Bright spots called white-matter hyperintensities (WMHs) often appear on MRI scans of people with familial or sporadic Alzheimer’s disease, and they tend to intensify as the disease progresses. Some scientists think they reflect cerebrovascular disease. However, researchers led by Jasmeer Chhatwal at Massachusetts General Hospital in Charlestown offer a different explanation. In the October 16 JAMA Neurology, they reported that WMHs worsened most in people with extensive neurodegeneration, amyloid plaques, or cerebral microbleeds, a sign of cerebral amyloid angiopathy (CAA), while WMH severity did not correlate with vascular risk. They concluded that WMHs are driven by AD pathology.

  • White-matter hyperintensities bloom as Alzheimer’s disease progresses.
  • People with rampant amyloid and neurodegeneration were most affected.
  • WMHs did not track with cardiovascular health.

“Conventional wisdom that white-matter disease is solely representative of vascular brain injury in older people is proven not to be the case,” Charles DeCarli at the University of California, Davis, told Alzforum. He said the data was convincing that neurodegeneration leads to WMHs, though he was less sure about plaques or CAA driving damage in all cases. Costantino Iadecola, Weill Cornell Medical College, New York, does think that amyloid’s assault on vessels underlies some of the pathology and that this work may refocus minds. “These results will get researchers out of the rut of thinking that hypertension and diabetes are the only things that damage the blood vessels in Alzheimer’s,” Iadecola said.

First author Zahra Shirzadi compared the total volume of WMHs to cardiovascular risk, as measured by the Framingham Heart Study cardiovascular disease risk score, and to the amount of amyloid, be it plaques or CAA. She drew data from clinical records and almost 4,000 brain scans of 1,141 people from three longitudinal cohorts: the Harvard Aging Brain Study, the Alzheimer’s Disease Neuroimaging Initiative, and the Dominantly Inherited Alzheimer Network. Participants included 252 carriers of pathogenic APP, PSEN1, and PSEN2 variants, average age 38; 458 healthy controls, averaging 72 years old; and 431 adults with MCI or AD, average age 73. Almost all were Caucasian and had better-than-average cardiovascular health for their age. Each had at least two structural MRI and amyloid PET scans over an average of three to five years.

At baseline, WMH volume was greatest among those who were oldest, had the least gray matter, the highest amyloid burden, or who had two or more cerebral microbleeds, a commonly used indicator of CAA that can only be diagnosed at autopsy. Over time, WMHs worsened more among these people than among their respective controls.

Amyloid and WMHs. MRI scans from people with a high amyloid load (left) predict more white-matter damage (red) than do scans from people with little amyloid (right). [Courtesy of Shirzadi et al., JAMA Neurology, 2023.]

Larger WMH volume might signify CAA before cerebral microbleeds do, the authors propose. Among the 527 older adults without such microbleeds at baseline, those with above-average WMH volume were 2.6 times likelier to develop microbleeds after six years than people with below-average WMH volume. Shirzadi had also seen WMHs worsen before microbleeds appeared in people who developed Dutch CAA, an autosomal-dominant form of the disease caused by a point mutation within the Aβ sequence of APP (Shirzadi et al., 2022). These results suggest that enlarged WMHs are an early sign of CAA.

To Iadecola, both studies fit with the idea that amyloid constricts blood vessels. He and others have found that soluble Aβ slowed cerebral blood flow in wild-type and amyloidosis mice and that vascular injury can be prevented if reactive oxygen species (ROS) scavengers are administered before the peptide settles into vessels as CAA. ROS are a major cause of vessel damage by Aβ (Mar 2011 news; Park et al., 2014). Iadecola thinks a similar window of opportunity for early intervention might exist in people with WMHs but no microbleeds. “By the time you see a single microbleed, it's probably too late to save those damaged blood vessels,” he said.

