Risk of Cognitive Impairment in Patients With Parkinson’s Disease With Visual Hallucinations and Subjective Cognitive Complaints

344 Copyright © 2023 Korean Neurological Association Risk of Cognitive Impairment in Patients With Parkinson’s Disease With Visual Hallucinations and Subjective Cognitive Complaints pISSN 1738-6586 / eISSN 2005-5013 / J Clin Neurol 2023;19(4):344-357 / https://doi.org/10.3988/jcn.2022.0186 Received May 7, 2022 Revised August 31, 2022 Accepted September 1, 2022 Correspondence Diego Santos-García, PhD, Department of Neurology, Hospital Universitario de A Coruña (HUAC), Complejo Hospitalario Universitario de A Coruña (CHUAC), C/As Xubias 84, A Coruña 15006, Spain Tel +34-646173341 E-mail [email protected] *Details of COPPADIS Study Group is presented in Supplementary Material. cc This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (https://creativecommons.org/ licenses/by-nc/4.0) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited. JCN Open Access ORIGINAL ARTICLE Diego Santos-Garcíaa, Teresa de Deus Fonticobab, Carlos Cores Bartoloméa, Maria J. Feal Painceirasa, Jose M. Paz Gonzáleza, Cristina Martínez Miróa, Silvia Jesúsc,d, Miquel Aguilare, Pau Pastore, Lluís Planellasf, Marina Cosgayag, Juan García Caldenteyh, Nuria Caballoli, Ines Legardaj, Jorge Hernández Varad,k, Iria Cabol, Lydia López Manzanaresm, Isabel González Aramburud,n, Maria A. Ávila Riverao, Víctor Gómez Mayordomop, Víctor Nogueiraq, Víctor Puenter, Julio Dotor García-Sotos, Carmen Borruét, Berta Solano Vilau, María Álvarez Saucov, Lydia Velaw, Sonia Escalantex, Esther Cuboy, Francisco Carrillo Padillaz, Juan C. Martínez CastrilloA, Pilar Sánchez AlonsoB, Maria G. Alonso LosadaC, Nuria López ArizteguiD, Itziar GastónE, Jaime Kulisevskyd,F, Marta Blázquez EstradaG, Manuel Seijol, Javier Rúiz MartínezH, Caridad ValeroI, Mónica KurtisJ, Oriol de Fábreguesk, Jessica González ArduraK, Ruben Alonso RedondoL, Carlos OrdásM, Luis M. López Díaz LN, Darrian McAfeeO, Pablo Martinez-Martind, Pablo Mirc,d, COPPADIS Study Group* aCHUAC, Complejo Hospitalario Universitario de A Coruña, A Coruña, Spain bCHUF, Complejo Hospitalario Universitario de Ferrol, A Coruña, Spain cUnidad de Trastornos del Movimiento, Servicio de Neurología y Neurofisiología Clínica, Instituto de Biomedicina de Sevilla, Hospital Universitario Virgen del Rocío/CSIC/ Universidad de Sevilla, Seville, Spain dCIBERNED (Centro de Investigación Biomédica en Red Enfermedades Neurodegenerativas), Madrid, Spain eHospital Universitari Mutua de Terrassa, Terrassa, Spain fClínica del Pilar, Barcelona, Spain gHospital Clínic de Barcelona, Barcelona, Spain hCentro Neurológico Oms 42, Palma de Mallorca, Spain iConsorci Sanitari Integral, Hospital Moisés Broggi, Sant Joan Despí, Spain jHospital Universitario Son Espases, Palma de Mallorca, Spain kHospital Universitario Vall d´Hebron, Barcelona, Spain lComplejo Hospitalario Universitario de Pontevedra (CHOP), Pontevedra, Spain mHospital Universitario La Princesa, Madrid, Spain nHospital Universitario Marqués de Valdecilla, Santander, Spain oConsorci Sanitari Integral, Hospital General de L’Hospitalet, L’Hospitalet de Llobregat, Barcelona, Spain pHospital Universitario Clínico San Carlos, Madrid, Spain qHospital Da Costa, Burela, Lugo, Spain rHospital del Mar, Barcelona, Spain sHospital Universitario Virgen Macarena, Sevilla, Spain tHospital Infanta Sofía, Madrid, Spain uInstitut d’Assistència Sanitària (IAS)-Institut Català de la Salut, Girona, Spain vHospital General Universitario de Elche, Elche, Spain wFundación Hospital de Alcorcón, Madrid, Spain xHospital de Tortosa Verge de la Cinta (HTVC), Tortosa, Spain yComplejo Asistencial Universitario de Burgos, Burgos, Spain zHospital Universitario de Canarias, San Cristóbal de la Laguna, Spain AHospital Universitario Ramón y Cajal, RYCIS, Madrid, Spain BHospital Universitario Puerta de Hierro, Madrid, Spain CHospital Álvaro Cunqueiro, Complejo Hospitalario Universitario de Vigo (CHUVI), Vigo, Spain DComplejo Hospitalario de Toledo, Toledo, Spain EComplejo Hospitalario de Navarra, Pamplona, Spain FHospital de Sant Pau, Barcelona, Spain GHospital Universitario Central de Asturias, Oviedo, Spain HHospital Universitario Donostia, San Sebastián, Spain IHospital Arnau de Vilanova, Valencia, Spain JHospital Ruber Internacional, Madrid, Spain KHospital de Cabueñes, Gijón, Spain LUniversitario Lucus Augusti (HULA), Lugo, Spain MHospital Rey Juan Carlos, Madrid, Spain NComplejo Hospitalario Universitario de Orense (CHUO), Orense, Spain OUniversity of Maryland School of Medicine, Baltimore, MD, USA www.thejcn.com 345 Santos-García D et al. JCN JCN Open Access INTRODUCTION Cognitive impairment (CI) is one of the most important nonmotor symptoms that can appear in patients with Parkinson’s disease (PD).1 The full spectrum of cognition appears in individuals with