https://doi.org/10.1093/braincomms/fcad074 BRAIN COMMUNICATIONS 2023: Page 1 of 16 | 1 BRAIN COMMUNICATIONS Association of biological sex with clinical outcomes and biomarkers of Alzheimer’s disease in adults with Down syndrome M. Florencia Iulita,1,2,3,* Alexandre Bejanin,1,2,* Eduard Vilaplana,1,2 Maria Carmona-Iragui,1,2,4 Bessy Benejam,2,4 Laura Videla,1,2,4 Isabel Barroeta,1,2 Susana Fernández,4 Miren Altuna,1,2 Jordi Pegueroles,1,2 Victor Montal,1,2 Silvia Valldeneu,1,2 Sandra Giménez,5 Sofía González-Ortiz,6 Soraya Torres,1,2 Shaimaa El Bounasri El Bennadi,1,2 Concepcion Padilla,1,2 Mateus Rozalem Aranha,1,2 Teresa Estellés,1,2 Ignacio Illán-Gala,1,2 Olivia Belbin,1,2 Natalia Valle-Tamayo,1,2 Valle Camacho,7 Esther Blessing,8 Ricardo S. Osorio,8 Sebastian Videla,9 Sylvain Lehmann,10 Anthony J. Holland,11,12 Henrik Zetterberg,13,14,15,16,17 Kaj Blennow,13,14 Daniel Alcolea,1,2 Jordi Clarimón,1,2 Shahid H. Zaman,11,12 Rafael Blesa,1,2 Alberto Lleó1,2 and Juan Fortea1,2,4 * These authors contributed equally to this work. The study of sex differences in Alzheimer’s disease is increasingly recognized as a key priority in research and clinical development. People with Down syndrome represent the largest population with a genetic link to Alzheimer’s disease (>90% in the 7th decade). Yet, sex differences in Alzheimer’s disease manifestations have not been fully investigated in these individuals, who are key candidates for preventive clinical trials. In this double-centre, cross-sectional study of 628 adults with Down syndrome [46% female, 44.4 (34.6; 50.7) years], we compared Alzheimer’s disease prevalence, as well as cognitive outcomes and AT(N) biomarkers across age and sex. Participants were recruited from a population-based health plan in Barcelona, Spain, and from a convenience sample recruited via services for people with intellectual disabilities in England and Scotland. They underwent assessment with the Cambridge Cognitive Examination for Older Adults with Down Syndrome, modified cued recall test and determinations of brain amyloidosis (CSF amyloid-β 42 / 40 and amyloid-PET), tau pathology (CSF and plasma phosphorylated-tau181) and neurodegeneration biomar­ kers (CSF and plasma neurofilament light, total-tau, fluorodeoxyglucose-PET and MRI). We used within-group locally estimated scat­ terplot smoothing models to compare the trajectory of biomarker changes with age in females versus males, as well as by apolipoprotein ɛ4 carriership. Our work revealed similar prevalence, age at diagnosis and Cambridge Cognitive Examination for Older Adults with Down Syndrome scores by sex, but males showed lower modified cued recall test scores from age 45 compared with females. AT(N) biomarkers were comparable in males and females. When considering apolipoprotein ɛ4, female ɛ4 carriers showed a 3-year earlier age at diagnosis compared with female non-carriers (50.5 versus 53.2 years, P = 0.01). This difference was not seen in males (52.2 versus 52.5 years, P = 0.76). Our exploratory analyses considering sex, apolipoprotein ɛ4 and biomarkers showed that female ɛ4 carriers tended to exhibit lower CSF amyloid-β 42/amyloid-β 40 ratios and lower hippocampal volume com­ pared with females without this allele, in line with the clinical difference. This work showed that biological sex did not influence clin­ ical and biomarker profiles of Alzheimer’s disease in adults with Down syndrome. Consideration of apolipoprotein ɛ4 haplotype, particularly in females, may be important for clinical research and clinical trials that consider this population. Accounting for, report­ ing and publishing sex-stratified data, even when no sex differences are found, is central to helping advance precision medicine. Received July 12, 2022. Revised December 07, 2022. Accepted March 16, 2023. Advance access publication March 17, 2023 © The Author(s) 2023. Published by Oxford University Press on behalf of the Guarantors of Brain. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. 2 | BRAIN COMMUNICATIONS 2023: Page 2 of 16 M. F. Iulita et al. 1 Sant Pau Memory Unit, Department of Neurology, Hospital de la Santa Creu i Sant Pau, Biomedical Research Institute Sant Pau, Universitat Autònoma de Barcelona, Barcelona 08025, Spain 2 Center of Biomedical Investigation Network for Neurodegenerative Diseases (CIBERNED), Madrid 28031, Spain 3 Women’s Brain Project, Guntershausen 8357, Switzerland 4 Barcelona Down Medical Center, Fundació Catalana Síndrome de Down, Barcelona 08029, Spain 5 Multidisciplinary Sleep Unit, Hospital de la Santa Creu i Sant Pau, Barcelona 08041, Spain 6 Hospital del Mar, Barcelona 08003, Spain 7 Nuclear Medicine Department, Hospital de la Santa Creu i Sant Pau, Barcelona 08041, Spain 8 Department of Psychiatry, NYU Grossman School of Medicine, New York, NY 10016, USA 9 Clinical Research Support Unit, Bellvitge Biomedical Research Institute (IDIBELL), Department of Clinical Pharmacology, University of Barcelona, Barcelona 08908, Spain 10 Institute for Neurosciences of Montpellier, Institute for Regenerative Medicine and Biotherapy, Université de Montpellier, CHU de Montpellier, INSERM, Montpellier 34295, France 11 Department of Psychiatry, Cambridge Intellectual and Developmental Disabilities Research Group, University of Cambridge, Douglas House, Cambridge CB2 8AH, United Kingdom 12 Cambridgeshire & Peterborough NHS Foundation Trust, Fulbourn Hospital, Cambridge CB21 5EF, United Kingdom 13 Department of Psychiatry and Neurochemistry, University of Gothenburg, Möndal 40530, Sweden 14 Clinical Neurochemistry Laboratory, Sahlgrenska University Hospital, Mölndal 40530, Sweden 15 UK Dementia Research Institute, University College London, London WC1E 6BT, United Kingdom 16 Department of Neurodegenerative Disease, University College London Institute of Neurology, London WC1E 6BT, United Kingdom 17 Hong Kong Center for Neurodegenerative Diseases, Clear Water Bay, Hong Kong 1512-1518, China Correspondence to: M. Florencia Iulita Memory Unit, Neurology Department Hospital de la Santa Creu i Sant Pau Biomedical Research Institute Sant Pau Sant Antoni Maria Claret, 167, Barcelona 08025, Spain E-mail: florencia.iulita@gmail.com Correspondence may also be addressed to: Juan Fortea Memory Unit, Neurology Department Hospital de la Santa Creu i Sant Pau Biomedical Research Institute Sant Pau Sant Antoni Maria Claret, 167, Barcelona 08025, Spain E-mail: jfortea@santpau.cat Keywords: Down syndrome; Alzheimer’s disease; sex; gender; precision medicine Abbreviations: APOE = apolipoprotein E; APP = amyloid precursor protein; Aβ = amyloid-β; CAMCOG-DS = Cambridge Cognitive Examination for Older Adults with Down Syndrome; CAMDEX-DS = Cambridge Examination for Mental Disorders of Older People with Down’s Syndrome and Others with Intellectual Disabilities; FDG-PET = 18F-fludeoxyglucose positron emission tomography; ID = intellectual disability; LOESS = locally estimated scatterplot smoothing; mCRT = modified cued recall test; NfL = neurofilament light chain; p-tau181 = phosphorylated tau at threonine 181; PSEN1 = presenilin 1; SUVR = standardized uptake value ratio; t-tau = total tau; TIV = total intracranial volume Clinical and biomarker changes by sex in DS-AD BRAIN COMMUNICATIONS 2023: Page 3 of 16 | 3 Graphical Abstract in females with mild cognitive impairment compared with Introduction males.10,11 Biomarkers including, CSF amyloid-β (Aβ)42, The topic of sex differences is now positioned as a top prior­ Aβ42/Aβ40 and amyloid-PET do not appear to differ by ity in neurology and translational neuroscience research, sex in cross-sectional studies across the Alzheimer’s disease particularly in the context of precision medicine and perso­ spectrum, as reviewed by Mielke.12 Similarly, no sex differ­ nalized care strategies.1,2 In Alzheimer’s disease, females ac­ ences in plasma and CSF phosphorylated tau at threonine count for around two-thirds of patients and caregivers 181 (p-tau181) concentrations were found,12 although worldwide.3 While this is partly explained by differences in tau-PET showed a greater extent of tau pathology in females longevity, increasing evidence shows that biological sex in Alzheimer’s disease-relevant regions compared with can influence key aspects of the disease, including molecular males, even when controlling for disease severity.13-15 This pathways, cognitive progression and risk factor profiles.4,5 is in line with autopsy studies, which reported greater neuro­ Clinically, females outperform males in verbal memory fibrillary tangle density in females.16,17 For neurofilament tests,6-8 but this ‘female advantage’ is lost at the dementia light chain (NfL), a marker of axonal damage, some studies stage,9 possibly due to faster rates of cognitive deterioration indicate that males exhibit higher concentrations than 4 | BRAIN COMMUNICATIONS 2023: Page 4 of 16 M. F. Iulita et al. females in CSF, but not in plasma.18 Evidence of sex differ­ (n = 584) and the other from Cambridge, UK (n = 44). ences in MRI biomarkers is mixed at the cross-sectional le­ Inclusion criteria were defined as individuals with Down syn­ vel; for a review, see Ferretti et al.4 drome of both sexes, aged 18 + years, in good general condi­ While some biomarkers might not differ by sex, certain risk tion, who understood and accepted the procedures of the factors can impact males and females differently. For example, study. Exclusion criteria were inability to provide informed the ɛ4 allele of the apolipoprotein E gene (APOE) confers a great­ consent, any significant unstable medical or psychiatric dis­ er risk of Alzheimer’s disease in females compared with males at ease affecting cognition, anticoagulant treatment or other ages 65–75 years.19 In line with this, several studies found signifi­ blood dyscrasias that contraindicated the lumbar puncture cant interactions between sex and APOE ɛ4 in CSF and tau-PET and contraindications for MRI (claustrophobia, pacemaker, biomarkers, brain hypometabolism [18F-fludeoxyglucose posi­ aneurysm clip and etc.) and/or pregnancy. tron emission tomography (FDG-PET)] and atrophy, with fe­ The study was approved by the Ethics Committee at males showing significantly greater pathological biomarker Hospital Sant Pau (Spain), as well as by the University of values at similar clinical stages.14,20-23 Cambridge Research Ethics Committees and the People with Down syndrome represent the largest popula­ Administration of Radioactive Substances Advisory tion with a genetic link to Alzheimer’s disease (>90% in the Committee (UK), following the standards of the Declaration 7th decade).24-26 Due to the trisomy of chromosome 21, of Helsinki. In Spain, all participants or their legal representa­ which harbours the amyloid precursor protein (APP) gene tives were required to give written informed consent. In the encoding for the amyloid precursor protein, people with UK, written consent was obtained from all adults with Down syndrome invariably develop the neuropathological Down syndrome who had the capacity to consent. For parti­ changes associated with Alzheimer’s disease by middle cipants who did not have the capacity to consent, the proce­ age.27,28 Down syndrome, together with autosomal domin­ dures in the Mental Capacity Act of 2005 were followed. ant Alzheimer’s disease, constitutes the main evidence for Participants were recruited between June 2009 and August the amyloid cascade hypothesis in which genetic mutations 2021. or gene dose loading results in increased amyloid produc­ In Barcelona, study participants were part of a prospective tion, leading to tau hyperphosphorylation, neurodegenera­ longitudinal cohort to screen for Alzheimer’s disease in tion and dementia.29 Indeed, the temporality of Down syndrome in Catalonia.25 The Cambridge cohort con­ Alzheimer’s disease biomarkers in Down syndrome is similar sisted of a convenience sample recruited via services for peo­ to the sporadic forms and nearly identical to autosomal dom­ ple with intellectual disabilities in England and Scotland.38 inant Alzheimer’s disease.25,30,31 Most but not all study participants had at least one biochem­ Despite the importance of studying Alzheimer’s disease in ical or neuroimaging biomarker assessment. A convenience Down syndrome, little is known about the effect of sex on sample of non-trisomic volunteers with no cognitive impair­ Alzheimer’s disease manifestations in this population. ment (n = 173) was selected from the Sant Pau Initiative on There are discrepant results from studies of Alzheimer’s dis­ Neurodegeneration cohort.39 This group was included only ease risk by sex in Down syndrome,32-36 and there are still no as a visual reference of the biomarker trajectories in euploid studies about sex differences across a range of biomarker individuals. The reference sample had a mean age of 55.6 modalities. Understanding the effect of sex in Down syn­ (9.9) years, and 61.8% (n = 107) were female. Differences drome is relevant for clinical research, particularly for up­ in biomarkers between people with Down syndrome and eu­ coming clinical trials on Alzheimer’s disease that consider ploid controls were described elsewhere.25 this population.37 We examined the association of biological sex with clinical Neurological and neuropsychological outcomes and biomarkers of Alzheimer’s disease in a dual- centre cohort of adults with Down syndrome. We used assessment biochemical and neuroimaging measures to assess amyloid Intellectual disability (ID) was stratified as mild, moderate, (A), tau (T) and neurodegeneration (N) biomarkers across severe or profound based on the DSM 5th Edition and using age and sex. We explored whether APOE ɛ4 influenced the the individuals’ best level of functioning, determined from association between sex and clinical and biomarker profiles. caregiver reports. Each participant further received a diag­ nostic evaluation of dementia in a consensus meeting be­ tween the neurologists and neuropsychologists who Materials and methods assessed them independently, masked to biomarker data. A diagnosis of asymptomatic was given when there was no This study followed the STROBE reporting guidelines for clinical or neuropsychological suspicion of Alzheimer’s dis­ cross-sectional studies. ease (i.e. absence of cognitive impairment beyond the intel­ lectual and developmental disabilities or functional decline compared to the previous functioning). A diagnosis of pro­ Study population dromal Alzheimer’s disease was given when there was suspi­ Participants were members of two clinical cohorts of adults cion of Alzheimer’s disease, but symptoms did not fulfill the with Down syndrome (n = 628), one from Barcelona, Spain criteria for dementia (i.e. cognitive impairment without Clinical and biomarker changes by sex in DS-AD BRAIN COMMUNICATIONS 2023: Page 5 of 16 | 5 functional changes). A diagnosis of Alzheimer’s disease de­ preprocess the structural 3DT1 sequence of the MRI and ex­ mentia required evidence of cognitive impairment beyond tract the hippocampal and total intracranial volumes (TIVs). the intellectual and developmental disabilities that interfered TIV was used to normalize differences in head size. with everyday activities (i.e. presence of a functional decline FDG-PET images were intensity-scaled by the pons-vermis compared to previous functioning). Functional status to dif­ region and spatially normalized using SPM12.39 ferentiate prodromal and dementia stages were assessed Standardized uptake value ratios (SUVRs) were extracted based on anamnesis, the Dementia Questionnaire for from Landau’s regions of interest.47 The amyloid-PET data Persons with Mental retardation and the Cambridge in Barcelona was collected using 18F-florbetapir (2013– 18 Examination for Mental Disorders of Older People with 2017) and F-flutemetamol (2018–2021), and 11 Down’s Syndrome and Others with Intellectual Disabilities C-Pittsburgh compound B was used in Cambridge. Images (CAMDEX-DS). Dementia and prodromal Alzheimer’s dis­ were spatially normalized using MRI transformations com­ ease were considered symptomatic Alzheimer’s disease. puted with the Advanced Normalization Tools 48 and scaled Our diagnostic procedures follow the recommendations of using the cerebellum as the reference region.49 The mean cor­ the Working Group for the Establishment of the Criteria tical SUVRs were then transformed into Centiloid units.50 for the Diagnosis of Dementia in Individuals with Developmental Disability.40 Global cognition was assessed with the Cambridge Statistical analysis Cognitive Examination for Older Adults with Down The R software v.4.04 was used. A chi-squared test (or Syndrome (CAMCOG-DS), which evaluates orientation, Fisher’s exact test when appropriate) was used to test differ­ language, memory, attention, praxis, abstract thinking and ences between categorical data, and Mann–Whitney was perception.41 Episodic memory (immediate and delayed re­ used for