In a multiethnic population-based sample, specific toxic and endogenous metabolic factors are correlated with smaller volume in brain segments linked with risk for neurodegenerative disease.
Abstract
Purpose
To determine in a large multiethnic cohort the cardiovascular and genetic risk factors associated with smaller volume in the hippocampus, precuneus, and posterior cingulate, and their association with preclinical deficits in cognitive performance in patients younger and older than 50 years.
Materials and Methods
The institutional review board approved the study and all participants provided written informed consent. Eligible for this study were 1629 participants (700 men and 929 women; mean age, 50.0 years ± 10.2 [standard deviation]) drawn from the population-based Dallas Heart Study who underwent laboratory and clinical analysis in an initial baseline visit and approximately 7 years later underwent brain magnetic resonance imaging with automated volumetry and cognitive assessment with the Montreal Cognitive Assessment (MoCA). Regression analysis showed associations between risk factors and segmental volumes, and associations between these volumes with cognitive performance in participants younger and older than 50 years.
Results
Lower hippocampal volume was associated with previous alcohol consumption (standardized estimate, −0.04; P = .039) and smoking (standardized estimate, −0.04; P = .048). Several risk factors correlated with lower total brain, posterior cingulate, and precuneus volumes. Higher total (standardized estimate, 0.06; P = .050), high-density lipoprotein (standardized estimate, 0.07; P = .003), and low-density lipoprotein (standardized estimate, 0.04; P = .037) cholesterol levels were associated with larger posterior cingulate volume, and higher triglyceride levels (standardized estimate, 0.06; P = .004) were associated with larger precuneus volume. Total MoCA score was associated with posterior cingulate volume (standardized estimate, 0.13; P = .001) in younger individuals and with hippocampal (standardized estimate, 0.06; P < .05) and precuneus (standardized estimate, 0.08; P < .023) volumes in older adults.
Conclusion
Smaller volumes in specific brain regions considered to be early markers of dementia risk were associated with specific cardiovascular disease risk factors and cognitive deficits in a predominantly midlife multiethnic population-based sample. Additionally, the risk factors most associated with these brain volumes differed in participants younger and older than 50 years, as did the association between brain volume and MoCA score.
© RSNA, 2015
Introduction
There is an increasing body of evidence that links cardiovascular risk factors to the development of Alzheimer disease (AD). As a result, there was a shift in the study of AD from genetics to potentially modifiable risk factors. The early underlying alterations in brain structure that are associated with and possibly mediate this relationship, however, remain poorly understood.
Diminished brain volumes in specific regions are identified as early preclinical markers of AD risk. Hippocampal atrophy, which is closely linked with AD, may be reflected by volume loss in the medial temporal lobe on magnetic resonance (MR) images. Volume loss was seen (1) in the region of the precuneus and posterior cingulate cortex in patients who develop AD at a notably younger age. Posterior cingulate atrophy has been identified very early in sporadic AD, and may occur before onset of hippocampal atrophy (2,3). Volume loss in these regions may serve as important early imaging biomarkers to identify risk for dementia.
Preclinical vascular insults that confer early risk for dementia are also known to begin long before clinical manifestations of the disease (4). The important biologic events that occur as a result of these risk factors, however, are not well understood. To better understand possible underlying disease processes, a first step is to identify environmental risk factors associated with the earliest structural brain changes and to determine the relation of these changes, if any, to cognition.
Previous studies showed that vascular risk factors that start in midlife raise risk of subsequent dementia (4–7). Additionally, apolipoprotein (apo) E4, the primary known genetic link to AD, was associated with accelerated volume loss in the hippocampus (8), which highlights that a combination of factors likely contributes to dementia risk. Total brain, gray matter, hippocampal, and cerebrospinal fluid volume have also all been associated with cognitive deficits identified by the Montreal Cognitive Assessment (MoCA), a validated indicator of mild cognitive impairment and preclinical AD (9). However, to our knowledge, a single study that linked smaller regional brain volumes with both cardiovascular risk factors and diminished cognitive performance in a preclinical setting is lacking.
