Received: 7 September 2020 Revised: 10 November 2020 Accepted: 2 January 2021 DOI: 10.1002/alz.12299 F E AT U R E D A R T I C L E Sex differences in the behavioral variant of frontotemporal dementia: A new window to executive and behavioral reserve Ignacio Illán-Gala1,2,3,4 Kaitlin B. Casaletto4 Sergi Borrego-Écija5 Eider M. Arenaza-Urquijo6 Amy Wolf4 Yann Cobigo4 Sheng Yang M. Goh4 Adam M. Staffaroni4 Daniel Alcolea1,2 Juan Fortea1,2 Rafael Blesa1,2 Jordi Clarimon1,2 Maria Florencia Iulita1,2,7 Anna Brugulat-Serrat3,6 Albert Lladó5 Lea T. Grinberg4,8 Katherine Possin4 Katherine P. Rankin4 Joel H. Kramer4 Gil D. Rabinovici4 Adam Boxer4 William W. Seeley4,8 Virginia E. Sturm4 Maria Luisa Gorno-Tempini4 Bruce L. Miller4 Raquel Sánchez-Valle5 David C. Perry4 Alberto Lleó1,2 Howard J. Rosen4 1 Sant Pau Memory Unit, Department of Neurology, Biomedical Research Institute Sant Abstract Pau, Hospital de la Santa Creu i Sant Pau, Introduction: Biological sex is an increasingly recognized factor driving clinical and Universitat Autònoma de Barcelona, Barcelona, Spain structural heterogeneity in Alzheimer’s disease, but its role in the behavioral variant 2 Centro de Investigación Biomédica en Red de of frontotemporal dementia (bvFTD) is unknown. Enfermedades Neurodegenerativas Methods: We included 216 patients with bvFTD and 235 controls with magnetic reso- (CIBERNED), Spain 3 Atlantic Fellow for Equity in Brain Health at nance imaging (MRI) from a large multicenter cohort. We compared the clinical charac- the University of California San Francisco, San teristics and cortical thickness between men and women with bvFTD and controls. We Francisco, California, USA followed the residuals approach to study behavioral and cognitive reserve. 4 Memory and Aging Center, Department of Neurology, University of California, San Results: At diagnosis, women with bvFTD showed greater atrophy burden in the fron- Francisco, California, USA totemporal regions compared to men despite similar clinical characteristics. For a sim- 5 Alzheimer’s Disease and Other Cognitive ilar amount of atrophy, women demonstrated better-than-expected scores on execu- Disorders Unit, Service of Neurology, Hospital Clínic de Barcelona, Institut d’Investigació tive function and fewer changes in apathy, sleep, and appetite than men. Biomèdica August Pi i Sunyer, University of Discussion: Our findings suggest that women might have greater behavioral and exec- Barcelona, Barcelona, Spain 6 utive reserve than men, and neurodegeneration must be more severe in women to pro- Barcelonaβeta Brain Research Center (BBRC), Pasqual Maragall Foundation, duce symptoms similar in severity to those in men. Barcelona, Spain 7 KEYWORDS Women’s Brain Project, Guntershausen, Switzerland cognitive reserve, diagnosis, frontotemporal dementia, magnetic resonance imaging, neuroimag- 8 Memory and Aging Center, Department of ing, progression, resilience, survival Pathology, University of California, San Francisco, California, USA Correspondence Dr. Ignacio Illán-Gala, Memory Unit, Depart- ment of Neurology, Hospital de la Santa Creu i Sant Pau, Sant Antoni Maria Claret 167, 08025 Barcelona, Spain. Email: iillan@santpau.cat This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made. © 2021 The Authors. Alzheimer’s & Dementia published by Wiley Periodicals LLC on behalf of Alzheimer’s Association Alzheimer’s Dement. 2021;17:1329–1341. wileyonlinelibrary.com/journal/alz 1329 1330 ILLÁN-GALA ET AL . 1 BACKGROUND RESEARCH IN CONTEXT The behavioral variant of frontotemporal dementia (bvFTD) is a clini- 1. Systematic review: The authors used PubMed to iden- cal syndrome characterized by progressive personality and behavioral tify previous studies examining sex differences in the change along with social, cognitive, and functional deterioration.1,2 behavioral variant of frontotemporal dementia (bvFTD). This syndrome is the most frequent presentation of frontotemporal The impact of sex in the clinical presentation, longitudi- lobar degeneration (FTLD). In contrast with typical Alzheimert’s dis- nal decline, and cortical thickness of bvFTD has not been ease (AD), patients with bvFTD show an anterior pattern of neurode- investigated previously. generation with selective involvement of the frontotemporal cortex.2–4 2. Interpretation: Women with bvFTD showed increased These structures are critical nodes for the sustainment of executive atrophy burden in frontotemporal regions compared to function, language, and social cognition. men despite similar clinical characteristics. For a given Structural imaging biomarkers, such as magnetic resonance imaging amount of atrophy, women with bvFTD demonstrated (MRI), are powerful tools for the in vivo study of neurodegenera- better-than-expected scores on executive function and tive diseases. The identification of frontotemporal atrophy on MRI fewer changes in apathy, sleep, and appetite than men. can increase diagnostic certainty of underlying FTLD in bvFTD, Our findings extend previous research on sex differences but patterns of atrophy and progression are heterogeneous.5,6 The in cognitive reserve in neurodegenerative dementias. observed variability in the patterns of atrophy, together with vari- 3. Future direction: Women with bvFTD might have a able clinical presentation and rates of longitudinal decline, hampers greater ability to cope with frontotemporal neurodegen- accurate predictions of treatment effects using clinical or imaging eration. Future studies should examine the specific mech- outcome measures.7 A better understanding of the factors determin- anisms underlying the observed differences and consider ing clinical and structural heterogeneity is an essential step for the biological sex for the design of research studies and clini- development of precision medicine approaches in neurodegenerative cal trials in bvFTD and other frontotemporal lobar degen- dementias.8,9 eration syndromes. Biological sex is an increasingly recognized factor driving clinical and structural heterogeneity at the single-subject level, and the impor- tance of sex-related differences has been recently highlighted as a research priority in the study of neurodegenerative dementias.8,10–13 To date, most studies investigating sex differences in neurodegen- in disease severity or clinical characteristics at diagnosis; and (3) men erative dementias have focused on AD.14 However, some studies in and women with bvFTD may show similar clinical features at diagnosis FTLD syndromes indicate that sex could influence clinical presen- with significant differences in the atrophy burden. tation and diagnosis.8,15–19 For instance, in large cohorts of FTLD, patients presenting with bvFTD included a higher proportion