Education plays a different role in Frontotemporal Dementia and Alzheimer's disease

INTERNATIONAL JOURNAL OF GERIATRIC PSYCHIATRY

Int J Geriatr Psychiatry 2008; 23: 796–800.

Published online 3 January 2008 in Wiley InterScience

*S

(www.interscience.wiley.com) DOI: 10.1002/gps.1974

Education plays a different role in Frontotemporal Dementiaand Alzheimer’s disease

Barbara Borroni*, Antonella Alberici, Chiara Agosti, Enrico Premi and Alessandro Padovani

Department of Neurology, University of Brescia, Italy

SUMMARY

Background The role of modifiable and non-modifiable variables in Frontotemporal Dementia (FTD) as compared toAlzheimer’s dDisease (AD) and to Progressive Supranuclear Palsy (PSP) or Corticobasal Degeneration Syndrome (CBDS)has not been extensively evaluated. In particular, low education levels have been reported to be a risk factor for AD, but theircontribution in FTD is yet not known.Objective To investigate the role of education, other modifiable and non-modifiable factors in FTD as compared to AD,PSP and CBDS patients.Methods One hundred and seventeen FTD patients, 400 AD, 55 PSP, and 55 CBDS entered the study. Demographic andclinical characteristics were carefully recorded. Age, gender, family history for dementia and Apolipoprotein E (APOE)genotype were considered as non-modifiable factors; education and comorbidities were included as modifiable variables.Regression analyses were applied in order to identify differences among groups.Results FTD differed from AD patients in terms of younger age, positive family history and gender status. In regard toAPOE genotype, no differences between FTD and AD were found, but FTD showed higher prevalence of epsilon 4 allelecompared to both CBDS and PSP patients (p< 0.05). When modifiable factors were considered, FTD were higher educatedthan AD patients (p< 0.001). Regression analysis identified younger age, positive family history, and education levels asindependently associated variables to FTD diagnosis compared to AD (F¼ 21.27, R2¼ 24.1, p¼ 0.036).Conclusion Our results highlight that the contribution of education and non-modifiable factors is likely different in FTDand AD. Further work is needed to completely establish the role of this modifiable variable as a potential area of interventionfor dementias. Copyright # 2008 John Wiley & Sons, Ltd.

key words— Frontotemporal Dementia; Alzheimer’s disease; education; Apolipoprotein E

INTRODUCTION

Frontotemporal Dementia (FTD) is one of the mostcommon forms of dementia after Alzheimer’s disease(AD) (McKhann et al., 2001). At the disease onset, themain clinical features are variably represented bysocial misconduct and behavioral disturbances,impairment in language and executive functions,and motor deficits, described as atypical extrapyr-amidal syndromes or motor neuron disease (Hodgeset al., 2004; Rosen et al., 2005). The disease rapidlyprogresses to global dementia. At present, a wide

Correspondence to: B. Borroni, Department of Neurology, Pzapedali Civili, 1, 25100 Brescia, Italy. E-mail: [email protected]

Copyright # 2008 John Wiley & Sons, Ltd.

t

spectrum of heterogeneous clinical and anatomicalconditions are referred under the term FTD, andtherefore the concept of Pick’s Complex was adopted(Kertesz et al., 2003). The sharing of a commongenetic and pathological overlap with ProgressiveSupranuclear Palsy (PSP) and Corticobasal Degener-ation Syndrome (CBDS) promoted the inclusion ofthese movement disorders belonging to the same labelof FTD (Kertesz et al., 2005).

Although FTD is a sporadic disease, a highproportion of patients have an important autosomaldominant familial trait of inheritance (Chow et al.,1999), and MAPT tau and Progranulin genes seem tohave the major role (Spillantini and Goedert, 2000;Baker et al., 2006; Cruts et al., 2006). However, after

Received 19 May 2007Accepted 12 November 2007

modifiable and non-modifiable variables in ftd and ad 797

excluding the strong genetic component, risk factorsfor FTD remain largely unknown, in particular incomparison with AD. As we consider APOE the mostcommon genetic risk factor for AD, its involvement inFTD pathogenesis it is still a matter of debate (Borroniet al., 2006; Srinivasan et al., 2006). Other candidategenes have been recognized as disease modifiers(Borroni et al., 2005), but the small sample size ofFTD cohorts, fractionated in different clinicalvariants, has limited the research in this field.

