New Frontiers in Understanding and Treating Age-Related Cognitive Decline

Table of Contents

• Key Points
• Background: Why New Approaches Are Needed
• Better Animal Models for Studying Cognitive Aging
• How Epigenetic Changes Contribute to Memory Loss
• Can the Aging Brain Be Rejuvenated?
• Validating Noninvasive Brain Stimulation
• New Neuron Growth Throughout Life
• Discussion and Key Takeaways
• What This Means for Patients
• Limitations and Open Questions
• Recommendations and Future Directions
• Frequently Asked Questions
• Source Information

Key Points

• Chronic inflammation may contribute to cognitive decline, connecting cognitive aging to conditions such as cardiovascular disease, obesity, diabetes, and cancer.
• In aging mice, HDAC3 acts as a brake on long-term memory formation; deleting or inhibiting it improved memory and synaptic plasticity.
• In mouse experiments, blood-borne factors influenced brain function: old blood impaired young brains, and young blood rejuvenated older brains.
• Noninvasive brain stimulation may improve memory, but researchers must show target engagement and use controls to rule out placebo effects.
• New neurons form throughout life in the hippocampus but decline with age; promoting neurogenesis may help memory discrimination.

Background: Why New Approaches Are Needed

Scientists are only beginning to understand the molecular mechanisms that underlie cognitive function, cognitive reserve, and resilience. Cognitive reserve refers to the brain's ability to withstand damage and still function well. Resilience refers to the brain's capacity to adapt and recover.

Despite significant progress in recent years, a tremendous unmet need remains for new approaches to delay or prevent dementia. The authors argue that progress will come from a better understanding of several key areas.

• How early-life and mid-life experiences contribute to the buildup of cognitive reserve and resilience
• How these mental capacities interact with ever-accumulating pathology (disease-related damage) in the aging brain
• The extent to which the brain retains plasticity (the ability to change and adapt) in later life

Research from both animal studies and human trials has identified physiological factors that influence higher-level brain functions. As has been seen in many other diseases — including cardiovascular disease, obesity, diabetes, neurological disease, and cancer — chronic inflammation may play a role in cognitive decline and other neurological diseases.

Identifying factors that contribute to cognitive decline, or conversely, factors that contribute to cognitive resilience, could represent new avenues of research and treatment. A better understanding of the causes of inflammation and ways to reduce it could benefit many people. The impact of chronic inflammation on gene expression, possibly resulting in epigenetic modifications, may be of particular significance.

Some key questions could also be addressed through better animal models of age-related cognitive decline. Rodent studies have immeasurably improved our understanding of human disease, but their applicability to humans is limited. This limitation is especially true when considering higher-order cognitive domains such as complex memory and decision-making.

Studying cognitive decline in animals that are more similar to humans in terms of social and cognitive abilities may be useful. Domestic dogs can learn and remember complex tasks for extended periods. Dogs socialize and interact well with humans. Dogs also seem to undergo cognitive decline as they age. For these reasons, dogs could be a valuable model system in which the effects of aging are studied.

In the end, research on cognition must be interpreted in the context of the whole individual. Studies of cognitive function will need to be performed and validated in older adults. Speakers in this session of the Cognitive Aging Summit III, chaired by Marilyn Albert, explored novel approaches and models that can capture the many nuances of cognitive aging.

Better Animal Models for Studying Cognitive Aging

The design and interpretation of experiments regarding cognitive function will require an appreciation of the complexity of the human cognitive experience across the life span. Researchers also must understand the cognitive and neural response to interventions designed to improve function.

Mammals commonly used in cognitive studies include mice, rats, dogs, and primates. It is important to take into account the differences among strains and individual animals. Indeed, there is no such thing as "the" mouse or "the" primate. All animal strains have unique idiosyncrasies that can confound interpretation of experimental results.

For example, the C57BL/6 mouse is the most commonly used inbred mouse strain for laboratory research. A study of mice with severe neurodegeneration identified a mutation in the GTPBP2 gene (Ishimura et al., 2014). However, the physical characteristics (phenotype) were only observed in C57BL/6 mice obtained from the Jackson Laboratory. The same strain obtained from other sources did not show these characteristics.

Analysis revealed that the Jackson strain had a brain-specific mutation in one of its transfer RNA molecules. This mutation, in combination with the GTPBP2 mutation, led to the observed phenotype only in this strain. This unexpected finding underscores the need to be vigilant for potential strain-specific idiosyncrasies when designing and interpreting studies of cognitive aging.

