Cognition and beta-amyloid in preclinical Alzheimer's disease: Data from the AIBL study

Neuropsychologia 49 (2011) 2384–2390

Contents lists available at ScienceDirect

Neuropsychologia journal homepage: www.elsevier.com/locate/neuropsychologia

Cognition and beta-amyloid in preclinical Alzheimer’s disease: Data from the AIBL study Kerryn E. Pike a,b,c,d,∗ , Kathryn A. Ellis e , Victor L. Villemagne a,b,c , Norm Good f , Gael Chételat a,g , David Ames e,h , Cassandra Szoeke i , Simon M. Laws j,k , Giuseppe Verdile j,k , Ralph N. Martins j,k , Colin L. Masters b,c , Christopher C. Rowe a,l a

Department of Nuclear Medicine, Centre for PET, Austin Health, Heidelberg, VIC, Australia Mental Health Research Institute, Parkville, VIC, Australia c Centre for Neurosciences, University of Melbourne, Parkville, VIC, Australia d School of Psychological Science, La Trobe University, Bundoora, VIC, Australia e Department of Psychiatry, University of Melbourne, Parkville, VIC, Australia f CSIRO Mathematics, Informatics and Statistics, Australian E-Health Research Centre, Herston, QLD, Australia g Inserm-EPHE-Université de Caen/Basse-Normandie, Unité U923, GIP Cyceron, CHU Côte de Nacre, Caen, France h National Ageing Research Institute, Parkville, VIC, Australia i CSIRO Neurodegenerative Disease, Mental Disorders & Brain Health, Preventative Health Flagship, CSIRO Molecular and Health Technologies, Parkville, VIC, Australia j Centre of Excellence for Alzheimer’s Disease Research & Care, School of Exercise Biomedical and Health Sciences, Edith Cowan University, Joondalup, Western Australia, Australia k Sir James McCusker Alzheimer’s Disease Research Unit (Hollywood Private Hospital), Perth, Western Australia, Australia l Department of Medicine, University of Melbourne, Parkville, VIC, Australia b

a r t i c l e

i n f o

Article history: Received 7 January 2011 Received in revised form 28 March 2011 Accepted 11 April 2011 Available online 16 April 2011 Keywords: Memory Ageing PiB-PET Amyloid imaging

a b s t r a c t The ‘preclinical’ phase of Alzheimer’s disease is a future target for treatment, but additional research is essential to understand the relationship between ␤-amyloid burden and cognition during this time. We investigated this relationship using a large sample of apparently healthy older adults (N = 177), which also enabled examination of whether the relationship differed according to age, gender, years of education, apolipoprotein E status, and the presence of subjective memory complaints. In addition to episodic memory, a range of cognitive measures (global cognition, semantic memory, visuospatial performance, and executive function) were examined. Participants were aged over 60 years with no objective cognitive impairment and came from the imaging arm of the Australian Imaging, Biomarkers, and Lifestyle (AIBL) study of ageing. 11 C-PiB PET was used to measure ␤-amyloid burden and a PiB ‘cut-off’ level of 1.5 was used to separate participants with low PiB retention from those with high PiB retention. Thirty-three percent of participants had a PiB positive scan. PiB positive participants were 5 years older, twice as likely to carry an apolipoprotein E ␧4 allele, and their composite episodic memory was 0.26 SD worse than PiB negative volunteers. Linear regressions with ␤-amyloid burden as a dichotomous predictor, revealed an interaction between ␤-amyloid burden and gender, as well as age and education effects, in predicting episodic memory and visuospatial performance. In females, but not in males, increased ␤-amyloid was related to worse episodic memory and visuospatial performance. Furthermore, an interaction between ␤-amyloid burden and APOE status was found in predicting visuospatial performance, whereby there was a trend for increased ␤-amyloid to relate to worse visuospatial performance for those without an APOE ␧4 allele. There were no other main or interaction effects of ␤-amyloid on any of the other composite cognitive measures. These cross-sectional findings suggest that ␤-amyloid burden does not have a large effect on cognition in this subset of apparently healthy older people. The finding of gender differences deserves further research to answer definitively the important question of gender susceptibility to adverse cognitive effects from ␤-amyloid. © 2011 Elsevier Ltd. All rights reserved.

1. Introduction

∗ Corresponding author at: School of Psychological Science, La Trobe University, Bundoora, VIC 3086, Australia. Tel.: +61 3 9479 5079; fax: +61 3 9479 1956. E-mail address: [email protected] (K.E. Pike). 0028-3932/$ – see front matter © 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.neuropsychologia.2011.04.012

The pathological processes underlying Alzheimer’s disease (AD) begin years before the onset of symptoms (Amieva et al., 2005). When available, disease-modifying treatments targeting these processes are likely to be most efficacious at this ‘preclinical’ stage.

