amnestic/non-amnestic), and MCI profile (possible/probable) were also included in the models and were considered as fixed factors. 1 X1 –X4 = Difference of last follow-up from the baseline. FIGURE 1 | Cognitive decline across clusters for verbal learning. Measures for each group were obtained using LMM means by calculating differences between baseline and follow-ups for each time point. Slopes for each cluster were calculated using the 2-known points approach. Negative values correspond to decrements: the larger the absolute values the steeper the line. Numbers at the end of the lines indicate the cognitive slopes for each class. of cognitive worsening in MCI patients, especially those with an resulting clusters are mainly constellations of NPS, in which one amnestic profile. symptom is the most manifest. For instance, Irritability cluster Although irritability is included among the so-called affective embraced irritable symptoms (0.93), but also anxiety and apathy NPS, there is no substantial literature reporting consistent results to a lesser extent (both 0.63). Therefore, it is possible that the on how individuals evolve in terms of cognitive decline, neither differential cognitive decline observed in individuals belonging in healthy controls (Lobo et al., 2008; Leoutsakos et al., 2015), to the Irritability cluster may be somehow influenced by anxious nor in patients with MCI (Forrester et al., 2016), and/or dementia and apathic symptoms. In any case, with the present results, (Moran et al., 2004). Indeed, most of the studies in MCI this is just speculative, but future studies should explore the have focused on other affective NPS such as anxiety, apathy mechanistic process underneath the effect of NPS on cognitive and depression (Penna, 2013). However, some authors have decline. It can be hypothesized that the presence of irritability postulated that irritability could be among the affective symptoms may confer extra vulnerability to a faster conversion to dementia. that foretell a faster decline in conversion to dementia (Ismail It is worth mentioning that the large sample size of this study et al., 2017; Jang et al., 2020), but to date none have provided allowed the detection of an Irritability class, and it is probable that data on that. Therefore, the present results partially support this previous studies failed to detect a consistent cluster comprising hypothesis, adding some novelty about which cognitive domains individuals with irritability due to the lower prevalence compared could be more affected, always taking into account that the to other affective NPS. Frontiers in Aging Neuroscience | www.frontiersin.org 6 December 2021 | Volume 13 | Article 718949 Roberto et al. Neuropsychiatric Profile Predicting Cognitive Decline FIGURE 2 | Cognitive decline across clusters for cued-recall. Measures for each group were obtained using LMM means by calculating differences between baseline and follow-ups for each time point. Slopes for each cluster were calculated using the 2-known points approach. Negative values correspond to decrements: the larger the absolute values the steeper the line. Numbers at the end of the lines indicate the cognitive slopes for each class. In contrast, several studies have explored the relationship decline for those two clusters, as reported by other researchers between apathy and cognitive decline. Some authors indicated an (Ismail et al., 2017; Martin and Velayudhan, 2020). These findings increased risk of progression from MCI to AD when apathy was may suggest that although anxiety and depression are the most presented in isolation (Vicini Chilovi et al., 2009; Richard et al., frequently detected and known affective NPS in patients with 2012), whereas others postulated the risk was even higher when cognitive decline, clinicians should bear in mind other affective combined