PD, ranging from normal cognition (NC) to mild cognitive impairment (MCI) to Parkinson’s disease dementia (PDD).2 About 50% of patients with Parkinson’s disease and normal cognition (PD-NC) develop MCI within 6 years,3 and almost 40% of patients with Parkinson’s disease and mild cognitive impairment (PD-MCI) subsequently develop PDD.4 Although findings vary among studies, the cumulative prevalence rates of PDD were found to be 17%, 46%, and 83% at 5, 10, and 20 years after diagnosis, respectively.5 Since dementia is a frequent and highly disabling complication in patients with PD, it is important to identify predictive factors for CI development. The most-established risk factors for early dementia are old age, motor symptom severity (particularly postural and gait disturbances), MCI, and visual hallucinations (VH).6 VH are a common symptom in PD, affecting up to 45% of patients without dementia and 65% of those with PDD.7 Importantly, the early presence of VH is a strong predictor of cognitive decline,8 as well as increased mortality and reduced quality of life (QoL) for patients and their caregivers.9 Subjective cognitive complaints (SCC) have more recently been suggested to be an independent predictor of MCI development in PD-NC.10,11 SCC is the subjective identification of cognitive decline in people who may or may not have had impairment detected in neuropsychological tests.12 The prevalence of SCC in PD reportedly varies from 30% to 60%,10,13 and up to 70% of patients with PD-NC who develop CI over time have SCC at baseline.14 However, the etiology of SCC can differ among patients, and they can be related to cognitive decline as well as to depression or other psychiatric symptoms.13,15-17 In this context, the relationships among SCC, VH, and CI are unknown, and it is unclear how VH and SCC contribute to CI development in patients with PD-NC. We hypothesized that the VH prevalence is higher in patients with PD-NC with SCC than in those without SCC and that both VH and SCC together could increase the risk of CI development in patients with PD-NC. Our aims were to determine the frequencies of VH and SCC in a PD-NC cohort, to perform comparison with a control group, and to determine the relationship between cognitive function and factors associated with VH and SCC and the risk of CI development at a 2-year follow-up. Moreover, we analyzed the values of different serum biomarkers (SB) regarding the presence of VH and SCC. Background and Purpose Visual hallucinations (VH) and subjective cognitive complaints (SCC) are associated with cognitive impairment (CI) in Parkinson’s disease. Our aims were to determine the association between VH and SCC and the risk of CI development in a cohort of patients with Parkinson’s disease and normal cognition (PD-NC). Methods Patients with PD-NC (total score of >80 on the Parkinson’s Disease Cognitive Rating Scale [PD-CRS]) recruited from the Spanish COPPADIS cohort from January 2016 to November 2017 were followed up after 2 years. Subjects with a score of ≥1 on domain 5 and item 13 of the Non-Motor Symptoms Scale at baseline (V0) were considered as “with SCC” and “with VH,” respectively. CI at the 2-year follow-up (plus or minus 1 month) (V2) was defined as a PD-CRS total score of <81. Results At V0 (n=376, 58.2% males, age 61.14±8.73 years [mean±SD]), the frequencies of VH and SCC were 13.6% and 62.2%, respectively. VH were more frequent in patients with SCC than in those without: 18.8% (44/234) vs 4.9% (7/142), p<0.0001. At V2, 15.2% (57/376) of the patients had developed CI. VH presenting at V0 was associated with a higher risk of CI at V2 (odds ratio [OR]=2.68, 95% confidence interval=1.05–6.83, p=0.039) after controlling for the effects of age, disease duration, education, medication, motor and nonmotor status, mood, and PD-CRS total score at V0. Although SCC were not associated with CI at V2, presenting both VH and SCC at V0 increased the probability of having CI at V2 (OR=3.71, 95% confidence interval=1.36–10.17, p=0.011). Conclusions VH were associated with the development of SCC and CI at the 2-year follow-up in patients with PD-NC. Key Words cognitive impairment; dementia; parkinson’s disease; subjective cognitive complaints; visual hallucinations. 