K.E. Pike et al. / Neuropsychologia 49 (2011) 2384–2390

This requires, however, a better understanding of the relationship between the pathophysiology and emergence of clinical symptoms. Much research has focused on identification of preclinical AD, culminating in a recent workgroup by the National Institute on Aging (NIA) and the Alzheimer’s Association to define research criteria for preclinical AD (Sperling et al., 2011). Integral to the proposed definition of preclinical AD is biomarker evidence of beta-amyloid (A␤) accumulation, reflecting the growth in knowledge regarding biomarkers of AD. Deposition of A␤ is believed to be the initial step in the disease process (Villemagne et al., 2006). Cerebral A␤ deposition is found in up to 45% of apparently healthy older people (Bennett et al., 2006) and prevalence increases with age (Braak & Braak, 1997; Davies et al., 1988). Until recently, cerebral A␤ burden could only be measured at autopsy, with some, but not all, studies suggesting that increased AD pathology in apparently healthy older adults is related to decreased cognition. Furthermore, these studies are limited by the time lag between the last cognitive assessment and autopsy, whereas A␤-neuroimaging techniques enable measurement of A␤ burden proximal to cognitive performance. A␤-imaging studies to date, however, have demonstrated inconsistent results regarding the relationship between A␤ burden and cognition in apparently healthy older adults. Previously, our group found that A␤ burden was related to decreased episodic memory performance in healthy older participants and that those participants with a PiB-positive scan performed 0.8 SD worse on memory tasks than those with a PiB-negative scan (Pike et al., 2007). We have also shown that apparently normal controls who decline on cognitive tasks over time are more likely to have a PiB positive scan than those with stable cognitive performance (Villemagne, Pike, et al., 2008). Mormino et al. (2009) found a relationship between PiB retention in hippocampal regions and episodic memory in apparently normal controls in one of their examined cohorts, although reported that this was mainly driven by 2/20 participants with high PiB retention. In addition, Braskie et al. (2008) found a relationship between a composite cognitive score and the retention of an alternative radiotracer that binds to A␤ as well as neurofibrillary tangles. In contrast, other studies (Aizenstein et al., 2008; Mintun et al., 2006) have not found a difference in cognitive performance between apparently healthy older participants with PiB-positive and PiB-negative scans. The small samples of these studies (from 10 to 43) may help explain the discrepant findings. Recently, a large study (N = 135) was published examining cognition and A␤ burden as measured by PiB (Storandt, Mintun, Head, & Morris, 2009). No relationship was found between concurrent cognition and A␤ burden. Participants in their study had annual cognitive assessments in various longitudinal studies beginning in 1985, thus they were also able to examine decline over time and found relationships between A␤ burden and decline on the visuospatial and working memory measures, and one of their episodic memory tests (associate learning). They did not examine the effect of age, gender, years of education, apolipoprotein E (APOE) status, or memory complaints on the findings. Since joining forces with the Australian Imaging Biomarkers and Lifestyle (AIBL) study (Ellis et al., 2009), our sample has increased nearly 6-fold. We recently reported the amyloid imaging results from the cohort, but the relationship between cognition and PiB retention was only briefly examined; we found no difference between PiB-positive and PiB-negative apparently healthy older controls on the long delay free recall from the California Verbal Learning Test—second edition (CVLT-II; Rowe et al., 2010). The present paper aims to examine the relationship between concurrent cognitive performance and A␤ burden in greater depth in this large two-site sample. A number of composite cognitive measures were constructed to consider cognitive domains in addition

2385

to episodic memory. Furthermore, the large sample provides sufficient power to enable examination of some individual differences that may affect the relationship between A␤ burden and cognitive performance—including age, gender, years of education, APOE status, and the presence of any subjective memory complaints. The main goal of the present study was thus to investigate if, and under what circumstances, increased cerebral A␤ burden is associated with lowered cognition in apparently healthy older adults. 2. Methods 2.1. Participants The participants for this study were the 177 (100 from Melbourne, 77 from Perth) apparently healthy older people (mean age = 72 ± 7, range 60–89) enrolled in the AIBL study, and reported in Rowe et al. (2010). All participants had no objective evidence of cognitive impairment, were fluent in English, and had no significant neurological history. Informed written consent was obtained prior to participation. Ethics approval was granted from the Human Research Ethics Committees at Austin Health, St Vincent’s Health, Hollywood Private Hospital, and Edith Cowan University. APOE genotype was determined by direct sequencing, and 43% carried an APOE ␧4 allele. Fifty percent of the volunteers were male, 54% had subjective complaints about their memory (established by the question: “do you have difficulties with your memory?”), and 54% had more than 12 years of education. Table 1 displays participant demographics according to PiB status.

2.2. Neuroimaging All participants underwent a PiB-PET scan as previously described (Pike et al., 2007; Rowe et al., 2007) at Austin Health Centre for PET, Melbourne or WA PET and Cyclotron service, Sir Charles Gairdner Hospital, Perth. Each participant received ∼370 MBq 11 C-PiB intravenously over 1 min. A 30-min acquisition in 3D mode starting 40 min after injection of PiB was performed with a Phillips AllegroTM PET camera. A transmission scan was performed for attenuation correction. PET images were reconstructed using a 3D RAMLA algorithm. Participants also received a 3D T1-weighed MRI acquisition with MP-RAGE, FLAIR, SWI, and DTI sequences, for screening and subsequent co-registration with the PET images. Co-registration of each individual’s MRI with the PET images was performed in PET native space with MilxView® , developed by the Australian e-Health Research Centre BioMedIA (Brisbane, Australia). A region of interest template was placed on the MR and transferred to the co-registered PET images. Standardized uptake values (SUV) were calculated for all brain regions examined. SUV ratios (SUVR) were generated by normalising the regional SUV to the cerebellar cortex SUV. Neocortical A␤ burden was expressed as the average SUVR of the area-weighted mean of frontal, superior parietal, lateral temporal, lateral occipital, and anterior and posterior cingulate regions. In accordance with previous studies reporting marked PiB retention in cognitively unimpaired healthy controls (Aizenstein et al., 2008; Mintun et al., 2006; Pike et al., 2007; Rowe et al., 2007), most participants demonstrated low levels of PiB retention, but there was a subset of participants with higher PiB retention. Consequently, to identify a PiB ‘cut-off’ level to separate participants with low versus high PiB retention, a hierarchical cluster analysis was performed on all elderly apparently healthy participants yielding a mean cut-off for neocortical SUVR of 1.5.

Table 1 Participant demographics and composite cognitive scores split by PiB result.

% female % complainers Age % Yrs ed < 13 % APOE ␧4 MMSE Global EM EF SM VS

PiB negative (N = 119)

PiB positive (N = 58)

50 53 69.8 (7.0) 46 33 28.9 (1.2) −0.21 (0.63) 0.07 (0.66) −0.13 (0.59) −0.96 (2.49) 0.12 (0.85)

50 55 75.2 (7.1)* 47 64* 28.5 (1.2) −0.32 (0.76) −0.19 (0.88)* −0.17 (0.57) −0.89 (2.42) −0.09 (1.15)

Values are means (SD) unless otherwise noted. Differences between groups were determined using t-tests for continuous variables, or 2 test for independence for dichotomous variables: MMSE = Mini Mental State Examination; Global = Composite Global Cognition Score; EM = Composite Episodic Memory Score; EF = Composite Executive Function Score; SM = Composite Semantic Memory Score; VS = Visuospatial Composite Score. * p < .05.

