Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Mar 3.
Published in final edited form as: Neurobiol Aging. 2020 May 21;94:81–88. doi: 10.1016/j.neurobiolaging.2020.05.007

Association of Hippocampal Volume Polygenic Predictor Score with Baseline and Change in Brain Volumes and Cognition Among Cognitively Healthy Older Adults

Nicole M Armstrong 1, Logan Dumitrescu 2,3, Chiung-Wei Huang 1,4, Yang An 1, Toshiko Tanaka 1, Dena Hernandez 1, Jimit Doshi 5, Guray Erus 5, Christos Davatzikos 5, Luigi Ferrucci 1, Susan M Resnick 1,*, Timothy J Hohman 2,3,*
PMCID: PMC8893954  NIHMSID: NIHMS1608189  PMID: 32593031

Abstract

A high hippocampal volume polygenic predictor score (HV-PPS), computed based on GWAS summary statistics (n=33,536), could be protective against declines in brain volume and cognition in older adults. Linear mixed effects models with random intercepts and slopes were used to estimate associations of HV-PPS with baseline and annual rate of change in both brain volumes (n=508) and cognitive performance (n=1,041) in Caucasian Baltimore Longitudinal Study of Aging participants. Higher HV-PPS was associated with greater baseline volumes of the hippocampus and parahippocampal gyrus, and slower rates of ventricular enlargement and volume loss in frontal and parietal white-matter, all adjusted for intracranial volume. Additionally, higher HV-PPS was associated with better executive function performance and slower rates of decline in verbal fluency scores over time. Individuals with a genetic predisposition towards larger hippocampal volumes show better baseline executive function, slower decline in verbal fluency performance, and slower rates of longitudinal brain atrophy.

Keywords: polygenic predictor score, hippocampus, cognitive function, genetics, APOE

1. Introduction

Alzheimer’s disease (AD) is the sixth leading cause of death in the United States and the only leading cause of death without a treatment or cure (Alzheimer’s Association, 2017). It is characterized by neuropathological and neurodegenerative changes that originate in the medial temporal lobe (Braak, 1993; Jack et al., 1997; Jobst et al., 1994; Pegueroles et al., 2017; Thal, Rüb, Orantes, & Braak, 2002; Visser, Verhey, Hofman, Scheltens, & Jolles, 2002). The hippocampus, located within the medial temporal lobe, is one of the first brain structures to show disease progression (Apostolova et al., 2006; C.R. Jack et al., 2000). Thus, hippocampal atrophy is considered to be a biomarker of neurodegeneration. Magnetic resonance imaging (MRI) measures of hippocampal atrophy have become a central component of the AD neuropathological cascade and are included in the “N” category of the “A/T/N” classification scheme (Jack et al., 2016). Further, lower baseline hippocampal volumes among cognitively normal older adults have been shown to predict longitudinal cognitive decline (Golomb et al., 1996; R. C. Petersen et al., 2000; Raz & Rodrigue, 2006) and an increased risk of conversion to cognitive impairment (Apostolova et al., 2006; C. R. Jack et al., 2000; Raz & Rodrigue, 2006). For that reason, individuals who have a high genetic risk for small hippocampal volumes could be at higher risk for neurodegeneration and subsequent cognitive impairment.

Hippocampal volumes are highly heritable (h2~70%) (den Braber et al., 2013; Kremen et al., 2010; Rentería et al., 2014); thus, the genetic architecture of hippocampal volume has been the focus of many analyses. In the largest study to date, a GWAS meta-analysis of mean bilateral hippocampal volume was performed in 26,814 individuals of European ancestry, and six independent loci were identified as being associated with hippocampal volume (Hibar et al., 2017). Apart from these six independent loci, there were many variants with small effect sizes that did not meet genome-wide significance but may still contribute to the overall genetic architecture of hippocampal volume. Indeed, Hibar et al. (2017) estimated that nearly ~19% of the variance in hippocampal volume was explained by common variants across the whole-genome. For complex traits that are affected by many genetic variants with small effects, a polygenic score that incorporates information from all potential causal loci may be a more powerful method to determine genetic risk than simply examining loci that survive GWAS-level significance thresholds (Escott-Price et al., 2015; Lee et al., 2013; Ridge, Mukherjee, Crane, Kauwe, & Alzheimer’s Disease Genetics Consortium, 2013). It is unknown whether genetic contributors to normal variation in hippocampal volumes are relevant to late-life risk for neurodegeneration and cognitive decline. Therefore, we sought to use a hippocampal volume polygenic predictor score (HV-PPS), based on results from the large meta-analysis of hippocampal volume by Hibar et al. (2017), to evaluate the late-life effects of these multiple identified genetic variants (Marden, Walter, Tchetgen Tchetgen, Kawachi, & Glymour, 2014; Verhaaren et al., 2013).

In this study of older adults who are cognitively normal at baseline, we examine associations of the HV-PPS with baseline and change in brain volumes, general cognitive performance, and domain-specific cognitive performance up to 25 years. We hypothesize that higher HV-PPS will correlate with higher baseline and slower change in temporal lobe structures, especially areas affected by early AD pathology, i.e., hippocampus, entorhinal cortex and parahippocampal gyrus (Convit et al., 2000; Driscoll et al., 2009; Du et al., 2004; Galton et al., 2001). Change in brain gray matter volume is associated with change in cognition in older adults, with atrophy in temporal lobe gray matter being a strong predictor of longitudinal cognitive decline (Fletcher et al., 2018). Since there is a strong association between brain and behavior, we hypothesize that higher HV-PPS will also be associated with slower declines in cognitive performance, especially memory.

2. Methods

2.1. Characteristics of the Study Sample

This study included 1,041 Caucasian participants, aged ≥50 years, from the Baltimore Longitudinal Study of Aging (BLSA) who had both genome-wide genetic data and cognitive testing completed up to 2016. Supplementary Table 1 has the dates of test administration for each cognitive test. Test administration generally occurred biennially until 2003, although a subset of participants involved in the neuroimaging substudy received annual cognitive assessments from 1994–2004. Since 2003, participants aged 50–59 years had visits every four years, while those aged 60–79 years had biennial visits and those aged 80 and older had annual visits. There were 6,510 total observations. Of the 1,041 participants, there were 508 participants with structural magnetic resonance imaging (MRI).