with depressive symptoms (Ruthirakuhan et al., 2019). NPS beyond anxious and depressed manifestations that could be Strikingly, low isolated depressive symptoms were not associated more relevant in the progression to dementia. with cognitive decline (Richard et al., 2012). Conversely, another The classification of individuals by symptomatic classes rather recent study demonstrated that both apathy and anxiety were than isolated symptoms seems to be more useful and informative associated with cognitive decline when presented comorbidly as it better reflects day-to-day reality in a memory clinic. (Johansson et al., 2020). Our results converged with these Among the different studies exploring MCI populations grouped findings, as we observed a sharper cognitive decline suffered by according to comorbid NPS, significant differences exist in patients in the Apathy class compared to Anxiety/Depression. obtained cluster solutions, probably due to the methodological Given that cognitive decline is one of the factors favoring approaches used. For instance, some works used a volunteer conversion to dementia and it was adjusted in our analyses, the sample (Leoutsakos et al., 2015; Forrester et al., 2016; Jang findings shed light on the NPS profiles that could entail an earlier et al., 2020), whereas others used clinical samples (David et al., risk of conversion, and thus act as isolated markers. 2016; Siafarikas et al., 2018; Liew, 2019). The statistical approach With regard to anxiety and depression, both are among the and designs were also different among studies (i.e., LCA vs. most prevalent affective NPS in MCI patients (Lyketsos et al., factor analysis, techniques that group individuals vs. grouping 2002; Zhang et al., 2012), but their influence on cognitive decline characteristics, respectively; or cross-sectional vs. longitudinal), is still controversial (Chan et al., 2011). Those symptoms have which could have undermined the importance of taking affective mostly been considered to be precursors of dementia, whereas NPS into consideration in diagnostic and prognostic evaluations. only a few studies considered anxiety and depression to be One of the abovementioned publications assessed cognitive a mere reaction to cognitive losses perceived by the patient decline across latent classes (David et al., 2016). However, (Simard et al., 2009; Di Iulio et al., 2010), which could be the authors did not include affective NPS per se, but rather a consequence of and not an early marker for conversion to their severity, and they only evaluated memory and executive dementia. The present findings provided evidence of no clear function domains, apart from the MMSE, to obtain a global association between anxious and depressive symptoms and faster cognitive measure. Therefore, the present study represents a step cognitive decline; in contrast to other studies, both symptoms did forward as cognitive decline was explored in assessing different not yield a worst prognosis in our sample. Note that the cognitive cognitive processes. trajectory of the Anxiety/Depression class was comparable to the A relevant finding of the present study is therefore that not Asymptomatic class in terms of showing no consistent cognitive all cognitive domains were affected equally at this early stage, Frontiers in Aging Neuroscience | www.frontiersin.org 7 December 2021 | Volume 13 | Article 718949 Roberto et al. Neuropsychiatric Profile Predicting Cognitive Decline but instead they behaved as isolated processes that showed subtle classes. The present findings emphasize the relevance of including differences in cognitive decline when NPS classes were taken an assessment of affective NPS when starting a diagnostic process into account. Likewise, the results also revealed that an accurate provided that such symptoms—and in particular irritability and