346 J Clin Neurol 2023;19(4):344-357 Visual Hallucinations and Subjective Cognitive Complaints in PD JCN METHODS Patients diagnosed with PD from January 2016 to November 2017 and controls were recruited from the COPPADIS cohort for inclusion in this study, and were evaluated again at a 2-year follow-up in 35 centers in Spain.18 The methodology of the COPPADIS-2015 study can be found at https://bmcneurol.biomedcentral.com/articles/10.1186/s12883-016-05489.19 This was a multicenter, observational, longitudinal-prospective, 5-year follow-up study designed to analyze disease progression in a Spanish population of patients with PD. All patients included were diagnosed according to the UK PD Brain Bank criteria.20 The same inclusion (except PD diagnosis) and exclusion criteria were applied to patients and to the controls (subjects with a disabling neurological or nonneurological condition were excluded). Only subjects from the COPPADIS cohort without CI at baseline (i.e., PD-NC, corresponding to a total score on the Parkinson’s Disease Cognitive Rating Scale [PD-CRS] of <81)21 were analyzed in this study. Information on sociodemographic aspects, and factors related to PD, comorbidities, and treatment were collected. The evaluations performed at V0 (baseline) and V2 (2-year followup, plus or minus 1 month) included motor assessments (Hoehn and Yahr, Unified Parkinson’s Disease Rating Scale [UPDRS] part III and part IV, Freezing of Gait Questionnaire [FOGQ]), nonmotor symptoms (Non-Motor Symptoms Scale [NMSS], Parkinson’s Disease Sleep Scale [PDSS], Visual Analog Scale–Pain [VAS-Pain], Visual Analog Scale–Fatigue [VASF]), cognition (PD-CRS), mood and neuropsychiatric symptoms (Beck Depression Inventory-II [BDI-II], Neuropsychiatric Inventory [NPI], Questionnaire for ImpulsiveCompulsive Disorders in Parkinson’s Disease Rating Scale [QUIP-RS]), disability (Schwab and England Activities of Daily Living Scale [ADLS]), and QoL (39-item Parkinson’s Disease Questionnaire Summary Index [PDQ-39SI], 8-item EUROHIS-QOL index [EUROHIS-QOL8]).19 A higher score indicates a more-severe finding on each scale/questionnaire except for PD-CRS, PDSS, ADLS, and EUROHIS-QOL8, for which the opposite holds. In patients with motor fluctuations, motor assessments were performed during the Off state (without medication in the last 12 hours) and On state. On the other hand, the assessment was only performed without medication in patients without motor fluctuations. The same evaluations (except for the motor assessment) were performed on the control subjects at V0 and V2.19 Cognitive assessment and cognitive status classification Cognitive statuses at V0 and V2 and the changes between these time points were assessed using the PD-CRS,22 which is a cognitive screening battery validated to assess the cognitive status of patients with PD. The total score is the sum of the frontal-subcortical (FS) subscore (item 1, immediate free recall verbal memory; item 3, sustained attention; item 4, working memory; item 5, unprompted drawing of a clock; item 7, delayed free recall verbal memory; item 8, alternating verbal fluency; and item 9, action verbal fluency) and the posteriorcortical (PC) subscore (item 2, confrontation naming; item 6, copy drawing of a clock). According to the PD-CRS total score, each patient was classified as cognitively preserved (PDNC; score of >80) or CI (score of <81).23 All patients had PDNC at baseline. VH and SCC definition Subjects were classified as with or without VH according to item 13 of the NMSS at baseline (V0).24 This is 1 of 30 items in this scale and is included in domain 4 (perception problems/ hallucinations). The symptoms refer to the those that occur during the 4 weeks prior to the assessment. The question asking about VH was “Does the patient indicate that he/she sees things that are not there?” The score ranges from 0 (without symptoms) to 12 (most frequent and severe symptoms). Subjects with a score of 0 on item 13 of the NMSS were considered as “without VH” whereas subjects with a score of 1–12 were considered as “with VH.” The same method was used at V2, and persistent VH were defined as having VH at both V0 and V2. Regarding SCC, subjects were classified as with or without SCC according to domain 5 (attention/memory) of the NMSS at baseline. This domain includes three questions about cognition perception: “Does the patient have problems sustaining concentration during activities?” (item 16), and “Does the patient forget things that he/she has been told a short time ago or events that happened in the last few years?” (item 17), and “Does the patient forget to do things?” (item 18). The score for each item ranges from 0 (without symptoms) to 12 (most frequent and severe symptoms), with the total score of domain 5 ranging from 0 (without symptoms) to 36 (score of 12 [most frequent and severe symptoms] in all items). Subjects with a score of 0 in domain 5 of the NMSS were considered as “without SCC” whereas subjects with a score of 1–36 were considered as “with SCC.” The same method was used at V2, and persistent SCC was defined as having SCC at both V0 and V2. SB determination SB were analyzed among a subgroup of the COPPADIS cohort.19 Collected blood samples were used to determine different SB, and included S100B protein, tumor necrosis factor (TNF)-α, interleukin (IL)-1, IL-2, IL-6, vitamin B12, methyl- www.thejcn.com 347 Santos-García D et al. JCN malonic acid, homocysteine, uric