continuous variables. Survival analysis with a log- call) was evaluated with the modified cued recall test rank test was conducted to assess sex differences in the age (mCRT), adapted for people with intellectual disabilities.42 at the first diagnosis of symptomatic Alzheimer’s disease. To account for the effect of ID on cognitive performance, Statistical significance was set at P < 0.05. Extreme values we excluded severe or profound cases to prevent floor effects (>5 IQR below Q1 or above Q3) were excluded from the and computed Z-scores in the mild and moderate ID groups analysis. The final sample size for each biomarker and clinic­ separately. al outcome is presented in Table 1. To compare changes in cognition and biomarkers across Apolipoprotein E genotyping age and sex, we fitted a first-degree locally estimated scatter­ plot smoothing (LOESS) curve in each group independent­ A total of 551 participants were screened for APOE geno­ ly.51 We used a first-order LOESS model with a tricubic type. DNA was extracted from peripheral blood, and geno­ weight function and a span parameter of 0.75. This ap­ typing was determined by Sanger sequencing of the proach was used to account for the non-linear trajectory of polymerase chain reaction (PCR) amplification of coding these measurements. The exact age at which the groups di­ exon 4, as previously described.43 Individuals were dichoto­ verge depends on the nature of the variable, the sensitivity mized as APOE ɛ4 carriers or non-carriers, based on the of the assay, the slope of the association and the sample sizes presence of at least one ɛ4 allele. for the different measurements. We defined a clinical or bio­ marker difference between groups as significant at the age at Plasma and CSF biomarker analyses which the curves diverged visually, and the 95% confidence Blood and CSF were obtained in a subset of participants with intervals did not overlap. We also conducted independent Down syndrome and processed as previously described.43 between-group comparisons across decades using Mann– CSF Aβ40 (n = 235), Aβ42 (n = 235), p-tau181 (n = 235) Whitney tests since most variables of interest did not follow and total tau protein (t-tau, n = 235) were quantified with a normal distribution, as assessed using the Shapiro–Wilk the Lumipulse automated platform (Fujirebio, Europe) fol­ test (Supplementary Material). lowing previously reported methods established in our la­ boratory.44,45 CSF NfL (n = 154) was measured with Data availability ELISA (UmanDx, Sweden), as previously described.45 Plasma p-tau181 (n = 514) and NfL (n = 499) were measured The authors may share de-identified data that underlie the re­ with Simoa® (Quanterix, USA), using validated assays.25,46 sults reported in this article. Data will be available upon re­ ceipt of a request detailing the study hypothesis and statistical analysis plan. All requests should be sent to the Brain imaging corresponding authors. The steering committee of this study A subset of participants underwent a 3T-MRI (n = 243), will discuss all requests and decide based on the novelty and FDG-PET (n = 147) and/or amyloid-PET (n = 119). We scientific rigor of the proposal whether data sharing is appro­ used the Computational Anatomy Toolbox (CAT12, http:// priate. All applicants will be asked to sign a data access dbm.neuro.uni-jena.de/cat) for the SPM12 software to agreement. 6 | BRAIN COMMUNICATIONS 2023: Page 6 of 16 M. F. Iulita et al. Table 1 Characteristics of the study population Females (n = 287) Males (n = 341) P-value Age (years) 44.5 [33.6; 50.8] 44.0 [35.1; 50.7] 0.91 APOE ɛ4 haplotype 0.63 ɛ4 carriers 54 (21.4%) 58 (19.4%) ɛ4 non-carriers 198 (78.6%) 241 (80.6%) Level of intellectual disability 0.09 Mild 72 (25.7%) 61 (18.4%) Moderate 144 (51.4%) 184 (55.6%) Severe or profound 64 (22.9%) 86 (26.0%) Diagnostic group 0.78 Asymptomatic 177 (66.0%) 218 (67.5%) Symptomatic Alzheimer’s disease 91 (34.0%) 105 (32.5%) Medical conditions Hypothyroidism (n = 406) 108 (58.7%) 87 (39.2%) <0.001 Epilepsy (n = 330) 17 (11.0%) 15 (8.6%) 0.58 Sleep apnoea (n = 403) 15 (8.2%) 34 (15.4%) 0.04 Depression (n = 433) 31 (16.0%) 36 (15.1%) 0.90 Cognition CAMCOG-DS score (n = 441) 72.0 [56.0; 84.0] 70.5 [57.0; 83.0] 0.83 mCRT immediate recall (n = 374) 35.0 [28.0; 36.0] 35.0 [31.0; 36.0] 0.69 mCRT delayed recall (n = 369) 11.5 [9.0; 12.0] 11.0 [10.0; 12.0] 0.76 Fluid biomarkers CSF Aβ42/Aβ40 (n = 235) 5.3 [4.2; 7.4] e−2 5.9 [4.3; 7.9] e−2 0.54 CSF p-tau181 (n = 232) 67.0 [29.7; 153] 55.6 [29.3; 120] 0.42 CSF total tau (n = 233) 519 [280; 886] 452 [256; 808] 0.53 CSF NfL (n = 150) 586 [356; 876] 560 [353; 976] 0.59 Plasma p-tau 181 (n = 514) 15.1 [9.4; 24.1] 12.9 [9.0; 22.2] 0.09 Plasma NfL (n = 499) 12.8 [7.3; 21.8] 11.1 [6.8; 19.1] 0.20 Imaging biomarkers Centiloid amyloid PET (n = 119) 13.9 [1.6; 59.9] 12.0 [1.7; 48.2] 0.66 FDG-PET SUVR (n = 147) 1.2 [0.9; 1.4] 1.3 [1.1; 1.4] 0.40 Hippocampal volume (n = 243) 5.5 [4.9; 6.0] 6.2 [5.5; 6.7] <0.001 Adj hippocampal volume (n = 243) 4.9 [4.2; 5.3] e−3 4.9 [4.3; 5.3] e−3 0.94 Unless otherwise indicated, values represent n (%) or median [Quartile 1 and Quartile 3]. All fluid biomarker concentration units, except for the CSF Aβ42/40 ratio, are pg/mL. P-values refer to analyses of chi-squared tests for categorical variables and Mann–Whitney tests for continuous variables. Aβ40, amyloid β peptide 40; Aβ42, amyloid β peptide 42; CAMCOG-DS, Cambridge Cognitive Examination for Older Adults with Down Syndrome; CSF, cerebrospinal fluid; FDG, ¹18F-fluorodeoxyglucose; mCRT, modified cued recall test; NfL, neurofilament light chain; PET, positron emission tomography; p-tau181, phosphorylated tau at threonine 181; SUVR, standardized uptake value ratio; TIV, total intracranial volume. males and females, based on assessments of each partici­ Results pant’s first clinical visit. In both groups, the prevalence of Alzheimer’s disease increased with age, and there were Participants no significant sex differences in prevalence across age A total of 628 adults with Down syndrome, including 287 brackets (Fig. 1A and Supplementary Table 2). A survival (46%) females, were included (Table 1). There were no sig­ analysis assessing the probability of developing symptom­ nificant differences in age, proportion of APOE ɛ4 carriers, atic Alzheimer’s disease at different ages also revealed level of ID or prevalence of medical comorbidities by sex, ex­ a similar risk between males and females with Down cept for hypothyroidism which was more common in females syndrome (Fig. 1B). Finally, the age at diagnosis of (58.7% versus 39.2%, P < 0.001). The different subsamples Alzheimer’s disease did not differ by sex (52.5 ± 5.5 in fe­ for each biomarker did not present differences in demograph­ males versus 52.7 ± 5.2 in males, P = 0.73, Fig. 1C and ic characteristics. Supplementary Table 2). When comparing recruiting sites, we found no demo­ We next compared CAMCOG-DS and mCRT scores graphic differences between the two cohorts. The core differ­ between males and females by age, as age can be used as a ence concerned ID (Supplementary Table 1), which is in line proxy of disease progression in this population.29 with the different sources of participant recruitment between CAMCOG-DS scores declined with age in both males and the two sites (population-based versus convenience sample). females with Down syndrome, with no apparent sex differ­ ences in the whole group (Fig. 1D and Supplementary Fig. Sex differences in Alzheimer’s 1) or when considering ID (Supplementary Fig. 2). For episodic memory, males presented lower mCRT scores disease clinical presentation compared with female from age 45 onwards, both for de­ We performed an age-stratified analysis to compare the layed and immediate recall (Fig. 1E–F and Supplementary prevalence of symptomatic Alzheimer’s disease between Fig. 1). Clinical and biomarker changes by sex in DS-AD BRAIN COMMUNICATIONS 2023: Page 7 of 16 | 7 Figure 1 Association of biological sex with Alzheimer’s disease prevalence and cognitive decline in adults with Down syndrome. (A) Point prevalence of symptomatic Alzheimer’s disease by 5-year age intervals. Group differences were assessed with Pearson’s chi-squared analysis (for details, see Supplementary Table 2). (B) Survival analysis showing the probability of developing symptomatic Alzheimer’s disease by age and sex. The P-value refers to log-rank test analysis. (C) Scatterplots illustrating the distribution of age at