assessment of MCI patients, contrary to previous work in already apathy—might act as aggravating factors. Our findings appear to diagnosed dementia patients (Escudero et al., 2019), should cover open a new avenue to use NPS assessment as a clinical tool of cognitive performance by domains as well as NPS as these can great value when it comes to detecting in advance which patients guide the prognosis of MCI, especially now, when diagnosis could suffer from a marked worsening in cognition. can be sought earlier than ever. Interestingly, different cognitive trajectories were observed according to early NPS instead of neurological symptoms, which could help clinicians be aware DATA AVAILABILITY STATEMENT of a possible diagnosis of dementia or other neurodegenerative diseases (Geda et al., 2013; Dietlin et al., 2019; Wise et al., 2019), The data that support this study may be provided upon and consider what is necessary to slow down progression of the reasonable request. Requests to access the datasets should be illness, where possible. directed to SV, svalero@fundacioace.org. There are also limitations to be considered in the current study. First, there are baseline differences in demographic characteristics and clinical variables among NPS clusters which ETHICS STATEMENT could undermine the findings, even though the analyses included The studies involving human participants were reviewed and these variables. Second, it is important to consider including approved by Hospital Clínic i Provincial of Barcelona, Spain. The medication records in future studies, as it could affect the patients/participants provided their written informed consent to evolution of an underlying neuropsychiatric condition. Also, a participate in this study. longer follow-up would be appropriate to determine how NPS and cognitive decline will interact in the long run, as well as to analyze the long-term stability of NPS classes. Also, the AUTHOR CONTRIBUTIONS presence of early AD-related biomarkers would help achieve a more accurate etiological diagnosis, and to benchmark NPS NR, MP, LT, MB, and SV conceived the idea, designed the study, observations. A third limitation is associated with the estimated and wrote the protocol and methodology. MM, MA, IH, AM, variances of parameter estimates; these may have been biased MR-R, CA, EE, SM-G, JT, LV, AE, GO, AP-C, ÁS, AO, IR, SM-G, because heteroscedasticity was not taken into account for the and LM acquired all the data, performed technical procedures, repeated measurements of individuals and consequently may and managed the data set. NR, MM, and AR managed previous have affected the precision of estimating the appropriate model. literature searches. NR, MP, and SV contributed to the statistical However, similar studies published so far have failed to account analysis, interpretation of the results and writing of the first for heteroscedasticity and the findings are consistent. Finally, draft of the manuscript. EA-M helped in the statistical analysis the last limitation is focused on sex perspective. In the present and interpretation of the results. MM, MA, AR, LT, and MB study it was necessary to explore over 2,100 MCI participants revised the manuscript critically for intellectual content. All to generate a consistent group of 134 irritable participants. In authors contributed to the writing of the final version and fact, this is one of the most relevant findings of our previous approved the manuscript. study (Roberto et al., 2021). Unfortunately, in this cluster, only 39 were women. The idea to analyze data considering sex as a main factor (in interaction with cluster and time) is, of course, FUNDING clinically interesting, but the distribution of this factor in the different clusters, and especially in the irritable group, prevented MP receives funding from the Centro de Investigación Biomédica