When we take in account the weight of demographiccharacteristics of the disease, very few data areavailable. It is known as a presenile dementia, butthe application of current criteria to epidemiologicstudies have given discordant reports, documented alsoa significantly prevalence of FTD also in old agedpopulation (Ratnavalli et al., 2002; Rosso et al., 2003).Moreover, the combination of different cohortsdemonstrated that gender and age-at onset showsimportant differences among FTD subgroups,suggesting the role of unknown biological determinantsin the disease pathogenesis (Johnson et al., 2005).

The identification of modifiable predictors ofdisease onset and its progression addresses the issueto plan decision making for patients’ families andpublic health policies. For AD, education has beenextensively evaluated in epidemiological studies,supporting the idea that cognitive reserve constitutesone of the major factors to cope with the pathology(Karp et al., 2004; Scarmeas et al., 2004; Roe et al.,2007). In FTD, the role of education has not beencompletely explored. PET studies demonstrated thatin FTD cerebral hypometamolism is inverselycorrelated to the length of school, and there isevidence of sex-related effects for the cognitivereserve hypothesis (Perneczky et al., 2007). To ourknowledge, however, educational level has not yetbeen directly compared between FTD and AD, andwithin the FTD spectrum.

We aimed to investigate this aspect in a largeclinical series of FTD as compared to AD patients,also including patients with PSP and CBDS to furtheranalyse the different role of education in Pick’sComplex.

METHODS

Subjects

One-hundred and seventeen FTD patients, 400 AD, 55PSP, and 55 CBDS, recruited from the Centre forNeurodegenerative Diseases, and the Centre for


Movement Disorders, University of Brescia, Italy,entered the study. All patients were assessed between1 January 1999 and 30 March 2007.

All subjects performed a somatic and neurologicalevaluation, and routine laboratory examinationincluding Apolipoprotein E (APOE) genotype, andreceived a brain structural imaging study (MagneticResonance Imaging or Computed Tomography).Demographic characteristics were carefully recordedby an informant interview with the proxy caregiver.Positive family history was considered when afirst-degree sibling/parent was reported as affectedby dementia. Education level was considered asnumber of years of attending school.

Current diagnostic criteria for FTD, PSP, CBDS,and probable AD were used for patients’ inclusion(McKhann et al., 1984; Litvan et al., 1996; Nearyet al., 1998; Lantos, 2000; McKhann et al., 2001). Inthe FTD group, 102 behavioural variant FTD and 15temporal variant FTD were included.

A standardized neuropsychological assessment,including Mini-Mental State-Examination, was admi-nistered. Instrumental Activities of Daily Living andBasic Activities of Daily Living were assessed as well.Behavioral and psychiatric disturbances were eval-uated by Neuropsychiatry Inventory.

Stringent exclusion criteria were applied, asfollows: (a) a history of schizophrenia, schizoaffectivedisorder, delusional disorder or mood disorder withpsychotic features, major depressive disorder, sub-stance use disorder, or mental retardation according toDSM-IV criteria; (b) cerebrovascular disorders,hydrocephalus, and intra-cranial mass, documentedby CT or MRI within the past 12 months; (c) abnor-malities in serum folate and vitamin B12, syphilisserology, or thyroid hormone levels; (d) a history oftraumatic brain injury or another neurological disease(e.g. Parkinson disease, Huntington disease, seizuredisorders); (e.g. poorly controlled diabetes or hyper-tension; cancer within the past 5 years; clinicallysignificant hepatic, renal, cardiac or pulmonarydisorders).

All participants were made fully aware of the aimsof the research and a signature of an informed consentwas sought from all subjects. The work was conductedin accordance with local clinical research regulationsand conformed to the Helsinki Declaration.

To evaluate the educational level in the samegeographical area as stratified for age, a query wassubmitted to the Department of Statistics andEpidemiology of Brescia Council and the data werekindly provided accordingly to current countyregulations.


DOI: 10.1002/gps

798 b. borroni ET AL.