of men, whereas patients presenting the non-fluent variant of primary pro- 2 METHODS gressive aphasia had a higher proportion of women (72% and 28% of men, respectively, in the largest pathology proven cohorts).3,20 2.1 Study population Under normal conditions, sex-related differences in brain structure have been reported consistently in the frontotemporal cortex with Participants with bvFTD were recruited from two cohorts, at three dif- higher cortical thickness in women.21,22 Moreover, the observed dif- ferent centers: 149 at the University of California, San Francisco Mem- ference in the ability to cope with pathology and sustain cogni- ory and Aging Center (MAC, San Francisco, California, USA), and 67 at tive function (known as cognitive reserve) can also be modulated the Catalan Frontotemporal Dementia Initiative (CATFI; 52 at Hospi- by sex.19,23 tal de Sant Pau, and 15 at the Hospital Clinic de Barcelona, Barcelona, In this multicenter study, we aimed to characterize the impact of Spain). Figure 1 shows a flowchart of the sample composition. Inclu- sex on clinical presentation, longitudinal decline, and cortical thick- sion criteria for bvFTD participants were: (1) meeting the International ness in bvFTD. We further explored the existence of sex differences in Behavioral Variant FTD Criteria Consortium (FTDC) revised guidelines cognitive reserve by following the residuals approach to operational- for the diagnosis of bvFTD25 and, (2) having an MRI available for anal- ize reserve as having better cognition and fewer behavioral changes ysis at the time of diagnosis. To avoid the inclusion of bvFTD “phe- than predicted by cortical thickness (a proxy of pathology). To do this, nocopies” (bvFTD caused by non-neurodegenerative conditions), we we modeled the relationship between cognition or behavior with cor- excluded 19 bvFTD participants classified as “possible bvFTD” with- tical thickness and used each individual’s residual as a proxy of their out evidence of progressive deterioration or an alternative diagnosis reserve.19,24 We anticipated three possible scenarios: (1) men and during follow-up (Figure 1). All patients underwent complete clinical women with bvFTD may show similar clinical presentation, progres- history taking, physical examination, neuropsychological evaluation, sion, and burden of frontotemporal atrophy; (2) men and women with and structural brain imaging. Neuropathological evaluation and genetic bvFTD may show different patterns of atrophy, reflecting differences analyses were also available in a subgroup of bvFTD participants ILLÁN-GALA ET AL . 1331 F I G U R E 1 Flowchart of the sample composition. Abbreviations: bvFTD, behavioral variant of frontotemporal dementia; CATFI, Catalan Frontotemporal Dementia Initiative; FTLD, frontotemporal lobar degeneration; M, men; MRI, magnetic resonance imaging; UCSF, University of California San Francisco; W, women (Supplementary Methods).3,26,27 A total of 225 age-matched healthy 2.3 Measures of cognitive function controls from the two cohorts were also included as imaging con- trols (168 UCSF and 57 CATFI). All healthy controls had normal cog- The Mini-Mental State Examination (MMSE) was recorded in all nitive performance according to normative data and did not have any centers as a general measure of cognitive impairment.31 The MMSE neurologic, psychiatric, or other major medical illnesses.2,28 Written was also recorded during follow-up. Detailed cognitive functioning informed consent was obtained from all participants or their assigned was assessed using two previously published neuropsychological surrogate decision-makers, and local institutional review boards for protocols.2,28 We selected cognitive measures from each protocol human research approved the study. measuring parallel neuropsychological constructs, and we defined four main neuropsychological domains (memory, executive functioning, language, and visuospatial functioning) as in previous multicenter 2.2 Measures of disease severity studies.32,33 Details on the neuropsychological protocols can be found in Supplementary Methods and Table S2. At presentation, the Clinical Dementia Rating Scale Sum of Boxes (CDRsb) and the CDR Dementia Staging Instrument PLUS National Alzheimer’s Coordinating Center Behavior and Language Domains 2.4 Behavioral measures (CDR plus NACC FTLD) were recorded as the main measures of dis- ease severity.29,30 During follow-up, longitudinal CDRsb and CDR plus We recorded the specific features of bvFTD diagnostic criteria present NACC FTLD were also recorded. at diagnosis. We selected the frequency-by-severity product of the 1332 ILLÁN-GALA ET AL . Neuropsychiatric Inventory (NPI) and the total score of the NPI (sum of matter hyperintensities volumes, as implemented in CAT12. Finally, we the frequency-by-severity scores) as measures of behavioral changes calculated the mean frontotemporal cortical thickness by averaging independent of diagnostic criteria recorded in all centers.34 Because the mean cortical thickness of all the frontal and temporal regions in the frequency-by-severity product of the NPI could be influenced by the Desikan atlas. the characteristics of the informant (gender, values, and so on) we also considered the count of abnormal behaviors (herein, NPI behav- ior count) as a measure of behavioral burden less dependent on the 2.7 Group comparison of cortical thickness informants’ characteristics. Finally, we included the Geriatric Depres- sion Scale (GDS) to quantify symptoms of depression at baseline.35 We performed separate group-comparisons analyses in men and women following the recommendation for the study of sex-related dif- ferences in neurodegenerative dementias.8 We introduced age, edu- cation, and MRI scanner as covariates in between-group (bvFTD vs 2.5 Progression and survival analyses healthy controls) cortical thickness comparisons. For group compar- isons involving healthy controls, we considered a significant statisti- We used linear mixed-effects analyses controlling for age, sex, base- cal threshold of P < .05, family-wise error (FWE)–corrected, using an line frontotemporal atrophy, and genetic status (presence or absence extent threshold of the expected vertices per cluster. To assess the of an FTLD-related mutation) to predict a longitudinal change as mea- sured by CDRsb (256 measurements:183 at baseline and 73 at year consistency of results across cohorts, we performed secondary anal- 1) and CDR plus NACC FTLD (212 measurements: 145 at baseline yses in UCSF and CATFI subgroups, separately. Due to the relatively and 67 at year 1). We included genetic status (mutation carrier vs small sample size in the CATFI cohort, we applied a less-restrictive sta- no