us to consider sex factor as a key effect. More longitudinal en Red de Salud Mental (CIBERSAM) and from the Generalitat studies are necessary, with larger samples, to obtain consistent de Catalunya through recognition of consolidate group or empirical groups of patients where cognitive trajectories in men research (SGR17/001343). MM received research support from and women could be estimated with precision. the Acción Estratégica en Salud, integrated in the Spanish National RCDCI Plan and financed by the Instituto de Salud Carlos III (ISCIII)—Subdirección General de Evaluación and the CONCLUSION Fondo Europeo de Desarrollo Regional (FEDER—“Una manera de hacer Europa”) grant PII9/00335. AR has received support The approach of this study explores specific cognitive decline from CIBERNED [Instituto de Salud Carlos III (ISCIII)], the trajectories based on affective NPS clusters in MCI patients from EU/EFPIA Innovative Medicines Initiative Joint Undertaking, a memory clinic, adding some novelty with respect to previous ADAPTED Grant No. 115975, from the EXIT project, and works. Specifically, and according to our results, Irritability EU Euronanomed3 Program JCT2017 Grant No. AC17/00100, and Apathy classes share a similar pattern of faster cognitive from PREADAPT project. Joint Program for Neurodegenerative decline in two memory domains (verbal learning and cued- Diseases (JPND) Grant No. AC19/00097, and from grants recall), compared to the Anxiety/Depression and Asymptomatic PI13/02434, PI16/01861 BA19/00020, and PI19/01301, and Frontiers in Aging Neuroscience | www.frontiersin.org 8 December 2021 | Volume 13 | Article 718949 Roberto et al. Neuropsychiatric Profile Predicting Cognitive Decline Acción Estratégica en Salud, integrated in the Spanish National ACKNOWLEDGMENTS RCDCI Plan and financed by Instituto de Salud Carlos III (ISCIII)—Subdirección General de Evaluación and the Fondo The authors would like to thank all participating patients and Europeo de Desarrollo Regional (FEDER—“Una manera de their caregivers for their essential contribution to this project. Hacer Europa”), by Fundación Bancaria “La Caixa” and Grífols SA (GR@ACE project). MB has received grants from La Caixa, IMI, and ISCIII. MB has also received funding from SUPPLEMENTARY MATERIAL Grifols, Biogen, Roche, Merck, Lilly, Nutricia, Cortexyme, Renew Research, Araclon. The funders are not involved in the study The Supplementary Material for this article can be found design, collection, analysis, interpretation of data, the writing of online at: https://www.frontiersin.org/articles/10.3389/fnagi. this article or the decision to submit it for publication. 2021.718949/full#supplementary-material REFERENCES Erzigkeit, H. (1989). “The SKT: A short cognitive performance test as an instrument for the assessment of clinical efficacy of cognitive enhancers,” in Alegret, M., Espinosa, A., Valero, S., Vinyes-Junqué, G., Ruiz, A., Hernández, I., Diagnosis And Treatment Of Senile Dementia, eds M. Berenger and B. Reisberg et al. (2013). Cut-off scores of a brief neuropsychological battery (NBACE) (Berlin: Springer Verlag), 164–174. for spanish individual adults older than 44 years old. PLoS One 8, 1–8. doi: Escudero, J. M. S., Beltrán, J., Palacios, Á, Chimbí, C. M., Matallana, D., Reyes, 10.1371/journal.pone.0076436 P., et al. (2019). Neuropsychiatric symptoms as predictors of clinical course Alegret, M., Espinosa, A., Vinyes-Junqué, G., Valero, S., Hernández, I., Tárraga, in neurodegeneration. A longitudinal study. Front. Aging Neurosci. 11:176. L., et al. (2012). Normative data of a brief neuropsychological battery for doi: 10.3389/fnagi.2019.00176 Spanish individuals older than 49. J. Clin. Exp. Neuropsychol. 34, 209–219. Forrester, S. N., Gallo, J. J., Smith, G. S., and Leoutsakos, J.