Statistical analysis

We assessed differences in continuous variablesamong diagnostic groups and associations with othervariables using analysis of variance methods, while weassessed the differences in discrete variables usinglogistic regression. A Bonferroni post-hoc analysiswas applied for evaluating the main effects amonggroups. Multiple regression analysis was also per-formed to evaluate which risk factors were indepen-dently associated. All values are expressed asmean� SD. The significant level was set atp< 0.05. Analyses were conducted using statisticalsoftware (SPSS Inc., Chicago, IL).

RESULTS

Demographic and clinical characteristics of thestudied population are shown in Table 1.

Differences between FTD and other diagnosticgroups in non-modifiable and modifiable associatedvariables were evaluated.

Among the former, one-way ANOVA performed onage and a Bonferroni post-hoc analysis revealed thatFTD patients were significantly younger than ADpatients (p< 0.0001) and PSP patients (p< 0.0001).The same results were obtained in regard to age atonset.

Table 1. Demographic and clinical characteristics of the studied po

Variable FTD (117) AD (

Mean� SD MeanAge, year 66.2� 8.0 76.8�Education, years 7.1� 3.3 5.5�Estimated age at onset, years 63.6� 8.0 73.8�

% (n) %Age� 75 years 13.7 (16) 69.8Education >5 years 43.6 (51) 19.0Gender, Female 53.0 (62) 69.0Family History# 42.0 (47) 23.5Cardiomiopathy 19.6 (23) 33.8Hypertension 33.4 (39) 52.0Hypecholesterolaemia 41.0 (48) 42.5Diabetes 8.5 (10) 14.5Apolipoprotein E, epsilon4# 38.1 (40) 41.3

AD¼Alzheimer Disease; CBDS¼Corticobasal Degeneration Syndrnuclear Palsy; SD¼ Standard Deviation.#Missing values are due to unknown data.*One-way ANOVA and Bonferroni post-hoc analysis or.**Chi-Square test.(a) FTD vs AD, p< 0.0001; (b) FTD vs PSP, p< 0.0001; (c) FTD vs A(f) FTD vs PSP, p< 0.05.


FTD patients significantly differed in genderdistribution compared to AD patients, female preva-lence being higher in the latter group (p< 0.01).Family history was found positive in a higher numberof cases of FTD than in AD (p< 0.0001).

In regard to APOE genotype, a higher prevalence ofepsilon 4 allele was present in FTD patients comparedto both CBDS and PSP patients (p< 0.05). Nosignificant differences between FTD and AD werefound.

When modifiable factors were considered, FTDshowed higher education than AD patients(p< 0.001).

Cardiomiopathy and hypertension were less preva-lent in FTD group than in AD (p< 0.01), andhypertension was less prevalent in FTD than in PSP(p< 0.0001).

To further evaluate the factors differentiating FTDfrom other dementing illnesses, regression analyseswere conducted considering significantly associatedfactors.

When FTD and AD patients were considered, lowerage (b¼ 0.357, p¼ 0.0001) and higher educationlevels (b¼�0.114, p¼ 0.015), and positive familyhistory (b¼�0.108, p¼ 0.017) were independentlyassociated with FTD diagnosis (F¼ 21.27, R2¼ 24.1,p¼ 0.036).

Being younger (b¼ 0.107, p¼ 0.037) and carryingAPOE epsilon 4 (b¼�0.184, p¼ 0.013) allele was

pulation

400) PSP (55) CBDS (55) P*

� SD Mean� SD Mean� SD8.2 73.2� 5.4 63.5� 9.2 a, b2.7 7.1� 4.4 7.7� 3.9 a9.3 69.9� 5.6 60.8� 9.4 a, b

(n) % (n) % (n) P**(279) 40.0 (22) 10.9 (6) a,b(76) 32.7 (18) 45.5 (25) a(276) 58.2 (32) 38.2 (21) c(74) 37.7 (20) 34.0 (18) a(135) 18.2 (10) 9.0 (5) d(208) 63.6 (36) 45.5 (25) b, d(170) 38.2 (21) 25.4 (14) —(58) 20 (11) 3.6 (2) —(147) 18.0 (9) 16.3 (7) e, f

ome; FTD¼ Frontotemporal Dementia; PSP¼ Progressive Supra-

D, p< 0.001; (d) FTD vs AD, p< 0.01; (e) FTD vs CBDS, p< .05;


DOI: 10.1002/gps

modifiable and non-modifiable variables in ftd and ad 799

independently correlated to FTD diagnosis in thecomparison with PSP patients (F¼ 12.43, R2¼ 0.183,p¼ 0.014). FTD and CBDS patients differed only inAPOE status (F¼ 4.69, R2¼ 31.4, b¼�0.201,p¼ 0.012).

