In one of the largest studies of its type, researchers confirm a link between midlife vascular disease and late-life dementia. As reported in the April 11 Journal of the American Medical Association, 50-year-olds with two or more vascular risk factors were almost three times more likely to have amyloid in the brain in their 70s as were those with no signs of cardiovascular disease in middle age. Researchers said that the study, led by Rebecca Gottesman at Johns Hopkins University School of Medicine, Baltimore, was impressive for the number of subjects and its prospective design. However, scientists are still unsure how vascular disease begets amyloid.

Gottesman and colleagues analyzed data collected from the Atherosclerosis Risk in Communities (ARIC) study, which has been following almost 16,000 volunteers recruited from four communities since 1987 (see ARIC Investigators, 1989). A subset of 346 people from three of those communities (in Maryland, North Carolina, and Mississippi) underwent 3T MRI and amyloid PET scanning with florbetapir between 2011 and 2013, when they were about 76 years old on average. Using a standard uptake value ratio (SUVR) of 1.2 as a cutoff for tracer uptake, Gottesman correlated amyloid positivity with earlier cardiovascular factors, including body mass index, high blood pressure, diabetes, plasma cholesterol, and smoking history.

When comparing individual risk factors, only people with a body mass index of 3.0 or higher at midlife were more likely to have amyloid in their 70s, with an odds ratio of 2.06. This included 38 percent of whites and 32 percent of blacks. A stronger correlation emerged when the researchers accounted for multiple cardiovascular risk factors. Those with two or more were 2.8 times as likely to have amyloid in late life as those with fewer than two. Almost 42 percent of people in the study had more than two risk factors.

The findings make a lot of sense, said Charles DeCarli, University of California, Davis. He noted that while prior studies have linked midlife cardiovascular disease to late-life dementia, researchers have debated whether vascular dementia occurred independently of amyloid (e.g., Launer et al., 2000Kivipelto et al., 2001Whitmer et al., 2005). “It’s interesting here in this ARIC cohort that midlife vascular risk factors increased amyloid retention in the brain.”  

William Jagust, University of California at Berkeley, agreed. “Other studies have measured vascular risk at the same time they measured amyloid,” he told Alzforum (see Reed et al., 2014; Rodrigue et al., 2013; Reed et al., 2012). “From epidemiological and public health perspectives, the midlife risk factor correlation here is very important,” he said.

Jagust also praised the community-based nature of the study, which makes it more representative of the general population than similar studies, which have tended to be more selective. He welcomed the high proportion of African-Americans (43 percent), for example, who are historically under-recruited relative to whites.

However, both Jagust and DeCarli emphasized the lack of a clear mechanism linking vascular risk factors to cortical amyloid. “We could never find a relation between cortical amyloid and measures of vascular disease such as stroke, infarct, and white-matter hyperintensities,” said Jagust. In fact, at this year’s Human Amyloid Imaging meeting, held in January in Miami Beach, Gottesman herself reported finding no correlation between elevated florbetapir SUVR and white-matter hyperintensities or infarcts in the ARIC cohort, though she noted that her study was probably underpowered to draw those links out. Likewise, the JAMA paper reports no statistically significant influence of race or ApoE genotype on the vascular-amyloid correlation, but here again, the study may have been underpowered, write the authors.

While growing evidence links vascular disease with dementia and now amyloid, the lack of a clear mechanism puzzles Jagust. While Gottesman and colleagues suggest that poor clearance of Aβ from the brain may be the link, a scenario DeCarli considers reasonable, hard evidence is wanting. Jagust thinks better animal models may help scientists understand Aβ clearance, while in people, high field strength imaging may reveal correlations between amyloid and microscopic damage that cannot be seen on 3T MRI (e.g.,  Bouvy et al., 2017).—Tom Fagan


  1. I read with interest the paper of Gottesman et al. concerning midlife vascular risk factors and accumulation of amyloid in the human brain. This report drew a lot of media interest. I am glad that this is now being brought to public attention. The authors stated that “previous studies have demonstrated inconsistent results evaluating associations between vascular risk factors and brain amyloid” and that “whether these risk factors directly increase the neurodegeneration specifically associated with AD (such as through increasing amyloid deposition) … is not yet known.”

    I’d like to point out, however, that the first report of high cholesterol as a midlife risk factor for amyloid accumulation in human brain was published by us in a large autopsy study 14 years ago (Pappolla et al., 2003). We have some minor discrepancies with this recent JAMA study, but all on the same direction of heightened risk for midlife hypercholesterolemia and amyloid accumulation (see discussion below).