Did vascular health factor into WMH severity? Surprisingly, the amount of WMHs had no correlation with cardiovascular risk score after accounting for age, gray-matter volume, amyloid burden, and cerebral microbleeds. The authors concluded that amyloid and gray-matter atrophy, i.e., neurodegeneration, drives the brain lesions rather than small-vessel disease.

Others partially agreed. DeCarli thinks amyloid may be to blame in the young, heart-healthy DIAN cohort but that there is insufficient evidence to exclude concurrent vascular disease causing WMHs in the older ADNI and HABS participants.

Adam Brickman of Columbia University, New York, shared similar sentiments. “Showing that WMHs are not related to vascular risk factors and that they are related to aspects of AD pathophysiology tells us two things: that WMHs are not due exclusively to (modifiable) risk factors and that there is a relationship between WMHs and other markers of AD,” he wrote (comment below). “But do these observations prove the etiology of WMHs? I do not think that question has been settled.”—Chelsea Weidman Burke


  1. I am delighted to see that consideration of the role and meaning of white-matter hyperintensities (WMH) continues to be part of our conversation about Alzheimer’s disease. The paper by Shirzadi and colleagues showcases the powerful approach of combining data from many cohorts with different participant characteristics to examine the correlates of WMH as they appear in AD. Their study confirms that WMH are a prominent feature of AD, that they are related to other pathophysiological indicators, like atrophy and cerebral microbleeds, and that they are not accounted for entirely by vascular risk factors. The authors conclude that, in the context of AD, WMH should be considered a reflection of “AD-intrinsic” factors like amyloid pathology and neurodegeneration, and not an indicator of small vessel cerebrovascular disease.

    There are many strengths to this study. I think at the top of the list is the utilization of data from individuals with autosomal-dominant forms of AD from the Dominantly Inherited Alzheimer Network (DIAN). Individuals with autosomal-dominant AD tend to be much younger than those with late-onset AD and, importantly, have low levels of vascular risk factors, so consideration of the role of WMH without the confound of exposure to vascular risk factors is possible. Even in this younger, healthier, and genetically “pure” form of AD, the authors observed increased WMH that seemed to covary with other aspects of AD.

    We took a similar approach in 2016, showing that WMH are increased among mutation carriers in DIAN about 20 years prior to expected symptom onset (Lee et al., 2016) and that these changes, while related, were not mediated entirely by cerebral microbleeds, a marker of cerebral amyloid angiopathy (Lee et al., 2018). Similarly, in older adults with Down’s syndrome, another population at genetic risk for developing AD with very low amounts of vascular risk factors, we observed increased WMH in addition to several other markers of cerebrovascular disease, like microbleeds, enlarged perivascular spaces, and frank infarcts that were not attributable entirely to vascular risk factors (Lao et al., 2020). The cerebrovascular indicators, rather, were related to protein markers that suggest inflammatory processes, particularly in early stages of AD (Moni et al., 2022). 

    These studies, along with myriad others in late-onset AD, converge toward a similar conclusion drawn by Shirzadi and colleagues: WMH are not entirely attributable to exposure to vascular risk factors in AD, so they must be due to “something else.”

    Showing that WMH are not related to vascular risk factors and that they are related to aspects of AD pathophysiology tells us two things: that WMH are not due exclusively to (modifiable) risk factors and that there is a relationship between WMH with other markers of AD. But do these observations prove the etiology of WMH? I do not think this question has been settled. Emerging work implicates inflammatory processes and vascular changes, including aspects of blood-brain barrier dysfunction, at the level of the endothelium in the pathogenesis and progression of AD. This emerging literature points to “vascular intrinsic factors” that may be part of AD pathogenesis; these changes could be amplified by exposure to vascular risk factors, but vascular risk factors are not necessary for there to be a vascular component to AD pathogenesis.

    A potential role of intrinsic vascular factors would induce correlations among factors implicated in AD pathogenesis and progression (and minimize correlations with risk factors). Demonstrating a relationship among biomarkers in AD, as in the paper by Shirzadi and colleagues, shows a codependency among pathophysiological factors related to AD and even highlights that WMH are a core feature of AD, but falls short of proving causality.