acid, ultrasensitive CRP (US-CRP), ferritin, and iron. SB levels were determined from frozen blood samples obtained from subjects who participated in the COPPADIS-2015 study from nine centers in Spain. Sampling was carried out no longer than 3 months after the first clinical assessment (V0) in the absence of infections and/ or fever. All analyses were conducted at the same laboratory: REFERENCE LABORATORY (www.reference-laboratory. es).19 Different methods were used: visible spectrophotometry (iron), immunoluminescence (S100B protein, ferritin, vitamin B12, and homocysteine), enzyme immunoassay (IL1, IL-2, and TNF-α), immunoassay (US-CRP), mass spectrometry (methylmalonic acid), and an enzymatic technique (uric acid). Outliers were excluded from the analysis. Data analysis Data were processed using SPSS software (version 20.0 for Windows, IBM Corp, Armonk, NY, USA). Comparisons between patients and controls and between patients with and without VH and/or SCC were performed using Student’s ttest, the Mann-Whitney U test, the chi-square test, or Fisher’s test as appropriate (distributions of variables were verified using the one-sample Kolmogorov-Smirnov test). The general linear model repeated-measures procedure was used to test whether the PD-CRS total score, subscores, and items differed significantly between the two visits (V0 and V2) in patients with PD regarding the presence of VH and SCC. The Bonferroni method was used as a post-hoc test after ANOVA. Interactions for visit and group were then tested before testing for group differences over time. Cohen’s d formula was applied to measure the effect size (in patients with PD), which was categorized into a small effect (=0.2), medium effect (=0.5), or large effect (=0.8). Age, disease duration, education, and levodopa equivalent daily dose (LEDD)25 at V0 were included as covariates. To explore the association between VH and/or SCC and the risk of CI, binary regression models were used with the presence of CI (PD-CRS score <81) as a dependent variable. The effect was controlled for age, sex, disease duration, education, LEDD, motor (UPDRS-III and UPDRS-IV) and nonmotor (NMSS) status, mood (BDI-II), cognitive function (PDCRS total score), REM behavior disorder (RBD), and taking a dopamine agonist at V0, which were included as covariates in the model. Our analysis was based on a clearly specified a-priori hypothesis and a well-planned regression model, as recommended by best-practice methods.26 All values were quoted to two decimal places, with the exception of percentage values for which a single decimal place was used (in the case of zero, it was omitted). A probability value of p<0.05 was considered significant. Standard protocol approvals, registrations, and patient consents We received approval for this study from the Comité de Ética de la Investigación Clínica de Galicia in Spain (2014/534; December 2, 2014). Written informed consents were obtained from all participants in this study. COPPADIS-2015 was classified by the Agencia Española del Medicamento y Productos Sanitarios (AEMPS) as a postauthorization prospective follow-up study with the code COH-PAK-2014-01. Data availability The protocol and statistical analysis plan are available on request. Deidentified participant data are not available for legal and ethical reasons. RESULTS VH and SCC in patients with PD vs controls At V0, VH and SCC were more frequent in patients with PD (n=376, 58.2% males, age 61.14±8.73 years) than in the controls (n=116, 48.3% males, age 62.40±8.46 years): 13.6% (51/376) vs 0% (0/116) (p<0.001) and 62.2% (234/376) vs. 50.9% (59/116) (p=0.019), respectively. SCC were significantly more frequent in patients with PD than in controls according to NMSS item 16 (44.4% vs. 21.6%, p<0.001) and item 17 (39.6% vs. 29.3%, p=0.028), but not in item 18 (29.5% vs. 25%, p=0.205). No VH cases were detected in the control group. In the PDNC group, SCC were present in 86.3% (44/51) of patients with VH compared with 58.5% (190/325) of patients without VH (p<0.001) (Fig. 1A). In the PD-NC subgroup with SCC, VH were more frequent than in the PD-NC group without SCC (18.8% [44/234] vs. 4.9% [7/142], p<0.001) (Fig. 1B). At the 2-year follow-up, 66 out of 376 patients with PD-NC (17.6%) presented VH, of which 35 were new cases (10.8% of the patients without VH at V0). Only four cases (3.4%) with VH at V2 were found in the control group. SCC were detected at V2 in 65.7% and 44% of the patients and controls, respectively. The frequencies of SCC at V2 in the PD-NC and control groups that did not report SCC at V0 were 42.3% (60/142) and 15.8% (9/57), respectively. Patients with PD-NC with vs without VH at baseline VH presenting at baseline were associated with a higher LEDD (661.49±390.14 vs. 529.51±395.02, p=0.011), higher scores on the UPDRS-III (24.57±11.45 vs. 20.51±9.78, p=0.022), UPDRS-IV (2.63±2.67 vs. 1.74±2.26, p=0.009), FOGQ (4.41± 4.45 vs. 3.14±4.41, p=0.005), NMSS (73.55±46.58 vs. 36.55± 29.28, p<0.001), NPI (7.95±9.01 vs. 4.70±6.88, p=0.010), VASF–mental (2.82±2.43 vs. 1.87±2.4, p=0.004), and PDQ- 348 J Clin Neurol 2023;19(4):344-357 Visual Hallucinations and Subjective Cognitive Complaints in PD JCN Fig. 1. SCCs and VH at baseline. A: Frequency at baseline (V0) of subjective cognitive complaints (SCC) in patients with Parkinson’s disease with normal cognition (PD-NC, defined as total score on the Parkinson’s Disease Cognitive Rating Scale [PD-CRS] of >80) with visual hallucinations (VH) (n=51) vs. without VH (n=325). B: Frequency at V0 of VH in patients with PD-NC with SCC (n=234) vs without SCC (n=142). 39SI (26.92±15.87 vs. 13.72±10.75, p<0.001), and lower scores on the PDSS (110.76±29.91 vs. 119.67±23.42, p=0.035), ADLS (86.86±9.05 vs. 90.31±8.49, p=0.004), and EUROHIS-QOL8 (3.67±0.55 vs. 3.86±0.52, p=0.026) (Table 1). Specifically, motor fluctuations (49.0% vs. 28.3%, p=0.003), FOGQ (41.2% vs. 28.1%, p=0.044), falls (21.6% vs. 8.0%, p=0.005), severe or very severe nonmotor symptoms burden (66.7% vs. 33.8%, p<0.001), and major depression (21.6% vs. 10.8%, p=0.031) were more frequent in patients with PD-NC with VH than in those without VH (Table 1). No differences were observed between patients with vs without VH in global cognition (PDCRS total score) or FS functions (PD-CRS FS subscore), but the PD-CRS PC subscores were significantly lower in the subgroup with VH (26.16±4.58 vs. 28.97±1.46, p=0.001). There were no differences between patients with and without VH in the drugs that they were taking: levodopa (72.5% vs. 67.4%, p=0.286), dopamine agonist (76.5% vs. 71.1%, p=0.269), MAOB inhibitor (86.3% vs. 74.5%, p=0.053), COMT inhibitor (23.5% vs. 16.6%, p=0.156), amantadine (5.9% vs. 8%, p= 0.426), or anticholinergic drugs (2% vs. 3.1%, p=0.547). Patients with PD-NC with vs without SCC at baseline SCC presenting at baseline were associated with higher scores on the UPDRS-III (22.25±10.57 vs. 19.2±9.05, p=0.008), UPDRS-IV (2.08±2.47 vs. 1.5±2.05, p=0.006), FOGQ (3.77±4.6 vs. 2.54±4.02, p=0.002), NMSS (51.94±37.05 vs. 24.49±20.64, p<0.001), BDI-II (9.00±6.92 vs. 4.85±4.42, p<0.001), NPI (5.93±7.81 vs. 3.68±5.94, p=0.002), QUIP-RS (5.55±9.80 vs. 2.79±6.99, p<0.001), VAS-Pain (2.72±2.86 vs. 2.12±2.79, p= 0.019), VASF–physical (3.19±2.67 vs. 2.00±2.45, p<0.001), VASF–mental (2.44±2.50 vs. 1.27±2.09, p<0.001), and PDQ39SI (18.80±13.36 vs. 10.09±8.18, p<0.001), and lower scores on the PD-CRS (97.44±11.15 vs. 99.92±11.45, p=0.026), PDCRS CP subscale (28.30±2.79 vs. 29.07±1.25, p=0.029), PDSS (113.38±27.12 vs. 126.85±16.54, p<0.001), ADLS (86.76± 9.20 vs. 91.62±7.30, p=0.004), and EUROHIS-QOL8 (3.74± 0.54 vs. 3.98±0.45, p<0.001) (Table 2). Specifically, FOGQ (34.6% vs. 22.0%, p=0.006), falls (12.8% vs. 4.9%, p=0.008), severe or very severe nonmotor symptoms burden (51.7% vs. 16.2%, p<0.001), major depression (17.1% vs 4.2%, p<0.001), pain (63.7% vs 47.9%, p=0.002), and functional dependency (6.8% vs 2.1%, p=0.032) were more frequent in patients with PD-NC with SCC than in those without SCC (Table 2). SB related to VH and SCC No significant differences were detected in SB levels between patients with PD-NC with (n=64) and without (n=68) SCC (data not shown). There were also no differences in SB levels 44/51 (86.3) 44/234 (18.8) 190/325 (58.5) 190/234 (81.2) 135/325 (41.5) 135/142 (95.1) 350 300 250 200 150 100 50 0 250 200 150 100 50 0 Number Number PD-NC patients with VH PD-NC patients without VH PD-NC patients with SCCs PD-NC patients without SCCs SCCs in PD-NC patients with vs. without VH at baseline VH in PD-NC patients with vs. without SCCs at baseline With SCCs Without SCCs With VH Without VH n=376 p<0.001 n=376 p<0.001 7/51 (13.7) 7/142 (4.9) A B www.thejcn.com 349 Santos-García D et al. JCN between patients with PD-NC with VH (n=14) and without VH (n=118) (data not shown). VH and SCC development at the 2-year follow-up related to baseline symptoms In the subgroup of patients with PD-NC without VH at baseTable 1. Disease-related characteristics, motor and nonmotor symptoms, autonomy in performing the activities of daily living, and quality of life in patients with PD-NC (defined as total score on the PD-CRS of >80) with and without VH at baseline (n=376) Entire sample ( n =376) Patients with PD-NC with VH ( n =51) Patients with PD-NC without VH ( n =325) p Age (yr) 61.14±8.73 62.16±8.26 60.98±8.8 0.396 Sex, male (%) 58.2 56.9 58.5 0.473 Disease duration (yr) 5.29±3.9 5.86±3.74 5.19±3.92 0.176 Levodopa equivalent daily dose (mg) 547.61±396.44 661.49±390.14 529.51±395.02 0.011 Motor phenotype (%) 0.129 Tremor dominant 49.5 58.8 48.0 PIGD 34.0 21.6 36.0 Indeterminate 16.5 19.6 16.0 Hoehn and Yahr stage 2 [1.5–2] 2 [1.5–2] 2 [1.5–2] 0.680 From 3 to 5 (%) 7.8 4.7 