carrier) as a covariate because mutations have been related to tistical threshold in the secondary analyses of this subgroup (P < .05, with false discovery rate [FDR] correction and an extent threshold of the rate of clinical decline in FTLD syndromes.36,37 We used a com- the expected vertices per cluster). Next, we calculated Cohen’s d as pound symmetry covariance matrix in all linear-mixed models, and we a measure of the effect size of cortical thinning. Cohen’s d takes into included random intercepts to account for the effect of baseline values. account sample size P-values and allows the comparison of the effect A term for biomarker by time interaction was used to study the associ- size between groups of different sizes. Effect size computation was ation between the baseline biomarker level and the outcome slope (eg, restricted to cortical regions showing statistically significant differ- CDR+NACC/FTLD-SB) over time. ences (according to predefined statistical thresholds) between bvFTD For survival analyses, we designed a Cox regression model to and healthy controls. Next, we determined the net effect size by sub- account for known factors affecting survival in bvFTD. Patients alive tracting the observed effect size in men to the one observed in women. at analysis were censored at the date of their last visit. We first intro- duced age at diagnosis, genetic status, disease severity (CDRsb), and education as covariates.36,38 Next, in a second Cox regression model, 2.8 Interaction between group and sex we added sex to test if the previous model was significantly improved. In a separate multiple regression model, we tested sex x group (bvFTD vs healthy control) interaction on cortical thickness in all the partici- 2.6 MRI pants (Supplementary Methods and Figure S1). 2.6.1 MRI acquisition 2.9 Cognitive and behavioral reserve analyses The images were acquired on scanners from seven different manu- facturers using different imaging protocols (Table S1). Magnetic field Next, we investigated the existence of sex differences in cognitive and strength varied between 1.5 T (n = 48 scans), 3.0 T (n = 385 scans), and behavioral reserve by following the residuals approach.19,24 First, we 4.0 T (n = 27 scans). For additional information on the MRI quality con- fitted different linear regression models with each of the cognitive z- trol, please refer to Supplementary Methods. scores and NPI scores as the response variable and age, education, and frontotemporal cortical thickness as predictors. This model provided an individual’s predicted cognition/behavior scores for a certain level 2.6.2 MRI processing of atrophy. Here, we operationalized “reserve” as having better func- tion than is predicted by the cortical thickness (a proxy of pathology) MRIs were processed with the CAT12 toolbox within SPM12 (running and used each individual’s residual as a proxy of their reserve. This pro- in MATLAB r2019b).39,40 It is notable that we performed cortical cedure was done for each cognitive composite and each NPI frequency thickness analyses instead of volume methods, because the cortical x severity scores as the response variable. Robust Welch t tests with thickness is not significantly affected by total intracranial volume, a bias-corrected accelerated bootstrapping on 1000 samples were then measure that is known to differ between men and women (Supple- performed to test for a difference in residuals (ie, reserve) between mentary Methods).41 We obtained total intracranial volume and white women and men. ILLÁN-GALA ET AL . 1333 2.10 Other statistical analyses associated with a longitudinal increase in CDR plus NACC FTLD (esti- mate increase of 2.5 points per year, 95% CI 0.23 to 4.8). Sex was Data were explored for normality using the Kolmogorov-Smirnov test. not an independent predictor of the rate of CDR plus NACC FTLD Between-group differences in baseline characteristics were assessed change (Time x sex interaction, P = .807), or changes of MMSE or using t test, analysis of variance (ANOVA), or Kruskal-Wallis test for CDRsb. continuous variables and chi-square test for categorical data. We com- Overall survival from diagnosis in the bvFTD group was similar com- puted the effect size (partial Eta Squares) in analysis of covariance pared to previous studies (mean survival of 3.5 years 95% CI 3.0 to (ANCOVA) analyses of the difference in cognitive composites between 4.0). Figure 3B shows Kaplan-Meier survival curves in men and women bvFTD participants and healthy controls after adjusting by age and with bvFTD. Mean survival in women was 2.7 years (95% CI 2.2 to education (Table S5). To obtain robust confidence intervals, we esti- 3.2) and mean survival in men 3.8 years (95% CI 3.2 to 4.3) from diag- mated 95% confidence intervals (95% CIs) with bootstrapping-based nosis. The difference in survival times between men and women with methods (bias-corrected and accelerated for 1000 samples). Statistical bvFTD was not statistically significant (Log Rank [Mantel Cox]: chi- significance for all tests was set at 5% (α = .05), and all statistical tests square = 2.982, P = .082). To take into account other factors that are were two sided. All analyses were performed using SPSS 25 (Armonk, known to affect survival in bvFTD, we designed different Cox regres- NY: IBM Corp.). sion models adjusted by age, education, disease severity at diagnosis (as measured by CDR plus NACC FTLD), and mutation status. We found that years of education (1.087, 95% CI 1.016 to 1.163, and P = .016) and 3 RESULTS CDR plus NACC FTLD (1.099, 95% CI 1.027 to 1.176, and P = .006) at diagnosis were independently associated with survival. However, the Age at MRI and education were similar between men and women addition of sex did not improve the model and was not an independent with bvFTD and between participants with bvFTD and healthy con- predictor of survival (Table S8). trols (Table 1). Of note, age at disease onset and measures of disease Men with bvFTD showed reduced cortical thickness in frontotem- severity (CDRsb and CDR plus NACC FTLD) were similar between men poral structures when compared to healthy controls of the same sex and women with bvFTD (Table 1). An FTLD mutation was identified (Figure 4A). Similarly, women with bvFTD showed reduced cortical in 22% of bvFTD participants (28 C9orf72, 12 GRN, 6 MAPT, and 2 thickness in frontotemporal structures when compared to healthy con- TARDBP). FTLD was confirmed on autopsy in 32% (19 FTLD-tau [includ- trols of the same sex (Figure 4B). However, we observed larger effect ing 3 MAPT], 43 FTLD-TAR DNA-binding protein 43 [FTLD-TDP; includ- sizes than in men, as illustrated in Figure 4C. Of note, we found similar ing 15 C9orf72 and 6 GRN], and 