-M. (2016). Patterns doi: 10.1080/13803395.2011.630652 of neuropsychiatric symptoms in MCI and risk of dementia. Am. J. Geriatr. Boada, M., Cejudo, J. C., Tàrraga, L., López, O. L., and Kaufer, D. (2002). Psychiatry 24, 87–92. doi: 10.1016/j.coviro.2015.09.001.Human Neuropsychiatric inventory questionnaire (NPI-Q): validación española de Gallagher, D., Fischer, C. E., and Iaboni, A. (2017). Neuropsychiatric symptoms in una forma abreviada del Neuropsychiatric Inventory (NPI). Neurología 17, mild cognitive impairment: an update on prevalence, mechanisms, and clinical 317–323. significance. Can. J. Psychiatry 62, 161–169. doi: 10.1177/0706743716648296 Boada, M., Tárraga, L., Hernández, I., Valero, S., Alegret, M., Ruiz, A., et al. Geda, Y. E., Roberts, R. O., Knopman, D. S., Petersen, R. C., Christianson, T. J. H., (2014). Design of a comprehensive Alzheimer’s disease clinic and research Pankratz, V. S., et al. (2008). Prevalence of neuropsychiatric symptoms in mild center in Spain to meet critical patient and family needs. Alzheimers Dement. cognitive impairment and normal cognitive aging: population-based study. 10, 409–415. doi: 10.1016/j.jalz.2013.03.006 Arch. Gen. Psychiatry 65, 1193–1198. doi: 10.1001/archpsyc.65.10.1193 Burhanullah, M. H., Tschanz, J. A. T., Peters, M. E., Leoutsakos, J. M., Matyi, J., Geda, Y. E., Schneider, L. S., Gitlin, L. N., Miller, D. S., Smith, G. S., Bell, J., et al. Lyketsos, C. G., et al. (2020). Neuropsychiatric symptoms as risk factors for (2013). Neuropsychiatric symptoms in Alzheimer’s disease: past progress and cognitive decline in clinically normal older adults: the cache county study. Am. anticipation of the future. Alzheimers Dement. 9, 602–608. doi: 10.1016/j.jalz. J. Geriatr. Psychiatry 28, 64–71. doi: 10.1016/j.jagp.2019.03.023 2012.12.001 Canevelli, M., Adali, N., Voisin, T., Soto, M. E., Bruno, G., Cesari, M., et al. (2013). Ginsberg, T. B., Powell, L., Emrani, S., Wasserman, V., Higgins, S., Chopra, A., Behavioral and psychological subsyndromes in Alzheimer’s disease using the et al. (2019). Instrumental activities of daily living, neuropsychiatric symptoms, neuropsychiatric inventory. Int. J. Geriatr. Psychiatry 28, 795–803. doi: 10.1002/ and neuropsychological impairment in mild cognitive impairment. J. Am. gps.3904 Osteopath. Assoc. 119, 96–101. doi: 10.7556/jaoa.2019.015 Chan, W. C., Lam, L. C. W., Tam, C. W. C., Lui, V. W. C., Leung, G. T. Y., Glisky, E. (2007). “Changes in cognitive function in human aging,” in Brain Aging, Lee, A. T. C., et al. (2011). Neuropsychiatric symptoms are associated with ed. D. R. Riddle (Boca Raton, FL: CRC Press/Taylor & Francis). doi: 10.1201/ increased risks of progression to dementia: a 2-year prospective study of 321 9781420005523.sec1 Chinese older persons with mild cognitive impairment. Age Ageing 40, 30–35. Golden, C. J. (1980). In reply to adams “in search of luria’s battery: a false start”. doi: 10.1093/ageing/afq151 J. Consult. Clin. Psychol. 48, 517–521. David, N. D., Feng Lin, R. N., and Porsteinsson, A. P. (2016). Trajectories of Gonzales, M. M., Insel, P. S., Nelson, C., Tosun, D., Mattsson, N., Mueller, S. G., neuropsychiatric symptoms and cognitive decline in mild cognitive impairment et al. (2017). Cortical atrophy is associated with accelerated cognitive decline Nicholas. Am. J. Geriatr. Psychiatry 24, 70–80. doi: 10.1016/j.jagp.2015.06.001 in mild cognitive impairment with subsyndromal depression. Am. J. Geriatr. Della Sala, S., Laiacona, M., Trivelli, C., and Spinnler, H. (1995). Poppelreuter- Psychiatry 25, 980–991. doi: 10.1016/j.jagp.2017.04.011 Ghent overlapping figures test. Its sensitivity to age, and its clinical use. Arch. Grober, E., Hall, C. B., Lipton, R. B., Zonderman, A. B., Resnick, S. M., and Kawas, Clin. Neuropsychol. 