diagnosis of symptomatic Alzheimer’s disease. The graph shows the individual data points, the means and the nonparametric bootstrapped 95% confidence intervals. The P-value refers to a t-test (for details, see Supplementary Table 2). (D–F) Neuropsychological performance by age and sex at the CAMCOG-DS (D), immediate (E), and delayed (F) recall at the mCRT, with bands representing the 95% confidence intervals. A significant difference between LOESS curves was defined as the age at which the curves diverged visually and the 95% confidence intervals did not overlap (P < 0.05). CAMCOG-DS, Cambridge Cognitive Examination for Older Adults with Down Syndrome; mCRT, modified cued recall test; F, female; M, male. Sex differences in Alzheimer’s showed a trend towards higher p-tau181 concentrations in CSF and plasma (Fig. 2C–D and Supplementary Fig. 3 and disease biomarkers Supplementary Table 3 for the analysis by decades). No The CSF Aβ42/Aβ40 ratio (Fig. 2A, Supplementary Fig. 3 and other trends or differences were seen at other age intervals. Supplementary Table 3) and Aβ42 (Supplementary Fig. 4 and We next compared biomarkers of neurodegeneration. Supplementary Table 3) declined with age with no significant Males and females showed overall similar increases in CSF differences between males and females with Down syndrome. and plasma NfL with age (Fig. 2E–F, Supplementary Fig. 3 CSF concentrations of Aβ40 did not change significantly with and Supplementary Table 3). The CSF and plasma NfL age and were comparable between males and females LOESS curves only tended to diverge between sexes at the (Supplementary Fig. 4 and Supplementary Table 3). We older ages (55–65 years). There were no sex differences in also found comparable increases in cortical amyloid burden the trajectories of CSF t-tau in CSF (Supplementary Fig. 4 (as assessed by amyloid-PET) with age by sex (Fig. 2B, and Supplementary Table 3). For FDG-PET, the glucose me­ Supplementary Fig. 3 and Supplementary Table 3). tabolism in typical Alzheimer’s disease brain regions de­ There were no significant differences in plasma or CSF creased similarly with age in both males and females with p-tau181 concentrations between males and females. Down syndrome (Fig. 2G, Supplementary Fig. 3 and However, between the ages of 40 and 50 years, female Supplementary Table 3). Likewise, hippocampal volumes, 8 | BRAIN COMMUNICATIONS 2023: Page 8 of 16 M. F. Iulita et al. Figure 2 Association of biological sex with Alzheimer’s disease biomarkers in adults with Down syndrome. Shading represents 95% confidence intervals. The dotted lines represent the age-related changes in euploid individuals (healthy controls) for visual reference. A significant difference between LOESS curves was defined as the age at which the curves diverged visually, and the 95% confidence intervals did not overlap (P < 0.05). The scatterplots display, for males and females separately, the relationship between age and CSF Aβ42/40 (A), amyloid-PET (B), CSF and plasma p-tau181 (C, D), CSF and plasma NfL (E, F), FDG-PET (G) and hippocampal volume (H). Aβ40, amyloid β peptide 40; Aβ42, amyloid β peptide 42; CSF, cerebrospinal fluid; FDG, ¹18F-fluorodeoxyglucose; NfL, neurofilament light chain; SUVR, standardized uptake value ratio; TIV, total intracranial volume; F, female; M, male. Clinical and biomarker changes by sex in DS-AD BRAIN COMMUNICATIONS 2023: Page 9 of 16 | 9 Figure 3 Association of APOE ɛ4 haplotype and biological sex with Alzheimer’s disease prevalence and cognitive decline. (A) Point prevalence of symptomatic Alzheimer’s disease by sex and APOE ɛ4 haplotype across 5-year age intervals. Group differences were assessed with Pearson’s chi-squared analysis (for details, see Supplementary Table 4). (B) Survival analysis showing the probability to develop symptomatic Alzheimer’s disease by sex and APOE haplotype across age. The P-value refers to log-rank test analysis. (C) Scatterplots illustrating the distribution of age at diagnosis of Alzheimer’s disease and showing the individual data points, the means and the nonparametric bootstrapped 95% confidence intervals. The P-values refer to t-tests (for details, see Supplementary Table 4). (D–I) Age-related changes in neuropsychological performance at the CAMCOG-DS (D, G), immediate (E, H) and delayed recall (F, I) at the mCRT, with bands representing the 95% confidence intervals in male and female separately. A significant difference between LOESS curves was defined as the age at which the curves diverged visually and the 95% confidence intervals did not overlap (P < 0.05). CAMCOG-DS, Cambridge Cognitive Examination for Older Adults with Down Syndrome; mCRT, modified cued recall test; F, female; M, male; ɛ4, APOE ɛ4 carriers. 10 | BRAIN COMMUNICATIONS 2023: Page 10 of 16 M. F. Iulita et al. Figure 4 Association of APOE ɛ4 haplotype and sex with Alzheimer’s disease biomarkers in adults with Down syndrome. Shading represents 95% confidence intervals. The dotted lines represent the age-related changes in euploid individuals for visual reference. The dashed line without confidence intervals was used in groups with a sample size inferior to 20 participants. A significant difference between LOESS curves was defined as the age at which the curves diverged visually and the 95% confidence intervals did not overlap (P < 0.05). The scatterplots display, for males and females separately, the relationship between age and APOE ϵ4 carriership with CSF Aβ42/40 (A), amyloid-PET (B), CSF and plasma p- tau181 (C, D), CSF and plasma NfL (E, F), FDG-PET (G) and hippocampal volume (H). Aβ40, amyloid β peptide 40; Aβ42, amyloid β peptide 42; APOE, apolipoprotein E; CSF, cerebrospinal fluid; FDG, ¹18F-fluorodeoxyglucose; NfL, neurofilament light chain; SUVR, standardized uptake value ratio; TIV, total intracranial volume; F, female; M, male; ɛ4, APOE ɛ4 carriers. Clinical and biomarker changes by sex in DS-AD BRAIN COMMUNICATIONS 2023: Page 11 of 16 | 11 when adjusted by total intracranial volume, decreased simi­ course and pathology burden.4,52 Whether these differences larly with age in males and females (Fig. 2H, Supplementary are consistent across Alzheimer’s disease populations, in­ Fig. 3 and Supplementary Table 3). cluding the genetic forms, is not yet well understood. We addressed this by examining a large clinical cohort of adults with Down syndrome. We found that males and fe­ Association of sex and APOE ɛ4 males with Down syndrome showed similar clinical presen­ haplotype in Down syndrome tation and comparable Alzheimer’s disease biomarker When we considered the APOE ɛ4 haplotype, we found profiles. Despite these similarities, we did observe better epi­ that the prevalence of symptomatic Alzheimer’s disease sodic memory in females from age 45 onwards, and a clear was different across groups at age 40–45 years (Fig. 3A, APOE ɛ4 and sex association on disease onset. Supplementary Table 4, P = 0.05), with the greatest Previous studies assessing sex differences in age at diagno­ prevalence in female ɛ4 carriers (44.4% versus 11.5% sis or Alzheimer’s disease risk in Down syndrome reported for female non-carriers, 9.5% for male non-carriers and conflicting findings. Some have found greater risk and earlier 25% for male ɛ4 carriers). Likewise, the probability of age at diagnosis in males,32,36 while others have found great­ developing symptoms of Alzheimer’s disease at younger er risk or earlier age at diagnosis in females,33,53 trends in ages was highest for females with an APOE ɛ4 allele (P = such direction but no significance35 or no sex differences 0.005, Fig. 3B). In addition, female ɛ4 carriers were diagnosed after adjusting for confounding variables.34 The reasons with symptomatic Alzheimer’s disease on average 3 years for these inconsistencies might be related to differences in earlier than female without this allele (50.5 versus 53.2 years, study design, with some studies using longitudinal, prospect­ P = 0.01, Fig. 3C and Supplementary Table 4). For males, ive, clinician-driven diagnosis of dementia at recurrent there was no difference in age at diagnosis between ɛ4 carriers follow-ups, while others reviewing medical records for previ­ and non-carriers (52.2 versus 52.5 years, P = 0.76). ous dementia diagnoses which could be potentially biased by When we explored the association between sex and APOE differences in referral to a diagnostic physician. Moreover, ɛ4 on cognitive performance, we found that APOE ɛ4 