2386

K.E. Pike et al. / Neuropsychologia 49 (2011) 2384–2390

2.3. Neuropsychological assessment Participants took part in a comprehensive neuropsychological assessment, as previously described (Ellis et al., 2009). The following composite measures of cognitive functioning were constructed using z scores in relation to a large normative database (Ellis et al., 2009): global cognition, episodic memory, semantic memory, visuospatial performance, and executive function. Global cognition was based on the average of delayed recall on the California Verbal Learning Test—second edition (CVLT-II; Delis, Kramer, Kaplan, & Ober, 2000), copy and delayed recall of the Rey Complex Figure Test (RCFT; Meyers & Meyers, 1995), delayed recall of the first Logical Memory (LM) story from the WMS (Wechsler, 1945), Stroop incongruent condition (Strauss, Sherman, & Spreen, 2006), category, letter, and switching fluency (Delis, Kaplan, & Kramer, 2001), Boston Naming Test (BNT; Saxton et al., 2000), and Digit Span and Digit Symbol Coding subtests from the WAIS-III (Wechsler, 1997). The composite episodic memory score was based on the average of delayed recall on the CVLT-II, RCFT, and first LM story. Semantic memory was based on the average of the BNT and category fluency. Visuospatial performance was based on RCFT copy. Executive function was based on the average of Stroop incongruent condition, category, letter, and switching fluency, and digit span. 2.4. Data analysis Preliminary analyses examined demographic and cognitive differences between groups (PiB positive; PiB negative) using t tests or chi-square analysis, as appropriate. Pearson’s correlations were used to examine the basic relationship between each cognitive composite score and A␤ burden. To investigate the effect of possible moderating variables (age, gender, years of education, APOE status, and subjective memory complaints), a series of multiple regressions was conducted to examine the relationship between A␤ burden and each of the composite cognitive measures (global cognition, episodic memory, semantic memory, visuospatial performance, and executive function). Years of education was coded as a dichotomous variable: up to 12 years; greater than 12 years. To represent the interaction between A␤ burden and each moderator variable, the variables were first centred and then multiplied together. Separate regressions were conducted treating A␤ burden as a continuous variable and as a dichotomous variable. The overall significance level (˛) was set to .05. Given the number of predictors, and the primary aim of exploring any relationship between cognition and A␤ burden, if a trend (p < .1) was observed for the effect of A␤ burden or any interactions involving A␤ burden, the regression was re-run with only those predictors with p < .1 in the original regression, plus main effects of A␤ burden and the moderating variable. Finally, any significant A␤ burden interactions were examined using follow-up t-tests.

3. Results 3.1. Aˇ burden Overall, 33% of the apparently healthy older volunteers had a PiB positive scan. As shown in Table 1, there were no differences between PiB positive and PiB negative participants in terms of gender, years of education, or the proportion of memory complainers. PiB positive participants were on average 5 years older, t(175) = 4.77, p < .001, were nearly twice as likely to carry an APOE ␧4 allele, 2 (1, n = 177) = 14.07, p < .001, and performed 0.26 SD

worse on the composite episodic memory measure t(88.7) = 1.93, p = .049 (using a t statistic not assuming homogeneity of variance, as Levene’s test for equality of variances was violated, F (1, 175) = 7.71, p = .006). There were no significant differences between PiB positive and PiB negative participants on MMSE or any of the other cognitive composite scores (Table 1). 3.2. Relationship between Aˇ burden and cognition Correlations were performed for each cognitive composite score to determine whether A␤ burden predicts cognitive performance, without accounting for possible moderating factors. These revealed a relatively weak, but significant, relationship between A␤ burden and episodic memory only (R = −.19, p = .012). No relationships were found between A␤ burden and global cognition, semantic memory, executive functions, or visuospatial performance. Next, a range of individual differences (age, gender, years of education, APOE status, and memory complaints) were examined to determine whether they had any influence on the relationship between A␤ burden (as a continuous variable) and cognition. As shown in Table 2, there were main effects of years of education on all cognitive variables, except semantic memory. Age had a main effect on global composite, episodic memory, and visuospatial performance, and a trend for a main effect on executive functions (p = .057). Memory complaints had a main effect on executive functions only. Gender, APOE status, and A␤ burden did not have main effects for any of the cognitive measures. The primary interest was to identify any main or interaction effects involving A␤ burden, but no significant effects were found for any of the cognitive measures. The same regression analyses were repeated, except that established cut-offs for a positive versus negative PiB scan were used to treat A␤ burden as a dichotomous, rather than as a continuous, variable. As shown in Table 3, the main effects were unchanged, including no main effect of A␤ burden on any of the cognitive measures. In this case, however, there were some interaction effects involving A␤ burden. For episodic memory, there was a trend for a gender × A␤ burden interaction (p = .062). Given that the main purpose of the study was to examine any effects of A␤ burden on cognition, this trend was further examined by conducting another regression analysis, removing any moderating demographic variables where p was >1. Thus the regression included: A␤ burden, age, education, gender, and gender × A␤, as shown in column EMR in Table 3. In this case, the interaction between gender and A␤ attained significance (p = .040). The interaction is demonstrated in Fig. 1, and follow-up t tests showed that PiB positive females performed 0.5 SD worse on

Table 2 Regression analyses examining the effect of demographic variables on the relationship between cognition and A␤ burden (continuous variable). Composite cognitive score

Global

EM

EF

VS

SM

Constant A␤ burden (SUVR) Age Age × A␤ Education Education × A␤ APOE APOE × A␤ Complaints Complaints × A␤ Gender Gender × A␤ R2