The purpose of this study was to determine in a large multiethnic cohort the cardiovascular and genetic risk factors associated with smaller volume in the hippocampus, precuneus, and posterior cingulate, and their association with preclinical deficits in cognitive performance in patients younger and older than 50 years.
Materials and Methods
Standard Protocol Approvals, Registrations, and Patient Consents
The institutional review board of this institution approved the study and all participants provided written informed consent.
Experimental Design
The Dallas Heart Study is a large, multiethnic, probability-based population study of Dallas County residents that was initiated in 1999 as previously discussed in published methods of study design (10). The initial Dallas Heart Study evaluated participants from 2000 to 2002 and a follow-up study was conducted from 2008 to 2009. At the follow-up, participants underwent MR imaging of the brain and cognitive testing by using MoCA at the University of Texas Southwestern Medical Center. The study used United States postal codes to produce a sample representative of the Dallas metropolitan area. As previously described (10), it was designed to produce population estimates of biologic and social variables with minimal bias while allowing for oversampling of African Americans to ensure approximately 50% representation.
In this study we included data from 1629 individuals (700 men and 929 women; mean age, 50.0 years ± 10.2 [standard deviation]) without a history of stroke who underwent laboratory evaluation of cardiovascular risk factors and apo E genetic testing at the time of the original Dallas Heart Study, and returned approximately 7 years later for the follow-up study and underwent brain imaging and cognitive assessment. Exclusion criteria were self-reported history of stroke, of which there were 37 individuals. Those with major structural defects that appeared on MR images were also excluded, including those with corpus callosum agenesis, imaging evidence of stroke, or hydrocephalus, and those with image-acquisition errors including metal or motion artifact or significant noise, of which there were 70 individuals.
MR Imaging Protocol
MR images of the brain were obtained on a 3-T MR imager (Achieva; Philips Healthcare, Best, the Netherlands) by using T1-weighted three-dimensional magnetization-prepared rapid acquisition of gradient echo imaging. Images were obtained from the vertex of the skull to the foramen magnum in true axial orientation. Specifications for three-dimensional magnetization-prepared rapid acquisition of gradient echo imaging were the following: repetition time msec/echo time msec, 9.6/5.8; inversion time, 1100 msec; flip angle, 12°; sensitivity encoding factor of two; 2-mm sections spaced at 1-mm centers; 288 rows × 288 columns × 140 sections; and voxel size, 1.0 × 0.9 × 0.9 mm.
Image Analysis
Quantification of total brain and intracranial volumes was performed by using the freely available Functional MR Imaging of the Brain (known as FMRIB; University of Oxford, Oxford, England) Software Library Brain Extraction Tool and FMRIB Automated Segmentation Tool. Segmental volumetric MR imaging quantification was performed by using an image analysis suite (FreeSurfer version 4.4; http://surfer.nmr.mgh.harvard.edu/). The fully automated analysis was run at the Texas Advanced Computing Center at the University of Texas at Austin, as previously described (11). Images of individuals with errors on automated analysis were reanalyzed and the masks generated by software (FreeSurfer) verified by a board-certified neuroradiologist (K.S.K., with 5 years of experience).