8.2 0.323 UPDRS-III score 21.09±10.11 24.57±11.45 20.51±9.78 0.022 UPDRS-IV score 1.86±2.33 2.63±2.67 1.74±2.26 0.009 Motor fluctuations (%) 31.1 49.0 28.3 0.003 Dyskinesia (%) 17.4 21.6 16.7 0.253 FOGQ score 3.31±4.34 4.41±4.45 3.14±4.41 0.005 Patients with freezing of gait (%) 29.9 41.2 28.1 0.044 Patients with falls (%) 9.8 21.6 8.0 0.005 PD-CRS total score 98.38±11.33 96.82±10.76 98.62±11.41 0.268 PD-CRS FS subscore 69.79±10.83 70.67±9.07 69.65±11.09 0.319 PD-CRS PC subscore 28.59±2.36 26.16±4.58 28.97±1.46 0.001 SCC (%) 62.2 86.3 58.5 <0.001 NMSS score 41.57±34.51 73.55±46.58 36.55±29.28 <0.001 Severe or very severe nonmotor symptoms burden (NMSS score >40) (%) 38.3 66.7 33.8 <0.001 BDI-II score 7.43±6.41 8.71±7.17 7.23±6.28 0.148 Major depression (%) 12.2 21.6 10.8 0.031 NPI score 5.14±7.28 7.95±9.01 4.7±6.88 0.010 QUIP-RS score 4.3±8.1 4.5±83.47±6.88 0.119 PDSS score 118.46±24.55 110.76±29.91 119.67±23.42 0.035 VAS-Pain score 2.49±2.84 2.64±2.9 2.47±2.84 0.791 Patients with pain (%) 57.7 60.8 57.2 0.375 VASF–physical score 2.74±2.65 3.35±2.57 2.65±2.65 0.072 VASF–mental score 2±2.42 2.82±2.43 1.87±2.4 0.004 ADLS score 89.84±8.64 86.86±9.05 90.31±8.49 0.004 Patients with functional dependency (%) 5.1 9.8 4.3 0.099 PDQ-39SI score 15.51±12.41 26.92±15.87 13.72±10.75 <0.001 EUROHIS-QOL8 score 3.83±0.53 3.67±0.55 3.86±0.52 0.026 Data are percentage, mean±SD, or median [interquartile range] values. Chi-square and Mann-Whitney-Wilcoxon tests were used to compare patients with and without SCC at baseline. Data on Hoehn and Yahr stages and UPDRS scores were obtained during the Off state (first hour in the morning without taking medication in the previous 12 hours). ADLS, Schwab and England Activities of Daily Living Scale; BDI-II, Beck Depression Inventory-II; EUROHIS-QOL8, 8-item EUROHIS-QOL index; FS, frontal-subcortical; NMSS, Non-Motor Symptoms Scale; NPI, Neuropsychiatric Inventory; PC, posterior-cortical; PD-CRS, Parkinson’s Disease Cognitive Rating Scale; PD-NC, Parkinson’s disease with normal cognition; PDQ-39SI, 39-item Parkinson’s Disease Quality of Life Questionnaire Summary Index; PDSS, Parkinson’s Disease Sleep Scale; PIGD, Postural Instability Gait Difficulty; QUIP-RS, Questionnaire for Impulsive Compulsive Disorders in Parkinson’s Disease Rating Scale; SCC, subjective cognitive complaints; UPDRS, Unified Parkinson’s Disease Rating Scale; VASF, Visual Analog Scale–Fatigue; VAS-Pain, Visual Analog Scale–Pain; VH, visual hallucinations. 350 J Clin Neurol 2023;19(4):344-357 Visual Hallucinations and Subjective Cognitive Complaints in PD JCN line (n=325), 13.7% of the group with SCC at baseline (26/190) and 6.7% of the group without SCC at baseline (9/135) had developed VH at V2 (p=0.032). In contrast, in the subgroup of patients without SCC at baseline (n=142), 42.7% of the group with VH at baseline (3/7) and 42.2% of the group without VH at baseline (57/135) had developed SCC at V2 (p=0.632). Like Table 2. Disease-related characteristics, motor and nonmotor symptoms, autonomy in performing the activities of daily living, and quality of life in patients with PD with and without SCC at baseline (n=499) Entire sample ( n =376) Patients with PD-NC with SCC ( n =234) Patients with PD-NC without SCC ( n =142) p Age (yr) 61.14±8.73 62.31±8.40 60.86±9.26 0.791 Sex, male (%) 58.2 60.3 54.9 0.182 Disease duration (yr) 5.29±3.9 5.53±3.99 4.91±3.72 0.144 Levodopa equivalent daily dose (mg) 547.61±396.44 569.35±390.69 512.39±404.5 0.099 Motor phenotype (%) 0.363 Tremor dominant 49.5 47.4 52.8 PIGD 34.0 36.8 29.6 Indeterminate 16.5 15.8 17.6 Hoehn and Yahr stage 2 [1.5–2] 2 [1.5–2] 2 [1.5–2] 0.412 From 3 to 5 (%) 7.8 8.6 6 .5 0.319 UPDRS-III score 21.09±10.11 22.25±10.57 19.2±9.05 0.008 UPDRS-IV score 1.86±2.33 2.08±2.47 1.5±2.05 0.006 Motor fluctuations (%) 31.1 33.3 27.5 0.141 Dyskinesia (%) 17.4 18.1 16.3 0.391 FOGQ score 3.31±4.34 3.77±4.6 2.54±4.02 0.002 Patients with FOG (%) 29.9 34.6 22.0 0.006 Patients with falls (%) 9.8 12.8 4.9 0.008 PD-CRS total score 98.38±11.33 97.44±11.15 99.92±11.45 0.026 PD-CRS FS subscore 69.79±10.83 69.15±10.6 70.85±11.16 0.112 PD-CRS PC subscore 28.59±2.36 28.3±2.79 29.07±1.25 0.029 VH (%) 13.6 18.8 4.9 <0.001 NMSS score 41.57±34.51 51.94±37.05 24.49±20.64 <0.001 Severe or very severe nonmotor symptoms burden (NMSS score >40) (%) 38.3 51.7 16.2 <0.001 BDI-II score 7.43±6.41 9±6.92 4.85±4.42 <0.001 Major depression (%) 12.2 17.1 4.2 <0.001 NPI score 5.14±7.28 5.93±7.81 3.68±5.94 0.002 QUIP-RS score 4.30±8.10 5.55±9.80 2.79±6.99 <0.001 PDSS score 118.46±24.55 113.38±27.12 126.85±16.54 <0.001 VAS-Pain score 2.49±2.84 2.72±2.86 2.12±2.79 0.019 Patients with pain (%) 57.7 63.7 47.9 0.002 VASF–physical score 2.74±2.65 3.19±2.67 2.00±2.45 <0.001 VASF–mental score 2.00±2.42 2.44±2.50 1.27±2.09 <0.001 ADLS score 89.84±8.64 86.76±9.20 91.62±7.30 0.004 Patients with