was the age range of studies included individuals between 30 associated with poorer episodic memory in both sexes, but and 70 years, but overall sample sizes varied from n = 21 not with global cognition (CAMCOG-DS scores). The differ­ to n = 100 or even over n = 400. The distribution of intellec­ ence in mCRT scores between ɛ4 carriers and non-carriers tual disability was also heterogeneous. Likewise, different was more pronounced in females compared with males proportions of APOE ɛ4 carriers or of females in meno­ (Fig. 3 E–F and H–I). pause, perimenopause and post-menopause, which are At the level of biomarkers, we performed an exploratory known factors that can impact age at diagnosis,54 might analysis stratifying by sex and APOE ɛ4 haplotype, with help explain these conflicting findings. the limitation that the sample sizes of some biomarker sub­ We found overlapping trajectories of amyloid and neuro­ groups were small (<20). We found that female ɛ4 carriers degeneration biomarkers between males and females with had lower levels of the CSF Aβ42/Aβ40 ratio between ages Down syndrome explored across different modalities (CSF, 20 and 50 years, overlapping afterward (Fig. 4A). In males, plasma and PET), indicating no main effect of sex on these ɛ4 carriers tended to show earlier increases in amyloid-PET biomarkers in this population. This is in line with the general uptake compared with non-carriers (Fig. 4B). The sample literature on sporadic Alzheimer’s disease, where no clear ef­ size of the female ɛ4 subgroup for amyloid-PET was too fect of sex on global amyloid burden has been re­ small for comparisons. Both males and females tended to ported.15,55,56 Contrary to some studies in the general show lower cerebral glucose metabolism between ages 40 population,18 males with Down syndrome did not show high­ and 50 years in ɛ4 carriers compared with non-carriers er NfL concentrations in CSF. The higher levels of NfL in (Fig. 4G). Hippocampal volumes also tended to be lower in healthy males might be attributed to the greater proportion female ɛ4 carriers between ages 45 and 60 years, compared of white matter compared with females, as reported by with females who did not carry this allele (Fig. 4H). No other some studies.57 It would be interesting to investigate whether relevant sex differences in Alzheimer’s disease biomarkers sex differences in white matter integrity or regional variations were detected after stratification by APOE ɛ4 haplotype in exist in Down syndrome that would help interpret this result. this population (Fig. 4). Supplementary Table 5 and The concentrations of plasma and CSF p-tau181 were also Supplementary Table 6 detail the relevant demographic, clin­ highly overlapping across both sexes. Although this suggests ical and biomarker outcomes and sample size of the sub­ a similar degree of tau pathology, we cannot rule out the ex­ groups included in the exploratory analyses segregated by istence of regional differences, as our study did not include APOE ɛ4 carriership and sex. tau-PET imaging data. It would be interesting to examine this, since an autopsy study found greater neurofibrillary tangle density in females with Down syndrome compared with males,53 and there is limited information from other Discussion neuropathology studies due to the absence of reporting of Growing evidence is pointing to sex differences in sex-stratified data28 or insufficient cases to allow for com­ Alzheimer’s disease, including clinical features, disease parisons.58-61 12 | BRAIN COMMUNICATIONS 2023: Page 12 of 16 M. F. Iulita et al. While we found similar ages at diagnosis of Alzheimer’s and females.67 This suggests that consideration of APOE disease between sexes and comparable scores on global cog­ ɛ4 haplotype and sex may have relevant pathophysiological nition, we noted that males exhibited lower mCRT scores implications in Alzheimer’s disease, both in sporadic and from age 45 compared with females. Because of the ceiling genetically determined populations. effect on mCRT scores observed in our study before this Several mechanisms driving sex differences in Alzheimer’s age, we cannot rule out that males and females experience disease have been considered, including the drop of sex hor­ a similar decline in mCRT scores in Down syndrome, with mones with menopause, sex differences in the manifestation females starting at a higher level at younger ages. Another of cerebrovascular and cardiovascular pathologies as well as possibility is that females manifest cognitive resilience to in immune system responses.4 A recent review by Andrews Alzheimer’s disease neuropathological changes. Vila-Castelar and colleagues, discusses such mechanisms in the context et al.62 have similarly reported poorer verbal episodic mem­ of Down syndrome.68 Females with Down syndrome experi­ ory performance in male carriers of the PSEN1 E280A vari­ ence menopause ∼5–7 years earlier than the general popula­ ant at the asymptomatic stage, with no sex differences in tion.69,70 Oestrogens (the primary female sex hormones) act hippocampal volume. These and our results show concord­ through receptors that are also present in the brain71 and ance between two genetically-at risk Alzheimer’s disease have been reported to reduce Aβ toxicity by stimulating the populations. Studies that support the resilience hypothesis metabolism of APP, inhibiting the aggregation of Aβ, or have also found higher cortical thickness or better verbal stimulating its degradation.72-74 Besides modulating amyloid memory scores in females compared with males at similar le­ pathology, oestradiol (the primary form of oestrogen) is vels of tau pathology within the sporadic Alzheimer’s disease known to promote cholinergic activity75 as well as to en­ spectrum.8,63 hance memory consolidation and synaptic plasticity in trans­ Our study also showed that even if the main effect of sex genic mice.76 Interestingly, this protective effect is not seen in on Alzheimer’s disease biomarkers was subtle or null, sex mice who are APOE ɛ4 carriers.76 Given the neuroprotective can interact with other factors, and thus impact the clinical functions of oestrogens and their drop in menopause, it has expression of Alzheimer’s disease in Down syndrome. We been postulated that such event may contribute to the found that both sexes were vulnerable to the genetic risk of damaging effects on cognition and higher vulnerability to APOE ɛ4, as we recently reported,64 but that females with Alzheimer’s disease in females.77,78 Indeed, studies in clinical Down syndrome showed earlier age at diagnosis by 3 years. cohorts of females with Down syndrome have reported This difference was not seen in males. Our exploratory ana­ significant associations between age at menopause,54,70,79 lyses considering sex, APOE ɛ4 and biomarker profiles differences in bioavailable oestrogen levels80 and poly­ showed that female ɛ4 carriers exhibited lower CSF Aβ42/ morphisms in oestrogen receptors81 and age at diagnosis of Aβ40 ratio and lower hippocampal volume compared with dementia. females without this allele, in line with the clinical difference. Outside the context of Alzheimer’s disease, sex differences Given that sample sizes after stratification were small, these in Down syndrome have also been studied. De exploratory results need further confirmation in other Gonzalo-Calvo et al.82 found sex-related differences in bio­ cohorts. chemical and haematological parameters in a clinical cohort Male ɛ4 carriers tended to show earlier increases in of adults with Down syndrome. Further, Startin et al.83 re­ amyloid-PET uptake compared with male non-carriers, but ported sex differences in the prevalence of comorbidities we cannot rule out a similar response in females given the known to occur commonly in people with Down syndrome, small sample size of this subgroup. Indeed, a study by finding higher prevalence rates of otitis and reflux in males Cacciaglia et al.65 recently showed that age, APOE ɛ4 and and higher hypothyroidism in females, in line with our re­ female sex were associated with higher amyloid-PET uptake sults. Both studies outlined recommendations for the incorp­ in posterior middle cortical regions in cognitively unim­ oration of such knowledge into clinical practice, highlighting paired individuals. Likewise, APOE ɛ4 carriers in both sexes that the study of sex differences is relevant for and beyond tended to show lower cerebral glucose metabolism compared