1.188** 0.093 −0.027** −0.012 0.324** −0.060 −0.007 0.095 −0.113 −0.079 −0.127 −0.144 0.145**

−1.399** −0.039 −0.027** −0.016 0.471** −0.113 0.049 −0.370 −0.056 −0.105 −0.177* 0.141 0.228**

0.345 0.102 −0.013* −0.019 0.283** −0.204 −0.028 0.189 −0.210** 0.156 −0.062 −0.191 0.146**

2.622** 0.025 −0.041** −0.002 0.299** 0.558 −0.108 0.464 −0.194 0.166 0.023 0.249 0.146**

1.319 0.239 −0.037 0.011 0.088 −0.173 0.061 0.536 −0.087 −0.783 −0.247 −0.269 0.020

Values are unstandardized coefficients (B): Global = Composite Global Cognition Score; EM = Composite Episodic Memory Score; EF = Composite Executive Function Score; SM = Composite Semantic Memory Score; VS = Visuospatial Composite Score. Note that a separate regression analysis was conducted for each composite cognitive measure. * p < .1. ** p < .05.

K.E. Pike et al. / Neuropsychologia 49 (2011) 2384–2390

2387

Table 3 Regression analyses examining the effect of demographic variables on the relationship between cognition and A␤ burden (dichotomous variable). Composite cognitive score

Global

EM

EMR

EF

VS

VSR

SM

Constant A␤ burden Age Age × A␤ Education Education × A␤ APOE APOE × A␤ Complaints Complaints × A␤ Gender Gender × A␤ R2

1.213** 0.061 −0.026** −0.025 0.325** 0.219 0.018 0.076 −0.087 −0.055 −0.135 0.133 0.161**

1.342** −0.052 −0.027** −0.026 0.477** −0.138 0.053 −0.277 −0.036 −0.104 −0.192* 0.418* 0.244**

1.402** −0.107 −0.028** – 0.471** – – – – – −0.179* 0.441** 0.227**

0.439 0.040 −0.012* −0.013 0.275** 0.065 −0.010 0.200 −0.198** 0.140 −0.063 −0.037 0.121**

2.377** −0.012 −0.038** −0.016 0.322** 0.583* −0.080 0.614* −0.151 0.206 0.048 0.607** 0.179**

2.140** −0.068 −0.037** – 0.341** 0.537* −0.029 0.617** – – 0.038 0.644** 0.168**

1.410 0.284 −0.034 −0.049 0.112 0.813 0.115 0.298 −0.025 −0.569 −0.290 0.205 0.026

Values are unstandardized coefficients (B): Global = Composite Global Cognition Score; EM = Composite Episodic Memory Score; EMR = Revised Regression Analysis for Composite Episodic Memory Score; EF = Composite Executive Function Score; VS = Visuospatial Composite Score; VSR = Revised Regression Analysis for Visuospatial Composite Score; SM = Composite Semantic Memory Score. Note that a separate regression analysis was conducted for each composite cognitive measure. * p < .1. ** p < .05.

episodic memory tasks than PiB negative females t(40.72) = 2.89, p = .006 (using a t statistic not assuming homogeneity of variance, as Levene’s test for equality of variances was violated, F (1, 86) = 7.63, p = .007), whereas there were no differences between PiB positive and negative males, t(87) = 0.02, ns. For visuospatial function, there was a significant gender × A␤ burden interaction, as well as trends for interactions between education × A␤ (p = .057) and APOE × A␤ (p = .065). The trends were further examined by conducting another regression analysis, removing any moderating demographic variables where p was >1. Thus the regression included: A␤ burden, age, education, education × A␤, APOE, APOE × A␤, gender, and gender × A␤, as shown in column VSR in Table 3. In this case, the statistically significant interaction between gender and A␤ remained, and the interaction between APOE and A␤ attained significance (p = .047), although the interaction between education and A␤ remained a trend only (p = .075). The gender interaction is demonstrated in Fig. 2, and follow-up t tests revealed that PiB positive females performed 0.6 SD worse on the visuospatial task than PiB negative females, t(38.3) = 2.15, p = .038 (using a t statistic not assuming homogeneity of variance, as Levene’s test for equality of variances was violated, F (1, 86) = 4.54, p = .036), whereas there was no significant difference between PiB positive and negative males t(87) = 0.96, ns. The APOE interaction is demonstrated in Fig. 3, and follow-up t tests revealed that there was a strong trend within those who did not carry the E4 allele for PiB positive to perform worse than PiB negative,

t(99) = 1.88, p = .063, whereas there were no differences between PiB positive and negative E4 carriers, t(74) = 0.17, ns. 4. Discussion This large study of A␤-imaging in apparently healthy older volunteers examined the effect of a number of individual differences on the relationship between cerebral A␤ burden and concurrent cognition. Consistent with previous work, 33% of the present sample had significant cerebral A␤ burden, and PiB positive participants were older, more likely to be APOE ␧4 carriers, and had lower episodic memory performance. The strongest predictors of cognition in most domains were age and education, consistent with the extensive literature demonstrating well-established age and education effects on most cognitive tasks (Strauss et al., 2006). As increased age was also associated with increased A␤ burden, it is important to determine that any apparent relationship between cognition and A␤ burden is not completely explained by age. When age and education were accounted for, no striking relationships were found between cognition and A␤ burden, consistent with another recent study (Storandt et al., 2009). Our study differs from previous work, however, by finding a modest relationship in females only between A␤ burden (when treated as a dichotomous variable) and both episodic memory and visuospatial performance. We also found a trend for A␤ burden (when treated as a dichotomous variable) to be asso-

Fig. 1. Interaction between gender and PiB status (positive: neocortical SUVR > 1.5) on episodic memory performance (unadjusted values).