10, 511–534. C. (2008). Memory impairment, executive dysfunction, and intellectual decline Di Iulio, F., Palmer, K., Blundo, C., Casini, A. R., Gianni, W., Caltagirone, C., et al. in preclinical Alzheimer’s disease. J. Int. Neuropsychol. Soc. 14, 266–278. (2010). Occurrence of neuropsychiatric symptoms and psychiatric disorders Hermida, A. P., McDonald, W. M., Steenland, K., and Levey, A. (2012). in mild Alzheimer’s disease and mild cognitive impairment subtypes. Int. The association between late-life depression, mild cognitive impairment and Psychogeriatrics 22, 629–640. doi: 10.1017/S1041610210000281 dementia: is inflammation the missing link? Expert Rev. Neurother. 12, 1339– Dietlin, S., Soto, M., Kiyasova, V., Pueyo, M., De Mauleon, A., Delrieu, J., et al. 1350. doi: 10.1586/ern.12.127 (2019). Neuropsychiatric symptoms and risk of progression to Alzheimer’s Huang, M. F., Lee, W. J., Yeh, Y. C., Liao, Y. C., Wang, S. J., Yang, Y. H., et al. disease among mild cognitive impairment subjects. J. Alzheimers Dis. 70, 25–34. (2020). Genetics of neuropsychiatric symptoms in patients with Alzheimer’s doi: 10.3233/JAD-190025 disease: a 1-year follow-up study. Psychiatry Clin. Neurosci. 74, 645–651. doi: Donaghy, P. C., Taylor, J. P., O’Brien, J. T., Barnett, N., Olsen, K., Colloby, 10.1111/pcn.13150 S. J., et al. (2018). Neuropsychiatric symptoms and cognitive profile in mild Ismail, Z., Agüera-Ortiz, L., Brodaty, H., Cieslak, A., Cummings, J., Fischer, C. E., cognitive impairment with lewy bodies. Psychol. Med. 48, 2384–2390. doi: 10. et al. (2017). The mild behavioral impairment checklist (MBI-C): a rating scale 1017/S0033291717003956 for neuropsychiatric symptoms in pre-dementia populations. J. Alzheimers Dis. Edwards, E. R., Spira, A. P., Barnes, D. E., and Yaffe, K. (2009). Neuropsychiatric 56, 929–938. doi: 10.3233/JAD-160979 symptoms in mild cognitive impairment: differences by subtype and Ismail, Z., Gatchel, J., Bateman, D. R., Barcelos-Ferreira, R., Chantillon, M., Jaeger, progression to dementia. Int. J. Geriatr. Psychiatry 24, 716–722. doi: 10.1002/ J., et al. (2018). Affective and emotional dysregulation as pre-dementia risk gps.2187 markers: exploring the mild behavioral impairment symptoms of depression, Frontiers in Aging Neuroscience | www.frontiersin.org 9 December 2021 | Volume 13 | Article 718949 Roberto et al. Neuropsychiatric Profile Predicting Cognitive Decline anxiety, irritability, and euphoria. Int. Psychogeriatrics 30, 185–196. doi: 10. non-depressed subjects for the Alzheimer’s disease neuroimaging initiative. 1017/S1041610217001880 Dement. Geriatr. Cogn. Disord. 33, 204–209. doi: 10.1159/000338239 Jang, J. Y., Ho, J. K., Blanken, A. E., Dutt, S., and Nation, D. A. (2020). Affective Roberto, N., Portella, M. J., Marquié, M., Alegret, M., Hernández, I., Mauleón, neuropsychiatric symptoms as early signs of dementia risk in older adults. A., et al. (2021). Neuropsychiatric profiles and conversion to dementia in mild J. Alzheimers Dis. 77, 1195–1207. doi: 10.3233/jad-200190 cognitive impairment, a latent class analysis. Sci. Rep. 11, 1–9. doi: 10.1038/ Johansson, M., Stomrud, E., Lindberg, O., Westman, E., Johansson, P. M., van s41598-021-83126-y Westen, D., et al. (2020). Apathy and anxiety are early markers of Alzheimer’s Ruthirakuhan, M., Herrmann, N., Vieira, D., Gallagher, D., and Lanctôt, K. L. disease. Neurobiol. Aging 85, 74–82. doi: 10.1016/j.neurobiolaging.2019.10.008 (2019). The roles of apathy and depression in predicting Alzheimer disease: a Kaplan, E. F., Goodglass, H., and Weintraub, S. (1983). The Boston Naming Test. longitudinal analysis in older adults with mild cognitive impairment. Am. J. Philadelphia, PA: Lea & Febiger. Geriatr. Psychiatry 27, 873–882. doi: 10.1016/j.jagp.2019.02.003 Karttunen, K., Karppi, P., Hiltunen, A., Vanhanen, M., Välimäki, T., Martikainen, Siafarikas, N., Selbaek, G., Fladby, T., Šaltyte Benth, J., Auning, E., and Aarsland, J., et al. (2011). Neuropsychiatric symptoms and