8 FTLD-FUS. But the proportion of results when we performed separate analyses in each cohort (Figure bvFTD participants with pathology-proven FTLD or FTLD mutations S1). Next, we directly compared cortical thickness between men and was similar between men and women (Table 1). Within bvFTD partici- women with bvFTD and found that women with bvFTD had reduced pants with pathology-proven FTLD, we observed a higher frequency of cortical thickness compared to men with bvFTD in several frontotem- FTLD-Fused in Sarcoma (FTLD-TDP; Table 1). Particularly, FTLD-TDP poral regions of both hemispheres (Figure S2). It is important to note type B was more frequent in women than in men with FTLD (Table S9). that we found a significant group (bvFTD or healthy control) by sex Women and men showed a similar frequency of bvFTD diagnostic interaction in frontotemporal structures, indicating that, at diagnosis, features (Figure 2A). In addition, on average, men and women showed women with bvFTD had lost more frontotemporal gray matter than the same number of features of the diagnostic criteria at diagnosis men with bvFTD after accounting for age, education, MMSE, and MRI (Mann-Whitney U = 3.404; P = .737). Disinhibition and delusions were scan (Figure 4D and Figure S3). more frequently reported (present vs absent) in women than in men To more directly evaluate potential behavioral and cognitive reserve with bvFTD (Table 1). However, we only observed higher frequency x differences by sex, we took the residuals approach. We first fitted severity scores on women in delusions (Figure 2B). The NPI total score, multiple linear regression models with NPI and cognitive z-scores as the NPI behavior count, and the GDS score were similar between men the response variables and with age, education, and cortical thickness and women with bvFTD (Table 1). as predictors (Table S6). For illustrative purposes, Figure 5 shows the Next, we compared the cognitive profile between men and women relationship between apathy NPI score, executive z-score, and fron- with bvFTD. As expected, bvFTD participants were more impaired totemporal cortical thickness (Figure 5A and D). The sex-comparison than age- and education-matched healthy controls in all four domains. of the residuals obtained in these linear regression models suggests Detailed information on full the neuropsychologic testing for each that women had lower-than-expected apathy (t(348) = 2.5, P = .012, cohort can be found in Table S3 to S4. However, no differences were Figure 5B) and higher-than-expected executive z-score (t(352) = −2.1, noted between men and women in any cognitive domain in the bvFTD P = .039, Figure 5D) than men, given their frontotemporal atrophy. group (Figure 2C). Of note, the MMSE scores were lower in women We also observed lower-than-expected sleep and appetite change with bvFTD compared to men. However, the effect size of the differ- NPI scores in women compared to men (t(361) = 3.3, P = .001; and ence was small (Table S5). t(333) = 2.3, P = .020, respectively). Finally, we observed lower- As shown in Figure 3A, disease progression over 1 year of follow- than-expected NPI total scores and NPI behavior counts in women up was similar between men and women with bvFTD. Time was (t(349) = 2.2, P = .026 and t(349) = 2.2, P = .029, respectively), but we 1334 ILLÁN-GALA ET AL . TA B L E 1 Characteristics of the participants bvFTD HC bvFTD-men bvFTD-women All bvFTD HC-men HC-women All HC n = 149 (69%) n = 67 (31%) n = 216 n = 131 (58%) n = 94 (42%) n = 225 Demographics and disease severity Age at MRI (y) 63.3 (10.4) 63.5 (9.6) 63.3 (10.1) 62.4 (14) 61 (10.5) 61.8 (12.6) Age at disease onset (y) 57.7 (10.9) 58.4 (9.8) 57.9 (10.5) – – – Education (y) 14.9 (4.2) 14.8 (4.5) 14.8 (4.3) 15.5 (3.6) 15.1 (4.1) 15.4 (3.8) † CDRsb* 6.6 (3.4) 6.7 (3.5) 6.6 (3.4) 0.0 (0.1) 0.0 (0.1) 0.0 (0.1)† CDR plus NACC FTLD‡ , 8.7 (4.1) 9.0 (4.1) 8.8 (4.1)† 0.0 (0.1) 0.0 (0.1) 0.0 (0.1)† Cohort, No. (UCSF/CATFI) 99/50 50/17 149/67 111/20 57/37 168/57 Diagnostic certainty Time of follow-up (y) 2.1 (2.0) 1.4 (1.4) 1.9 (1.9) – – – § FTLD mutation, % 19 30 22 – – – FTLD confirmed on autopsy, % 36 25 32 – – – Neuropathological diagnosis,¶ 16 FTLD-Tau 3 FTLD-Tau 19 FTLD-Tau – – – No. of major neuropathological 29 FTLD-TDP 14 FTLD-TDP 43 FTLD-TDP categories 8 FTLD-FUS 0 FTLD-FUS 8 FTLD-FUS Behavioral measures Number of bvFTD diagnostic 4.6 (1.2) 4.5 (1.4) 4.6 (1.2)† 0 (0) 0 (0) 0 (0)† features* Delusions, % 11# 26# 16† 0 0 0† Hallucinations, % 8 13 9† 0 0 0† Dysphoria/Aggression, % 56 56 56† 5 6 6† † Anxiety, % 39 34 37 2 2 2† Euphoria/Elation, % 47 53 49† 0 0 0† † Apathy/Indifference, % 93 89 92 2 1 1† Disinhibition, % 78# 92# 83† 0 0 0† Irritability/Lability, % 47 48 47† 2 3 3† † Aberrant motor behavior, % 64 72 66 0 0 0† Sleep changes, % 48 42 46† 5 0 3† † Appetite changes, % 81 76 80 7 5 3† NPI total score 40.3 (21.5) 41.6 (20.3) 40.7 (21.0)† 2.7 (5.2) 2.3 (4.5) 2.5 (4.9) NPI behavior count 6.2 (2.4) 6.5 (2.4) 6.3 (2.4)† 0.2 (0.7) 0.1 (0.6) 0.2 (0.6) Geriatric depression scale 9.3 (6.8) 9.7 (7.2) 9.4 (6.9)† 3.9 (3.7) 4.9 (5.4) 4.3 (4.5) Cognitive measures MMSE 23.8 (6.3)# 21.4 (7.4)# 23.1 (6.7)† 29.1 (0.9) 28.8 (1.3) 29.0 (1.1)† † zMemory −0.4 (0.8) −0.6 (0.9) −0.5 (0.8) 0.6 (0.5) 0.7 (0.5) 0.6 (0.5)† zLanguage −0.6 (0.8) −0.9 (0.9) −0.7 (0.9)† 0.6 (0.4) 0.6 (0.4) 0.6 (0.4)† † zExecutive functions −0.7 (0.7) −0.7 (0.9) −0.7 (0.7) 0.5 (0.5) 0.5 (0.6) 0.5 (0.5)† zVisuospatial −0.4 (1.1) −0.7 (1.7) −0.5 (1.3)† 0.3 (0.5) 0.3 (0.4) 0.3 (0.4)† Imaging characteristics Quality of MRI,** 80 (7) 86 (1) 82 (6)† 84 (3) 84 (2) 84 (3)† (Continues) ILLÁN-GALA ET AL . 1335 TA B L E 1 (Continued) bvFTD HC bvFTD-men bvFTD-women All bvFTD HC-men HC-women All HC n = 149 (69%) n = 67 (31%) n = 216 n = 131 (58%) n = 94 (42%) n = 225 Total intracranial volume, mL 1487# 1335# 1440 (153) 1503# 1328# 1430 (135) White matter hyperintensities, mL 6.0 (10.5) 4.2 (3.4) 5.5 (8.9)† 2.8 (4.0)# 2.1 (3.2)# 2.5 (3.7)† NPI measures were available in 389 (88%) of participants. All cognitive measures were available in at least 85% of participants. Abbreviations: bvFTD, behavioral variant of frontotemporal dementia; CATFI, Catalan Frontotemporal Dementia Initiative; CDR plus NACC FTLD, CDR Dementia Staging Instrument PLUS National Alzheimer’s Coordinating Center Behavior and Language Domains; CDRsb, Clinical Dementia Rating scale sum of boxes; MMSE, Mini-Mental State Examination; MRI, magnetic resonance image; NPI, Neuropsychiatric Inventory; UCSF, University of California San Fran- cisco. *CDRsb was available in 183 (85%) bvFTD participants and 188 (84%) of healthy controls. † Difference between all bvFTD group and HC group (P < .05). ‡ CDR plus NACC FTLD was available in 