2388

K.E. Pike et al. / Neuropsychologia 49 (2011) 2384–2390

Fig. 2. Interaction between gender and PiB status (positive: neocortical SUVR > 1.5) on visuospatial performance (unadjusted values).

ciated with visuospatial performance in people who did not carry the APOE ␧4 allele. The finding of such a modest interaction effect of A␤ burden helps clarify the discrepancies in previous studies. It suggests that sample composition – particularly gender, APOE genotype, and the number of PiB positive cases – may determine whether a relationship between A␤ burden and cognition is found. Additionally, previous studies may vary on where their participants fell within the preclinical spectrum. The preclinical stage of AD, before cognitive impairment becomes apparent, extends for many years. It is likely that there is a stronger relationship between A␤ burden and cognition in the late preclinical stage (close to mild cognitive impairment) than in the earlier stages. Indeed, previous studies have demonstrated that apparently healthy volunteers who decline are often PiB positive (Storandt et al., 2009; Villemagne, Pike, et al., 2008), and that there is a stronger negative relationship between episodic memory and A␤ burden in mild cognitive impairment than in healthy ageing (Fripp et al., 2008; Pike et al., 2007; Rowe et al., 2010). Together, these findings suggest that PiB positive individuals with even subtle decline in cognition are likely to develop AD. On the other hand, significant deposition of A␤ does not guarantee conversion to AD within the person’s lifetime. A␤ appears necessary, but not sufficient for the development of AD (Villemagne, Fodero-Tavoletti, et al., 2008). The finding that episodic memory was one of the domains affected (in females only) by increased A␤ is consistent with it being affected early in the preclinical stages of AD (Bäckman,

Jones, Berger, Jonsson Laukka, & Small, 2005; Twamley, Legendre Ropacki, & Bondi, 2006). Although episodic memory performance is considered the best predictor of AD in non-demented individuals, other cognitive functions are also predictive, with some reports of changes in visuospatial abilities (Artero, Tierney, Touchon, & Ritchie, 2003; Johnson, Storandt, Morris, & Galvin, 2009). Moreover, the involvement of visuospatial functions (for females and those without the APOE ␧4 allele), thought to involve the temporo-parietal regions, fits with temporo-parietal hypometabolism frequently associated with conversion to AD (Chételat et al., 2005). PiB positive participants were more likely to carry an APOE ␧4 allele, consistent with previous studies (Reiman et al., 2009; Rowe et al., 2007, 2010; Storandt et al., 2009). In contrast, no main effect of APOE genotype on cognition was found. Careful examination of the literature, however, suggests that this result is not unexpected as APOE appears to have a modest effect at best on cognition, which is only evident in large studies, and not always then (Christensen et al., 2008; Luciano et al., 2009; Small, Rosnick, Fratiglioni, & Backman, 2004; Welsh-Bohmer et al., 2009). There was an interaction effect of APOE and PiB status on visuospatial performance, however, where for those without an APOE ␧4 allele, a positive PiB scan was associated with worse performance. This finding was not expected—instead we might have expected to see that the combination of an APOE ␧4 allele and a PiB positive scan (the greatest risk for developing AD) would negatively affect cognition. One possible explanation is that this finding may only show

Fig. 3. Interaction between APOE genotype (non-E4 carrier; E4 carrier) and PiB status (positive: neocortical SUVR > 1.5) on visuospatial performance (unadjusted values).

K.E. Pike et al. / Neuropsychologia 49 (2011) 2384–2390

up in the non-carriers, because there are few people carrying the APOE ␧4 allele with a PiB positive scan whose cognition fell within the normal range. This explanation is not consistent with the data, however, as more than half of the APOE ␧4 carriers in this study were PiB positive (39/76). Another possibility is that there may be a three way interaction between APOE, gender, and A␤ burden, as has been reported in the literature (Corder et al., 2004). Although the present study has a large sample compared to other studies of amyloid imaging, the sample size was not large enough to warrant developing a regression model with three-way interaction terms, particularly considering the number of different demographic predictors. We did, however, explore whether there were any cognitive differences between male and female PiB positive APOE ␧4 carriers, but no significant differences were found (data not shown). The lack of differences may relate to sample size; the study by Corder et al. contained over 5000 participants, thus had much greater power to detect a three-way interaction effect. It would be interesting for future work to study a much larger sample (perhaps combined across multiple research groups) to determine whether these kinds of interactions are present. In this study, memory complaints were not related to PiB status or cognitive performance. The lack of association with PiB status is at odds with previous research suggesting that memory complaints in older adults are associated with an increased risk of AD neuropathology (Barnes, Schneider, Boyle, Bienias, & Bennett, 2006; Jorm et al., 2004), increased association between A␤ and neurodegeneration (Chételat et al., 2010), and cognitive decline or development of dementia (Glodzik-Sobanska et al., 2007; Jessen et al., 2010), though other studies have not found a relationship (for reviews see Jonker, Geerlings, & Schmand, 2000; Reisberg & Gauthier, 2008). Similarly, some studies have reported a relationship between memory complaints and objective cognitive performance (Archer et al., 2006; Snitz, Morrow, Rodriguez, Huber, & Saxton, 2008), whereas others have not (Jungwirth et al., 2004; Weaver Cargin, Collie, Masters, & Maruff, 2008). A major issue with this emerging area is defining memory complaints (Adulrab & Heun, 2008), including overlap with mild cognitive impairment and relationship to affective status. Years of education is often used as a proxy measure of cognitive reserve, which is thought to help explain why some people with substantial brain pathology demonstrate no or minimal clinical manifestation of disease (Stern, 2002; Scarmeas & Stern, 2003). Previous research has demonstrated a moderating effect of cognitive reserve on the relationship between A␤ burden and cognition, where a stronger relationship is observed in people with lower cognitive reserve (e.g., Rentz et al., 2010; Yaffe et al., 2011). In contrast, we did not observe any significant interactions between education and A␤ burden for any of the cognitive measures. It may be that the use of more sophisticated proxies of cognitive reserve, such as occupational attainment or engagement in cognitively challenging leisure pursuits, may result in different findings, although years of education was used by Rentz et al., who found a moderating effect. One limitation in terms of the generalisation of our findings to the general population is that the sample consists of volunteers. On the one hand, some people volunteered because they have a family member with AD, or because someone was concerned about their memory—suggesting the cohort may be over-representative of preclinical AD compared to the general community. On the other hand, volunteer cohorts tend to be more highly educated and mentally active (greater cognitive reserve) than a general sample, which could provide protection against the development of AD. We believe this is the first A␤ imaging study to identify gender differences in the pattern between cognition and A␤ burden. Ideally, this finding should be examined in other groups’ cohorts. Gender differences in AD prevalence are reported: females are 1.5

