Researchers have been interested in the connections between stress and Alzheimer’s disease for years, but hard data remained elusive at first. “It's only been in the last decade, with more sensitive assay techniques and more insight into the mechanisms of AD, that supportive evidence has emerged,” Robert Sapolsky at Stanford University, Palo Alto, California, wrote to ARF. Researchers now have better tools, such as knockout mice and pharmacological manipulations, to investigate the mechanisms and pathways involved in stress, agreed John Trojanowski at the University of Pennsylvania, Philadelphia. In the last five years in particular, mounting animal data indicate that stress hormones have the ability to worsen both Aβ and tau pathology. At this point, “There’s pretty good evidence that the impact of glucocorticoids on the brain is somehow intimately involved with the development of unhealthy brain aging, and maybe AD,” said Philip Landfield at the University of Kentucky, Lexington.

Early studies showed that acute stress could trigger AD pathology, as administering glucocorticoids to wild-type rodents raises amyloid-β precursor protein (APP) levels and tau phosphorylation in the brain (see Elliott et al., 1993; Budas et al., 1999). Kim Green and Frank LaFerla at the University of California, Irvine, extended this finding to the triple transgenic AD mouse (3xTgAD), reporting that treating young mice for seven days with glucocorticoids raised levels of BACE1, APP, and Aβ, and, as a downstream consequence of Aβ changes, boosted total tau (see Green et al., 2006 and ARF related news story).

Other studies showed that Aβ levels also rise in models of chronic stress, such as extended periods of restraint and isolation (see Jeong et al., 2006; Lee et al., 2009; and Huang et al., 2011). John Csernansky and David Holtzman at Washington University, St. Louis, Missouri, reported that chronic isolation stress decreased hippocampal neurogenesis, impaired memory, and accelerated Aβ plaque deposition in Tg2576 transgenic mice (see Dong et al., 2004; Kang et al., 2007; and ARF related news story). Administering corticotrophin-releasing factor (CRF), but not glucocorticoids, mimicked this effect, suggesting it is mediated by CRF, the authors noted.

Osborne Almeida at the Max Planck Institute of Psychiatry, Munich, Germany, wanted to see how chronic stress would affect very early pathological events in AD. He decided to use wild-type animals to better model sporadic disease. As described in the May 25 Journal of Neuroscience, first authors Ioannis Sotiropoulos and Caterina Catania subjected healthy, middle-aged rats to a month of daily stresses, including overcrowding, restraint, or rocking motion. Serum corticosterone levels shot up approximately sevenfold. Then, the researchers examined tau hyperphosphorylation in the hippocampus and prefrontal cortex, areas that show early AD pathology, and found that stressed animals developed about 50 percent more tau phosphorylation and insoluble tau aggregates than did controls. They also showed memory problems. Administering exogenous glucocorticoids led to similar changes in tau as the stressors did, suggesting hormones were mediating the harmful effects of stress (see Sotiropoulos et al., 2011).

Almeida told Alzforum that previous in-vitro studies performed by his group suggest that glucocorticoid treatment leads to tau hyperphosphorylation through an Aβ-mediated mechanism (see Sotiropoulos et al., 2008). Although the wild-type rats did not develop Aβ plaques, Almeida believes tau phosphorylation in these animals also occurs downstream of changes in soluble Aβ. To test if glucocorticoids exacerbate Aβ pathology, the authors injected wild-type rats with both Aβ and corticosterone and saw higher levels of tau phosphorylation than in rats receiving Aβ alone. The effect was greatest in animals that had been previously stressed. The results suggest that stress can accelerate AD pathology, even in otherwise healthy animals, Almeida said.

Trojanowski and colleagues at UPenn took a different approach. They looked at chronic stress in a tau mutant background. As reported in the October 5 Journal of Neuroscience, first author Jenna Carroll subjected Tg2576 APP mutant mice and PS19 tau mutant mice to one month of severe stress by housing the animals in isolation (mice like company) and restraining them in a small conical tube for six hours each day. As expected from previous studies, this treatment increased Aβ levels in the Tg2576 mice and worsened fear and spatial memory. Similarly, in the tau mice, but not in wild-type, chronic stress increased tau hyperphosphorylation, tau aggregation, memory problems, and neurodegeneration. This showed that stress is equally harmful on a tau mutant background as in APP mutant mice. Moreover, both Tg2576 and PS19 mice were more sensitive to stress than were wild-type animals; they released more corticosterone, and chronic stress did not blunt the response as it does in normal mice. The authors found that glucocorticoid receptor expression stayed high in the hippocampus of tau mice during chronic stress, not falling as it does in wild-type mice. In toto, these data suggest that AD pathology can alter the way the brain responds to stress. Other lines of evidence also indicate that AD disrupts the normal functioning of the hypothalamic-pituitary-adrenal (HPA) axis (see Part 1).