145 (67%) of bvFTD participants and 127 (56%) of healthy controls. § A total of 48 FTLD-related mutations were identified: 28 C9orf72 (15 with neuropathology available), 12 GRN (six with neuropathology available), six MAPT (three with neuropathology available) and two TARDBP. ¶ The proportion of FTLD-TDP was higher in women with FTLD diagnosis than in men (82% vs 55%, P < .05, Bonferroni adjusted). For additional details on the neuropathological findings please refer to Table S9. # Difference between men and women (P < .05, Bonferroni adjusted). **Imaging quality rating obtained from CAT12 software (percentages closer to 100% indicating a better quality of MRI). did not observe a significant difference in the residuals for the rest of in women because of higher misdiagnosis with psychiatric diseases the NPI and cognitive scores (Table S7). compared to men.16 However, we found similar age at disease onset and similar cognitive performance between men and women with bvFTD in both cohorts. The sole difference for cognitive measures was 4 DISCUSSION found for the MMSE (discussed below). It is notable that both disease severity measured by CDR plus NACC FTLD, clinical decline, and sur- In this multicenter study we explored the impact of biological sex on vival were similar between men and women. Taken together our results the clinical presentation, prognosis, and cerebral structure of bvFTD, suggest that, in our sample, men and women with bvFTD were diag- a neurodegenerative syndrome characterized by the selective vulner- nosed at a similar disease stage and showed a similar trajectory after ability of frontal and temporal lobes.3,4 We found that women with diagnosis. Additional prospective studies are needed to more precisely bvFTD showed a higher burden of frontal atrophy at diagnosis com- define the role of sex in the early identification of FTLD syndromes pared to men despite very subtle differences in cognitive and func- (ie, need for sex-adjusted norms) and the effect of sex on prognosis. tional impairment, progression, and survival. Next, we followed the For this purpose, large multicenter cohorts of both sporadic FTLD residuals approach to investigate sex-related differences in reserve. syndromes and FTLD-related mutation carriers would be of particular We found that for a given amount of atrophy, women performed interest.42,43 better-than-expected on executive function and displayed fewer gen- Despite similar diagnostic features at diagnosis, men and women eral behavioral changes, and particularly fewer changes in apathy, displayed some subtle differences. We replicated previous obser- sleep, and appetite than men. This observation suggests that women vations of lower MMSE scores in women with bvFTD.2 However, with bvFTD have higher behavioral and executive reserve than men. the effect size of the observed difference in MMSE scores was small Although previous works have reported significant sex differences in (inferior to 0.2). Because the MMSE combines language, memory, and Alzheimer’s disease (or AD), this is the first study investigating this visuospatial tasks, we hypothesize that this combination may have an important topic in bvFTD, one of the first causes of early onset demen- additive effect to unveil small differences in general cognition between tia and the first clinical presentation of FTLD. men and women with bvFTD that were not statistically significant We found that at diagnosis, women with bvFTD had significantly when assessing independent tests or cognitive domains. We also found more atrophy than men, both in the whole sample and when segre- that a slightly higher frequency of delusions and disinhibition in women gating participants in two cohorts. It is important to note that we first with bvFTD.2 However, only delusions (and not disinhibition) showed performed stratified analyses by sex as recommended in recent con- a higher frequency by severity score in women. This finding could be sensus paper for the investigation of sex differences.8 By doing this, we related to the higher frequency of FTLD-TDP in women with bvFTD avoided the potential mitigation of sex-related differences in cerebral in our sample. However, a recent study evidenced a higher frequency structure when introducing sex as a covariate in structural analyses.8 of hallucinations (and not delusions) in FTLD-TDP compared to other We next investigated if the observed sex differences in cortical thin- FTLD subtypes.44 Moreover, a higher frequency of delusions has also ning could be related to disease severity at diagnosis. Indeed, previous been observed in women with AD.45 Although sex may play a role in the observations suggested that the diagnosis of bvFTD could be delayed higher frequency of delusions, more studies are needed to explore the 1336 ILLÁN-GALA ET AL . F I G U R E 2 Women and men with bvFTD have similar behavioral and cognitive impairment at diagnosis. Behavioral and cognitive characteristics of bvFTD participants by sex. (A) Percentage of bvFTD participants with features of diagnostic criteria at diagnosis. Of note, no differences between the frequency of bvFTD diagnostic features were noted between men and women. (B) NPI frequency x severity scores by sex. (C) Cognitive composites by sex. * Different between men and women (P < .05, Bonferroni adjusted); # Superior to both bvFTD-Men and bvFTD-Women (P < .05). Error bars represent 95% confidence intervals. Abbreviations: bvFTD, behavioral variant of frontotemporal dementia; NPI, neuropsychiatric inventory F I G U R E 3 Men and women with bvFTD show similar progression and survival. (A) CDR plus NACC FTLD estimates were obtained from linear mixed-effects models adjusted for age and sex. For illustrative purposes, we display the results for men and women (blue and pink lines, respectively). Error bars represent 95% confidence intervals. (B) Kaplan-Meier Survival Curves for men and women with bvFTD. Abbreviations: bvFTD, behavioral variant of frontotemporal dementia ILLÁN-GALA ET AL . 