functional dependency (%) 5.1 6.8 2.1 0.032 PDQ-39SI score 15.51±12.41 18.8±13.36 10.09±8.18 <0.001 EUROHIS-QOL8 score 3.83±0.53 3.74±0.54 3.98±0.45 <0.001 Data are percentage, mean±SD or median [interquartile range] values. Chi-square and Mann-Whitney-Wilcoxon tests were applied to compare patients with and without SCC at baseline. Data on Hoehn and Yahr stages and UPDRS-III scores were obtained from during the Off state. ADLS, Schwab and England Activities of Daily Living Scale; BDI-II, Beck Depression Inventory-II; EUROHIS-QOL8, 8-item EUROHIS-QOL index; FS, frontal-subcortical; NMSS, Non-Motor Symptoms Scale; NPI, Neuropsychiatric Inventory; PC, posterior-cortical; PD-CRS, Parkinson’s Disease Cognitive Rating Scale; PD-NC, Parkinson’s disease with normal cognition; PDQ-39SI, 39-item Parkinson’s Disease Quality of Life Questionnaire Summary Index; PDSS, Parkinson’s Disease Sleep Scale; PIGD, Postural Instability Gait Difficulty; QUIP-RS, Questionnaire for Impulsive Compulsive Disorders in Parkinson’s Disease Rating Scale; SCC, subjective cognitive complaints; UPDRS, Unified Parkinson’s Disease Rating Scale; VASF, Visual Analog Scale– Fatigue; VAS-Pain, Visual Analog Scale–Pain; VH, visual hallucinations. www.thejcn.com 351 Santos-García D et al. JCN at baseline, no relationship was observed between drug treatment and VH at V2 in patients with and without VH: levodopa (89.4% vs. 87.4%, p=0.378), dopamine agonist (71.2% vs. 75.8%, p=0.393), MAO-B inhibitor (78.8% vs. 77.1%, p=0.432), COMT inhibitor (33.3% vs. 26.8%, p=0.253), amantadine (12.1% vs. 10%, p=0.668), or anticholinergic drugs (4.5% vs. 2.6%, p=0.328). Change in cognition related to VH and SCC At the 2-year follow-up, the PD-CRS total score had significantly decreased in both patients with PD-NC with VH (n=51, from 96.82±10.76 to 91.08±17.85, Cohen’s d=-0.71, p=0.001) and with PD-NC but without VH (n=325, from 98.62±11.41 to 96.26±14.6, Cohen’s d=-0.31, p<0.001). The FS and PC subscores of PD-CRS also decreased significantly in both groups (Table 3). Comparing both groups, in patients with PD-NC with VH, the score on the PD-CRS PC subscale decreased significantly more than that in patients with PD-NC without VH (Cohen’s d=-0.41 vs. -0.23, p<0.001). A significant trend was observed in the reduction of the PD-CRS total score in the group with VH compared with the group without VH (Cohen’s d=-0.71 vs. -0.31, p=0.061), but no differences were detected in the PD-CRS FS subscore (Table 3). At V2, the frequency of CI was significantly higher in patients with than without VH at V0: 33.3% (17/51) vs. 12.3% (40/325), p< 0.001 (Fig. 2A). Moreover, CI was significantly more frequent at V2 in patients with PD with VH at V2 compared with those without VH at V2 (38.6% vs. 13.8%, p<0.001), and in those patients with persistent VH (i.e., at both V0 and V2) than in those with no VH or with VH at only one of the two visits (24.6% vs. 5.3%, p<0.001). Regarding SCC, global cognitive function (PD-CRS total score, Cohen’s d=-0.43 vs. -0.27, p=0.008) and FS function (PD-CRS FS subscore, Cohen’s d=-0.53 vs. -0.23, p=0.035) were significantly impaired in patients with PD-NC with SCC compared with those without SCC (Table 4). However, patients with PD-NC without SCC exhibited a significantly larger decrease in PD-CRS PC subscore compared with patients with PD-NC and SCC (Cohen’s d=-0.31 vs. -0.23, p=0.003). A significant trend (p=0.066) was observed in the higher frequency of CI at V2 in patients with PD-NC with SCC compared with those without SCC (17.5% [41/234] vs. 11.3% [16/142]) (Fig. 2B). Similar results were observed when SCC at V2 were considered (17.4% [43/247] vs. 10.9% [14/129], p=0.061). When patients with persistent SCC (at V0 and V2) were compared with patients without persistent SCC, CI was more frequent in the former (19.3% [36/187] vs. 11.1% [21/189], p=0.020). VH and SCC as predictors of CI at the 2-year follow-up In patients with PD-NC (n=376), VH presenting at V0 was associated with CI at the 2-year follow-up (odds ratio [OR]= 3.56, 95% confidence interval=1.82–6.96, p<0.001). After adTable 3. Changes in cognition in patients with PD-NC with vs without VH at V0 (baseline) from V0 to V2 (2-year follow-up, plus or minus 1 month) Patients with PD with VH at V0 ( n =51) Patients with PD with VH at V2 ( n =51) Cohen’s d p * Patients with PD without VH V0 ( n =325) Patients with PD without VH V2 ( n =325) Cohen’s d p † p ‡ p § PD-CRS total score 96.82±10.76 91.08±17.85 -0.71 0.001 98.62±11.41 96.26±14.6 -0.31 <0.001 0.100 0.061 PD-CRS FS subscore 70.67±9.07 66.61±14.32 -0.69 0.001 69.65±11.09 67.67±13.8 -0.28 <0.001 0.114 0.927 Immediate verbal memory 8.96±1.80 8.53±1.94 -0.32 0.103 8.45±1.86 8.64±2.75 0.10 0.179 0.153 0.324 Sustained attention 9.33±1.03 8.65±1.36 -0.69 0.001 9.03±1.21 8.66±1.70 -0.29 <0.001 0.310 0.218 Working memory 8.41±1.75 