The procedures for determination of subsequent cognitive impairment (mild cognitive impairment [MCI] or dementia/AD) have been detailed previously (S. Resnick, Pham, Kraut, Zonderman, & Davatzikos, 2003). Research diagnoses of cognitive impairment were determined using consensus case conference procedures. MCI was determined by Petersen criteria (R. Petersen, 2004). Dementia and AD, respectively, were based on DSM-III-R criteria (Millon, 1981) and the NINCDS-Alzheimer’s Disease and Related Disorders Association criteria (McKhann et al., 1984). The study protocol for both studies were reviewed and approved by the Internal Review Board of the National Institute for Environmental Health Sciences and all participants provided written informed consent.

2.2. GWAS Genotyping and Processing

Participants were genotyped on either the Illumina 550K or NeuroChip (Blauwendraat et al., 2017). All genotype data were on Build 37, and they went through the same genotyping quality control (QC) and imputation pipeline (Supplementary Figure 1). Variants were filtered for poor call rate (missing>1%), for violations of Hardy-Weinberg Equilibrium (p>1×10−6), and for minor allele frequency (<1%). Samples were excluded for poor genotyping efficiency (missing>2%), sex inconsistences, cryptic relatedness (pihat>0.25), or if they were not Caucasian by self-report or by principal component detection. Imputation was then performed on each individual dataset by the Michigan Imputation Server (https://imputationserver.sph.umich.edu/) using the HRC r1.1.2016 reference panel (Das et al., 2016). Post-imputation, the two datasets were merged, overlapping samples were removed (as some participants were genotyped on both platforms), and SNPs with low imputation quality (R2<0.9), minor allele frequencies <1%, and Hardy-Weinberg Equilibrium p-values <1×10−6 were excluded, along with SNPs that did not overlap between the two datasets (missingness <99%). The final genetic dataset included 5,439,477 SNPs and 1184 individuals, of which 1041 and 508 individuals also had cognitive assessments and neuroimaging data, respectively. BLSA data were not included in the ENIGMA consortium.

2.3. Computation of the Hippocampal Volume Polygenic Predictor Score

The HV-PPS was generated in PLINK (Chang et al., 2015) (www.cog-genomics.org/plink/1.9/) using summary results from a published hippocampal volume GWAS (Hibar et al., 2017), which included 26,814 individuals of European ancestry with mean bilateral hippocampal volume. Prior to score calculation, all strand-ambiguous SNPs and SNPs absent from the BLSA cohort were removed. SNPs were then pruned using PLINK’s linkage disequilibrium based clumping procedure based on: an r2>0.25, a 200-kb sliding window, and a hippocampal volume GWAS association p-value<0.01. We selected a p-value threshold of 0.01 following Mormino et al. (2016). A total of 12,148 SNPs was included in the final score. HV-PPS were then calculated for each individual by summing the reference allele counts at each SNP weighted by the beta from the hippocampal volume GWAS. Polygenic predictor scores were calculated both including and excluding the APOE region (i.e., 1 MB upstream and downstream of the gene; chr19:44409039–46412650).

2.4. Cognitive Assessments

We assessed the composite z-scores of memory, attention, executive function, verbal fluency, and visuospatial domains. Z-scores were created using the baseline means and standard deviations of each neuropsychological test score. Memory consisted of the immediate and long-delayed free recall scores from the California Verbal Learning Test (CVLT) (Delis, Kramer, Kaplan, & Ober, 1987). The delay time on the CVLT was approximately 20 minutes. Executive function consisted of completion time for Trail-Making Test Part B (TMT-B) (Reitan, 1958) and the total score from Digit Span Backward (Wechsler, 1981). Attention was comprised of TMT Part A (TMT-A) (Reitan, 1958) and Digit Span Forward (Wechsler, 1981). Completion times of TMT-A and TMT-B were truncated at 300 seconds. Due to the distributions of TMT-A and TMT-B, we log-transformed the completion times, standardized these tests, and reversed the signs so that higher scores reflected better performance. The composite for verbal fluency consisted of the total numbers of correct words generated in 60 seconds for the letter (Benton, 1968) and category fluency (Newcombe, 1969) tests. Visuospatial composite consisted of a modified version of the Educational Testing Service Card Rotations test (Wilson et al., 1975), based on the total number correct minus total number incorrect and the Clock Drawing Test (Rouleau, Salmon, Butters, Kennedy, & McGuire, 1992), using times of 11:10 and 3:25. A global composite score was created from the average z-scores of all domains. All tests were administered by highly trained psychometricians at each study visit.

2.5. Image Acquisition

MRI scanning was performed on a General Electric (GE) Signa 1.5-T scanner (Milwaukee, WI) or a 3-T Philips Achieva. GE 1.5-T scans used a high-resolution volumetric spoiled gradient recalled acquisition in a steady state (GRASS) series (axial acquisition, repetition time=35msec, echo time=5msec, flip angle=45°, field of view=24 cm, matrix=256×256, number of excitations=1, voxel dimensions=0.94×0.94×1.5 mm slice thickness). T1-weighted magnetization-prepared rapid gradient echo (MPRAGE) scans were acquired on a 3-T Philips Achieva (repetition time [TR]=6.8msec, echo time [TE]=3.2msec, flip angle=8°, image matrix=256×256, 170 slices, pixel size=1×1mm, slice thickness=1.2mm). There were 156 participants (472 scans) with 3-T MPRAGE images and 111 participants (854 scans) with 1.5-T SPGR images at baseline with available GWAS data. Participants receiving 1.5-T scans comprised enrollees in the original BLSA neuroimaging substudy dating back to 1994 (S. M. Resnick et al., 2000).