quality of life in patients with D. (2018). Frequency and subgroups of neuropsychiatric symptoms in mild very mild and mild Alzheimer’s disease. Int. J. Geriatr. Psychiatry 26, 473–482. cognitive impairment and different stages of dementia in Alzheimer’s disease. doi: 10.1002/gps.2550 Int. Psychogeriatrics 30, 103–113. doi: 10.1017/S1041610217001879 Leoutsakos, J. M. S., Forrester, S. N., Lyketsos, C. G., and Smith, G. S. (2015). Latent Simard, M., Hudon, C., and van Reekum, R. (2009). Psychological distress and risk classes of neuropsychiatric symptoms in NACC controls and conversion to mild for dementia. Curr. Psychiatry Rep. 11, 41–47. cognitive impairment or dementia. Handb. Depress. Alzheimers Dis. 48, 81–92. Singh-Manoux, A., Dugravot, A., Fournier, A., Abell, J., Ebmeier, K., Kivimäki, doi: 10.3233/978-1-61499-542-5-81 M., et al. (2017). Trajectories of depressive symptoms before diagnosis of Liew, T. M. (2019). Symptom clusters of neuropsychiatric symptoms in mild dementia: a 28-year follow-up study. JAMA Psychiatry 74, 712–718. doi: 10. cognitive impairment and their comparative risks of dementia: a cohort study 1001/jamapsychiatry.2017.0660 of 8530 older persons. J. Am. Med. Dir. Assoc. 20, 1054–1054.e9. doi: 10.1016/j. Tsunoda, K., Yamashita, T., Osakada, Y., Sasaki, R., Tadokoro, K., Matsumoto, jamda.2019.02.012 N., et al. (2020). Early emergence of neuropsychiatric symptoms in cognitively Lobo, A., López-Antón, R., De-La-Cámara, C., Quintanilla, M. Á, Campayo, A., normal subjects and mild cognitive impairment. J. Alzheimers Dis. 73, 209–215. and Saz, P. (2008). Non-cognitive psychopathological symptoms associated doi: 10.3233/JAD-190669 with incident mild cognitive impairment and dementia, Alzheimer’s type. Valero, S., Marquié, M., De Rojas, I., Espinosa, A., Moreno-Grau, S., Orellana, Neurotox. Res. 14, 263–272. doi: 10.1007/BF03033815 A., et al. (2020). Interaction of neuropsychiatric symptoms with APOE ε4 and Lyketsos, G. C., López, O., Jones, B., Fitzpatrick, A. L., Breitner, J., and Dekosky, S. conversion to dementia in MCI patients in a memory clinic. Sci. Rep. 10, 1–10. (2002). Prevalence of neuropsychiatric symptoms in dementia and mci: results doi: 10.1038/s41598-020-77023-z from the cardiovascular health study. JAMA 288, 1475–1483. doi: 10.1001/jama. Vicini Chilovi, B., Conti, M., Zanetti, M., Mazzù, I., Rozzini, L., and Padovani, 288.12.1475 A. (2009). Differential impact of apathy and depression in the development of Ma, L. (2020). Depression, anxiety, and apathy in mild cognitive impairment: dementia in mild cognitive impairment patients. Dement. Geriatr. Cogn. Disord. current perspectives. Front. Aging Neurosci. 12:9. doi: 10.3389/fnagi.2020. 27, 390–398. doi: 10.1159/000210045 00009 Wechsler, D. (1997a). WAIS–III. Wechsler Adult Intelligence Scale–Third Edition. Martin, E., and Velayudhan, L. (2020). Neuropsychiatric symptoms in mild Technical Manual. San Antonio, TX: The Psychological Corporation. cognitive impairment: a literature review. Dement. Geriatr. Cogn. Disord. 49, Wechsler, D. (1997b). WMS–III. Wechsler Memory Scale–Third Edition. 146–155. doi: 10.1159/000507078 Administration and Scoring Manual. San Antonio, TX: The Psychological Moran, M., Walsh, C., Lynch, A., Coen, R. F., Coakley, D., and Lawlor, B. A. (2004). Corporation. Syndromes of behavioural and psychological symptoms in mild Alzheimer’s Weintraub, D., Simuni, T., Caspell-Garcia, C., Coffey, C., Lasch, S., Siderowf, A., disease. Int. J. Geriatr. Psychiatry 19, 359–364. doi: 10.1002/gps.1091 et al. (2015). Cognitive performance and neuropsychiatric symptoms in early, Mortby, M. E., and Anstey, K. J. (2015). “Mental health and aging,” in Encyclopedia untreated Parkinson’s disease. Physiol. Behav. 30, 919–927. doi: 10.1002/mds. of Geropsychology ed. A. N. Pachana Singapore: Springer Singapore. doi: 10. 