7.57±1.70 -0.62 0.003 7.65±1.90 7.24±1.93 -0.29 <0.001 0.180 0.013 Clock drawing 9.65±0.62 8.86±1.45 -0.79 <0.001 9.33±1.51 9.18±1.32 -0.11 0.154 0.022 N.A. Delayed verbal memory 7.06±2.90 6.69±2.50 -0.25 0.210 5.97±2.63 6.3±2.88 0.17 0.026 0.105 0.016 Alternating verbal fluency 11.41±4.45 11.08±4.06 -0.43 0.032 12.9±3.96 12.1±4.47 -0.27 0.001 0.857 0.148 Action verbal fluency 12.2±3.17 14.24±7.04 0.28 0.160 16.33±5.18 15.56±5.50 -0.22 0.004 0.698 0.054 PD-CRS PC subscore 26.16±4.58 25.47±5.23 -0.41 0.046 28.97±1.46 28.59±2.19 -0.23 0.003 0.412 <0.001 Confrontation naming 16.27±4.57 16.12±4.81 -0.13 0.485 19.19±1.20 18.95±1.96 -0.17 0.030 0.787 <0.001 Clock copying 9.59±1.19 9.35±1.38 -0.60 0.004 9.78±0.81 9.64±0.95 -0.16 0.033 0.041 N.A. Data are mean±SD values. p values were computed using the GLM repeated-measures procedure. *p, change over time (V2 vs. V0) in patients with PD-NC with VH at V0; †p, change over time (V2 vs. V0) in patients with PD-NC without VH at V0; ‡p, group–visit interaction; §p, patients with PD-NC with VH vs patients with PD-NC without VH. Age, disease duration, education, and LEDD at V0 were included as covariates. Patients with PD-NC with VH vs. patients with PD-NC without VH were not applicable if the interaction test was significant (indicating that the rates of change over time differ between the two groups). GLM; general linear model; LEDD, levodopa equivalent daily dose; N.A., not applicable; PC, posterior-cortical; PD, Parkinson’s disease; PD-CRS, Parkinson’s Disease Cognitive Rating Scale; PD-NC, Parkinson’s disease with normal cognition; PIGD, Postural Instability Gait Difficulty; VH, visual hallucinations. 352 J Clin Neurol 2023;19(4):344-357 Visual Hallucinations and Subjective Cognitive Complaints in PD JCN 17/57 (29.8) 41/234 (17.5) 34/319 (10.7) 16/142 (11.3) 40/57 (70.2) 193/234 (82.5) 285/319 (89.3) 126/142 (98.7) 350 300 250 200 150 100 50 0 250 200 150 100 50 0 Number Number PD-NC patients with VH PD-NC patients without VH PD-NC patients with SCCs PD-NC patients without SCCs CI at V2 in PD-NC patients with vs. without VH at V0 CI at V2 in PD-NC patients with vs. without SCCs at V0 With CI Without CI With CI Without CI n=376 p<0.001 n=376 p<0.001 A B Fig. 2. CI at V2 regarding VH and SCCs. A: Frequency of cognitive impairment (CI, defined as Parkinson’s Disease Cognitive Rating Scale [PD-CRS] total score of <81) in patients with Parkinson’s disease with normal cognition (PD-NC) with vs without visual hallucinations (VH) at the 2-year follow-up (V2). B: Frequency of CI in patients with PD-NC with vs without subjective cognitive complaints (SCC) at V2. Table 4. Changes in cognition in patients with PD-NC with vs without SCC at V0 from V0 to V2 Patients with PD with SCC V0 ( n =234) Patients with PD with SCC V2 ( n =234) Cohen’s d p * Patients with PD without SCC V0 ( n =142) Patients with PD without SCC V2 ( n =142) Cohen’s d p † p ‡ p § PD-CRS total score 97.44±11.15 94.18±15.39 -0.43 <0.001 99.92±11.45 97.83±14.52 -0.27 0.023 0.434 0.008 PD-CRS FS subscore 69.15±10.6 66.31±13.81 -0.53 <0.001 70.85±11.16 69.16±13.84 -0.23 0.050 0.391 0.035 Immediate verbal memory 8.44±1.91 8.41±2.23 -0.02 0.785 8.64±1.76 8.99±3.21 0.16 0.163 0.106 0.022 Sustained attention 9.14±1.17 8.57±1.72 -0.46 <0.001 8.96±1.22 8.81±1.55 -0.12 0.287 0.026 N.A. Working memory 7.73±1.83 7.18±1.78 -0.42 <0.001 7.78±2.02 7.45±2.09 -0.22 0.064 0.305 0.645 Clock drawing 9.38±1.48 9.51±1.19 0.25 0.007 9.36±1.33 9.32±1.08 -0.03 0.754 0.111 0.432 Delayed verbal memory 6.06±2.80 6.22±2.82 0.09 0.310 6.21±2.50 6.56±2.86 0.17 0.149 0.729 0.233 Alternating verbal fluency 12.49±3.70 11.74±4.53 -0.26 0.004 13.32±4.08 12.31±4.26 -0.33 0.006 0.362 0.078 Action verbal fluency 15.90±5.57 15.16±5.81 -0.22 0.016 16.58±4.92 15.73±5.62 -0.24 0.045 0.718 0.177 PD-CRS PC subscore 28.30±2.79 27.86±3.44 -0.23 0.010 29.07±1.25 28.67±1.94 -0.31 0.010 0.976 0.003 Confrontation naming 18.50±2.69 18.35±3.12 -0.09 0.284 19.29±1.03 18.92±1.80 -0.30 0.010 0.199 0.007 Clock copying 9.80±0.88 9.51±1.19 -0.29 0.002 9.78±0.52 9.75±0.64 -0.06 0.571 0.041 N.A. Data are mean±SD values. p values were computed using a GLM repeated-measures procedure. *p, change over time (V2 vs V0) in patients with PD-NC with SCC at V0; †p, change over time (V2 vs V0) in patients with PD-NC without SCC at V0; ‡p, group–visit interaction; §p, patients with PD-NC with SCC vs patients with PD-NC without SCC. Age, disease duration, education, and LEDD at V0 were included as covariates. Patients with PD-NC with SCC vs patients with PDNC without SCC were not applicable if the interaction test was significant. FS, frontal-subcortical; GLM; general linear model; LEDD, levodopa equivalent daily dose; N.A., not applicable; PD, Parkinson’s disease; PD-CRS, Parkinson’s Disease Cognitive Rating Scale; PD-NC, Parkinson’s disease with normal cognition.