Carroll and colleagues wondered if glucocorticoids were mediating the effects of stress in this tau model. However, administering exogenous glucocorticoids did not mimic stress. In contrast, when the researchers blocked CRF receptors with an antagonist, they prevented tau accumulation and neurodegeneration, and rescued learning in the PS19 mice. This suggested to them that local CRF might bring about tau pathology. Strengthening this idea, the researchers found that transgenic mice that overexpress CRF in the brain produce more hyperphosphorylated tau than their wild-type littermates do.

Other animal studies, such as the one from Holtzman’s group mentioned above, implicate CRF as a key stress mediator as well (see also ARF related news story on Rissman et al., 2007). However, the picture is complicated because CRF has been shown to be neuroprotective (see, e.g., Bayatti and Behl, 2005), suggesting the effects of CRF may depend on the context. The CRF results also conflict with experiments in other animal models that have fingered glucocorticoids as the main stress mediator. Mark Mattson at the National Institute on Aging in Baltimore, Maryland, noted that the Trojanowski study does not rule out the possibility that glucocorticoids affect tau, since global inhibition of CRF in the brain will also lower glucocorticoid levels. An interesting follow-up experiment would be to use adrenalectomized mice that cannot produce corticosterone, or to administer a CRF antagonist directly to the hippocampus, to more clearly exclude a glucocorticoid effect, Mattson suggested. Green speculated that perhaps CRF is the most important stress mediator early in the development of pathology, while glucocorticoids may be the more important hormone in established disease when cortisol levels are up (see Part 1).

Trojanowski told ARF that his study tried to dissect out which types of stress are most harmful. The scientists put a separate group of mice through a month of variable stress, where every day the animals randomly experienced one of several stressors. They were either forced to swim for 20 minutes, restrained for 15 minutes, in cold water for 2.5 minutes, housed alone all day, or had the lights on all day. These forms of stress increased glucocorticoid secretion just as much as the lengthy restraint and isolation paradigm did. However, in contrast to the latter, mice exposed to variable stressors showed few negative effects, and no increase in Aβ or phosphorylated tau compared to unstressed mice.

The results emphasize the idea that not all stressors are equally bad. In particular, acute stress may have very different effects than chronic stress, Carroll told ARF, noting that a single 15-minute period of restraint in an otherwise unstressed tau mouse actually lowered tau phosphorylation. This fits with the classic idea that acute stress is an adaptive response that benefits an organism, in contrast to the negative effects of chronic stress. It is also possible that certain behaviors or environmental factors can counteract the effects of stress, Carroll suggested. She speculated that the mice who had variable stress may have inadvertently experienced some benefits from exercise or novel environments. Exercise has been consistently linked to lower risk of AD (see tau mutant mice AlzRisk data), and environmental enrichment has been shown to lessen neuropathology in numerous animal models (see Nithianantharajah and Hannan, 2006). In follow-up work, Carroll is looking at whether housing the PS19 mice in an enriched environment can lessen tau phosphorylation, and whether this environment can counteract the effects of previous stress.

How do these findings in animal models relate to the stresses humans experience? People regularly face all kinds of psychological stress, for example, from deadlines, interpersonal conflict, traffic jams, and demanding jobs, but is any of this bad for you? The animal results suggest that pathological effects may come primarily from chronic exposure to very high-stress situations, Carroll said, such as active military duty in a war zone. Indeed, epidemiologic evidence has linked post-traumatic stress disorder (a dysregulation of the HPA system) to a twofold higher risk of getting dementia (see ARF related news story). This, then, might relegate everyday stress in what would be considered normal proportions to a rather marginal role in AD. For a look at stress mechanisms and possible therapies, see Part 3.—Madolyn Bowman Rogers

This is Part 2 of a three-part series. See also Part 1 and Part 3. Download a PDF of the entire series.