Statistical Analysis
Statistical analyses were performed by using software (SAS version 9.3.0; SAS Institute, Cary, NC). For analysis of risk factors associated with segmental atrophy, separate multivariate regression models that used the stepwise procedure (α to enter and leave was set at P = .15) with multiple comparison correction were used to explore the following dependent variables: posterior cingulate, precuneus, hippocampal, and total brain volume. An additional similar analysis was performed to determine predictors of MoCA score by using precuneus, posterior cingulate, hippocampal, and total brain volume as independent variables. In addition to the demographic predictor variables, the final model for each dependent variable included all independent variables from stepwise significance indicated by P values less than .05 to increase the number of observations in the analysis. Independent variables considered in all regression analyses included the following: number of apo E2 and apo E4 alleles with apo E3 used as the reference point, presence of diabetes, past or present smoker, presence of hypertension, hypertension duration in years, current treatment of hypertension, presence of left ventricular hypertrophy normalized by body surface area, alcohol use in grams consumed per week, resting heart rate, total serum cholesterol, serum high-density lipoprotein (HDL), serum low-density lipoprotein, serum very low–density lipoprotein, serum triglycerides, fasting blood glucose, body mass index, and systolic and diastolic blood pressure. Adjustment for total intracranial volume was performed to normalize brain segmental volumes and reduce bias from interindividual variation in head size. The regression models were also adjusted for demographic variables (ie, age, sex, and ethnicity). The regression analysis performed for MoCA scores was additionally corrected for education level, which was found to be a significant predictor of cognitive performance in univariate analysis (data not shown). Analysis was performed for 1629 individuals with successful brain volumetry. Additional regression models were used to examine predictions for individuals who were younger than 50 years and individuals who were 50 years and older. Participant age at the time of the follow-up visit was used to define these groups. Not all clinical measures were available for every individual.
Results
The younger group consisted of 805 participants younger than 50 years (median age, 42 years [age range, 25–49 years]; lower quartile to upper quartile, 38–46 years), and the older group consisted of 824 participants who were 50 years and older (median age, 58 years [age range, 50–73 years]; lower quartile to upper quartile, 53–63 years). Demographic measures are shown in Table 1, which lists the dichotomous and continuous variables investigated in regression analysis. Intergroup comparison showed no statistically significant differences in the sex or ethnicity distribution between groups. The older age group had a higher prevalence of diabetes and hypertension with higher mean systolic and diastolic blood pressures, body mass index, cholesterol levels, and fasting blood glucose, and the difference between groups was statistically significant (P < .05).
Table 1.
Sample Demographics and Cardiovascular and Genetic Risk Factors
Note.—Data are means. LDL = low-density lipoprotein, VLDL = very low–density lipoprotein.
*The difference between the younger group and older group was statistically significant (P < .05).
†Data are ± standard deviation.
Table 2 shows average segmental brain volumes for the two groups. Intergroup comparison shows a statistically significant difference in mean hippocampal, precuneus, posterior cingulate, and total brain volumes, with smaller volumes in the older age group (P < .05). No statistically significant difference in intracranial volume was seen between groups (P = .342).
Table 2.
Sample Segmental Brain Volumes
Note.—Other than P values, data are means (mL) ± standard deviation.
*P value calculated by using t test for comparison of difference in mean volume between the younger group and older group.
The final regression models included only those predictor variables that reached a statistical significance level (P value) of less than .05. All other risk factors were not statistically significant, and thus were excluded from the final model because they did not well predict brain volumes. The final models produced by the stepwise regression analyses are shown in Table 3 and they delineate the statistically significant associations between risk factors and brain volumes for each age group.
Table 3.
Risk Factors Associated with Total and Segmental Brain Volumes
Note.—All models included the following risk factors as covariates with only those with P values less than .05 shown: number of apo E2 alleles, number of apo E4 alleles, diabetes, smoking status, hypertension including duration and treatment, left ventricular hypertrophy, alcohol use, heart rate, total serum cholesterol, serum HDL, serum low-density lipoprotein, serum very low–density lipoprotein, serum triglycerides, fasting blood glucose, body mass index, and systolic and diastolic blood pressure. Variables forced in to each model included age, sex, ethnicity, and intracranial volume (data not shown). BMI = body mass index, LDL = low-density lipoprotein, NA = no significant association.
*The risk factor demonstrated a positive association with brain volume.
Notably, the presence of apo E4 or the apo E2 allele was not a significant predictor of brain volume in any of the regions analyzed in our model. Likewise, hypertension and serum very low–density lipoprotein were not found to be statistically significant predictors of volume in any of the brain segments analyzed.