1337 F I G U R E 4 Women with bvFTD have higher atrophy than men at diagnosis. (A) Group difference in cortical thickness between men with bvFTD (bvFTD-men) and healthy controls (HC) in the whole cohort. Only the effect sizes of regions with statistically significant results (family-wise error [few], P < .05) are shown. (B) Group difference in cortical thickness between women with bvFTD (bvFTD-women) and HC in the whole cohort. Only the effect sizes of regions with statistically significant results (FWE, P < .05) are shown. (C) We show the net effect size illustrating the greater degree of gray matter loss in women with bvFTD compared to men. The net effect size was obtained by subtracting the effect size map showed in panel B (effect size of cortical thinning in women with bvFTD compared to healthy controls of the same sex) from the effect showed in panel A (effect size of cortical thinning in men with bvFTD compared to healthy controls of the same sex). Only the effect sizes of regions with statistically significant results (FWE, P < .05) are shown. Higher values represent a higher net effect size favoring bvFTD-women (ie, women have a higher effect size for the observed cortical thinning). (D) Colored regions denote areas where we observed a significant group by sex interaction (in green FDR, P < .05, and in red FWE, P < .05). In this model, age at MRI, education, MMSE, and MRI scan were introduced as covariates. Right hemisphere surfaces are shown on the right half of each panel. Abbreviations: bvFTD, behavioral variant of frontotemporal dementia; HC, healthy control; FDR, false discovery rate; FWE, family-wise error; MMSE, Mini-Mental State Examination; MRI, magnetic resonance imaging neural and pathological substrate of delusions in neurodegenerative expected scores on executive function and less than expected apathy, dementias. sleep, and appetite behavior changes. We interpret these results as The inclusion of a large group of age- and sex-matched healthy con- indirect evidence of higher brain reserve in women. However, the con- trols allowed us to study the specific contribution of the diagnostic cept of reserve encompasses different dynamic processes that could group (healthy control vs bvFTD), sex, and the interaction between also play a significant role in the observed differences.19 It should be group and sex on frontotemporal thickness. We found a significant noted that many of these processes cannot be captured with structural sex by group interaction in frontotemporal regions after accounting MRI. Thus, other functional biomarkers and imaging techniques are for the effect of age, education, MMSE, and MRI scan. This is a criti- needed to investigate the specific mechanisms underlying sex-related cal finding of the study, indicating that women have lost more cortical differences in cognitive reserve in bvFTD and other FTLD syndromes thickness at the moment of bvFTD diagnosis than men after account- characterized by prominent frontotemporal neurodegeneration. ing for potential confounding factors. These results, together with the Regardless of the specific mechanism, our results open a new win- observed null to small effect sizes in the cognitive features and lon- dow to the study of executive and behavioral reserve in neurodegen- gitudinal decline between men and women with bvFTD, suggested erative dementias and add to previous evidence in FTLD indicating that women could better tolerate frontotemporal pathology. To fur- the importance of other sociobehavioral proxies of reserve to improve ther test this hypothesis, we followed the residuals approach as a proxy the diagnosis and prognosis of FTLD syndromes. The concept of brain to investigate potential sex differences in reserve. We showed that, reserve emerged from the observation of a mismatch between cere- for a given amount of atrophy, women with bvFTD had better-than- bral pathology and cognitive changes in AD.23,46 Previous studies in 1338 ILLÁN-GALA ET AL . F I G U R E 5 Women with bvFTD have higher behavioral and cognitive reserve than men. (A) Scatter plot between NPI apathy score (higher values represent more apathy) and frontotemporal cortical thickness (negative values represent higher atrophy). (B) Comparison of predicted residuals for NPI apathy score between men and women. Residuals were obtained in a general linear model with NPI apathy score as response variable and age, education, and frontotemporal cortical thickness as predictors. (C) Scatter plots between executive z-score (negative values represent worst performance) and frontotemporal cortical thickness (negative values represent higher atrophy). (D) Comparison of predicted residuals for executive z-score between men and women. Residuals were obtained in a general linear model with NPI apathy score as response variable and age, education, and frontotemporal cortical thickness as predictors. * P < .05 (Robust Welch t test with bias-corrected accelerated bootstrapping on 1000 samples). Abbreviations: bvFTD, behavioral variant of frontotemporal dementia; NPI, Neuropsychiatric Inventory AD have identified the anterior cingulum as an important region for on established criteria, and participants with decreased diagnostic resilience.47 Because an anterior pattern of neurodegeneration char- certainty of underlying FTLD were excluded to avoid the inclusion of acterizes the bvFTD syndrome, the mechanisms and structures sup- participants with the so-called “phenocopy syndrome.” Of note, the porting reserve in bvFTD may be different than in AD, a disease survival of included bvFTD participants was similar to that reported in characterized by a posterior pattern of neurodegeneration. other large studies after excluding non-neurodegenerative causes of bvFTD.48 Second, in this study, we focused on biological sex, but more studies are needed to understand the role of gender (understood as 4.1 Strengths and limitations a social construct) in the diagnosis of bvFTD and other neurodegen- erative dementias. Finally, the rating of behavioral change was based This study has some limitations. First, most participants in this study on questionnaires administered to the caregiver, and it is possible that did not have pathological confirmation of neurodegenerative disease. some of the informants’ characteristics (including biological sex) could However, the diagnosis of bvFTD was made by expert clinicians based have meaningfully influenced the ratings. However, we also obtained ILLÁN-GALA ET AL . 