    We have been emphasizing this issue in numerous subsequent publications (some examples are Petanceska et al., 2003; Pappolla 2008; and Zambón et al., 2010). In fact, our very first study published in a transgenic mouse model of AD showed that diet-induced hypercholesterolemia dramatically accelerated amyloid deposition (Refolo et al., 2000). 

    In our human autopsy study, we showed (in agreement with Gottesman et al.), that high levels of cholesterol correlate with presence of amyloid deposition in human brain only in the youngest subjects of the cohort (40 to 55 years; p = 0.000 for all ApoE isoforms; p = 0.009 for ApoE3/3 subjects) but not in the older subjects. We also showed that elevations of cholesterol above 180 mg/dl, not 200 mg/dl as reported by Gottesman et al., already lead to abnormal amyloid accumulation in the brain. This is not a trivial point; even apparently harmless elevations in cholesterolemia from 181 to 200 almost tripled the odds for finding amyloid in the brain tissue, independent of other risk factors, such as ApoE isoform (Pappolla et al., 2003). This latter discrepancy can easily be explained because in our study, we used microscopy and immunohistochemistry, which are much more sensitive methods for detection of amyloid accumulation than PET scans. This is of importance because more efforts should be placed on early disease prevention (as further explained below) and because cholesterol-lowering therapy should be aggressive if necessary.

    Gottesman et al. claimed that several risk factors drive amyloid accumulation. However, we believe that among those mentioned, hypercholesterolemia is the main one that pertains to amyloid accumulation (but not neuronal degeneration). In fact, this suggestion was recently confirmed in paper published by Vemuri et al. (2017). They also used amyloid PET scans and showed that other vascular risk factors (i.e., hypertension, smoking, obesity, etc.) were related to what they called "AD neurodegeneration," but not to amyloid deposition. To quote from Vemuri et al.: "Apart from demographics and the APOE genotype, only midlife dyslipidemia was associated with amyloid deposition. Obesity, smoking, diabetes, hypertension, and cardiac and metabolic conditions, but not intellectual enrichment, were associated with greater AD-pattern neurodegeneration."

    Concerning the association between hypercholesterolemia and AD, the existing literature shows that most positive studies (i.e., studies showing a relationship between serum cholesterol levels and AD) examined cholesterol levels at midlife (cohorts' ages ranged from 40 to 59 years) and then correlated these levels to later development of dementia (reviewed in Kivipelto and Solomon, 2006; Kivipelto et al., 2002; Notkola et al., 1998). In contrast, most negative reports included participants of much more advanced ages (reviewed in Kivipelto and Solomon, 2006; Kivipelto et al., 2002; Notkola et al., 1998). Additionally, longitudinal studies have shown gradual declines in cholesterol serum levels, which precede the development of dementia by years in most AD patients, obscuring abnormalities that may have occurred earlier in life. This is an interesting and puzzling phenomenon of unknown mechanism.

    In our neuropathologic study (Pappolla et al., 2003), this interesting age-related relationship observed in epidemiological studies was substantiated. Hypercholesterolemia strongly correlated with presence of brain amyloid, but only in subjects aged 40 to 55. Strikingly, the differences in cholesterol between amyloid-bearing and amyloid-free brains disappeared as the subjects' age increased beyond 55 years. In fact, patients with clinical late-onset AD showed no difference in serum cholesterol levels with the control population. These observations confirmed that hypercholesterolemia is only an early (not a late) risk factor for AD. Unfortunately, studies that focus only on old subjects continue to appear in the literature from time to time claiming that there is no association between hypercholesterolemia and AD; these studies ignore the mentioned age-related dynamics and the risk exerted by hypercholesterolemia in earlier years. It appears that elevations of cholesterol that occur early in life lead to lifelong changes in several factors (ApoE expression? LDLr changes?) that result in increased risk for AD later in life. Another important factor in the relationship between cholesterolemia and AD neuropathology is a complex biphasic effect of cholesterol. Only mild elevations may lead to maximum increases in amyloid deposition (Pappolla et al., 2003). 

    I don't want to make it sound like cholesterol is the end of it all. Midlife hypercholesterolemia is just one of several risk factors for sporadic AD; the strongest ones are advanced age and inheritance of the apolipoprotein E4 isoform, with a number of other genetic polymorphisms adding smaller, yet substantive, degrees of risk.