Data presented in Table 4 shows associations of MoCA scores with segmental volumes in these same regions. All models included posterior cingulate, precuneus, and hippocampal volumes as covariates to predict total score on the MoCA with correction factors that include age, sex, ethnicity, and education level.
Table 4.
Regression Analysis of Clinical Outcomes Assessed by MoCA
Note.—NA denotes that the included covariate did not achieve statistical significance level of P < .05 in stepwise regression analysis and thus was not included in the final regression model. NA = not applicable.
*R2 = 0.02
†R2 = 0.02
‡R2 = 0.32
§All regression models were corrected for age, sex, ethnicity, and education level.
Discussion
Our findings reveal that lower total brain, hippocampal, precuneus, and posterior cingulate volumes are associated with cardiovascular risk factors and with impaired cognitive performance before the onset of clinical dementia. Importantly, we also show that smaller volumes in these brain segments may indicate susceptibility for cognitive insult. Additionally, our findings suggest lower volumes in the posterior cingulate may serve as an early indicator of brain changes associated with cognitive insult even when the patient is younger than 50 years.
Our large sample size allowed us to control for a number of different risk factors. This helps to identify the association of each risk factor independent of their relation to other risk factors. By dividing our cohort into younger and older samples, we were able to investigate how these associations may differ on the basis of age, which expanded on previous studies that identified risk factor profiles in cohorts older than ours (12–14). Additionally, few other studies exist that demonstrate risk factor profiles in the posterior cingulate and precuneus regions. Large studies that included these brain regions mainly used the Framingham cardiovascular risk profile to link atrophy in these regions with cardiovascular risk factors (12,13,15,16). The findings of these studies suggested that vascular risk factors may contribute to development of dementia in those with AD. These studies did not identify the specific risk factor associations, however, and further did not use segmental precuneus and posterior cingulate volumes. Our study contributes to this literature by showing which specific vascular risk factors are most associated with gray matter volume in three regions altered early in the course of AD, and by showing differences in these associations among younger and older individuals.
We have shown in our population that alcohol use, diabetes, and cardiac left ventricular hypertrophy are associated with smaller total brain volume detectable even before age 50 years. Additionally, smoking and obesity were associated with smaller volumes in the posterior cingulate, and alcohol use, obesity, and elevated fasting blood glucose were associated with smaller volumes in the precuneus region. In line with previous studies, we also show that lower hippocampal volume is associated with both alcohol consumption and smoking, and we add that both of these associations are significant in participants older than 50 years, but were not identifiable in participants younger than 50 years (17–20). Further, smaller hippocampal volumes in the older age group were associated with diminished performance on the MoCA. These findings may reflect a susceptibility of these brain regions to insults by these risk factors and, importantly, may serve as potential targets for preventative intervention, which warrants further study.
The association of posterior cingulate, precuneus, and hippocampal volumes with cognitive performance suggests their utility as biomarkers for brain changes linked with preclinical cognitive deficits. In addition, our data suggest that lower posterior cingulate volume may be an early risk marker for cognitive insult because of the association seen in patients younger than 50 years, while hippocampal and precuneus volumes may be better markers in patients older than 50 years. In numerous studies (21,22), the cardiovascular risk factors associated with volume differences in these brain regions were implicated with accelerated cognitive decline. Further longitudinal study is needed to establish whether these risk factors promote atrophy in these brain regions, and whether these changes may in part mediate the link between these risk factors and cognitive decline.
Although many cardiovascular risk factors were implicated in brain atrophy, little data exist regarding the relationship between measured cholesterol levels and segmental brain atrophy. Previous studies (23) demonstrated an association between lower serum HDL levels and cognitive decline. More recently, lower levels of brain cholesterol markers, such as 24S-hydroxycholesterol, a prominent metabolite of brain cholesterol (24), were identified as a risk factors for cognitive decline in elderly populations (25,26). One study (27) showed that the ratio of 24S-hydroxycholesterol to cholesterol was positively correlated to gray matter volume and diminished in patients with AD, and we previously found a correlation with gray matter volume in our cohort (28). Our data demonstrate that specific lipids, including total cholesterol, HDL, low-density lipoprotein, and triglycerides, have various positive associations with posterior cingulate and precuneus volumes. Because these lipids are associated with larger brain volumes, these findings are of particular interest and warrant longitudinal study.