DISCUSSION

FTD is a general diagnostic label currently includingdifferent clinical and pathological entities, amongwhich demographic and genetic characteristics havenot yet fully explored, in particular in comparison withAD. In this study, we tried to disentangle these issuesby investigating a large sample of FTD clinical seriesas compared to 400 AD patients and more than 100PSP or CBDS patients. The analyses were focused onnon-modifiable and modifiable variables. Age, gender,family history for dementia, and APOE status wereconsidered among non-modifiable characteristics,education and comorbidities as modifiable variables.

From present results, younger age differentiatesFTD from AD and PSP patients, further demonstratingdemographic differences among FTD spectrum(Johnson et al., 2005). As widely demonstrated, ahigher prevalence of females was reported among ADpatients. Further, the strong genetic background ofFTD was documented by a significant positive familyhistory (Rosso et al., 2003).

In regard to APOE status, no differences were foundbetween FTD and AD, confirming literature data.We might argue that epsilon 4 allele can modifycognitive performances also in FTD, likely influen-cing the disease duration with the neurodegenerationin specific brain areas (Borroni et al., 2006; Srinivasanet al., 2006). Remarkably, we demonstrated firstly thatin our sample, the extrapyramidal components ofFTD spectrum with CBDS and PSP patients had asignificant lower prevalence of epsilon 4 allele,probably sharing different genetic risk factors inrespect to other FTD variants.

Interestingly, among modifiable variables, edu-cation, cardiomiopathy and hypertension in firstinstance distinguished FTD from AD, but afteradjustment for other variables, the model gaveeducation as the solely different factor between thetwo pathologies.

We can speculate that differences in comorbilities,i.e. hypertension and cardiomyopathy, are likely to bemainly the result of a different age at disease onset.

What needs to be completely elucidated is whyeducation is higher in FTD as compared to AD. Atpresent, from epidemiologic studies, it has beenwidely recognized that low educational level is a risk


factor for AD (Scarmeas et al., 2004). Although it isnot possible to perform a direct comparison withextracted data from our epidemiologic survey inBrescia county, the mean educational level ofage-matched population was comparable to FTD,and higher than AD (age 65–69, education >5 years48%; age 70–74, education >5 years 48%). Literaturedata showed that in spite of the presence of abundantAD pathology, higher education confers a sort ofprotection to the onset of dementia, but at the sametime, it predicts a faster cognitive decline when ADtakes over (Scarmeas et al., 2004; Roe et al., 2007).Many studies have been devoted to explain themeaning of the phenomenon, contributing to theconceptualization of cognitive reserve hypothesis inAD.

Functional imaging studies have also suggested arole of cognitive reserve in FTD, the highereducational level the lower hypoperfusion pattern(Perneczky et al., 2007).

Further research should attempt to replicate thesefindings in population-based sample to completelyestablish the role of a modifiable variable as a potentialarea of intervention to dementias.

CONFLICT OF INTEREST

None known.

ACKNOWLEDGEMENTS

The authors would like to thank Marina Zanetti forvaluable neuropsychological assessment of thepatients, and the Staff of Ufficio Diffusione Informa-zione Statistica, Comune di Brescia, Italy, for theepidemiology data on education in Brescia county.

REFERENCES

Baker M, Mackenzie IR, Pickering-Brown SM, et al. 2006.Mutations in progranulin cause tau-negative frontotemporaldementia linked to chromosome 17. Nature 442: 916–919.

Borroni B, Perani D, Archetti S, et al. 2006. Functional correlates ofApolipoprotein E genotype in Frontotemporal Lobar Degener-ation. BMC Neurol 6: 31.

Borroni B, Yancopoulou D, Tsutsui M, et al. 2005. Associationbetween tau H2 haplotype and age at onset in frontotemporaldementia. Arch Neurol 62: 1419–1422.