    Shirzadi and colleagues caution us not to overinterpret increases of WMH as evidence of mixed vascular and AD pathology, but the autopsy literature would suggest that the vast majority of AD cases in fact comprise mixed pathology (Kapasi et al., 2017), suggesting we should amplify our consideration of multiple pathologies in AD. We need more specific biomarkers to detect the different vascular factors that may be playing a role in AD and increases in efforts to understand how vascular factors contribute to AD.

    The paper by Shirzadi and colleagues is a wonderful contribution to the literature, which I hope inspires continued conversation around these critical questions.


    . White matter hyperintensities are a core feature of Alzheimer's disease: Evidence from the dominantly inherited Alzheimer network. Ann Neurol. 2016 Jun;79(6):929-39. Epub 2016 Apr 27 PubMed.

    . White matter hyperintensities and the mediating role of cerebral amyloid angiopathy in dominantly-inherited Alzheimer's disease. PLoS One. 2018;13(5):e0195838. Epub 2018 May 9 PubMed.

    . Alzheimer-Related Cerebrovascular Disease in Down Syndrome. Ann Neurol. 2020 Dec;88(6):1165-1177. Epub 2020 Oct 9 PubMed.

    . Probing the proteome to explore potential correlates of increased Alzheimer's-related cerebrovascular disease in adults with Down syndrome. Alzheimers Dement. 2022 Feb 24; PubMed.

    . Impact of multiple pathologies on the threshold for clinically overt dementia. Acta Neuropathol. 2017 Aug;134(2):171-186. Epub 2017 May 9 PubMed.

  2. This very important study emphasizes the danger of equating WMH with systemic vascular risk in older persons. It also highlights cerebral amyloid angiopathy as an important factor in the evolution of WMH, and that WMHs may appear before the evolution of imaging evidence of CMBs.

    It is also important to note caveats, as reported by the authors. 

    First, the cohorts do not reflect a generalized population of vascular risk, especially diabetes. Of note, the DIAN cohort is young and with low vascular risk. This is noted as a limitation by the authors, who also did sensitivity analyses investigating the higher-tiered group for vascular risk.

    Second, the FRS is a scale derived to determine vascular risk in middle-aged persons but, in a study of FRS and brain vascular pathology in a community cohort, there was a relatively low correlation of FRS and actual brain vascular pathologies, and the FRS had a low discrimination on an individual level.  (Oveisgharan et al., 2021). 

    Use of a new vascular risk score in an older age groups will be important.

    Finally, these findings also emphasize that CAA, a common but variable feature of AD, is often included in the “vascular disease” category; however, risk factors for CAA are linked to AD, rather than traditional systemic vascular risk factors.  Nonetheless, CAA is related to downstream “vascular” disease effects like WMH, microinfarcts and microbleeds.”


    . Late-Life Vascular Risk Score in Association With Postmortem Cerebrovascular Disease Brain Pathologies. Stroke. 2021 Jun;52(6):2060-2067. Epub 2021 Apr 12 PubMed.

  3. When considering the causes of AD development, it should be taken into account that this disease occurs not only due to impaired metabolism of Aβ and tau protein in the cerebral tissue and vascular wall, but also because of specific disorders of cerebral blood supply.

    In our earlier studies, we identified vascular and microvascular changes that appear in the brain in AD, which were named dyscirculatory angiopathy of Alzheimer's type (DAAT). Regardless of the nature of AD, DAAT affects the microcirculatory arterial as well as venous bed and is classified as cerebral small vessel disease (CSVD). The lesion occurs at an early age and is congenital and hereditary in nature.

    DAAT develops only in patients with AD, not in other neurodegenerative and ischemic cerebral lesions. It is not associated with the atherosclerotic process. The lesion is manifested by increased tortuosity of the distal intracerebral arterial branches, the specific reduction of the capillary bed in the temporal and frontoparietal region, the development of arteriovenous shunts, early discharge of arterial blood through these shunts into the venous bed, the development of abnormal large venous trunks, and subsequent stagnation of venous blood. Our data are confirmed by the studies of others.