Finally, although it was suggested that apo E4 is the most influential genetic determinant of AD (17,29,30), its association with brain volumes before the onset of symptoms is unclear. Additionally, apo E2 was suggested (31) as a protective gene in studies of brain atrophy. Our cohort did not demonstrate an association between either apo E4 or apo E2 and brain volume in any of the regions studied.
Our study had several limitations. We assessed only cross-sectional differences in regional brain structure and function, but it is uncertain whether these differences will have the same associations in a longitudinal study. Healthy aging was associated with regional brain volume loss in numerous studies, and age contributed to the segmental brain volumes seen here (32). Furthermore, our study used only T1-weighted images for determination of regional volumes, while previous studies in the Dallas Heart Study population used fluid-attenuated inversion recovery images and showed that small-vessel ischemic disease, measured by white matter hyperintensity volume, does not have a statistically significant role in cognitive impairment in this population (9).
In conclusion, we showed that in a multiethnic population-based sample, smaller brain volumes were linked with specific risk factors and cognitive deficits before the onset of dementia. Importantly, the associations of brain volumes with risk factors and cognitive deficits were observed even in participants younger than 50 years. This may corroborate the importance of clinical strategies to preserve brain health beginning in midlife, which was suggested by longitudinal studies (22,33).
Advances in Knowledge
■ Smaller regional brain volumes are associated with previous presence of specific toxic factors, such as alcohol consumption (P < .05) and smoking (P < .05), and endogenous metabolic factors, such as obesity (P < .05) and diabetes (P < .05); however, larger volumes are associated with specific lipids, including total cholesterol, low-density lipoprotein, high-density lipoprotein, and triglycerides (all with P < .05), and risk factor associations differed by age.
■ In the Dallas Heart Study cohort, cognitive performance measured by the Montreal Cognitive Assessment correlates with smaller volume in the posterior cingulate cortex, precuneus, and hippocampus (P < .05 for each region).
Implication for Patient Care
■ Subtle differences in regional brain volumes in midlife are associated with specific vascular risk factors and with cognitive decline and may serve as a biomarker for brain insult before onset of dementia.
Received November 4, 2014; revision requested December 17; revision received February 4, 2015; accepted February 20; final version accepted May 8.
R.N.R. supported by the University of Texas Alzheimer’s Disease Center (grant P30AG12300-19).
The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Current address: Department of Radiology, University of Southern California, Keck School of Medicine, Los Angeles, Calif.
Funding: This research was supported by the National Institutes of Health (grants UL1TR001105 and KL2TR000453).
Disclosures of Conflicts of Interest: R.N.S. disclosed no relevant relationships. H.C.R. disclosed no relevant relationships. M.K.G. disclosed no relevant relationships. R.N.R. disclosed no relevant relationships. M.F.W. disclosed no relevant relationships. R.M.P. Activities related to the present article: author disclosed a grant from the Reynolds Foundation. Activities not related to the present article: disclosed no relevant relationships. Other relationships: disclosed no relevant relationships. R.W.M. disclosed no relevant relationships. L.S.H. disclosed no relevant relationships. R.T.L. Activities related to the present article: author disclosed a grant and travel support from the Doris Duke Charitable Foundation. Activities not related to the present article: disclosed no relevant relationships. Other relationships: disclosed no relevant relationships. K.S.K. disclosed no relevant relationships.