Chow TW, Miller BL, Hayashi VN, et al. 1999. Inheritance offrontotemporal dementia. Arch Neurol 56: 817–822.

Cruts M, Gijselinck I, van der Zee J, et al. 2006. Null mutations inprogranulin cause ubiquitin-positive frontotemporal dementialinked to chromosome 17q21. Nature 442: 920–924.


DOI: 10.1002/gps

800 b. borroni ET AL.

Hodges JR, Davies RR, Xuereb JH, et al. 2004. Clinicopathologicalcorrelates in frontotemporal dementia. Ann Neurol 56: 399–406.

Johnson JK, Diehl J, Mendez MF, et al. 2005. Frontotemporal lobardegeneration: demographic characteristics of 353 patients. ArchNeurol 62: 925–930.

Karp A, Kareholt I, Qiu C, et al. 2004. Relation of education andoccupation-based socioeconomic status to incident Alzheimer’sdisease. Am J Epidemiol 159(2): 175–183.

Kertesz A, Hillis A, Munoz D. 2003. Frontotemporal degeneration,Pick’s Disease, Pick Complex and Ravel.Ann Neurol 54(S5): 1–2.

Kertesz A, McMonagle P, Blair M, et al. 2005. The evolution andpathology of frontotemporal dementia. Brain 128: 1996–2005.

Lantos PL. 2000. Diagnostic criteria for corticobasal degeneration.J Neurol Neurosurg Psychiatry 69(5): 705–706.

Litvan I, Agid Y, Calne D, et al. 1996. Clinical research criteria forthe diagnosis of progressive supranuclear palsy (Steele-Richardson-Olszewski syndrome): report of the NINDS-SPSPinternational workshop. Neurology 47(1): 1–9.

McKhann GM, Albert MS, Grossman M, et al. 2001. Clinical andpathological diagnosis of frontotemporal dementia: report of theWork Group on Frontotemporal Dementia and Pick’s Disease.Arch Neurol 58(11): 1803–1809.

McKhann G, Drachman D, Folstein M, et al. 1984. Clinical diag-nosis of Alzheimer’s disease: report of the NINCDS-ADRDAWork Group under the auspices of Department of Health andHuman Services Task Force on Alzheimer’s Disease. Neurology34: 939–944.


Neary D, Snowden J, Gustafson L, et al. 1998. FrontotemporalLobar degeneration: a consensus on clinical diagnostic criteria.Neurology 51: 1546–1554.

Perneczky R, Diehl-Schmid J, Forstl H, et al. 2007. Male gender isassociated with greater cerebral hypometabolism in frontotem-poral dementia: evidence for sex-related cognitive reserve. Int JGeriatr Psychiatry, in press.

Ratnavalli E, Brayne C, Dawson K, et al. 2002. The prevalence offrontotemporal dementia. Neurology 58(11): 1615–1621.

Roe CM, Xiong C, Miller JP, et al. 2007. Education and Alzheimerdisease without dementia: support for the cognitive reservehypothesis. Neurology 68: 223–228.

Rosen HJ, Allison SC, Schauer GF, et al. 2005. Neuroanatomicalcorrelates of behavioural disorders in dementia. Brain 128:2612–2625.

Rosso SM, Donker Kaat L, et al. 2003. Frontotemporal dementia inThe Netherlands: patient characteristics and prevalence estimatesfrom a population-based study. Brain 126: 2016–2022.

Scarmeas N, Zarahn E, Anderson KE, et al. 2004. Cognitivereserve-mediated modulation of positron emission tomographicactivations during memory tasks in Alzheimer’s disease. ArchNeurol 61(1): 73–78.

Spillantini MG, Goedert M. 2000. Tau mutations in familial fron-totemporal dementia. Brain 123: 857–859.

Srinivasan R, Davidson Y, Gibbons L, et al. 2006. The apolipopro-tein E epsilon4 allele selectively increases the risk of frontotem-poral lobar degeneration in males. J Neurol Neurosurg Psychiatry77(2): 154–158.


DOI: 10.1002/gps

Documents

Education plays a different role in Frontotemporal Dementia and Alzheimer's disease