There is a need for better animal models of human physical and cognitive health assessment. This need is especially important when one realizes that an individual's goal is often not to live a longer life but to live a healthier one. People want mobility, cognition, and sensory function preserved for as long as possible. One of the more promising animals in which this "healthspan" is studied is the domestic dog.

Dogs offer several advantages over other animals in the study of human disease and health:

• Dogs are diurnal (active during the day, like humans) rather than nocturnal (active at night, like rodents)
• Dogs show tremendous physical variability
• Dogs have been selected for a wide range of physical traits and behaviors
• Dogs are comfortable with and respond to humans
• Detailed genetic and genomic information about the dog is now being developed (Ostrander, 2012)

Dogs can be trained to perform cognitively complex tasks that are similar to those performed by humans. Service dogs, for example, are trained to remember specific routes, open doors, turn lights on and off, find and retrieve specific objects, and help people get dressed and undressed.

Service dogs retire at various ages, perhaps because cognitive decline impairs their ability to carry out their service role. Retired service dogs could therefore represent an outstanding research resource for scientists studying aging and cognitive function. Different dog breeds cognitively age at different rates, or perform better or worse in different cognitive domains (Ostrander, 2012). Research to uncover the mechanisms underlying these processes could be relevant to humans.

Scientists who might collaborate in studies of aging in dogs could include cognitive scientists, dog behaviorists, dog service trainers, and canine geneticists. The Dog Aging Study (dogagingproject.com) at the University of Washington aims to enroll 10,000 dogs in a long-term study of aging. Interested researchers could access this resource to collaborate and generate a registry of people with specific dog breeds who are interested in participating in a research study. An additional resource might be people who are already participating in aging studies and are interested in enrolling their dogs in the study.

How Epigenetic Changes Contribute to Memory Loss

One research laboratory is working to understand the mechanisms underlying long-term memory processes in the young adult and aging brain. Most of this work is at the level of the epigenome — the chemical modifications to histone proteins and DNA that modify and regulate chromatin structure.

The epigenome represents a powerful signal transduction platform. It reflects the interaction of the genome (a person's genetic material) and experience or environment, and how this interaction ultimately produces observable characteristics. The overall hypothesis is that enzymes that modify chromatin play a role in controlling gene expression. This gene expression is required for stable changes in neuronal function that give rise to long-term changes in behavior.

Histone acetylation is a classic epigenetic modification of chromatin. This process has been implicated in hippocampal memory formation (Levenson et al., 2004) and synaptic plasticity (Vecsey et al., 2007). The hippocampus is a brain region critical for forming new memories. Synaptic plasticity is the brain's ability to strengthen or weaken connections between neurons.

Histone acetylation is a dynamic process. The addition of acetyl groups to histones is balanced by their removal via histone deacetylases, or HDACs. One HDAC, called HDAC3, has been identified as a critical negative regulator of long-term memory formation (McQuown et al., 2011).

Deletion or inhibition of HDAC3 can transform a subthreshold learning event into a robust long-term memory. A subthreshold learning event is an experience that would not normally lead to a short-term or long-term memory. With HDAC3 inhibition, this weak learning event produces a long-term memory that persists past the point at which normal memory would fail.

HDAC3 inhibition can gate the information that is processed into long-term memory. It can also affect the specificity and strength of memory, and potentially even the number of neurons recruited for memory representation (Bieszczad et al., 2015; Campbell and Wood, 2019).

These observations led researchers to examine the possibility that the impaired ability to form new memories in the aging brain might be a consequence of dysregulated chromatin modification. This dysregulation results in a more repressive chromatin structure that impedes gene expression. This hypothesis was originally put forth by Sweatt and Barnes (Penner et al., 2010).

Evidence for this hypothesis comes from several observations:

• Gene transcription is disrupted in normal aging (Berchtold et al., 2008; Rowe et al., 2007)
• Age-related memory problems seem to be dependent on hippocampal function (Driscoll and Sutherland, 2005)
• The epigenome is altered in aged neurons, resulting in a repressive chromatin structure that prevents normal gene expression required for long-term memory formation (Peleg et al., 2010; Snigdha et al., 2016)

The researchers found that young adult mice (2–4 months old) perform better on an object location memory task than aging mice (18–20 months old). However, older animals lacking HDAC3 perform like younger animals (Kwapis et al., 2018). Focal genetic deletion of HDAC3 ameliorates both synaptic plasticity and long-term memory impairments observed in aging mice. HDAC3 may therefore act as a molecular "brake pad," constraining the formation of new long-term memories in the aging brain.