2389

times more likely to develop AD than males—partly, but not solely, because they live longer (Baum, 2005; Gao, Hendrie, Hall, & Hui, 1998; Gilleard, 1997). This is thought to relate to differences in vascular risk factors, higher mortality in men with AD, and potential neuroprotective effects of sex hormones such as testosterone (Azad, Al Bugami, & Loy-English, 2007; Baum, 2005; Pike, Carroll, Rosario, & Barron, 2009). Furthermore, it has been suggested that the link between AD neuropathology and clinical symptoms of AD is stronger in women than in men (Barnes et al., 2005; Perneczky, Drzezga, Diehl-Schmid, Li, & Kurz, 2007). For example, one study found that each unit of AD pathology was related to a 20-fold increase in the odds of having clinical AD for women, but only a 3-fold increase for men (Barnes et al., 2005). We have observed that males with AD have significantly higher PiB binding than females (Rowe et al., 2010), possibly indicating that a greater amyloid load is necessary before men manifest the disease. The finding of gender differences deserves further research in a random sample to answer definitively the important question of gender differences in AD risk, and examine the contribution of sex hormones and vascular risk factors. Furthermore, longitudinal follow-up of the present cohort will help elucidate how A␤ relates to cognitive decline and the onset of AD. Acknowledgements Core funding for the study was provided by the Australian Commonwealth Scientific Industrial Research Organization (CSIRO) through the Australian Imaging, Biomarkers and Lifestyle flagship study of aging (AIBL), which was supplemented by “in kind” contributions from AIBL research partner organisations: University of Melbourne, Neurosciences Australia Ltd (NSA), Edith Cowan University (ECU), Mental Health Research Institute (MHRI), Alzheimer’s Australia (AA), National Ageing Research Institute (NARI), Austin Health, University of WA (UWA), CogState Ltd., Macquarie University, Hollywood Private Hospital, and Sir Charles Gairdner Hospital. The funding source had no direct role in the design and conduct of the study; collection, management, analysis, and interpretation of the data, or preparation, review, and approval of the manuscript. We would like to acknowledge the assistance of the volunteers who gave their time to participate in this study. In addition, we would like to thank the AIBL research group (see www.aibl.csiro.au), and staff of our AIBL research partner organisations as listed above. References Adulrab, K., & Heun, R. (2008). Subjective memory impairment. A review of its definitions indicates the need for a comprehensive set of standardised and validated criteria. European Psychiatry, 23, 321–330. Aizenstein, H. J., Nebes, R. D., Saxton, J. A., Price, J. C., Mathis, C. A., Tsopelas, N. D., et al. (2008). Frequent amyloid deposition without significant cognitive impairment among the elderly. Archives of Neurology, 65, 1509–1517. Amieva, H., Jacqmin-Gadda, H., Orgogozo, J.-M., Le Carret, N., Helmer, C., Letenneur, L., et al. (2005). The 9 year cognitive decline before dementia of the Alzheimer type: A prospective population-based study. Brain, 128(5), 1093–1101. Archer, H. A., Macfarlane, F., Price, S., Moore, E. K., Pepple, T., & Cutler, D. (2006). Do symptoms of memory impairment correspond to cognitive impairment: A cross sectional study of clinical cohort. International Journal of Geriatric Psychiatry, 21, 1206–1212. Artero, S., Tierney, M. C., Touchon, J., & Ritchie, K. (2003). Prediction of transition from cognitive impairment to senile dementia: A prospective longitudinal study. Acta Psychiatrica Scandinavica, 107, 390–393. Azad, N. A., Al Bugami, M., & Loy-English, I. (2007). Gender differences in dementia risk factors. Gender Medicine, 4(2), 120–129. Bäckman, L., Jones, S., Berger, A. K., Jonsson Laukka, E., & Small, B. J. (2005). Cognitive impairment in preclinical Alzheimer’s disease: A meta-analysis. Neuropsychology, 19(4), 520–531. Barnes, L. L., Schneider, J. A., Boyle, P. A., Bienias, J. L., & Bennett, D. A. (2006). Memory complaints are related to Alzheimer disease pathology in older persons. Neurology, 67, 1581–1585.