Alzheimer’s disease. with non-carriers, although the sample size of the subgroups Our study has strengths and limitations. We recruited the was small to compare by sex. Studies in the general popula­ largest clinical cohort of adults with Down syndrome with tion indicate that female APOE ɛ4 carriers manifest wide­ clinical and multimodal biomarker assessments (plasma, spread clusters of hypometabolism, while males show an CSF, MRI and PET) to compare clinical and biomarker tra­ isolated cluster of hypometabolism in the precuneus com­ jectories between males and females. Its limitations include pared with non-carriers.20 Moreover, a recent study analys­ the lack of tau-PET imaging, its cross-sectional design and ing longitudinal data from ADNI showed that APOE ɛ4 the relatively small sample size for some biomarkers, espe­ status and sex modified the progression of disease (from pre- cially for the investigation of sex differences across APOE clinical to dementia stages), with female APOE ɛ4 carriers ɛ4 haplotype (e.g. FDG-PET and amyloid-PET in females). showing the fastest cognitive decline.66 The APOE ɛ4-sex as­ Nevertheless, the uniform development of Alzheimer’s dis­ sociation is also supported by single-nucleus transcriptome ease pathology in Down syndrome and its consistent age at studies where differential transcriptional responses to onset29,84 allow quasi-longitudinal studies in this popula­ Alzheimer’s disease neuropathology were shown in males tion. Of note, our study did not consider menopause or Clinical and biomarker changes by sex in DS-AD BRAIN COMMUNICATIONS 2023: Page 13 of 16 | 13 hormonal status, which can impact risk in a sex-specific Alzheimer Drug Discovery Foundation (#201809-2016862), manner.54,80,85,86 We also could not conduct sub-analysis the Alzheimer’s Disease Strategic Fund and the Alzheimer’s by race, as DABNI is a population-based cohort reflecting Association (#ADSF-21-831376-C, #ADSF-21-831381-C the population of Catalonia, which is predominantly White. and #ADSF-21-831377-C), the Olav Thon Stiftelsen, the Taken together, our results indicate that sex does not Erling-Persson Family Foundation, Stiftelsen för Gamla modify Alzheimer’s disease clinical outcomes and biomarker Tjänarinnor, Hjärnfonden, Sweden (#FO2019-0228), the trajectories in adults with Down syndrome, except for the as­ European Union’s Horizon 2020 research and innovation pro­ sociation between sex and APOE ɛ4 haplotype. In light of gram under the Marie Skłodowska-Curie grant agreement no. the growth of personalized healthcare, we urge the scientific 860197 (MIRIADE: Multi-omics Interdisciplinary Research and medical community to account for, report and publish Integration to Address DEmentia diagnosis), the EU Joint sex-stratified data, even when no sex differences are found. Programme Neurodegenerative Disease Research (JPND2021-00694) and the UK Dementia Research Institute at University College London. K.B. is supported by the Supplementary material Swedish Research Council (#2017-00915), the Alzheimer Drug Discovery Foundation, USA Supplementary material is available at Brain Communications (#RDAPB-201809-2016615), the Swedish Alzheimerfonden online. (#AF-742881), Hjärnfonden, Sweden (#FO2017-0243), the Swedish state under the agreement between the Swedish gov­ ernment and the County Councils, the ALF agreement Acknowledgements (#ALFGBG-715986), the EU Joint Programme The authors would like to thank all study participants and Neurodegenerative Disease Research (JPND2019-466-236), caregivers for their support and dedication. We also would the National Institutes of Health grant like to thank the Fundació Catalana Síndrome de Down (#1R01AG068398-01) and the Alzheimer’s Association 2021 (https://fcsd.org/) and the UK Down Syndrome Research Zenith Award (ZEN-21-848495). The authors are grateful to Foundation (https://www.dsrf-uk.org/) for global support; the Medical Research Council, UK for generously providing a Laia Muñoz and Raúl Núñez for laboratory management; grant to support this research (#98480). Additional support Reyes Alcoverro, Marta Salinas and Tania Martínez for ad­ came from the NIHR Cambridge Biomedical Research ministrative support; and Concepción Escola and Diana Centre, the National Institute for Health and Care Research Garzón for nurse support. We also thank the PET imaging Collaborations in Leadership for Applied Health Research technicians at the Wolfson Brain Imaging Centre, UK, as and Care for the East of England, the National Institute for well as the involved clinicians for their help in acquiring Health and Care Research Cambridge Dementia Biomedical the data reported in this article. The authors also wish to ac­ Research Unit, the Down Syndrome Association and the knowledge the scientific input from Maria Teresa Ferretti, Health Foundation. A.J.H. was supported during the prepar­ PhD, CSO Women’s Brain Project, and the members of the ation of this manuscript by the National Institute for Health scientific committee at the Women’s Brain Project (https:// and Care Research Collaborations in Leadership for Applied www.womensbrainproject.com/). Health Research and Care East of England at Cambridgeshire and Peterborough National Health Service Foundation Trust. M.F.I. acknowledges support from the Jérôme Lejeune Funding postdoctoral award and pilot grant (#1941). A.B. acknowl­ This study was supported by the Fondo de Investigaciones edges support from a Miguel Servet grant (CP20/00038) Sanitario, Carlos III Health Institute (PI14/01126, PI17/ from the Carlos III Health Institute. M.C.I. acknowledges 01019 and PI20/01473 to J.F., PI13/01532 and PI16/01825 support from the Alzheimer’s Association and Global to R.B., PI18/00335 to M.C.I., PI18/00435 to D.A., PI14/ Brain Health Institute (GBHI_ALZ-18-543740), the 1561 and PI20/01330 to A.L. and P21/00791 to I.I.G.) and Jérôme Lejeune Foundation (#1913 Cycle 2019B) and the the Centro de Investigación Biomédica en Red sobre Societat Catalana de Neurologia (Premi Beca Fundació Enfermedades Neurodegenerativas (CIBERNED) Program SCN 2020). V.M. acknowledges support from a predoctoral 1. This work was also supported by the National Institutes of grant from the Carlos III Health Institute (FI18/00275). Health grants (1R01AG056850-01A1; R21AG056974 and M.A. and T.E. were supported by a Río Hortega R01AG061566 to J.F.), the Department de Salut de la Fellowship (CM19/00066 and CM19/0001) from the Generalitat de Catalunya, Pla Estratè gic de Recerca i Carlos III Health Institute. C.P. acknowledges support Innovació en Salut (SLT002/16/00408 to A.L.) and the from a Sara Borrell Fellowship (CP20/00133) from the Fundació La Marató de TV3 (20141210 to J.F. and 044412 Carlos III Health Institute. I.I.G. is supported by the Juan to R.B.). H.Z. is a Wallenberg Scholar supported by grants Rodés grant (JR20/00018) from the Carlos III Health from the Swedish Research Council (#2018-02532), the Institute, the Global Brain Health Institute, the Alzheimer´s European Research Council (#681712), the Swedish State Association and Alzheimer´s UK grants (GBHI ALZ Support for Clinical Research (#ALFGBG-720931), the UK-21-720973 and AACSF-21-850193). 14 | BRAIN COMMUNICATIONS 2023: Page 14 of 16 M. F. Iulita et al. 9. Laws KR, Irvine K, Gale TM. Sex differences in cognitive impair­ Competing interests ment in Alzheimer’s disease. World J Psychiatry. 2016;6(1):54-65. 10. Sohn D, Shpanskaya K, Lucas JE, et al. Sex differences in cognitive O.B. reported receiving personal fees from ADx decline in subjects with high likelihood of mild cognitive impair­ NeuroSciences outside the submitted work. H.Z. declares ment due to Alzheimer’s disease. Sci Rep. 2018;8(1):7490. that he has served on scientific advisory boards and/or as 11. Lin KA, Choudhury KR, Rathakrishnan BG, et al. Marked gender a consultant for Abbvie, Alector, Annexon, Artery differences in progression of mild cognitive impairment over 8 Therapeutics, AZTherapies, CogRx, Denali, Eisai, years. Alzheimers Dement (N Y). 2015;1(2):103-110. 12. Mielke MM. Consideration of sex differences in the measurement Nervgen, Pinteon Therapeutics, Red Abbey Labs, Passage and interpretation of Alzheimer disease-related biofluid-based bio­ Bio, Roche, Samumed, Siemens Healthineers, Triplet markers. J Appl Lab Med. 2020;5(1):158-169. Therapeutics and Wave, has given lectures in symposia spon­ 13. Buckley RF, Mormino EC, Rabin JS, et al. Sex differences in the as­ sored by Cellectricon, Fujirebio, Alzecure, Biogen and Roche sociation of global amyloid and regional tau deposition measured and is a co-founder of Brain Biomarker Solutions in by positron emission tomography in clinically normal older adults. JAMA Neurol. 