The relationship between late-onset Alzheimer’s disease and cardiovascular problems such as hypertension, heart disease, and diabetes is complicated, to say the least. However, according to a large, longitudinal study published August 15 in JAMA Neurology, one such association is straightforward: People with a history of stroke have double the risk of developing late-onset AD (LOAD). This correlation held up independently of other risk factors in people with a family history of LOAD. On the other hand, hypertension lowered the risk of LOAD, and neither Type II diabetes nor heart disease made a difference. It is unclear whether stroke somehow hastened AD, or whether the ischemic attacks served as markers of more general cerebrovascular unrest. The authors, led by Richard Mayeux of Columbia University in New York, favor the idea that stroke leads to the worsening of AD pathology in the brain. The study used only clinical criteria, no amyloid biomarkers, to diagnose LOAD.

The literature is full of observational studies linking a variety of cardiovascular problems with dementia and/or AD in particular (see Luchsinger et al., 2005Reitz et al., 2010Cheng et al., 2011). For hypertension, scientific consensus that mid-life (though not late-life) hypertension increases the risk of dementia has prompted the National Institutes of Health to launch a public awareness campaign. However, some studies have failed to find a correlation between vascular risk factors and AD protein pathology, raising the question of whether vascular problems lead to vascular or mixed dementia, not necessarily AD (see Arvanitakis et al., 2006Chui et al., 2012). To muddy the waters even more, cardiovascular risk factors interact with each other. For example, hypertension can lead to stroke in some people, but not others.

To try to tease out these relationships, first author Giuseppe Tosto and colleagues looked to the National Institutes on Aging multi-center LOAD cohort (aka NIA-LOAD). Participants have LOAD and/or a family history of the disease, which makes them particularly useful to analyze, Tosto said. Just over half of the study’s 6,553 participants, who average 77 years of age, were clinically diagnosed with LOAD either at baseline or during a follow-up period. The researchers found that hypertension at baseline nearly halved a person’s likelihood of having LOAD, while baseline Type II diabetes or heart disease did not alter it. A history of stroke at baseline doubled LOAD risk. Results were similar among 5,972 participants (average age 76.5) from the Washington Heights-Inwood Columbia Aging Project (WHICAP) cohort, which the researchers used as a replication sample: Stroke doubled AD risk, while diabetes and heart disease had no effect, though in this cohort neither did hypertension.

In both cohorts, the associations held steady even when controlling for ApoE4 genotype. This is an important finding because in many families hit hard by LOAD, ApoE4 explains a large fraction of their risk. In addition to ApoE, a subset of NIA-LOAD participants also had genome-wide association study (GWAS) data. From this, the researchers calculated a risk score that predicted a person’s genetic risk of developing AD, independent of his or her ApoE4 genotype. Accounting for this genetic risk score also failed to budge the association between stroke and LOAD risk. This indicated that stroke stood out as a risk factor for clinically diagnosed AD regardless of a person’s genetic predisposition to the disease.

The researchers next examined relationships among the different cardiovascular problems, and how each might alter AD risk. Paradoxically, the researchers found that stroke was positively associated with hypertension, diabetes, and heart disease, even though these latter three factors either lowered or did not alter AD risk when considered independently. Because hypertension is known to cause stroke, the researchers were able to conduct a statistical test to determine how that relationship would ultimately affect AD risk. They concluded that while hypertension lowered AD risk independently, if stroke occurs, all bets are off. In other words, late-life hypertension lowers AD risk, unless it leads to a stroke, in which case AD risk doubles.

Tosto told Alzforum that the results underscore the complex interactions between cardiovascular risk factors. “The relationship between all these interconnected factors and AD is not as straightforward as considered before,” he said. “Except stroke seems to be a solid risk.”

How might stroke raise the risk of AD? Tosto acknowledged that stroke could be a marker for cerebrovascular disease, or could worsen clinical symptoms caused by a given amount of AD pathology. However, he favors the idea that ischemic events worsen AD pathology. Tosto and colleagues previously reported that white-matter hyperintensities, which signify blood vessel leaks, were correlated not only with worse clinical outcomes, but also with brain atrophy and elevated tau in the cerebrospinal fluid (see Tosto et al., 2014Tosto et al., 2015). While the present study is not mechanistic, Tosto speculated that stroke could somehow trigger an exacerbation of AD pathology. Other researchers have proposed that AD pathology and vascular injury affect cognition separately (see Feb 2013 news). 

David Knopman of the Mayo Clinic in Rochester, Minnesota, found the results of the study compelling, but interpreted them differently. He pointed out that members of the NIA-LOAD cohort were diagnosed based only on clinical symptoms, and that recent imaging studies suggest that around a third of those people could lack AD pathology. Therefore, Knopman said the findings support the idea that stroke correlates with elevated dementia risk, but not necessarily AD risk. He also questioned the idea that stroke could directly cause cognitive problems after the initial consequences of the attack had subsided.

“It seems implausible that a stroke that had little or no cognitive consequence in its immediate aftermath would be the cause of cognitive impairment several years later,” Knopman said. “More likely stroke is a proxy for a more generalized process.”

Frank Wolters of Erasmus University in the Netherlands took a similar tack. “A lot of clinically defined LOAD is a mixture of classical AD pathology and vascular disease, such as microinfarcts,” he told Alzforum. “So while a person might be clinically diagnosed with AD, they might have other things going on that influence symptoms.”

Tosto told Alzforum that in the absence of effective treatments for AD, prevention by way of protecting vascular health remains the best medicine.—Jessica Shugart


No Available Comments

Make a Comment

To make a comment you must login or register.


News Citations

  1. Vascular Dementia or Alzheimer’s: Is the Delineation Emerging?

Paper Citations

  1. . Aggregation of vascular risk factors and risk of incident Alzheimer disease. Neurology. 2005 Aug 23;65(4):545-51. PubMed.
  2. . A summary risk score for the prediction of Alzheimer disease in elderly persons. Arch Neurol. 2010 Jul;67(7):835-41. PubMed.
  3. . Type 2 diabetes and late-onset Alzheimer's disease. Dement Geriatr Cogn Disord. 2011 Jan 1;31(6):424-30. PubMed.
  4. . Diabetes is related to cerebral infarction but not to AD pathology in older persons. Neurology. 2006 Dec 12;67(11):1960-5. PubMed.
  5. . Vascular risk factors and Alzheimer's disease: are these risk factors for plaques and tangles or for concomitant vascular pathology that increases the likelihood of dementia? An evidence-based review. Alzheimers Res Ther. 2012 Jan 4;3(6):36. PubMed.
  6. . Predicting aggressive decline in mild cognitive impairment: the importance of white matter hyperintensities. JAMA Neurol. 2014 Jul 1;71(7):872-7. PubMed.
  7. . The effect of white matter hyperintensities on neurodegeneration in mild cognitive impairment. Alzheimers Dement. 2015 Jun 13; PubMed.

External Citations

  1. public awareness campaign

Further Reading

Primary Papers

  1. . The Role of Cardiovascular Risk Factors and Stroke in Familial Alzheimer Disease. JAMA Neurol. 2016 Oct 1;73(10):1231-1237. PubMed.