    Despite this caveat, it is important to understand the complex relationship between serum cholesterol and AD risk, because poor knowledge of such age-related dynamics has led to many sub-optimally designed clinical trials (i.e., statin trials conducted in older subjects only), literally throwing hundreds of millions of research dollars out of the window. Worse yet, many investigators have abandoned this mechanism altogether. It is thus not surprising that most studies negating a role for statins in AD prevention have been conducted in populations older than 65 years, overlooking the mentioned age-related risk relationship (reviewed in Pappolla et al., 2003; Kivipelto and Solomon, 2006; Kivipelto et al., 2002; Notkola et al., 1998). These include the CRISP, the PROSPER, one community-based study, the Cache County and the Honolulu studies. The Honolulu study, for example, found no effect for statins but the participants' average age was 80. If this relationship had been better understood by the investigators planning for these trials, studies would have been designed to include individuals who began therapy much earlier in life, when there is evidence that hypercholesterolemia may impact the disease neuropathology.


    . Mild hypercholesterolemia is an early risk factor for the development of Alzheimer amyloid pathology. Neurology. 2003 Jul 22;61(2):199-205. PubMed.

    . Changes in apolipoprotein E expression in response to dietary and pharmacological modulation of cholesterol. J Mol Neurosci. 2003;20(3):395-406. PubMed.

    . Statins, incident Alzheimer disease, change in cognitive function, and neuropathology. Neurology. 2008 Dec 9;71(24):2020; author reply 2020-1. PubMed.

    . Higher incidence of mild cognitive impairment in familial hypercholesterolemia. Am J Med. 2010 Mar;123(3):267-74. PubMed.

    . Hypercholesterolemia accelerates the Alzheimer's amyloid pathology in a transgenic mouse model. Neurobiol Dis. 2000 Aug;7(4):321-31. PubMed.

    . Evaluation of Amyloid Protective Factors and Alzheimer Disease Neurodegeneration Protective Factors in Elderly Individuals. JAMA Neurol. 2017 Jun 1;74(6):718-726. PubMed.

    . Cholesterol as a risk factor for Alzheimer's disease - epidemiological evidence. Acta Neurol Scand Suppl. 2006;185:50-7. PubMed.

    . Apolipoprotein E epsilon4 allele, elevated midlife total cholesterol level, and high midlife systolic blood pressure are independent risk factors for late-life Alzheimer disease. Ann Intern Med. 2002 Aug 6;137(3):149-55. PubMed.

    . Serum total cholesterol, apolipoprotein E epsilon 4 allele, and Alzheimer's disease. Neuroepidemiology. 1998;17(1):14-20. PubMed.

    View all comments by Mike Miguel Pappolla

Make a Comment

To make a comment you must login or register.


Paper Citations

  1. The Atherosclerosis Risk in Communities (ARIC) Study: design and objectives. The ARIC investigators. Am J Epidemiol. 1989 Apr;129(4):687-702. PubMed.
  2. . Midlife blood pressure and dementia: the Honolulu-Asia aging study. Neurobiol Aging. 2000 Jan-Feb;21(1):49-55. PubMed.
  3. . Midlife vascular risk factors and Alzheimer's disease in later life: longitudinal, population based study. BMJ. 2001 Jun 16;322(7300):1447-51. PubMed.
  4. . Midlife cardiovascular risk factors and risk of dementia in late life. Neurology. 2005 Jan 25;64(2):277-81. PubMed.
  5. . Associations between serum cholesterol levels and cerebral amyloidosis. JAMA Neurol. 2014 Feb;71(2):195-200. PubMed.
  6. . Risk Factors for β-Amyloid Deposition in Healthy Aging: Vascular and Genetic Effects. JAMA Neurol. 2013 May 1;70(5):600-6. PubMed.
  7. . Coronary risk correlates with cerebral amyloid deposition. Neurobiol Aging. 2012 Sep;33(9):1979-87. PubMed.
  8. . Abnormalities of Cerebral Deep Medullary Veins on 7 Tesla MRI in Amnestic Mild Cognitive Impairment and Early Alzheimer's Disease: A Pilot Study. J Alzheimers Dis. 2017;57(3):705-710. PubMed.

Further Reading

No Available Further Reading

Primary Papers

  1. . Association Between Midlife Vascular Risk Factors and Estimated Brain Amyloid Deposition. JAMA. 2017 Apr 11;317(14):1443-1450. PubMed.