2390

K.E. Pike et al. / Neuropsychologia 49 (2011) 2384–2390

Barnes, L. L., Wilson, R. S., Bienias, J. L., Schneider, J. A., Evans, D. A., & Bennett, D. A. (2005). Sex differences in the clinical manifestations of Alzheimer disease pathology. Archives of General Psychiatry, 62, 685–691. Baum, L. W. (2005). Sex, hormones, and Alzheimer’s disease. Journals of Gerontology. Series A, Biological Sciences and Medical Sciences, 60A(6), 736–743. Bennett, D. A., Schneider, J. A., Arvanitakis, Z., Kelly, J. F., Aggarwal, N. T., Shah, R. C., et al. (2006). Neuropathology of older persons without cognitive impairment from two community-based studies. Neurology, 66, 1837–1844. Braak, H., & Braak, E. (1997). Frequency of stages of Alzheimer-related lesions in different age categories. Neurobiology of Aging, 18(4), 351–357. Braskie, M. N., Klunder, A. D., Hayashi, K. M., Protas, H., Kepe, V., Miller, K. J., et al. (2008). Plaque and tangle imaging and cognition in normal aging and Alzheimer’s disease. Neurobiology of Aging, doi:10.1016/j.neurobiolaging. 2008.09.012 Chételat, G., Eustache, F., Viader, F., De La Sayette, V., Pelerin, A., Mezenge, F., et al. (2005). FDG-PET measurement is more accurate than neuropsychological assessments to predict global cognitive deterioration in patients with mild cognitive impairment. Neurocase, 11, 14–25. Chételat, G., Villemagne, V. L., Bourgeat, P., Pike, K. E., Jones, G., Ames, D., et al. (2010). Relationship between atrophy and ␤-amyloid deposition in Alzheimer disease. Annals of Neurology, 67, 317–324. Christensen, H., Batterham, P. J., Mackinnon, A. J., Jorm, A. F., Mack, H. A., Mather, K. A., et al. (2008). The association of APOE genotype and cognitive decline in interaction with risk factors in a 65–69 year old community sample. BMC Geriatrics, 8, 14. doi:10.1186/1471-2318-8-14 Corder, E. H., Ghebremedhin, E., Taylor, M. G., Thal, D. R., Ohm, T. G., & Braak, H. (2004). The biphasic relationship between regional brain senile plaque and neurofibrillary tangle distributions: Modification by age, sex, and APOE polymorphism. Annals of the New York Academy of Sciences, 1019, 24–28. Davies, L., Wolska, B., Hilbich, C., Multhaup, G., Martins, R., Simms, G., et al. (1988). A4 amyloid protein deposition and the diagnosis of Alzheimer’s disease: Prevalence in aged brains determined by immunocytochemistry compared with conventional neuropathologic techniques. Neurology, 38, 1688–1693. Delis, D., Kramer, J., Kaplan, E., & Ober, B. (2000). California Verbal Learning Test—second edition. San Antonio, TX: The Psychological Corporation. Delis, D. C., Kaplan, E., & Kramer, J. H. (2001). The Delis-Kaplan Executive Function System (D-KEFS). San Antonio, TX: The Psychological Corporation. Ellis, K. A., Bush, A. I., Darby, D., De Fazio, D., Foster, J., Hudson, P., et al. (2009). The Australian Imaging, Biomarkers and Lifestyle (AIBL) study of aging: Methodology and baseline characteristics of 1112 individuals recruited for a longitudinal study of Alzheimer’s disease. International Psychogeriatrics, 21(4), 672–687. Fripp, J., Bourgeat, P., Acosta, O., Raniga, P., Modat, M., Pike, K. E., et al. (2008). Appearance modeling of 11C PiB PET images: Characterizing amyloid deposition in Alzheimer’s disease, mild cognitive impairment and healthy aging. NeuroImage, 43, 430–439. Gao, S., Hendrie, H., Hall, K. S., & Hui, S. (1998). The relationships between age, sex, and the incidence of dementia and Alzheimer disease: A meta-analysis. Archives of General Psychiatry, 55(9), 809–815. Gilleard, C. J. (1997). Education and Alzheimer’s disease: A review of recent international epidemiological studies. Ageing and Mental Health, 1(1), 33–46. Glodzik-Sobanska, L., Reisberg, B., De Santi, S., Babb, J. S., Pirraglia, E., Rich, K. E., et al. (2007). Subjective memory complaints: Presence, severity and future outcome in normal older subjects. Dementia and Geriatric Cognitive Disorders, 24, 177–184. Jessen, F., Wiese, B., Bachmann, C., Eifflaender-Gorfer, S., Haller, F., Kolsch, H., et al. (2010). Prediction of dementia by subjective memory impairment: Effects of severity and temporal association with cognitive impairment. Archives of General Psychiatry, 67, 414–422. Johnson, D. K., Storandt, M., Morris, J. C., & Galvin, J. E. (2009). Longitudinal study of the transition from healthy aging to Alzheimer disease. Archives of Neurology, 66(10), 1254–1259. Jonker, C., Geerlings, M. I., & Schmand, B. (2000). Are memory complaints predictive for dementia? A review of clinical and population-based studies. International Journal of Geriatric Psychiatry, 15, 983–991. Jorm, A. F., Masaki, K. H., David, D. G., Hardman, J., Nelson, J., Markesbery, W. R., et al. (2004). Memory complaints in nondemented men predict future pathologic diagnosis of Alzheimer disease. Neurology, 63, 1960–1961. Jungwirth, S., Fischer, P., Weissgram, S., Kirchmeyr, W., Bauer, P., & Tragl, K. (2004). Subjective memory complaints and objective memory impairment in the Vienna-Transdanube aging community. Journal of the American Geriatrics Society, 52, 263–268. Luciano, M., Gow, A. J., Taylor, M. D., Hayward, C., Harris, S. E., Campbell, H., et al. (2009). Apolipoprotein E is not related to memory abilities at 70 years of age. Behavior Genetics, 39, 6–14. Meyers, J. E., & Meyers, K. R. (1995). Rey Complex Figure Test and Recognition Trial. Professional Manual. Psychological Assessment Resource, Inc. Mintun, M. A., Larossa, G. N., Sheline, Y. I., Dence, C. S., Lee, S. Y., Mach, R. H., et al. (2006). [11C]PIB in a nondemented population: Potential antecedent marker of Alzheimer disease. Neurology, 67, 446–452. Mormino, E. C., Kluth, J. T., Madison, C. M., Rabinovici, G. D., Baker, S. L., Miller, B. L., et al. (2009). Episodic memory loss is related to hippocampal-mediated beta-amyloid deposition in elderly subjects. Brain, 132, 1310–1323.