Abbreviations:
- AD
- Alzheimer disease
- HDL
- high-density lipoprotein
- MoCA
- Montreal Cognitive Assessment
References
- 1.Lehmann M, Koedam EL, Barnes J, et al. Posterior cerebral atrophy in the absence of medial temporal lobe atrophy in pathologically-confirmed Alzheimer’s disease. Neurobiol Aging 2012;33(3):627.e1–627.e12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Pengas G, Hodges JR, Watson P, Nestor PJ. Focal posterior cingulate atrophy in incipient Alzheimer’s disease. Neurobiol Aging 2010;31(1):25–33. [DOI] [PubMed] [Google Scholar]
- 3.Choo IH, Lee DY, Oh JS, et al. Posterior cingulate cortex atrophy and regional cingulum disruption in mild cognitive impairment and Alzheimer’s disease. Neurobiol Aging 2010;31(5):772–779. [DOI] [PubMed] [Google Scholar]
- 4.Kivipelto M, Helkala EL, Laakso MP, et al. Midlife vascular risk factors and Alzheimer’s disease in later life: longitudinal, population based study. BMJ 2001;322(7300):1447–1451. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Debette S, Seshadri S, Beiser A, et al. Midlife vascular risk factor exposure accelerates structural brain aging and cognitive decline. Neurology 2011;77(5):461–468. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Whitmer RA, Sidney S, Selby J, Johnston SC, Yaffe K. Midlife cardiovascular risk factors and risk of dementia in late life. Neurology 2005;64(2):277–281. [DOI] [PubMed] [Google Scholar]
- 7.Gorelick PB, Scuteri A, Black SE, et al. Vascular contributions to cognitive impairment and dementia: a statement for healthcare professionals from the American Heart Association/American Stroke Association. Stroke 2011;42(9):2672–2713. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Schuff N, Woerner N, Boreta L, et al. MRI of hippocampal volume loss in early Alzheimer’s disease in relation to ApoE genotype and biomarkers. Brain 2009;132(Pt 4):1067–1077. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Gupta M, King KS, Srinivasa R, et al. Association of 3.0-T brain magnetic resonance imaging biomarkers with cognitive function in the Dallas Heart Study. JAMA Neurol 2015;72(2):170–175. [DOI] [PubMed] [Google Scholar]
- 10.Victor RG, Haley RW, Willett DL, et al. The Dallas Heart Study: a population-based probability sample for the multidisciplinary study of ethnic differences in cardiovascular health. Am J Cardiol 2004;93(12):1473–1480. [DOI] [PubMed] [Google Scholar]
- 11.Lucarelli RT, Peshock RM, McColl R, et al. MR imaging of hippocampal asymmetry at 3T in a multiethnic, population-based sample: results from the Dallas Heart Study. AJNR Am J Neuroradiol 2013;34(4):752–757. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Tamashiro-Duran JH, Squarzoni P, de Souza Duran FL, et al. Cardiovascular risk in cognitively preserved elderlies is associated with glucose hypometabolism in the posterior cingulate cortex and precuneus regardless of brain atrophy and apolipoprotein gene variations. Age (Dordr) 2013;35(3):777–792. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.de Toledo Ferraz Alves TC, Scazufca M, Squarzoni P, et al. Subtle gray matter changes in temporo-parietal cortex associated with cardiovascular risk factors. J Alzheimers Dis 2011;27(3):575–589. [DOI] [PubMed] [Google Scholar]
- 14.Seshadri S, Wolf PA, Beiser A, et al. Stroke risk profile, brain volume, and cognitive function: the Framingham Offspring Study. Neurology 2004;63(9):1591–1599. [DOI] [PubMed] [Google Scholar]
- 15.Cardenas VA, Reed B, Chao LL, et al. Associations among vascular risk factors, carotid atherosclerosis, and cortical volume and thickness in older adults. Stroke 2012;43(11):2865–2870. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.de la Torre JC. Vascular risk factors: a ticking time bomb to Alzheimer’s disease. Am J Alzheimers Dis Other Demen 2013;28(6):551–559. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.den Heijer T, van der Lijn F, Ikram A, et al. Vascular risk factors, apolipoprotein E, and hippocampal decline on magnetic resonance imaging over a 10-year follow-up. Alzheimers Dement 2012;8(5):417–425. [DOI] [PubMed] [Google Scholar]