2.6. Harmonization of MUSE Anatomical Labels across 1.5-T SPGR and 3-T MPRAGE

A new automated labeling method specifically designed to achieve a consistent parcellation of brain anatomy in longitudinal MRI studies with scanner and imaging protocol differences was used to harmonize BLSA MRI data. This method combines the MUSE anatomical labeling approach (Doshi et al., 2016) with harmonized acquisition-specific atlases (Erus et al., 2018). The approach is described in more detail in Erus et al. (2018). Briefly, using 35 labeled 3-T MPRAGE brain MRIs from the OASIS data set (available for download at https://masi.vuse.vanderbilt.edu/workshop2012) as atlases, we first performed the MUSE labeling method on 3-T MPRAGE images for 32 BLSA participants with 1.5-T SPGR at an earlier time point. Then, for each participant, we deformably registered their 1.5-T SPGR image to their 3-T MPRAGE image using a robust registration strategy that combines an ensemble of registrations obtained using two different algorithms and multiple smoothness parameters. From these steps, we obtained 32 pairs of 1.5-T SPGR and 3-T MPRAGE images in the same space with common anatomical labels, which served as atlases in the MUSE approach to obtain labels on the entire BLSA collection of 1.5-T SPGR and 3-T MPRAGE images. This workflow for anatomical labeling has been extensively validated on BLSA MRI data (Erus et al., 2018), and the harmonized values were used in previous studies (Armstrong et al., 2019a; Armstrong et al., 2019b).

2.7. Regions of Interest

First, we examined global regions, i.e., total brain, gray matter (GM), white matter (WM) and ventricles, and lobar GM and WM (frontal, temporal, parietal, and occipital). Then, we examined volumes in medial temporal regions previously reported to show early neurodegeneration in AD (Convit et al., 2000; Driscoll et al., 2009; Du et al., 2004; Galton et al., 2001). These regions of interest (ROIs) were hippocampus, entorhinal cortex, and parahippocampal gyrus.

2.8. Statistical Analyses

Means for continuous covariates and frequencies for categorical covariates were reported for the overall sample. Linear mixed effects models with unstructured variance-covariance were used to examine the association of the HV-PPS with baseline and change in volumes in brain ROIs and cognitive performance (general and domain-specific). These models also included the fixed effects of mean-centered baseline age, sex (male vs. female), years of education, subsequent impaired status (vs. cognitively normal), time since baseline and the following two-way interactions: age*time and sex*time. For brain volume analyses, we further adjusted for the fixed effects of mean-centered intracranial volume (ICV) at age 70 and image type (1.5-T SPGR vs. 3-T MPRAGE) (Erus et al., 2018). For the cognitive analysis, baseline was defined as the first cognitive assessment once the participant was at least 50 years of age. Cognitive assessments prior to age 50 were excluded. We excluded array platform (Illumina 550K or NeuroChip) as a covariate, since estimates did not change with the inclusion of this covariate in either analysis. Also, HV-PPS scores were not associated with genotyping platform (p=0.75 using generalized linear models) (Supplementary Figure 2).

To allow for baseline and slope differences between individuals, we included random intercepts and slopes for time. All analyses were performed in SAS (Cary, NC) or R version 3.4.1 (https://www.r-project.org/). We examined the collinearity of baseline hippocampal volume and HV-PPS through variance inflation factors (VIFs) to determine whether baseline hippocampal volume could be added to the models. The VIFs were estimated in multivariable linear regression models with ventricular size and WM volumes as outcomes and HV-PPS as a predictor. These models were adjusted by ICV at 70, scanner type, education, mean-centered baseline age, sex, subsequent impaired status and baseline hippocampal volume. Type I error level was set to 0.05 for analyses, although with Bonferroni correction for multiple comparisons for the 6 cognitive domains and 15 ROIs, results would be considered significant at p<0.008 and p<0.003 respectively. Three-way interactions were considered significant at p<0.10.

2.9. Sensitivity Analyses

We conducted two sets of sensitivity analyses for each outcome. Analyses were re-run with HV-PPS excluding the APOE region. Additionally, analyses were re-run excluding all participants who developed cognitive impairment during the follow-up period. We also compared the HV-PPS with clumping parameters r2>0.1 with 1MB sliding window to the HV-PPS used in the main analysis to see if inferences remained.

3. Results

3.1. Characteristics of the study sample

Table 1 shows the characteristics of the study sample (N=1,041). There were 578 (55.5%) females and 237 (22.8%) participants who developed cognitive impairment during follow-up. Mean baseline age was 68.3 (Standard Deviation, [SD]=9.3) years, and mean length of follow-up was 15.0 (SD=7.2) years. The mean years of education were 16.5 (SD=2.5) years. There were 508 participants who also had serial MRI assessments (n=2,018). Of these, 84 (16.5%) developed cognitive impairment. Characteristics of the subsample with MRI assessments were similar to those of the overall sample with the exception of follow-up time (Table 1). The mean follow-up time for the MRI subsample was 4.7 (SD=5.1) years.

Table 1.

Participant characteristics of the study sample from Baltimore Longitudinal Study of Aging

Characteristics Participants with Cognitive Assessments N = 1041 Participants with Both Cognitive Assessments and Neuroimaging N = 508
Number of Observations 6510 2018
Baseline Age, in years, mean (SD) 68.3 (9.3) 72.3 (9.2)
Female, n (%) 578 (55.5) 246 (48.4)
Years of Education, mean (SD) 16.5 (2.5) 17.0 (2.6)
Subsequent Cognitive Impairment*, n (%) 237 (22.8) 84 (16.5)
Intracranial Volume at Age 70, in cm3, mean (SD) --- 1415.8 (134.8)
NeuroChip, n (%) 451 (43.3) 236 (46.5)
3-T MPRAGE Images, n (%) --- 388 (76.38)
Number of Visits, mean (SD) 6.3 (4.3) 4.0 (3.7)
Range of Number of Visits 1-25 1-21
Follow-up Time, in years, mean (SD) 15.0 (7.2) 4.7 (5.1)
Follow-up Time for those with ≥2 assessments, in years, mean (SD) 15.2 (6.9) 6.2 (5.0)
Open in a new tab

SD = standard deviation

*

Subsequent Cognitive Impairment is defined as an adjudicated diagnosis of either Mild Cognitive Impairment or dementia.