1339 significant results with the NPI behavior count, a measure that is less A1 Markers of synaptopathy in neurodegenerative disease) and partici- dependent on the subjective assessment of symptom severity. More- pated in advisory boards from Fujirebio-Europe and Roche Diagnostics over, our main results were not limited to behavioral measures, since and received speaker honoraria from Fujirebio-Europe, Nutricia, we also found evidence for higher executive reserve in women. Future and Krka Farmacéutica S.L. JF reports no disclosures; RB reports no studies should examine the effect of biological sex and gender of the disclosures; JC reports no disclosures; MFI reports no disclosures; informant on NPI ratings and assess the impact of using clinician-based AB-S reports no disclosures; AL reports no disclosures; LTG reports instruments (such as the NPI-clinician version).49 no disclosures; KP reports no disclosures; KPR reports no disclosures; JHK reports no disclosures. GDR reports receiving grants from the NIH and Tau Consortium during the conduct of the study; grants from 5 CONCLUSION Avid Radiopharmaceuticals, Eli Lilly and Company, GE Healthcare, and Life Molecular Imaging outside the submitted work; and personal fees We describe striking differences in frontotemporal atrophy between from GE Healthcare, Axon Neurosciences, Merck, Eisai, Roche, and men and women with bvFTD despite showing less evident differences Genentech outside the submitted work. AB receives research support in baseline clinical characteristics and trajectories. It is notable that from NIH, the Tau Research Consortium, the Association for Fron- our results also support the existence of sex-related differences in the totemporal Degeneration, Bluefield Project to Cure Frontotemporal ability to cope with frontotemporal neurodegeneration and add to the Dementia, Corticobasal Degeneration Solutions, the Alzheimer’s Drug growing evidence highlighting the importance of considering sex for Discovery Foundation, and the Alzheimer’s Association. He has served the design of research studies and clinical trials in neurodegenerative as a consultant for Aeton, Abbvie, Alector, Amgen, Arkuda, Arvinas, dementias. Asceneuron, Ionis, Lundbeck, Novartis, Passage BIO, Samumed, Third Rock, Toyama, and UCB; and received research support from Avid, Bio- ACKNOWLEDGMENTS gen, BMS, C2N, Cortice, Eli Lilly, Forum, Genentech, Janssen, Novartis, The authors thank the patients and their relatives for their support for Pfizer, Roche, and TauRx. WWS reports no disclosures; VES reports no this study. The Catalan frontotemporal initiative (CATFI) was funded by disclosures; MGT reports no disclosures; BLM reports no disclosures. the Health Department of the Government of Catalonia (grant PERIS RSV has served at scientific advisory boards from Wave pharmaceu- SLT002/16/00408 to AL and RS-V). This work was supported by NIH ticals and Ionis-Biogen. DCP reports no disclosures. AL has served grants AG019724, AG032306, AG045390, NS092089, AG045333, at scientific advisory boards from Fujirebio-Europe, Nutricia, Biogen, AG056749, AG062422, AG061253 (AMS), and AG058752 (KBC), the and Roche Diagnostics and has filed a patent application of synaptic Larry. L. Hillblom Foundation (2018-A-025-FEL (AMS), 2017-A-004- markers in neurodegenerative diseases. HJR reports no disclosures. FEL (KBC), 2014-A-004-NET(JHK)), and the Association for Fron- totemporal Degeneration. LTG is supported by NIH K24AG053435. ORCID DCP funded by R01AG059794. This study was also supported by the Ignacio Illán-Gala https://orcid.org/0000-0002-5418-2052 Fondo de Investigaciones Sanitario (FIS), Instituto de Salud Carlos REFERENCES III (PI14/01126 and PI17/01019 to JF, PI13/01532 and PI16/01825 1. Elahi FM, Miller BL. A clinicopathological approach to the diagnosis of to RB, PI18/00335 to MCI, INT19/00016 and PI18/00435 to DA, dementia. Nat Rev Neurol. 2017;13:457-476. PI15/01618 to RR, PI14/1561 and PI18/00326 to J.C., AC14/00013 2. Ranasinghe KG, Rankin KP, Lobach IV, et al. Cognition and neu- to RS-V, and PI17/01896 to AL), the CIBERNED program (Program 1, ropsychiatry in behavioral variant frontotemporal dementia by dis- Alzheimer Disease to AL and SIGNAL study, www.signalstudy.es), the ease stage. Neurology. 2016;86:600-610. 3. Perry DC, Brown JA, Possin KL, et al. Clinicopathological cor- “Marató TV3” grant (20141210 to JF, 044412 to RB, and 20143810 relations in behavioural variant frontotemporal dementia. Brain. to RS-V), by Generalitat de Catalunya (2014SGR-0235 to AL, PERIS 2017;140:3329-3345. SLT006/17/125 to DA, and SLT006/17/00119 to JF), by BBVA founda- 4. Seeley WW, Crawford R, Rascovsky K, et al. Frontal paralimbic net- tion (grant to AL), and by a grant from the Fundació Bancaria La Caixa work atrophy in very mild behavioral variant frontotemporal demen- tia. Arch Neurol. 2008;65:249-255. to RB. II-G and AB-S are supported by the Global Brain Health Institute 5. Dong A, Toledo JB, Honnorat N, et al. Heterogeneity of neuroanatom- (Atlantic Fellows for Equity in Brain Health) and II-G was supported the ical patterns in prodromal Alzheimer’s disease: links to cognition, pro- Rio Hortega grant (CM17/00074) from Instituto de Salud Carlos III. SB- gression and biomarkers. Brain. 2017;140:735-747. E is the recipient of Emili Letang post-residency research grant from 6. Edmonds EC, Delano-Wood L, Clark LR, et al. Susceptibility of the con- ventional criteria for mild cognitive impairment to false-positive diag- Hospital Clínic de Barcelona. MFI acknowledges support from the Sys- nostic errors. Alzheimers Dement. 2015;11:415-424. ley D’Ornano and Jérôme Lejeune Foundations. 7. Ranasinghe KG, Rankin KP, Pressman PS, et al. Distinct subtypes of behavioral variant frontotemporal dementia based on patterns of net- DECLARATION OF INTEREST work degeneration. JAMA Neurol. 2016;73:1078. 8. for the Women’s Brain Project and the Alzheimer Precision Medicine II-G reports no disclosures; KBC reports no disclosures; SBE reports no Initiative, Ferretti MT, Iulita MF, Cavedo E, Chiesa PA, Schumacher disclosures; EA-U reports no disclosures; AW reports no disclosures; Dimech A, et al, for the Women’s Brain Project and the Alzheimer Pre- YC reports no disclosures; SYG reports no disclosures; AMS reports no cision Medicine Initiative. Sex differences in Alzheimer disease — the disclosures. DA declares a filed patent application (WO2019175379 gateway to precision medicine. Nat Rev Neurol. 2018;14:457-469. 1340 ILLÁN-GALA ET AL . 9. Frisoni GB, Boccardi M, Barkhof F, et al. Strategic roadmap for an early 29. Morris JC. The Clinical Dementia Rating (CDR): current version and diagnosis of Alzheimer’s disease based on biomarkers. The Lancet Neu- scoring rules. Neurology. 1993;43:2412-2414. rology. 2017;16:661-676. 30. Miyagawa T, Brushaber D, Syrjanen J, et al. Utility of the global CDR 10. Snyder HM, Asthana S, Bain L, et al. Sex biology contributions plus NACC FTLD rating and development of scoring rules: data from to vulnerability to Alzheimer’s disease: a think tank convened by the ARTFL/LEFFTDS Consortium. Alzheimers Dement. 2020;16:106- the Women’s Alzheimer’s Research Initiative. Alzheimers Dement. 117. 2016;12:1186-1196. 31. Folstein MF, Folstein SE, McHugh PR. Mini-mental state. J Psychiatr 11. Nebel RA, Aggarwal NT, Barnes LL, et al. Understanding the impact Res. 1975;12:189-198. of sex and gender in Alzheimer’s disease: a call to action. Alzheimers 32. Ossenkoppele R, Pijnenburg YAL, Perry DC, et al. The Dement. 2018;14:1171-1183. behavioural/dysexecutive variant of Alzheimer’s disease: clinical, 12. Mauvais-Jarvis F, Bairey Merz N, Barnes PJ, et al. Sex and gen- neuroimaging and pathological features. Brain. 