- 18.Fotuhi M, Do D, Jack C. Modifiable factors that alter the size of the hippocampus with ageing. Nat Rev Neurol 2012;8(4):189–202. [DOI] [PubMed] [Google Scholar]
- 19.Middleton LE, Yaffe K. Promising strategies for the prevention of dementia. Arch Neurol 2009;66(10):1210–1215. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Morris JK, Vidoni ED, Perea RD, et al. Insulin resistance and gray matter volume in neurodegenerative disease. Neuroscience 2014;270:139–147. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Alzheimer’s Association . 2010 Alzheimer’s disease facts and figures. Alzheimers Dement 2010;6(2):158–194. [DOI] [PubMed] [Google Scholar]
- 22.Nordström P, Nordström A, Eriksson M, Wahlund LO, Gustafson Y. Risk factors in late adolescence for young-onset dementia in men: a nationwide cohort study. JAMA Intern Med 2013;173(17):1612–1618. [DOI] [PubMed] [Google Scholar]
- 23.Singh-Manoux A, Gimeno D, Kivimaki M, Brunner E, Marmot MG. Low HDL cholesterol is a risk factor for deficit and decline in memory in midlife: the Whitehall II study. Arterioscler Thromb Vasc Biol 2008;28(8):1556–1562. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Hughes TM, Rosano C, Evans RW, Kuller LH. Brain cholesterol metabolism, oxysterols, and dementia. J Alzheimers Dis 2013;33(4):891–911. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Prospective Studies Collaboration , Lewington S, Whitlock G, et al. Blood cholesterol and vascular mortality by age, sex, and blood pressure: a meta-analysis of individual data from 61 prospective studies with 55,000 vascular deaths. Lancet 2007;370(9602):1829–1839. [DOI] [PubMed] [Google Scholar]
- 26.van den Kommer TN, Dik MG, Comijs HC, et al. The role of extracerebral cholesterol homeostasis and ApoE e4 in cognitive decline. Neurobiol Aging 2012;33(3):622.e17–622.e28. [DOI] [PubMed] [Google Scholar]
- 27.Solomon A, Leoni V, Kivipelto M, et al. Plasma levels of 24S-hydroxycholesterol reflect brain volumes in patients without objective cognitive impairment but not in those with Alzheimer’s disease. Neurosci Lett 2009;462(1):89–93. [DOI] [PubMed] [Google Scholar]
- 28.Stiles AR, Kozlitina J, Thompson BM, McDonald JG, King KS, Russell DW. Genetic, anatomic, and clinical determinants of human serum sterol and vitamin D levels. Proc Natl Acad Sci U S A 2014;111(38):E4006–E4014. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.van de Pol LA, van der Flier WM, Korf ES, Fox NC, Barkhof F, Scheltens P. Baseline predictors of rates of hippocampal atrophy in mild cognitive impairment. Neurology 2007;69(15):1491–1497. [DOI] [PubMed] [Google Scholar]
- 30.Corder EH, Saunders AM, Strittmatter WJ, et al. Gene dose of apolipoprotein E type 4 allele and the risk of Alzheimer’s disease in late onset families. Science 1993;261(5123):921–923. [DOI] [PubMed] [Google Scholar]
- 31.Rebeck GW, Kindy M, LaDu MJ. Apolipoprotein E and Alzheimer’s disease: the protective effects of ApoE2 and E3. J Alzheimers Dis 2002;4(3):145–154. [DOI] [PubMed] [Google Scholar]
- 32.Hedman AM, van Haren NE, Schnack HG, Kahn RS, Hulshoff Pol HE. Human brain changes across the life span: a review of 56 longitudinal magnetic resonance imaging studies. Hum Brain Mapp 2012;33(8):1987–2002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Launer LJ, Ross GW, Petrovitch H, et al. Midlife blood pressure and dementia: the Honolulu-Asia aging study. Neurobiol Aging 2000;21(1):49–55. [DOI] [PubMed] [Google Scholar]