Note: 921 participants with cognitive assessments had two or more assessments. 384 participants had both cognitive assessments and neuroimaging.

3.2. Associations of hippocampal volume polygenic predictor score with baseline and change in brain volumes

Table 2 shows the associations of the HV-PPS with baseline and change in brain volume. As expected, greater HV-PPS was associated with greater baseline hippocampal volume (estimate=0.1616, SE=0.0292, p<0.0001), which survived Bonferroni correction. Figure 1 shows the association between HV-PPS and hippocampal volume. Greater HV-PPS also was associated with greater baseline volumes in parahippocampal gyrus (estimate=0.0826, SE=0.0265, p=0.0020). In longitudinal analyses, HV-PPS was associated with slower annual rates of change of ventricles and WM regions, as shown in Figure 2. Greater HV-PPS was associated with slower rates of ventricular dilation (estimate=−0.1407, SE=0.0460, p=0.0024) and slower rates of WM atrophy (estimate=0.1854, SE=0.0696, p=0.0090), specifically in the frontal WM (estimate=0.0738, SE=0.0293, p=0.0132) and parietal WM (estimate=0.0475, SE=0.0170, p=0.0062). Greater HV-PPS was associated with slower rates of GM volumetric decline (estimate=0.3059, SE=0.1382, p=0.0280), but this did not survive Bonferroni correction.

Table 2.

Associations of hippocampal volume polygenic predictor score (HV-PPS) with baseline and change in brain volumes (N=508)

Association of HV-PPS with Baseline Brain Volumes Association of HV-PPS with Change in Brain Volumes
Brain Regions of Interest Estimate SE P Estimate SE P
Total brain 1.5487 1.7698 0.38 0.3381 0.1737 0.05
Gray matter 0.8702 1.1908 0.47 0.3059 0.1382 0.028
 Frontal 0.1124 0.4488 0.80 0.0892 0.0470 0.06
 Temporal 0.0670 0.2845 0.81 0.0522 0.0272 0.06
 Parietal 0.0194 0.2715 0.94 0.0481 0.0268 0.07
 Occipital 0.0229 0.2478 0.93 0.0167 0.0232 0.47
White matter 1.1298 0.9766 0.25 0.1854 0.0696 0.0090
 Frontal 0.3320 0.4425 0.45 0.0738 0.0293 0.0132
 Temporal 0.1176 0.2466 0.63 0.0285 0.0176 0.11
 Parietal 0.1329 0.2374 0.58 0.0475 0.0170 0.0062
 Occipital 0.0989 0.1641 0.55 0.0128 0.0106 0.23
Ventricles −0.8788 0.6974 0.21 −0.1407 0.0460 0.0024*
Hippocampus 0.1616 0.0292 <0.0001* 0.0035 0.0025 0.15
Parahippocampal gyrus 0.0826 0.0265 0.0020* 0.0015 0.0021 0.48
Entorhinal cortex 0.0249 0.0220 0.26 0.0023 0.0025 0.34
Open in a new tab

HV-PPS = Hippocampal Volume Polygenic Predictor Score; SE = standard error

Bolded values indicate p<0.05.

*

Significant at Bonferroni corrected p<0.003.

Note: All analyses are adjusted by baseline mean-centered age, sex (male vs. female), subsequent cognitive impairment status, mean-centered years of education, mean-centered intracranial volume at age 70, image type (3-T vs. 1.5-T), and two-way interactions of age and sex with time since baseline.

Figure 1.

Figure 1.

Open in a new tab

Baseline association between HV-PPS and hippocampal volume after covariate adjustment.

Figure 2. Longitudinal associations of time on study with ventricles and frontal and parietal white matter (WM) by standard deviations of hippocampal volume polygenic predictor score (HV-PPS).

Figure 2.

Open in a new tab

Note: HV-PPS categories mean the following: −1 is one standard deviation below the mean Z-score for the HV-PPS; 0 is the mean Z-score for HV-PPS; 1 is one standard deviation above the mean Z-score for HV-PPS. All analyses are adjusted by baseline mean-centered age, sex, subsequent cognitive impairment status, years of education, mean-centered intracranial volume at age 70, image type (3-T vs. 1.5-T images) and two-way interactions of age and sex with time. Those with one standard deviation above the mean Z-score had greater frontal and parietal white matter volumes and less declines over time as well as lower ventricular size and less steep ventricular enlargement over time, as compared to the other two groups.

Secondary analyses (covarying for both baseline hippocampal volume and baseline hippocampal volume*time interaction) indicated that these associations with HV-PPS were independent of baseline hippocampal volume (Supplementary Table 2). After baseline hippocampal volume adjustment, HV-PPS was no longer associated with greater baseline parahippocampal gyrus volumes (estimate=0.0058, SE=0.0238, p=0.81). Additionally, greater HV-PPS was associated with slower rates of volume loss in total GM (estimate=0.3129, SE=0.1383, p=0.0237). Furthermore, there was no evidence of collinearity among HV-PPS and baseline hippocampal volume (VIF=1.04).

3.3. Associations of hippocampal volume polygenic predictor score with baseline and change in cognitive performance

Table 3 contains the results from the associations of HV-PPS with cognitive performance. Higher HV-PPS was associated with better baseline executive function performance (estimate=0.0458, SE=0.0212, p=0.031), but not with change in executive function performance. While HV-PPS was not associated with baseline verbal fluency scores, higher HV-PPS was associated with slower rates of decline in verbal fluency scores over time (estimate=0.0048, SE=0.0018, p=0.0071). No other cognitive associations were observed, including null associations with memory performance (Table 3).

Table 3.