Perneczky, R., Drzezga, A., Diehl-Schmid, J., Li, Y., & Kurz, A. (2007). Gender differences in brain reserve: An 18F-FDG PET study in Alzheimer’s disease. Journal of Neurology, 254, 1395–1400. Pike, C. J., Carroll, J. C., Rosario, E. R., & Barron, A. M. (2009). Protective actions of sex steroid hormones in Alzheimer’s disease. Frontiers in Neuroendocrinology, 30, 239–259. Pike, K. E., Savage, G., Villemagne, V. L., Ng, S., Moss, S. A., Maruff, P., et al. (2007). Beta-amyloid imaging and memory in non-demented individuals: Evidence for preclinical Alzheimer’s disease. Brain, 130, 2837–2844. Reiman, E. M., Chen, K., Liu, X., Bandy, D., Yu, M., Lee, W., et al. (2009). Fibrillar amyloid-beta burden in cognitively normal people at three levels of genetic risk for Alzheimer’s disease. Proceedings of the National Academy of Sciences of the United States of America, 106, 6820–6825. Reisberg, B., & Gauthier, S. (2008). Current evidence for subjective cognitive impairment (SCI) as the pre-mild cognitive impairment (MCI) stage of subsequently manifest Alzheimer’s disease. International Psychogeriatrics, 20, 1–16. Rentz, D. M., Locascio, J. J., Becker, J. A., Moran, E. K., Eng, E., Buckner, R. L., et al. (2010). Cognition, reserve, and amyloid deposition in normal aging. Annals of Neurology, 67, 353–364. Rowe, C. C., Ellis, K. A., Rimajova, M., Bourgeat, P., Pike, K. E., Jones, G., et al. (2010). Amyloid imaging results from the Australian imaging, biomarkers and lifestyle study of ageing (AIBL). Neurobiology of Aging, doi:10.1016/j.neurobiolaging.2010.04.007 Rowe, C. C., Ng, S., Ackermann, U., Gong, S. J., Pike, K., Savage, G., et al. (2007). Imaging beta-amyloid burden in aging and dementia. Neurology, 68, 1718–1725. Saxton, J., Ratcliff, G., Munro, C. A., Coffey, E., Becker, J. T., Fried, L., et al. (2000). Normative data on the Boston Naming Test and two equivalent 30-item short forms. The Clinical Neuropsychologist, 14, 526–534. Scarmeas, N., & Stern, Y. (2003). Cognitive reserve and lifestyle. Journal of Clinical and Experimental Neuropsychology, 25(5), 625–633. Small, B. J., Rosnick, C. B., Fratiglioni, L., & Backman, L. (2004). Apolipoprotein E and cognitive performance: A meta-analysis. Psychology and Aging, 19, 592–600. Snitz, B. E., Morrow, L. A., Rodriguez, E. G., Huber, K. A., & Saxton, J. A. (2008). Subjective memory complaints and concurrent memory performance in older patients of primary care providers. Journal of the International Neuropsychological Society, 14, 1008–1013. Sperling, R. A., Aisen, P. S., Beckett, L. A., Bennet, D. A., Craft, S., Fagan, A. M., et al. (2011). Toward defining the preclinical stages of Alzheimer’s disease: Recommendations from the National Institute on Aging and the Alzheimer’s Association workgroup. Alzheimer’s & Dementia, doi:10.1016/j.jalz. 2011.03.003 Stern, Y. (2002). What is cognitive reserve? Theory and research application of the reserve concept. Journal of the International Neuropsychological Society, 8, 448–460. Storandt, M., Mintun, M. A., Head, D., & Morris, J. C. (2009). Cognitive decline and brain volume loss as signatures of cerebral amyloid-␤ peptide deposition identified with Pittsburgh compound B: Cognitive decline associated with A␤ deposition. Archives of Neurology, 66, 1476–1481. Strauss, E., Sherman, E., & Spreen, O. (2006). A compendium of neuropsychological tests: Administration, norms, and commentary (3rd Ed.). New York: Oxford University Press. Twamley, E. W., Legendre Ropacki, S. A., & Bondi, M. W. (2006). Neuropsychological and neuroimaging changes in preclinical Alzheimer’s disease. Journal of the International Neuropsychological Society, 12, 707–735. Villemagne, V. L., Cappai, R., Barnham, K. J., Cherny, R., Opazo, C., Novakovic, K. E., et al. (2006). The Abeta centric pathway of Alzheimer’s disease. In C. J. Barrow, & B. J. Small (Eds.), Abeta peptide and Alzheimer’s disease (pp. 5–32). London: Springer-Verlag. Villemagne, V. L., Fodero-Tavoletti, M. T., Pike, K. E., Cappai, R., Masters, C. L., & Rowe, C. C. (2008). The ART of loss: A␤ imaging in the evaluation of Alzheimer’s disease and other dementias. Molecular Neurobiology, 38, 1–15. Villemagne, V. L., Pike, K. E., Darby, D., Maruff, P., Savage, G., Ng, S., et al. (2008). Abeta deposits in older non-demented individuals with cognitive decline are indicative of preclinical Alzheimer’s disease. Neuropsychologia, 46, 1688–1697. Weaver Cargin, J., Collie, A., Masters, C., & Maruff, P. (2008). The nature of cognitive complaints in healthy older adults with and without objective memory decline. Journal of Clinical and Experimental Neuropsychology, 30, 245–257. Wechsler, D. (1945). A standardized memory scale for clinical use. Journal of Psychology, 19, 87–95. Wechsler, D. (1997). Wechsler Adult Intelligence Scale—Third edition (WAIS-III). San Antonio, TX: Psychological Corporation. Welsh-Bohmer, K. A., Ostbye, T., Sanders, L., Pieper, C. F., Hayden, K. M., Tschanz, J. T., et al. (2009). Neuropsychological performance in advanced age: Influences of demographic factors and apolipoprotein E: Findings from the Cache county memory study. The Clinical Neuropsychologist, 23, 77–99. Yaffe, K., Weston, A., Graff-Radford, N. R., Satterfield, S., Simonsick, E. M., Younkin, S. G., et al. (2011). Association of plasma ␤-amyloid level and cognitive reserve with subsequent cognitive decline. Journal of the American Medical Association, 305(3), 261–266.