2015;138:2732-2749. der: modifiers of health, disease, and medicine. Lancet North Am Ed. 33. Illan-Gala I, Pegueroles J, Montal V, et al. APP-derived peptides reflect 2020;396:565-582. neurodegeneration in frontotemporal dementia. Ann Clin Transl Neurol. 13. Ferretti MT, Martinkova J, Biskup E, et al. Sex and gender differ- 2019;6:2518-2530. ences in Alzheimer’s disease: current challenges and implications for 34. Cummings JL, Mega M, Gray K, Rosenberg-Thompson S, Carusi DA, clinical practice: position paper of the Dementia and Cognitive Dis- Gornbein J. The neuropsychiatric inventory: comprehensive assess- orders Panel of the European Academy of Neurology. Eur J Neurol. ment of psychopathology in dementia. Neurology. 1994;44:2308-2308. 2020;27:928-943. 35. Jongenelis K, Pot AM, Eisses AMH, et al. Diagnostic accuracy of the 14. Podcasy JL, Epperson CN. Considering sex and gender in Alzheimer original 30-item and shortened versions of the Geriatric Depression disease and other dementias. Clin Res. 2016;18:10. Scale in nursing home patients. Int J Geriat Psychiatry. 2005;20:1067- 15. Gossink FT, Dols A, Kerssens CJ, et al. Psychiatric diagnoses under- 1074. lying the phenocopy syndrome of behavioural variant frontotemporal 36. Caswell C, McMillan CT, Xie SX, et al. Genetic predictors of sur- dementia. J Neurol Neurosurg Psychiatry. 2016;87(1):64-68. vival in behavioral variant frontotemporal degeneration. Neurology. 16. Woolley JD, Khan BK, Murthy NK, Miller BL, Rankin KP. The diag- 2019;93:e1707-e1714. nostic challenge of psychiatric symptoms in neurodegenerative dis- 37. Agarwal S, Ahmed RM, D’Mello M, et al. Predictors of survival and pro- ease: rates of and risk factors for prior psychiatric diagnosis in patients gression in behavioural variant frontotemporal dementia. Eur J Neurol. with early neurodegenerative disease. J Clin Psychiatry. 2011;72:126- 2019;26:774-779. 133. 38. Massimo L, Xie SX, Rennert L, et al. Occupational attainment influ- 17. Onyike CU, Diehl-Schmid J. The epidemiology of frontotemporal ences longitudinal decline in behavioral variant frontotemporal degen- dementia. Int Rev Psychiatry. 2013;25:130-137. eration. Brain Imaging and Behavior. 2019;13:293-301. 18. Coyle-Gilchrist ITS, Dick KM, Patterson K, et al. Prevalence, character- 39. Dahnke R, Yotter RA, Gaser C. Cortical thickness and central surface istics, and survival of frontotemporal lobar degeneration syndromes. estimation. Neuroimage. 2013;65:336-348. Neurology. 2016;86:1736-1743. 40. Seiger R, Ganger S, Kranz GS, Hahn A, Lanzenberger R. Cortical thick- 19. Stern Y, Arenaza-Urquijo EM, Bartrés-Faz D, et al. Whitepaper: defin- ness estimations of FreeSurfer and the CAT12 toolbox in patients ing and investigating cognitive reserve, brain reserve, and brain main- with Alzheimer’s disease and healthy controls: cortical thickness of tenance. Alzheimers Dement. 2018:S1552526018334915. https://doi. FreeSurfer and CAT12. J Neuroimaging. 2018;28:515-523. org/10.1016/j.jalz.2018.07.219 41. Schwarz CG, Gunter JL, Wiste HJ, et al. A large-scale comparison of 20. Spinelli EG, Mandelli ML, Miller ZA, et al. Typical and atypical pathol- cortical thickness and volume methods for measuring Alzheimer’s dis- ogy in primary progressive aphasia variants. Ann Neurol. 2017;81:430- ease severity. NeuroImage: Clinical. 2016;11:802-812. 443. 42. Rosen HJ, Boeve BF, Boxer AL. Tracking disease progression in familial 21. Ruigrok ANV, Salimi-Khorshidi G, Lai M-C, et al. A meta-analysis and sporadic frontotemporal lobar degeneration: recent findings from of sex differences in human brain structure. Neurosci Biobehav Rev. ARTFL and LEFFTDS. Alzheimers Dement. 2020;16:71-78. 2014;39:34-50. 43. Tavares TP, Mitchell DGV, Coleman KK, et al. Early symptoms in symp- 22. Lotze M, Domin M, Gerlach FH, et al. Novel findings from 2,838 tomatic and preclinical genetic frontotemporal lobar degeneration. J adult brains on sex differences in gray matter brain volume. Sci Rep. Neurol Neurosurg Psychiatry. 2020;91:975-984. 2019;9:1671. 44. Scarioni M, Gami-Patel P, Timar Y, et al. Frontotemporal dementia: 23. Arenaza-Urquijo EM, Vemuri P. Resistance vs resilience to Alzheimer correlations between psychiatric symptoms and pathology. Ann Neu- disease: clarifying terminology for preclinical studies. Neurology. rol. 2020. ana.25739. 2018;90:695-703. 45. Karttunen K, Karppi P, Hiltunen A, et al. Neuropsychiatric symptoms 24. Digma LA, Madsen JR, Rissman RA, et al. Women can bear a bigger and Quality of Life in patients with very mild and mild Alzheimer’s burden: ante- and post-mortem evidence for reserve in the face of tau. disease: neuropsychiatric symptoms and Quality of Life in patients Brain Communications. 2020;2:fcaa025. with very mild and mild Alzheimer’s disease. Int J Geriatr Psychiatry. 25. Rascovsky K, Hodges JR, Knopman D, et al. Sensitivity of revised diag- 2011;26:473-482. nostic criteria for the behavioural variant of frontotemporal dementia. 46. Stern Y. Cognitive reserve in ageing and Alzheimer’s disease. The Brain. 2011;134:2456-2477. Lancet Neurology. 2012;11:1006-1012. 26. Illan-Gala I, Montal V, Borrego-Ecija S, et al. Cortical microstructure in 47. Arenaza-Urquijo EM, Przybelski SA, Lesnick TL, et al. The metabolic the behavioural variant of frontotemporal dementia: looking beyond brain signature of cognitive resilience in the 80+: beyond Alzheimer atrophy. Brain. 2019;142:1121-1133. pathologies. Brain. 2019;142:1134-1147. 27. Mackenzie IRA, Neumann M, Bigio EH, et al. Nomenclature and nosol- 48. Garcin B, Lillo P, Hornberger M, et al. Determinants of sur- ogy for neuropathologic subtypes of frontotemporal lobar degenera- vival in behavioral variant frontotemporal dementia. Neurology. tion: an update. Acta Neuropathol. 2010;119:1-4. 2009;73:1656-1661. 28. Alcolea D, Clarimon J, Carmona-Iragui M, et al. The Sant Pau Initiative 49. de Medeiros K, Robert P, Gauthier S, et al. The Neuropsychiatric on Neurodegeneration (SPIN) cohort: a data set for biomarker discov- Inventory-Clinician rating scale (NPI-C): reliability and validity of a ery and validation in neurodegenerative disorders. Alzheimers Dement. revised assessment of neuropsychiatric symptoms in dementia. Int Psy- 2019;5:597-609. chogeriatr. 2010;22:984-994. ILLÁN-GALA ET AL . 1341 SUPPORTING INFORMATION How to cite this article: Illán-Gala I, Casaletto KB, Additional supporting information may be found online in the Support- Borrego-Écija S, et al. Sex differences in the behavioral variant ing Information section at the end of the article. of frontotemporal dementia: A new window to executive and behavioral reserve. Alzheimer’s Dement. 2021;17:1329–1341. https://doi.org/10.1002/alz.12299