Associations of hippocampal polygenic predictor score (HV-PPS) with baseline and change in cognition

Association of HV-PPS with Baseline Cognition Association of HV-PPS with Change in Cognition
Outcome N Estimate SE P Estimate SE P
Memory Composite 1025 0.0039 0.0226 0.86 0.0038 0.0024 0.12
Attention Composite 986 0.0187 0.0203 0.36 0.0018 0.0016 0.25
Executive Function Composite 986 0.0458 0.0212 0.031 0.0005 0.0016 0.77
Verbal Fluency Composite 1041 −0.0238 0.0224 0.29 0.0048 0.0018 0.0071*
Visuospatial Composite 926 −0.0052 0.0189 0.78 0.0019 0.0018 0.28
Global Composite 904 0.0252 0.0156 0.11 0.0004 0.0013 0.75
Open in a new tab

HV-PPS = Hippocampal Volume Polygenic Predictor Score; SE = standard error

Bolded values indicate p<0.05.

*

Significant at Bonferroni corrected p<0.008.

Note: All analyses are adjusted by baseline mean-centered age, sex, subsequent cognitive impairment status, mean-centered years of education, and two-way interactions of age and sex with time since baseline.

3.4. Sensitivity Analyses

To preclude any influence of the APOE gene, SNPs 1MB upstream and downstream of the APOE locus were excluded from the HV-PPS, and analyses were repeated. All associations excluding the APOE region, whether with brain volumes or cognitive performance, were similar to those including the APOE region (Supplementary Tables 3 and 4). Analyses were also restricted to those who remained cognitively normal. Inferences did not change (Supplementary Tables 5 and 6). For the comparison between the HV-PPS with clumping parameters r2>0.1 with 1MB sliding window to the HV-PPS used in the main analysis, cross-sectional association between HV-PPS and baseline executive function was no longer significant, yet the results for verbal fluency still held (Supplementary Table 7).

4. Discussion

As expected, higher HV-PPS, regardless of inclusion or exclusion of APOE region, was associated with greater baseline hippocampal volumes, validating the integrity of the predictor score in our sample. Additionally, in support of our cross-sectional hypotheses, HV-PPS was associated with higher volumes in parahippocampal gyrus, a structure affected by early AD pathology. We found that higher HV-PPS was related to slower rates of ventricular dilation and volume loss in frontal and parietal WM in the overall sample, but there were no significant longitudinal associations between HV-PPS and change in temporal lobe structures. Regarding the associations between HV-PPS and domain-specific cognition, our hypotheses that HV-PPS was associated with memory were not supported. Higher HV-PPS was associated with higher executive function cross-sectionally and slower longitudinal rates of decline in verbal fluency.

The associations of greater HV-PPS with higher volumes in the hippocampus and parahippocampal gyrus highlights the specificity of the score in relation to the medial temporal lobe. The hippocampus has extensive connections with neighboring cortical areas, i.e., perirhinal cortex and parahippocampal cortex, for memory-guided behavior (Ranganath & Ritchey, 2012). The parahippocampal cortex is part of the posterior medial system, which includes the retrosplenial cortex, posterior cingulate, precuneus, and angular gyrus among other regions (Ranganath & Ritchey, 2012). There were no other baseline associations between HV-PPS and other volumes of interest.

In contrast, the associations of greater HV-PPS with slower rates of declines in ventricular size and WM, especially frontal and parietal WM, in the overall sample highlight that HV-PPS may reflect a protective effect, independent of baseline hippocampal volumes. Ventricular size is a well-established indicator of global atrophy, including in BLSA data from cognitively normal older adults (Resnick et al., 2000), and is highly sensitive to age. HV-PPS effect on WM may reflect the broad WM connections between cortical regions that support cognitive performance. However, more detailed WM measures are needed to further elucidate the tracts underlying the observed associations. HV-PPS was not associated with GM after correction for multiple comparisons. These findings suggest that genetic predictors of hippocampal volume at midlife are predictive of global age-related atrophy outside of the medial temporal lobe. Notably, these global associations are not dependent on baseline hippocampal volume, suggesting pleiotropic effects of HV-PPS on brain structures.

Regarding cognitive performance, higher HV-PPS was associated with higher baseline executive function performance and slower rates of decline in verbal fluency. HV-PPS was not significantly associated with memory, although the magnitude of the estimate was positive. The association between higher HV-PPS and both executive function performance and change over time in frontal and parietal volumes may reflect established relationships between executive function and WM volumes (Bennett & Madden, 2014; Madden et al., 2012). Additionally, signal abnormalities in WM are most often related to poorer performance on measures of executive function (Gunning-Dixon & Raz, 2000). In terms of the association between HV-PPS and verbal fluency performance, we have shown that verbal fluency performance is associated with volumes of many brain regions, including the hippocampus, reflecting the fact that many different cognitive functions contribute to verbal fluency (Armstrong et al., 2019).

The HV-PPS score may be useful in elucidating contributors to brain reserve. Brain reserve is defined by having higher neuronal or synaptic counts (Stern, 2009), which may allow an individual to withstand more severe disease pathology to prolong the preclinical stage of AD until a critical threshold is reached (Satz, Cole, Hardy, & Rassovsky, 2011). Presumably, individuals who have greater brain reserve have larger brain volumes, more neurons, and/or elaborate synaptic networks that serve as protection from cognitive decline (Katzman et al., 1988; Satz, 1993; Stern, 2002; Whalley, Deary, Appleton, & Starr, 2004). Genetic variation could play a role in brain reserve, and our findings suggest that the HV-PPS not only predicts volumes of hippocampal and related structures that are affected in early AD, but also may be protective against some aspects of cognitive decline and brain atrophy. A future direction would be to compare the rates of change in hippocampal volume between young and older adults, since this may have further implications on brain reserve.

There were several strengths and limitations of the study. This study consists of an extensively characterized sample of older adults with repeated measures. Second, our image processing pipeline uses state-of-the-art and validated multi-atlas approaches for regional definition, yielding high measurement stability over time. However, our sample is highly educated, relatively healthy, and Caucasian, thus limiting generalizability. Other limitations include relatively low sample size and the absence of replication.

In conclusion, our findings indicate that higher HV-PPS was associated with greater baseline volumes in the hippocampus and parahippocampal gyrus, yet it was not associated with change in these brain volumes over time. Thus, genetic loading for greater hippocampal volume may have a protective effect on overall level of hippocampal volume and may help maintain volume over time. Additionally, the HV-PPS was associated with slower rates of decline in verbal fluency over time in cognitively normal older adults. Our findings should motivate additional research to determine the contribution of the HV-PPS to the brain reserve hypothesis.