2019;76(5):542-551. Gothenburg AB, which is a part of the GU Ventures 14. Buckley RF, Scott MR, Jacobs HIL, et al. Sex mediates relationships Incubator Program, outside the submitted work. K.B. de­ between regional tau pathology and cognitive decline. Ann Neurol. clares that he has served as a consultant, on advisory boards, 2020;88(5):921-932. or data monitoring committees for Abcam, Axon, BioArctic, 15. Edwards L, La Joie R, Iaccarino L, et al. Multimodal neuroimaging Biogen, JOMDD/Shimadzu, Julius Clinical, Lilly, MagQu, of sex differences in cognitively impaired patients on the Alzheimer’s continuum: Greater tau-PET retention in females. Novartis, Pharmatrophix, Prothena, Roche Diagnostics Neurobiol Aging. 2021;105:86-98. and Siemens Healthineers and is a co-founder of Brain 16. Liesinger AM, Graff-Radford NR, Duara R, et al. Sex and age inter­ Biomarker Solutions in Gothenburg AB, which is a part of act to determine clinicopathologic differences in Alzheimer’s dis­ the GU Ventures Incubator Program, outside the submitted ease. Acta Neuropathol. 2018;136(6):873-885. work. D.A. reported receiving personal fees for advisory 17. Oveisgharan S, Arvanitakis Z, Yu L, Farfel J, Schneider JA, Bennett DA. Sex differences in Alzheimer’s disease and common neuro­ board services and/or speaker honoraria from pathologies of aging. Acta Neuropathol. 2018;136(6):887-900. Fujirebio-Europe, Roche, Nutricia, Krka Farmacéutica and 18. Bridel C, van Wieringen WN, Zetterberg H, et al. Diagnostic value Esteve, outside the submitted work. A.L. has served as a con­ of cerebrospinal fluid neurofilament light protein in neurology: A sultant or on advisory boards for Fujirebio-Europe, Roche, systematic review and meta-analysis. JAMA Neurol. 2019;76(9): Biogen and Nutricia, outside the submitted work. J.F. re­ 1035-1048. 19. Neu SC, Pa J, Kukull W, et al. Apolipoprotein E genotype and sex ported receiving personal fees for service on the advisory risk factors for Alzheimer disease: A meta-analysis. JAMA Neurol. boards, adjudication committees or speaker honoraria 2017;74(10):1178-1189. from AC Immune, Novartis, Lundbeck, Roche, Fujirebio 20. Sampedro F, Vilaplana E, de Leon MJ, et al. APOE-by-sex interac­ and Biogen, outside the submitted work. O.B., D.A., A.L. tions on brain structure and metabolism in healthy elderly controls. and J.F. report holding a patent for markers of synaptopathy Oncotarget. 2015;6(29):26663-26674. 21. Hohman TJ, Dumitrescu L, Barnes LL, et al. Sex-specific association in neurodegenerative disease (licensed to ADx, of apolipoprotein E with cerebrospinal fluid levels of tau. JAMA EPI8382175.0). No other competing interests were reported. Neurol. 2018;75(8):989-998. 22. Wang YT, Pascoal TA, Therriault J, et al. Interactive rather than in­ dependent effect of APOE and sex potentiates tau deposition in wo­ men. Brain Commun. 2021;3(2):fcab126. References 23. Yan S, Zheng C, Paranjpe MD, et al. Sex modifies APOE epsilon4 dose effect on brain tau deposition in cognitively impaired indivi­ 1. Hentzen NB, Ferretti MT, Santuccione A, et al. Mapping of duals. Brain. 2021;144(10):3201-3211. European activities on the integration of sex and gender factors in 24. McCarron M, McCallion P, Reilly E, Dunne P, Carroll R, Mulryan neurology and neuroscience. Eur J Neurol. 2022;29(9):2572–2579. N. A prospective 20-year longitudinal follow-up of dementia in per­ 2. Spires-Jones TL. Let’s talk about sex (in translational neuroscience). sons with Down syndrome. J Intellect Disabil Res. 2017;61(9): Brain Commun. 2022;4(2):fcac028. 843-852. 3. Gauthier S, Rosa-Neto P, Morais JA, Webster C. World Alzheimer 25. Fortea J, Vilaplana E, Carmona-Iragui M, et al. Clinical and bio­ Report 2021: Journey through the diagnosis of dementia. 2021. marker changes of Alzheimer’s disease in adults with Down syn­ 4. Ferretti MT, Iulita MF, Cavedo E, et al. Sex differences in Alzheimer drome: A cross-sectional study. Lancet. 2020;395(10242): disease—The gateway to precision medicine. Nat Rev Neurol. 1988-1997. 2018;14(8):457-469. 26. Snyder HM, Bain LJ, Brickman AM, et al. Further understanding 5. Nebel RA, Aggarwal NT, Barnes LL, et al. Understanding the im­ the connection between Alzheimer’s disease and Down syndrome. pact of sex and gender in Alzheimer’s disease: A call to action. Alzheimers Dement. 2020;16(7):1065-1077. Alzheimers Dement. 2018;14(9):1171-1183. 27. Lemere CA, Blusztajn JK, Yamaguchi H, Wisniewski T, Saido TC, 6. McCarrey AC, An Y, Kitner-Triolo MH, Ferrucci L, Resnick SM. Selkoe DJ. Sequence of deposition of heterogeneous amyloid beta- Sex differences in cognitive trajectories in clinically normal older peptides and APO E in Down syndrome: Implications for initial events adults. Psychol Aging. 2016;31(2):166-175. in amyloid plaque formation. Neurobiol Dis. 1996;3(1):16-32. 7. Sundermann EE, Maki P, Biegon A, et al. Sex-specific norms for ver­ 28. Wisniewski KE, Wisniewski HM, Wen GY. Occurrence of neuro­ bal memory tests may improve diagnostic accuracy of amnestic pathological changes and dementia of Alzheimer’s disease in MCI. Neurology. 2019;93(20):e1881-e1889. Down’s syndrome. Ann Neurol. 1985;17(3):278-282. 8. Digma LA, Madsen JR, Rissman RA, et al. Women can bear a bigger 29. Fortea J, Zaman SH, Hartley S, Rafii MS, Head E, Carmona-Iragui burden: Ante- and post-mortem evidence for reserve in the face of M. Alzheimer’s disease associated with Down syndrome: A genetic tau. Brain Commun. 2020;2(1):fcaa025. form of dementia. Lancet Neurol. 2021;20(11):930-942. Clinical and biomarker changes by sex in DS-AD BRAIN COMMUNICATIONS 2023: Page 15 of 16 | 15 30. Bateman RJ, Xiong C, Benzinger TL, et al. Clinical and biomarker 49. Zammit MD, Laymon CM, Betthauser TJ, et al. Amyloid accumu­ changes in dominantly inherited Alzheimer’s disease. N Engl J lation in Down syndrome measured with amyloid load. Alzheimers Med. 2012;367(9):795-804. Dement (Amst). 2020;12(1):e12020. 31. Rafii MS, Ances BM, Schupf N, et al. The AT(N) framework for 50. Klunk WE, Koeppe RA, Price JC, et al. The Centiloid project: Alzheimer’s disease in adults with Down syndrome. Alzheimers Standardizing quantitative amyloid plaque estimation by PET. Dement (Amst). 2020;12(1):e12062. Alzheimers Dement. 2015;11(1):1-15 e1–4. 32. Schupf N, Kapell D, Nightingale B, Rodriguez A, Tycko B, Mayeux 51. Hastie TJ. Generalized additive models. In: Chambers JM, Hastie R. Earlier onset of Alzheimer’s disease in men with Down syndrome. TJ, eds. Statistical models in S. Wadsworth & Neurology. 1998;50(4):991-995. Brooks; 1992:309-376. 33. Lai F, Kammann E, Rebeck GW, Anderson A, Chen Y, Nixon RA. 52. Ferretti MT, Martinkova J, Biskup E, et al. Sex and gender differ­ APOE genotype and gender effects on Alzheimer disease in 100 ences in Alzheimer’s disease: Current challenges and implications adults with Down syndrome. Neurology. 1999;53(2):331-336. for clinical practice: Position paper of the Dementia and Cognitive 34. Sinai A, Mokrysz C, Bernal J, et al. Predictors of age of diagnosis Disorders Panel of the European Academy of Neurology. Eur J and survival of Alzheimer’s disease in Down syndrome. J Neurol. 2020;27(6):928-943. Alzheimers Dis. 2018;61(2):717-728. 53. Raghavan R, Khin-Nu C, Brown AG, et al. Gender differences in the 35. Lai F, Mhatre PG, Yang Y, Wang MC, Schupf N, Rosas HD. Sex phenotypic expression of Alzheimer’s disease in Down’s syndrome differences in risk of Alzheimer’s disease in adults with Down syn­ (trisomy 21). Neuroreport. 1994;5(11):1393-1396. drome. Alzheimers Dement (Amst). 2020;12(1):e12084. 54. Schupf N, Pang D, Patel BN, et al. Onset of dementia is associated 36. Mhatre PG, Lee JH, Pang D, et al. The association between sex and with age at menopause in women with Down’s syndrome. Ann risk of Alzheimer’s disease in adults with Down syndrome. J Clin Neurol. 2003;54(4):433-438. Med. 2021;10(13):2966. 55. Jansen WJ, Ossenkoppele R, Knol DL, et al. Prevalence of cerebral 37. Rafii MS, Zaman S, Handen BL. Integrating biomarker outcomes amyloid pathology in persons without dementia: A meta-analysis. into clinical trials for Alzheimer’s disease in Down syndrome. J JAMA. 2015;313(19):1924-1938. Prev Alzheimers Dis. 2021;8(1):48-51. 