Why some seniors develop sporadic Alzheimer’s disease and others do not is still largely a mystery. Now, two new human studies implicate loss of oxygen in the brain as a possible culprit that touches off early AD pathology. In the December 14 PLoS One, researchers led by Henrik Zetterberg at the University of Gothenburg, Mölndal, Sweden, report that cardiac arrest, an extreme form of hypoxia, causes a massive surge in blood Aβ levels. Furthermore, the size and duration of the Aβ spike correlates with worse clinical outcomes, suggesting that more severe hypoxia results in greater Aβ release. Meanwhile, researchers led by Matej Oresic at the VTT Technical Research Centre of Finland, Espoo, analyzed panels of metabolites in the blood of people with mild cognitive impairment (MCI) or AD. In a Translational Psychiatry paper published online December 13, they report that a marker of hypoxia distinguishes people with MCI who go on to develop AD from those who remain stable. Together, these studies suggest that hypoxia is a factor that tilts an aging brain toward AD.

The hypothesis might help explain the consistent link seen between cardiovascular health and AD risk (see, e.g., homocysteine and hypertension; Solomon et al., 2009; Li et al., 2011), and even the association of age with AD. “Oxygen levels in the brain can drop with aging as cerebral blood flow decreases, which could be driving production of Aβ,” noted Kim Green at the University of California, Irvine.

Green was not involved in either study, but has seen similar effects in 3xTgAD mice. In these animals, temporarily decreasing blood flow to the brain jacked up expression of β-secretase (BACE), one of two enzymes that produce Aβ, and kept Aβ levels high for several weeks (see Koike et al., 2010). Low atmospheric oxygen also pushes mice into the disease state through this BACE-mediated mechanism (see ARF related news story on Sun et al., 2006). Similarly, other studies reported that reducing blood flow increases brain Aβ in rodents (see Wang et al., 2010). Brain Aβ is also up in humans who died after strokes or heart attacks (see Qi et al., 2007 and Wisniewski and Maslinska, 1996). And a recent large prospective epidemiological study showed that elderly women with sleep apnea have an almost doubled risk of developing cognitive impairment within five years compared to those who sleep normally, again implicating hypoxia in early neurodegeneration (see ARF related news story on Yaffe et al., 2011).

Zetterberg and colleagues wanted to look at the relationship between hypoxia and Aβ levels in living humans. To do this, they measured Aβ levels in blood serum using an ultrasensitive technology called Single Molecule Arrays, developed by coauthor David Wilson and colleagues at Quanterix Corporation, Cambridge, Massachusetts (see Rissin et al., 2010 and Rissin et al., 2011). The scientists captured serum Aβ with antibodies fixed to beads, then analyzed the samples in tiny, femtoliter-sized wells that held one bead apiece. They quantified Aβ levels by measuring the intensity of a fluorescent reaction produced by a reporter enzyme. Because the signal is concentrated in a minute volume, this method can measure serum concentrations as low as 0.04 pg/mL, making it far more sensitive than conventional enzyme-linked immunosorbent assay (ELISA) methods, the authors report.

The researchers took blood samples from 25 patients who were resuscitated after cardiac arrest, collecting blood at regular intervals from one hour to about four days after their heart attack. The patients’ average age was over 60, but ranged as low as 25. Regardless of age, all patients showed a massive increase in blood Aβ levels over the course of sampling. The hike averaged around sevenfold, but went as high as 70-fold. This is a far greater upsurge in Aβ levels than seen in other conditions, such as in familial AD, Zetterberg noted. The magnitude and duration of the Aβ surge correlated with patient outcome, meaning the spikes were shorter and smaller in patients who recovered good cognitive function, while the brains of patients who died or remained in a vegetative state continued to pump out massive and rising levels of the peptide. Notably, the researchers saw no relationship between clinical outcome and initial levels of Aβ taken after cardiac arrest, nor any interaction with ApoE genotype or gender.