Supplementary Material

1

Highlights.

  • Higher HV-PPS predicts larger hippocampal and parahippocampal gyrus volumes.

  • Higher HV-PPS predicts slower declines in white matter (WM) and verbal fluency.

  • Higher HV-PPS predicts slower rates of ventricular enlargement (VE).

  • Higher HV-PPS predicts higher baseline executive function.

  • Association of HV-PPS with WM and VE change were independent of hippocampal volume.

Acknowledgments

We would like to thank the participants and staff of the Baltimore Longitudinal Study of Aging, the neuroimaging staff of the Laboratory of Behavioral Neuroscience, and the staff of the Johns Hopkins and National Institute on Aging MRI facilities.

Funding: This research was supported fully by the Intramural Research Program of the National Institutes of Health, National Institute on Aging.

Role of Sponsor: The authors of this manuscript include employees of the Intramural Research Program of the National Institute on Aging, who participated in all aspects of the project.

Sources of Financial Support: This research was supported by the Intramural Research Program of the National Institutes of Health, National Institute on Aging.

Footnotes

Conflicts of Interest: The authors report no conflicts of interest.

Statement of Data Not Being Submitted, Previously Published, and Not Being Submitted Elsewhere: Data contained in the manuscript being submitted had not been previously published, have not been submitted elsewhere and will not be submitted elsewhere while under consideration at Neurobiology of Aging.

Appropriate Approval and Procedures Concerning Human Subjects and/or Animals: Appropriate approval and procedures were used concerning human subjects. The local Institutional Review Board approved the research protocol for this study, and written informed consent was obtained at each visit from all participants.

Verification of Final Approval and Review from All Authors: All authors have reviewed the contents of the manuscript being submitted, approve of its contents, and validate the accuracy of the data.