    As a result, the cerebral blood supply is completely rebuilt, which leads to damage to tissue structures. In the brain, the neurovascular unit (NVU) is damaged. In cells, mitochondria die, and neurons degenerate and die. This is accompanied by impaired metabolism of Aβ, which leads to a decrease in its excretion and an increase in its accumulation. In its turn, the deposition of Aβ in cerebral tissue and vascular wall reduces the elasticity of microvessels, causing an even greater narrowing of their lumen, which, secondarily, further reduces cerebral blood flow. A decrease in blood flow contributes to an even greater deposition of Aβ, which jointly contributes to the development of progressive neurovascular dysfunction, neurodegeneration, cerebral atrophy, dementia and the development of AD.

    DAAT, in combination with progressive deposition of amyloid beta, defines white-matter hyperintensities (WMHs).


    2023 Alzheimer's disease facts and figures. Alzheimers Dement. 2023 Apr;19(4):1598-1695. Epub 2023 Mar 14 PubMed.

    . A human brain vascular atlas reveals diverse mediators of Alzheimer's risk. Nature. 2022 Mar;603(7903):885-892. Epub 2022 Feb 14 PubMed.

    . The overlap between neurodegenerative and vascular factors in the pathogenesis of dementia. Acta Neuropathol. 2010 Sep;120(3):287-96. PubMed.

    . The vascular factor in Alzheimer's disease: A study in Golgi technique and electron microscopy. J Neurol Sci. 2012 Nov 15;322(1-2):117-21. PubMed.

    . Brain capillaries in Alzheimer's disease. Hell J Nucl Med. 2015 Sep-Dec;18 Suppl 1:152. PubMed.

    . Dyscirculatory Angiopathy of Alzheimer's Type. J Behav Brain Sci 2011;1(2):57-68.

    . Certain new aspects of etiology and pathogenesis of Alzheimer’s disease. Advances in Alzheimer’s Disease, 2012. Advances in Alzheimer's Disease Scientific Researcher

    . Vascular factors in Alzheimer’s disease. Health. 2012 Sep; 4(9A):735-42.

    . Differences in Cerebral Angioarchitectonics in Alzheimer’s Disease in Comparison with Other Neurodegenerative and Ischemic Lesions. World Journal of Neuroscience. 2018 Scientific Researcher

    . Cerebrovascular Changes and Cerebral Atrophy in the Development of Dementia during Alzheimer's Disease. European Society of Medicine, May 26, 2023 European Society of Medicine

  4. Agree that it is no longer correct to assume that white matter rarefaction is equivalent to or completely accounted for by small vessel disease. Our comprehensive postmortem study suggests that neuofibrillary tangles are statistically a much greater predictor of white matter rarefaction.


    . Cerebral white matter rarefaction has both neurodegenerative and vascular causes and may primarily be a distal axonopathy. J Neuropathol Exp Neurol. 2023 May 25;82(6):457-466. PubMed.

    View all comments by Thomas Beach

Make a Comment

To make a comment you must login or register.


News Citations

  1. Bad Blood—Scavenger Receptor Links Aβ to Oxidative Stress in Mice

Paper Citations

  1. . Progressive White Matter Injury in Preclinical Dutch Cerebral Amyloid Angiopathy. Ann Neurol. 2022 Sep;92(3):358-363. Epub 2022 Jun 25 PubMed.
  2. . Age-dependent neurovascular dysfunction and damage in a mouse model of cerebral amyloid angiopathy. Stroke. 2014 Jun;45(6):1815-21. Epub 2014 Apr 29 PubMed.

Further Reading

Primary Papers

  1. . Etiology of White Matter Hyperintensities in Autosomal Dominant and Sporadic Alzheimer Disease. JAMA Neurol. 2023 Dec 1;80(12):1353-1363. PubMed.