The data add to the evidence that hypoxia may be a causative factor in sporadic AD, Zetterberg told ARF. He suggested that future studies might examine serum Aβ levels in other conditions associated with hypoxia, such as sleep apnea, as well as look more closely at brain vasculature in AD patients. The Aβ upswell in blood probably reflects increased production of the peptide in brain, but it is also possible that the surge is due simply to greater leakage from the brain, perhaps through breakdown of the blood-brain barrier, the authors note. For his part, Green speculated that after hypoxia, Aβ production may ramp up throughout the body, not just in the brain.

The second study, by Oresic and colleagues, also ties markers in human blood to hypoxia in the brain. While other groups have identified protein panels in blood that help predict progression to AD (see ARF related news story on Ray et al., 2007; O’Bryant et al., 2010), Oresic and colleagues focused instead on lipids and other small metabolites. They analyzed fasting blood samples from more than 140 people with MCI, nearly 40 people with AD, and almost 50 healthy controls. The researchers followed up with the people who had MCI roughly 2.5 years later to see who had developed AD and who had remained stable.

In agreement with previous research, the authors found that people with AD tend to have lower levels of many lipids, including sterols, phosphatidylcholines, and sphingolipids, compared to controls (see, e.g., Han et al., 2001; Puglielli et al., 2003; He et al., 2010; Han et al., 2011). Intriguingly, however, Oresic and colleagues saw this altered lipid profile in only a minority of the people with MCI who progressed to AD. In other words, lipid profile did not predict who would develop Alzheimer’s.

Instead, they identified three metabolites in blood that together distinguished between progressors and non-progressors in the MCI group with a sensitivity and specificity of about 80 and 70 percent, respectively. The most predictive of the three was an organic acid, 2,4-dihydroxybutanoic acid, which is more abundant in cerebrospinal fluid (CSF) than in blood by two orders of magnitude. Its levels were higher in people whose disease progressed. Little is known about its biochemistry, Oresic said, except that it is overproduced in low oxygen conditions and generally considered a marker of hypoxia. Further supporting a link between disease progression and hypoxia, the researchers found that the pentose phosphate pathway, which metabolizes glucose to lactate under hypoxic conditions (see Hakim et al., 1976), was more active in MCI progressors compared to non-progressors.

“There is a biochemical signature, a specific set of metabolites, that seems to predict the onset of the disease years in advance,” Oresic told ARF. In the next year, he is planning to do pilot studies in the clinic to see if these blood markers could serve as a simple, inexpensive way to flag people with early cognitive problems who are most at risk for progressing to AD. He will also try to replicate these findings in a larger cohort, as well as look at metabolite profiles in CSF and biopsy samples.—Madolyn Bowman Rogers


  1. This is a fantastic study with important implications. It illustrates the link between oxygen levels and the production of Aβ—something on which I conducted my Ph.D. over 10 years ago. It also suggests that this effect is probably body-wide, rather than restricted to the brain, since large increases of Aβ are found in the serum as a result of cardiac arrest rather than something more brain-specific, such as a stroke. I believe this speaks to Aβ (or at least some fragment of APP) being a hypoxia adaptation molecule, and it could perform this function body-wide, in addition to in the brain. It’s also important to note that oxygen levels in the brain can drop with aging as cerebral blood flow decreases, which could drive production of Aβ as an adaptation.

    View all comments by Kim Green
  2. I agree with Kim Green. My lab recently showed that the upregulation of the genes required for Aβ production (APP, PSEN1, PSEN2, and the γ-secretase-dependent upregulation of BACE1, as observed under hypoxia in human cells) has been conserved in zebrafish, i.e., since the divergence of these lineages approximately 400 million years ago. Thus, these genes appear to be involved in a highly conserved (and thus selectively advantageous/protective) response to low oxygen. See the reference below.


    . The BACE1-PSEN-AβPP Regulatory Axis has an Ancient Role in Response to Low Oxygen/Oxidative Stress. J Alzheimers Dis. 2011 Nov 1; PubMed.