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. Alzheimer’s Association. (2017). 2017 Alzheimer’s disease facts and figures. Alzheimer’s & Dementia, 13(4), 325–373. [Google Scholar]
  2. Apostolova LG, Dutton RA, Dinov ID, Hayashi KM, Toga AW, Cummings JL, & Thompson PM (2006). Conversion of mild cognitive impairment to Alzheimer disease predicted by hippocampal atrophy maps. Archives of Neurology, 63(5), 693–699. [DOI] [PubMed] [Google Scholar]
  3. Armstrong N, An Y, Williams OA, Doshi J, Erus G, Davatzikos C, & Resnick SM (2019). Change in cognition is correlated with change in brain volumes in cognitively normal older adults. Alzheimer’s & Dementia, 15(7), P882–P883. doi: 10.1016/j.jalz.2019.06.4638 [DOI] [Google Scholar]
  4. Bennett IJ, & Madden DJ (2014). Disconnected aging: cerebral white matter integrity and age-related differences in cognition. Neuroscience, 276, 187–205. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Benton AL (1968). Differential behavioral effects in frontal lobe disease. Neuropsychologia, 6(1), 53–60. [Google Scholar]
  6. Blauwendraat C, Faghri F, Pihlstrom L, Geiger JT, Elbaz A, Lesage S, … Scholz SW (2017). NeuroChip, an updated version of the NeuroX genotyping platform to rapidly screen for variants associated with neurological diseases. Neurobiology of Aging, 57, 247.e249–247.e213. doi: 10.1016/j.neurobiolaging.2017.05.009 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Braak H (1993). Neuropathological staging of Alzheimer-related changes correlates with psychometrically assessed intellectual status. Paper presented at the Alzheimer’s Disease: Advances in Clinical and Basic Research. Third International Conference on Alzheimer’s Disease and Related Disorders, 1993. [Google Scholar]
  8. Chang CC, Chow CC, Tellier LC, Vattikuti S, Purcell SM, & Lee JJ (2015). Second-generation PLINK: rising to the challenge of larger and richer datasets. GigaScience, 4, 7. doi: 10.1186/s13742-015-0047-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Convit A, de Asis J, de Leon MJ, Tarshish CY, De Santi S, & Rusinek H (2000). Atrophy of the medial occipitotemporal, inferior, and middle temporal gyri in nondemented elderly predict decline to Alzheimer’s disease☆. Neurobiology of Aging, 21(1), 19–26. doi: 10.1016/S0197-4580(99)00107-4 [DOI] [PubMed] [Google Scholar]
  10. Das S, Forer L, Schönherr S, Sidore C, Locke AE, Kwong A, … Fuchsberger C (2016). Next-generation genotype imputation service and methods. Nature genetics, 48(10), 1284–1287. doi: 10.1038/ng.3656 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Delis D, Kramer J, Kaplan E, & Ober B (1987). California Verbal Learning Test, Research edition. New York, NY: Psychological Corporation. [Google Scholar]
  12. den Braber A, Bohlken MM, Brouwer RM, van ‘t Ent D, Kanai R, Kahn RS, … Boomsma DI (2013). Heritability of subcortical brain measures: a perspective for future genome-wide association studies. NeuroImage, 83, 98–102. doi: 10.1016/j.neuroimage.2013.06.027 [DOI] [PubMed] [Google Scholar]
  13. Doshi J, Erus G, Ou Y, Resnick SM, Gur RC, Gur RE, … The Alzheimer’s Neuroimaging Initiative. (2016). MUSE: MUlti-atlas region Segmentation utilizing Ensembles of registration algorithms and parameters, and locally optimal atlas selection. NeuroImage, 127, 186–195. doi: 10.1016/j.neuroimage.2015.11.073 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Driscoll I, Davatzikos C, An Y, Wu X, Shen D, Kraut M, & Resnick SM (2009). Longitudinal pattern of regional brain volume change differentiates normal aging from MCI. Neurology, 72(22), 1906–1913. doi: 10.1212/WNL.0b013e3181a82634 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Du AT, Schuff N, Kramer JH, Ganzer S, Zhu XP, Jagust WJ, … Weiner MW (2004). Higher atrophy rate of entorhinal cortex than hippocampus in AD. Neurology, 62(3), 422–427. Retrieved from http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1820859/ [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Erus G, Doshi J, An Y, Verganelakis D, Resnick SM, & Davatzikos C (2018). Longitudinally and inter-site consistent multi-atlas based parcellation of brain anatomy using harmonized atlases. NeuroImage, 166, 71–78. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Escott-Price V, Sims R, Bannister C, Harold D, Vronskaya M, Majounie E, … Williams J (2015). Common polygenic variation enhances risk prediction for Alzheimer’s disease. Brain, 138(Pt 12), 3673–3684. doi: 10.1093/brain/awv268 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Fletcher E, Gavett B, Harvey D, Farias ST, Olichney J, Beckett L, … Mungas D (2018). Brain volume change and cognitive trajectories in aging. Neuropsychology, 32(4), 436–449. doi: 10.1037/neu0000447 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Galton CJ, Patterson K, Graham K, Lambon-Ralph MA, Williams G, Antoun N, … Hodges JR (2001). Differing patterns of temporal atrophy in Alzheimer’s disease and semantic dementia. Neurology, 57(2), 216–225. doi: 10.1212/wnl.57.2.216 [DOI] [PubMed] [Google Scholar]
  20. Golomb J, Kluger A, de Leon MJ, Ferris SH, Mittelman MP, Cohen J, & George AE (1996). Hippocampal formation size predicts declining memory performance in normal aging. Neurology, 47(3), 810–813. doi: 10.1212/wnl.47.3.810 [DOI] [PubMed] [Google Scholar]
  21. Gunning-Dixon FM, & Raz N (2000). The cognitive correlates of white matter abnormalities in normal aging: a quantitative review. Neuropsychology, 14(2), 224–232. doi: 10.1037/0894-4105.14.2.224 [DOI] [PubMed] [Google Scholar]
  22. Hibar DP, Adams HHH, Jahanshad N, Chauhan G, Stein JL, Hofer E, … Ikram MA (2017). Novel genetic loci associated with hippocampal volume. Nature Communications, 8, 13624. doi: 10.1038/ncomms13624 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Jack CR, Bennett DA, Blennow K, Carrillo MC, Feldman HH, Frisoni GB, … Dubois B (2016). A/T/N: An unbiased descriptive classification scheme for Alzheimer disease biomarkers. Neurology, 87(5), 539–547. doi: 10.1212/WNL.0000000000002923 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Jack CR, Petersen RC, Xu Y, O’Brien PC, Smith GE, Ivnik RJ, … Kokmen E (2000). Rates of hippocampal atrophy correlate with change in clinical status in aging and AD. Neurology, 55(4), 484–489. Retrieved from http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2724764/ [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Jack CR, Petersen RC, Xu Y, O’Brien PC, Smith GE, Ivnik RJ, … Kokmen E (2000). Rates of hippocampal atrophy correlate with change in clinical status in aging and AD. Neurology, 55(4), 484–490. doi: 10.1212/wnl.55.4.484 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Jack CR, Petersen RC, Xu YC, Waring SC, O’Brien PC, Tangalos EG, …Kokmen E (1997). Medial temporal atrophy on MRI in normal aging and very mild Alzheimer’s disease. Neurology, 49(3), 786–794. doi: 10.1212/wnl.49.3.786 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Jobst KA, Smith AD, Szatmari M, Esiri MM, Jaskowski A, Hindley N, … Molyneux AJ (1994). Rapidly progressing atrophy of medial temporal lobe in Alzheimer’s disease. The Lancet, 343(8901), 829–830. doi: 10.1016/S0140-6736(94)92028-1 [DOI] [PubMed] [Google Scholar]