  1. Sotiropoulos et al. investigated the impact of stress, amyloid, and glucocorticoids on tau phosphorylation and behavioral performance in rats. Although I think the paper is relevant to current research directions, the data as presented are troubling. The method for administering amyloid is unconventional and completely uncontrolled. The latter is a big problem because in this case, the peptide was administered by ICV mini-pump for 14 days. There is no mention that control animals received mock pumps or manipulations (e.g., anesthesia, etc.) of any kind. The photomicrographs do not demonstrate the effects discussed, and methods for scoring them are inadequate (score 0.5-5 by visual examination by independent investigator).

    Unfortunately, this paper does little to advance our understanding of how stress can contribute to AD.

    View all comments by RR Robert Rissman
  2. In rereading the methods, I withdraw my statement about the pumps not being controlled. The methods mention "or vehicle" which comprised sterile distilled water. I apologize for this mistake. Other than this, I stand by the remainder of my comments.

    View all comments by RR Robert Rissman

Make a Comment

To make a comment you must login or register.


News Citations

  1. SfN: Return of the Other—Tau Is Back, Part 1
  2. Stress and Aβ—A Fluid Connection in Mice
  3. Stress and AD: Does One Beget the Other?
  4. Stress and AD—Taking a Toll on Tau
  5. Stress and Trauma: The Puzzle of Post-traumatic Stress Disorder
  6. Stress and AD: How the Brain Responds Makes All the Difference

Paper Citations

  1. . Corticosterone exacerbates kainate-induced alterations in hippocampal tau immunoreactivity and spectrin proteolysis in vivo. J Neurochem. 1993 Jul;61(1):57-67. PubMed.
  2. . The effect of corticosteroids on amyloid beta precursor protein/amyloid precursor-like protein expression and processing in vivo. Neurosci Lett. 1999 Nov 26;276(1):61-4. PubMed.
  3. . Glucocorticoids increase amyloid-beta and tau pathology in a mouse model of Alzheimer's disease. J Neurosci. 2006 Aug 30;26(35):9047-56. PubMed.
  4. . Chronic stress accelerates learning and memory impairments and increases amyloid deposition in APPV717I-CT100 transgenic mice, an Alzheimer's disease model. FASEB J. 2006 Apr;20(6):729-31. PubMed.
  5. . Behavioral stress accelerates plaque pathogenesis in the brain of Tg2576 mice via generation of metabolic oxidative stress. J Neurochem. 2009 Jan;108(1):165-75. PubMed.
  6. . Long-term social isolation exacerbates the impairment of spatial working memory in APP/PS1 transgenic mice. Brain Res. 2011 Jan 31;1371:150-60. PubMed.
  7. . Modulation of hippocampal cell proliferation, memory, and amyloid plaque deposition in APPsw (Tg2576) mutant mice by isolation stress. Neuroscience. 2004;127(3):601-9. PubMed.
  8. . Acute stress increases interstitial fluid amyloid-beta via corticotropin-releasing factor and neuronal activity. Proc Natl Acad Sci U S A. 2007 Jun 19;104(25):10673-8. PubMed.
  9. . Stress acts cumulatively to precipitate Alzheimer's disease-like tau pathology and cognitive deficits. J Neurosci. 2011 May 25;31(21):7840-7. PubMed.
  10. . Glucocorticoids trigger Alzheimer disease-like pathobiochemistry in rat neuronal cells expressing human tau. J Neurochem. 2008 Oct;107(2):385-97. PubMed.
  11. . Corticotropin-releasing factor receptors differentially regulate stress-induced tau phosphorylation. J Neurosci. 2007 Jun 13;27(24):6552-62. PubMed.
  12. . The neuroprotective actions of corticotropin releasing hormone. Ageing Res Rev. 2005 May;4(2):258-70. PubMed.
  13. . Enriched environments, experience-dependent plasticity and disorders of the nervous system. Nat Rev Neurosci. 2006 Sep;7(9):697-709. PubMed.

Other Citations

  1. 3xTgAD

External Citations

  1. PS19 tau mutant mice
  2. tau mutant mice AlzRisk data

Further Reading

Primary Papers

  1. . Chronic stress exacerbates tau pathology, neurodegeneration, and cognitive performance through a corticotropin-releasing factor receptor-dependent mechanism in a transgenic mouse model of tauopathy. J Neurosci. 2011 Oct 5;31(40):14436-49. PubMed.
  2. . Stress acts cumulatively to precipitate Alzheimer's disease-like tau pathology and cognitive deficits. J Neurosci. 2011 May 25;31(21):7840-7. PubMed.