    View all comments by Michael Lardelli
  3. I think this is a wonderful study. It shows us a strong relation between oxygen and Aβ. We can also study what happens with people living at a high elevation, such as the people in La Paz. Is there more Alzheimer's disease there? Or would adaptation to low oxygen lead to less production of Aβ? It will be interesting to me to see if APP is a hypoxia protein molecule, or if Aβ is induced after hypoxia.

    View all comments by Soraya Valles

Make a Comment

To make a comment you must login or register.


News Citations

  1. Environmental Impact Statement: Hypoxia, Stress Boost APP Processing
  2. Breathe Deep—Nighttime Oxygen Loss Linked to Dementia
  3. A Blood Test for AD?

Paper Citations

  1. . Midlife serum cholesterol and increased risk of Alzheimer's and vascular dementia three decades later. Dement Geriatr Cogn Disord. 2009;28(1):75-80. PubMed.
  2. . Vascular risk factors promote conversion from mild cognitive impairment to Alzheimer disease. Neurology. 2011 Apr 26;76(17):1485-91. PubMed.
  3. . Oligemic hypoperfusion differentially affects tau and amyloid-{beta}. Am J Pathol. 2010 Jul;177(1):300-10. PubMed.
  4. . Hypoxia facilitates Alzheimer's disease pathogenesis by up-regulating BACE1 gene expression. Proc Natl Acad Sci U S A. 2006 Dec 5;103(49):18727-32. PubMed.
  5. . Cerebrovascular hypoperfusion induces spatial memory impairment, synaptic changes, and amyloid-β oligomerization in rats. J Alzheimers Dis. 2010;21(3):813-22. PubMed.
  6. . Cerebral ischemia and Alzheimer's disease: the expression of amyloid-beta and apolipoprotein E in human hippocampus. J Alzheimers Dis. 2007 Dec;12(4):335-41. PubMed.
  7. . Beta-protein immunoreactivity in the human brain after cardiac arrest. Folia Neuropathol. 1996;34(2):65-71. PubMed.
  8. . Sleep-disordered breathing, hypoxia, and risk of mild cognitive impairment and dementia in older women. JAMA. 2011 Aug 10;306(6):613-9. PubMed.
  9. . Single-molecule enzyme-linked immunosorbent assay detects serum proteins at subfemtomolar concentrations. Nat Biotechnol. 2010 Jun;28(6):595-9. PubMed.
  10. . Simultaneous detection of single molecules and singulated ensembles of molecules enables immunoassays with broad dynamic range. Anal Chem. 2011 Mar 15;83(6):2279-85. PubMed.
  11. . Classification and prediction of clinical Alzheimer's diagnosis based on plasma signaling proteins. Nat Med. 2007 Nov;13(11):1359-62. PubMed.
  12. . A serum protein-based algorithm for the detection of Alzheimer disease. Arch Neurol. 2010 Sep;67(9):1077-81. PubMed.
  13. . Plasmalogen deficiency in early Alzheimer's disease subjects and in animal models: molecular characterization using electrospray ionization mass spectrometry. J Neurochem. 2001 May;77(4):1168-80. PubMed.
  14. . Alzheimer's disease: the cholesterol connection. Nat Neurosci. 2003 Apr;6(4):345-51. PubMed.
  15. . Deregulation of sphingolipid metabolism in Alzheimer's disease. Neurobiol Aging. 2010 Mar;31(3):398-408. PubMed.
  16. . Metabolomics in early Alzheimer's disease: identification of altered plasma sphingolipidome using shotgun lipidomics. PLoS One. 2011;6(7):e21643. PubMed.
  17. . The effect of hypoxia on the pentose phosphate pathway in brain. J Neurochem. 1976 Apr;26(4):683-8. PubMed.

Other Citations

  1. 3xTgAD mice

External Citations

  1. homocysteine
  2. hypertension

Further Reading

Primary Papers

  1. . Metabolome in progression to Alzheimer's disease. Transl Psychiatry. 2011 Dec 13;1:e57. PubMed.
  2. . Hypoxia due to cardiac arrest induces a time-dependent increase in serum amyloid β levels in humans. PLoS One. 2011;6(12):e28263. PubMed.