26170 1007/978-981-287-080-3_1-1 Wise, E. A., Rosenberg, P. B., Lyketsos, C. G., and Leoutsakos, J. M. (2019). Palmer, K., Berger, A. K., Monastero, R., Winblad, B., Bäckman, L., and Fratiglioni, Time course of neuropsychiatric symptoms and cognitive diagnosis in national L. (2007). Predictors of progression from mild cognitive impairment to Alzheimer’s coordinating centers volunteers. Alzheimers Dement. Diagnosis Alzheimer disease. Neurology 68, 1596–1602. doi: 10.1212/01.wnl.0000260968. Assess. Dis. Monit. 11, 333–339. doi: 10.1016/j.dadm.2019.02.006 92345.3f Zhang, M., Wang, H., Li, T., and Yu, X. (2012). Prevalence of neuropsychiatric Park, D. C., and Reuter-Lorenz, P. (2009). The adaptive brain: aging and symptoms across the declining memory continuum: an observational study neurocognitive scaffolding. Annu. Rev. Psychol. 60, 173–196. doi: 10.1146/ in a memory clinic setting. Dement. Geriatr. Cogn. Dis. Extra 2, 200–208. annurev.psych.59.103006.093656 doi: 10.1159/000338410 Park, D. C., Lautenschlager, G., Hedden, T., Davidson, N. S., Smith, A. D., and Smith, P. K. (2002). Models of visuospatial and verbal memory across the adult Conflict of Interest: The authors declare that the research was conducted in the life span. Psychol. Aging 17, 299–320. doi: 10.1037/0882-7974.17.2.299 absence of any commercial or financial relationships that could be construed as a Penna, S. (2013). Cognitive and emotional dysfunction in mild cognitive potential conflict of interest. impairment. Clin. Geriatr. Med. 29, 773–789. doi: 10.1016/j.cger.2013.07.004 Peters, M. E., Rosenberg, P. B., Steinberg, M., Tschanz, J. T., Norton, M. C., Welsh- Publisher’s Note: All claims expressed in this article are solely those of the authors Bohmer, K. A., et al. (2012). Prevalence of neuropsychiatric symptoms in CIND and do not necessarily represent those of their affiliated organizations, or those of and its subtypes: the cache county study. Am. J. Geriatr. Psychiatry 20, 416–424. the publisher, the editors and the reviewers. Any product that may be evaluated in doi: 10.1097/JGP.0b013e318211057d this article, or claim that may be made by its manufacturer, is not guaranteed or Petersen, R. C. (2004). Mild cognitive impairment as a diagnostic entity. J. Intern. endorsed by the publisher. Med 256, 183–194. doi: 10.1111/j.1365-2796.2004.01388.x/pdf Petersen, R. C., Smith, G. E., Waring, S. C., Ivnik, R. J., Tangalos, E. G., and Copyright © 2021 Roberto, Portella, Marquié, Alegret, Hernández, Mauleón, Kokmen, E. (1999). Mild cognitive impairment: clinical characterization and Rosende-Roca, Abdelnour, Esteban de Antonio, Tartari, Vargas, López-Cuevas, outcome. Arch. Neurol. 56, 303–309. Bojaryn, Espinosa, Ortega, Pérez-Cordón, Sanabria, Orellana, de Rojas, Moreno- Pirogovsky-Turk, E., Moore, R. C., Filoteo, J. V., Litvan, I., Song, D. D., Lessig, Grau, Montrreal, Alarcón-Martín, Ruíz, Tárraga, Boada and Valero. This is an S. L., et al. (2017). Neuropsychiatric predictors of cognitive decline in parkinson open-access article distributed under the terms of the Creative Commons Attribution disease: a longitudinal study. Am. J. Geriatr. Psychiatry 25, 279–289. doi: 10. License (CC BY). The use, distribution or reproduction in other forums is permitted, 1016/j.jagp.2016.10.004 provided the original author(s) and the copyright owner(s) are credited and that the Richard, E., Schmand, B., Eikelenboom, P., Yang, S. C., Ligthart, S. A., Moll original publication in this journal is cited, in accordance with accepted academic Van Charante, E. P., et al. (2012). Symptoms of apathy are associated practice. No use, distribution or reproduction is permitted which does not comply with progression from mild cognitive impairment to Alzheimer’s disease in with these terms. Frontiers in Aging Neuroscience | www.frontiersin.org 10 December 2021 | Volume 13 | Article 718949