  28. Katzman R, Terry R, DeTeresa R, Brown T, Davies P, Fuld P, … Peck A (1988). Clinical, pathological, and neurochemical changes in dementia: a subgroup with preserved mental status and numerous neocortical plaques. Annals of Neurology, 23(2), 138–144. [DOI] [PubMed] [Google Scholar]
  29. Kremen WS, Prom-Wormley E, Panizzon MS, Eyler LT, Fischl B, Neale MC, … Fennema-Notestine C (2010). Genetic and environmental influences on the size of specific brain regions in midlife: the VETSA MRI study. NeuroImage, 49(2), 1213–1223. doi: 10.1016/j.neuroimage.2009.09.043 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Lee SH, Harold D, Nyholt DR, ANZGene Consortium, International Endogene Consortium, Genetic Environmental Risk for Alzheimer’s disease Consortium, … Visscher PM (2013). Estimation and partitioning of polygenic variation captured by common SNPs for Alzheimer’s disease, multiple sclerosis and endometriosis. Human molecular genetics, 22(4), 832–841. doi: 10.1093/hmg/dds491 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Madden DJ, Bennett IJ, Burzynska A, Potter GG, Chen N. k., & Song AW (2012). Diffusion tensor imaging of cerebral white matter integrity in cognitive aging. Biochimica et Biophysica Acta (BBA)-Molecular Basis of Disease, 1822(3), 386–400. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Marden JR, Walter S, Tchetgen Tchetgen EJ, Kawachi I, & Glymour MM (2014). Validation of a polygenic risk score for dementia in black and white individuals. Brain and behavior, 4(5), 687–697. doi: 10.1002/brb3.248 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. McKhann G, Drachman D, Folstein M, Katzman R, Price D, & Stadlan E (1984). Clinical diagnosis of Alzheimer’s disease: report of the NINCDS-ADRDA Work Group* under the auspices of Department of Health and Human Services Task Force on Alzheimer’s Disease. Neurology, 34(7), 939–939. [DOI] [PubMed] [Google Scholar]
  34. Millon T (1981). Disorders of Personality, DSM-III: Axis II. New York, NY: John Wiley & Sons Inc. [Google Scholar]
  35. Mormino EC, Sperling RA, Holmes AJ, Buckner RL, De Jager PL, Smoller JW, & Sabuncu MR (2016). Polygenic risk of Alzheimer disease is associated with early- and late-life processes. Neurology, 87(5), 481–488. doi: 10.1212/wnl.0000000000002922 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Newcombe F (1969). Missile wounds of the brain: A study of psychological deficits. Oxford, England: Oxford University Press. [Google Scholar]
  37. Pegueroles J, Vilaplana E, Montal V, Sampedro F, Alcolea D, Carmona-Iragui M, … Fortea J (2017). Longitudinal brain structural changes in preclinical Alzheimer’s disease. Alzheimer’s & Dementia, 13(5), 499–509. doi: 10.1016/j.jalz.2016.08.010 [DOI] [PubMed] [Google Scholar]
  38. Petersen R (2004). Mild cognitive impairment as a diagnostic entity. Journal of Internal Medicine, 256(3), 183–194. [DOI] [PubMed] [Google Scholar]
  39. Petersen RC, Jack CR, Xu Y-C, Waring SC, O’Brien PC, Smith GE, … Kokmen E (2000). Memory and MRI-based hippocampal volumes in aging and AD. Neurology, 54(3), 581–581. doi: 10.1212/wnl.54.3.581 [DOI] [PubMed] [Google Scholar]
  40. Ranganath C, & Ritchey M (2012). Two cortical systems for memory-guided behaviour. Nature Reviews Neuroscience, 13, 713–726. doi: 10.1038/nrn3338 [DOI] [PubMed] [Google Scholar]
  41. Raz N, & Rodrigue KM (2006). Differential aging of the brain: patterns, cognitive correlates and modifiers. Neuroscience & Biobehavioral Reviews, 30(6), 730–748. doi: 10.1016/j.neubiorev.2006.07.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Reitan RM (1958). Validity of the Trail Making Test as an indicator of organic brain damage. Perceptual and Motor Skills, 8(271–276). [Google Scholar]
  43. Rentería ME, Hansell NK, Strike LT, McMahon KL, de Zubicaray GI, Hickie IB, … Wright MJ (2014). Genetic architecture of subcortical brain regions: common and region-specific genetic contributions. Genes, brain, and behavior, 13(8), 821–830. doi: 10.1111/gbb.12177 [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Resnick S, Pham D, Kraut M, Zonderman A, & Davatzikos C (2003). Longitudinal magnetic resonance imaging studies of older adults: a shrinking brain. The Journal of Neuroscience, 23(8), 3295–3301. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Resnick SM, Goldszal AF, Davatzikos C, Golski S, Kraut MA, Metter EJ, … Zonderman AB (2000). One-year age changes in MRI brain volumes in older adults. Cerebral cortex, 10(5), 464–472. [DOI] [PubMed] [Google Scholar]
  46. Ridge PG, Mukherjee S, Crane PK, Kauwe JSK, & Alzheimer’s Disease Genetics Consortium. (2013). Alzheimer’s disease: analyzing the missing heritability. PLoS ONE, 8(11), e79771–e79771. doi: 10.1371/journal.pone.0079771 [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Rouleau I, Salmon DP, Butters N, Kennedy C, & McGuire K (1992). Quantitative and qualitative analyses of clock drawings in Alzheimer’s and Huntington’s Disease. Brain and Cognition, 18, 70–87. [DOI] [PubMed] [Google Scholar]
  48. Satz P (1993). Brain reserve capacity on symptom onset after brain injury: a formulation and review of evidence for threshold theory. Neuropsychology, 7(3), 273–295. [Google Scholar]
  49. Satz P, Cole MA, Hardy DJ, & Rassovsky Y (2011). Brain and cognitive reserve: mediator(s) and construct validity, a critique. Journal of Clinical and Experimental Neuropsychology, 33(1), 121–130. [DOI] [PubMed] [Google Scholar]
  50. Stern Y (2002). What is cognitive reserve? Theory and research application of the reserve concept. Journal of the International Neuropsychological Society, 8(3), 448–460. doi: 10.1017/S1355617702813248 [DOI] [PubMed] [Google Scholar]
  51. Stern Y (2009). Cognitive reserve. Neuropsychologia, 47(10), 2015–2028. doi: 10.1016/j.neuropsychologia.2009.03.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Thal DR, Rüb U, Orantes M, & Braak H (2002). Phases of Aβ-deposition in the human brain and its relevance for the development of AD. Neurology, 58(12), 1791–1800. doi: 10.1212/wnl.58.12.1791 [DOI] [PubMed] [Google Scholar]
  53. Verhaaren BFJ, Vernooij MW, Koudstaal PJ, Uitterlinden AG, van Duijn CM, Hofman A, … Ikram MA (2013). Alzheimer’s disease genes and cognition in the nondemented general population. Biological psychiatry, 73(5), 429–434. doi: 10.1016/j.biopsych.2012.04.009 [DOI] [PubMed] [Google Scholar]
  54. Visser PJ, Verhey FRJ, Hofman PAM, Scheltens P, & Jolles J (2002). Medial temporal lobe atrophy predicts Alzheimer’s disease in patients with minor cognitive impairment. Journal of Neurology, Neurosurgery, & Psychiatry, 72(4), 491–497. doi: 10.1136/jnnp.72.4.491 [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Wechsler D (1981). Wechsler Adult Intelligence Scale-Revised (Vol. 1). New York, NY: Psychological Corporation. [Google Scholar]
  56. Whalley LJ, Deary IJ, Appleton CL, & Starr JM (2004). Cognitive reserve and the neurobiology of cognitive aging. Ageing research reviews, 3(4), 369–382. doi: 10.1016/j.arr.2004.05.001 [DOI] [PubMed] [Google Scholar]
  57. Wilson JR, De Fries JC, Mc Clearn GE, Vandenberg SG, Johnson RC, & Rashad MN (1975). Cognitive abilities: use of family data as a control to assess sex and age differences in two ethnic groups. The International Journal of Aging and Human Development, 6(3), 261–276. doi: 10.2190/BBJP-XKUG-C6EW-KYB7 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS1608189-supplement-1.docx (81.3KB, docx)

RESOURCES