56. Buckley RF, Mormino EC, Amariglio RE, et al. Sex, amyloid, and 38. Annus T, Wilson LR, Hong YT, et al. The pattern of amyloid accu­ APOE epsilon4 and risk of cognitive decline in preclinical mulation in the brains of adults with Down syndrome. Alzheimers Alzheimer’s disease: Findings from three well-characterized cohorts. Dementia. 2016;12(5):538-545. Alzheimers Dement. 2018;14(9):1193-1203. 39. Alcolea D, Clarimon J, Carmona-Iragui M, et al. The Sant Pau ini­ 57. Gur RC, Turetsky BI, Matsui M, et al. Sex differences in brain gray tiative on neurodegeneration (SPIN) cohort: A data set for biomark­ and white matter in healthy young adults: Correlations with cogni­ er discovery and validation in neurodegenerative disorders. tive performance. J Neurosci. 1999;19(10):4065-4072. Alzheimers Dement (N Y). 2019;5:597-609. 58. Wisniewski KE, Dalton AJ, McLachlan C, Wen GY, Wisniewski 40. Aylward EH, Burt DB, Thorpe LU, Lai F, Dalton A. Diagnosis of de­ HM. Alzheimer’s disease in Down’s syndrome: Clinicopathologic mentia in individuals with intellectual disability. J Intellect Disabil studies. Neurology. 1985;35(7):957-961. Res. 1997;41(Pt 2):152-164. 59. Mann DM. The pathological association between Down syndrome 41. Esteba-Castillo S, Dalmau-Bueno A, Ribas-Vidal N, Vila-Alsina M, and Alzheimer disease. Mech Ageing Dev. 1988;43(2):99-136. Novell-Alsina R, Garcia-Alba J. [Adaptation and validation of 60. Ball MJ, Nuttall K. Topography of neurofibrillary tangles and gran­ CAMDEX-DS (Cambridge examination for mental disorders of ulovacuoles in hippocampi of patients with Down’s syndrome: older people with Down’s syndrome and others with intellectual dis­ Quantitative comparison with normal ageing and Alzheimer’s dis­ abilities) in Spanish population with intellectual disabilities]. Rev ease. Neuropathol Appl Neurobiol. 1981;7(1):13-20. Neurol. 2013;57(8):337-346. Adaptacion y validacion del 61. Motte J, Williams RS. Age-related changes in the density and Cambridge Examination for Mental Disorders of Older People morphology of plaques and neurofibrillary tangles in Down syn­ with Down’s Syndrome and Others with Intellectual Disabilities drome brain. Acta Neuropathol. 1989;77(5):535-546. (CAMDEX-DS) en poblacion espanola con discapacidad 62. Vila-Castelar C, Guzman-Velez E, Pardilla-Delgado E, et al. intelectual. Examining sex differences in markers of cognition and neurodegen­ 42. Benejam B, Videla L, Vilaplana E, et al. Diagnosis of prodromal and eration in autosomal dominant Alzheimer’s disease: Preliminary Alzheimer’s disease dementia in adults with Down syndrome using findings from the Colombian Alzheimer’s prevention initiative bio­ neuropsychological tests. Alzheimers Dement (Amst). 2020;12(1): marker study. J Alzheimers Dis. 2020;77(4):1743-1753. e12047. 63. Ossenkoppele R, Lyoo CH, Jester-Broms J, et al. Assessment of 43. Fortea J, Carmona-Iragui M, Benejam B, et al. Plasma and CSF bio­ demographic, genetic, and imaging variables associated with brain markers for the diagnosis of Alzheimer’s disease in adults with resilience and cognitive resilience to pathological tau in patients Down syndrome: A cross-sectional study. Lancet Neurol. 2018;17­ with Alzheimer disease. JAMA Neurol. 2020;77(5):632-642. (10):860-869. 64. Bejanin A, Iulita MF, Vilaplana E, et al. Association of apolipopro­ 44. Alcolea D, Pegueroles J, Munoz L, et al. Agreement of amyloid PET tein E varepsilon4 allele with clinical and multimodal biomarker and CSF biomarkers for Alzheimer’s disease on Lumipulse. Ann changes of Alzheimer disease in adults with Down syndrome. Clin Transl Neurol. 2019;6(9):1815-1824. JAMA Neurol. 2021;78(8):937-947. 45. Delaby C, Estelles T, Zhu N, et al. The Abeta1–42/Abeta1–40 ratio 65. Cacciaglia R, Salvado G, Molinuevo JL, et al. Age, sex and in CSF is more strongly associated to tau markers and clinical pro­ APOE-epsilon4 modify the balance between soluble and fibrillar gression than Abeta1–42 alone. Alzheimers Res Ther. 2022;14(1): beta-amyloid in non-demented individuals: Topographical patterns 20. across two independent cohorts. Mol Psychiatry. 2022;27(4): 46. Lleo A, Zetterberg H, Pegueroles J, et al. Phosphorylated tau181 in 2010–2018. plasma as a potential biomarker for Alzheimer’s disease in adults 66. Cho SH, Woo S, Kim C, et al. Disease progression modelling from with Down syndrome. Nat Commun. 2021;12(1):4304. preclinical Alzheimer’s disease (AD) to AD dementia. Sci Rep. 47. Landau SM, Harvey D, Madison CM, et al. Associations between 2021;11(1):4168. cognitive, functional, and FDG-PET measures of decline in AD 67. Mathys H, Davila-Velderrain J, Peng Z, et al. Single-cell transcriptomic and MCI. Neurobiol Aging. 2011;32(7):1207-1218. analysis of Alzheimer’s disease. Nature. 2019;570(7761):332-337. 48. Avants BB, Tustison NJ, Song G, Cook PA, Klein A, Gee JC. A re­ 68. Andrews EJ, Martini AC, Head E. Exploring the role of sex differ­ producible evaluation of ANTs similarity metric performance in ences in Alzheimer’s disease pathogenesis in Down syndrome. brain image registration. Neuroimage. 2011;54(3):2033-2044. Front Neurosci. 2022;16:954999. 16 | BRAIN COMMUNICATIONS 2023: Page 16 of 16 M. F. Iulita et al. 69. Schupf N, Zigman W, Kapell D, Lee JH, Kline J, Levin B. Early 77. Brinton RD. Cellular and molecular mechanisms of estrogen regula­ menopause in women with Down’s syndrome. J Intellect Disabil tion of memory function and neuroprotection against Alzheimer’s Res. 1997;41(Pt 3):264-267. disease: Recent insights and remaining challenges. Learn Mem. 70. Coppus A, Evenhuis H, Verberne GJ, et al. Early age at 2001;8(3):121-133. menopause is associated with increased risk of dementia and mor­ 78. Garcia-Segura LM, Azcoitia I, DonCarlos LL. Neuroprotection by tality in Down syndrome. J Appl Res Intellect Disabil. 2010; estradiol. Prog Neurobiol. 2001;63(1):29-60. 23(5):408-408. 79. Cosgrave MP, Tyrrell J, McCarron M, Gill M, Lawlor BA. Age at 71. Almey A, Milner TA, Brake WG. Estrogen receptors in the central onset of dementia and age of menopause in women with Down’s nervous system and their implication for dopamine-dependent cog­ syndrome. J Intellect Disabil Res. 1999;43(Pt 6):461-465. nition in females. Horm Behav. 2015;74:125-138. 80. Schupf N, Winsten S, Patel B, et al. Bioavailable estradiol and age at 72. Spampinato SF, Molinaro G, Merlo S, et al. Estrogen receptors and onset of Alzheimer’s disease in postmenopausal women with Down type 1 metabotropic glutamate receptors are interdependent in pro­ syndrome. Neurosci Lett. 2006;406(3):298-302. tecting cortical neurons against beta-amyloid toxicity. Mol 81. Schupf N, Lee JH, Wei M, et al. Estrogen receptor-alpha variants in­ Pharmacol. 2012;81(1):12-20. crease risk of Alzheimer’s disease in women with Down syndrome. 73. Liang K, Yang L, Yin C, et al. Estrogen stimulates degradation of Dement Geriatr Cogn Disord. 2008;25(5):476-482. beta-amyloid peptide by up-regulating neprilysin. J Biol Chem. 82. de Gonzalo-Calvo D, Barroeta I, Nan MN, et al. Evaluation of bio­ 2010;285(2):935-942. chemical and hematological parameters in adults with Down syn­ 74. Amtul Z, Wang L, Westaway D, Rozmahel RF. drome. Sci Rep. 2020;10(1):13755. Neuroprotective mechanism conferred by 17beta-estradiol on the 83. Startin CM, D’Souza H, Ball G, et al. Health comorbidities and cog­ biochemical basis of Alzheimer’s disease. Neuroscience. 2010; nitive abilities across the lifespan in Down syndrome. J Neurodev 169(2):781-786. Disord. 2020;12(1):4. 75. Batallan Burrowes AA, Olajide OJ, Iasenza IA, Shams WM, Carter 84. Iulita MF, Garzon D, Christensen MK, et al. Association of F, Chapman CA. Ovariectomy reduces cholinergic modulation of Alzheimer disease with life expectancy in people with Down syn­ excitatory synaptic transmission in the rat entorhinal cortex. drome. JAMA Network Open. 2022;5(5):e2212910. PLoS One. 2022;17(8):e0271131. 85. Patel BN, Seltzer GB, Wu HS, Schupf N. Effect of menopause on 76. Taxier LR, Philippi SM, Fleischer AW, York JM, LaDu MJ, Frick cognitive performance in women with Down syndrome. KM. APOE4 homozygote females are resistant to the beneficial ef­ Neuroreport. 2001;12(12):2659-2662. fects of 17beta-estradiol on memory and CA1 dendritic spine dens­ 86. Coppus AMW, Evenhuis HM, Verberne GJ, et al. Early age at meno­ ity in the EFAD mouse model of Alzheimer’s disease. Neurobiol pause is associated with increased risk of dementia and mortality in Aging. 2022;118:13-24. women with Down syndrome. J Alzheimers Dis. 2010;19(2):545-550.