In trying to identify possible mediators of this effect, researchers identified 4 genes whose expression was increased in younger mice during memory consolidation. These genes failed to increase in older mice. Their expression was restored in older mice after HDAC3 disruption (Kwapis et al., 2018). One of these genes encodes a core component of the circadian molecular machinery. Circadian machinery controls the body's internal clock and daily rhythms.

Additional experiments demonstrated how HDAC3 may be regulating this circadian gene in the hippocampus during memory formation. Overall, the results suggest that circadian gene regulation represents a potential mechanism to link aberrant epigenetic repression to memory impairments in the aging brain. Circadian genes may have an autonomous role in the hippocampus for memory function beyond their role in circadian rhythm function in the suprachiasmatic nucleus. The suprachiasmatic nucleus is the brain's master clock region.

Can the Aging Brain Be Rejuvenated?

Another research group is addressing the question of whether the effects of aging — in particular, its effects on regeneration and cognitive aging — can be reversed. These researchers are examining the features of the young brain (around age 20s) with its adaptive, resilient properties. They are asking whether there is a way to restore these functions in the older brain (over age 65).

If this approach could be realized, it would represent a transformative way of thinking about aging and addressing its effects on cognition in later life. The ultimate goal is to extend the number of cognitively healthy years in the life span.

Researchers use mouse models to explore the potential for animals to repair and restore lost cognitive function in old age. To examine these processes, they administer tasks that are associated with regenerative capacity and age-related impairment. One of these tasks is a test of spatial learning and memory.

In this test, researchers measure the ability of mice to find a hidden platform in a pool of water based on its location with respect to surrounding visual cues. A young animal can learn this task quickly. An older animal takes longer to learn the task and makes many more errors while doing so (Villeda et al., 2011, 2014).

The researchers explored systemic factors that might influence brain function. They looked for factors in the blood, because blood touches all organs and systems. They used parabiosis — the anatomic joining of 2 animals so that they share a common blood supply — to test whether there are systemic factors present in old animals that dispose them toward cognitive decline. They also tested whether there are factors present in young animals that protect them from cognitive decline. In other words, they asked whether cognitive decline was the consequence of loss of a protective factor or the gain of a harmful one.

Using this approach, the researchers conducted a targeted proteomic screen. A proteomic screen is a method that examines many proteins at once. They identified several immune factors that were especially prominent in "old blood" (Villeda et al., 2011). This finding suggested that inflammation at a systemic level may be an important contributor to cognitive decline.

The researchers looked more closely at beta-2 microglobulin (B2M), whose levels increase in blood during aging in mice and humans. Genetic studies of B2M knockout mice (mice bred to lack this protein) show that young animals are normal but older animals have more regenerative capacity.

Conversely, young mice who received injections of systemic B2M performed more poorly on tests of cognitive function. However, this effect was reversible if B2M was discontinued (Smith et al., 2015).

Ongoing studies are exploring whether there are "pro-youthful" factors that can help rejuvenate neural regeneration and restore cognitive function. Ultimately, researchers hope to better understand the mechanisms by which cognitive function can be preserved. They also hope to identify factors that may mediate this ability to rejuvenate neural functioning and plasticity.

Validating Noninvasive Brain Stimulation

In recent years, there has been an explosion of researchers trying to find ways to influence the brain using noninvasive methods. These methods include noninvasive brain stimulation such as transcranial magnetic stimulation, transcranial direct current stimulation, and transcranial alternating current stimulation. They also include activities such as cognitive training.

Current methods to evaluate the effectiveness of these approaches are limited. They may not differentiate placebo from treatment. This is because the interventions themselves are complex and involve a variety of subjective experiences. These subjective experiences limit the ability to blind participants and researchers to experimental conditions. They also limit the ability to control various nonspecific factors that could mediate the response of outcome variables.

Because of this complexity and the relative lack of information regarding potential mechanisms of action on brain and cognitive function, special precautions must be taken to validate these interventions.

One laboratory is evaluating the utility of transcranial magnetic stimulation to influence the human hippocampal brain network and thereby alter memory function. This is an important target because, with age, connections between the hippocampus and its cortical networks decrease. This decrease is associated with age-related memory impairment (Salami et al., 2014).

The studies exemplify the approach the researchers advocate for validating neural stimulation and other putative interventions for cognition. This approach includes the following 3 requirements:

1. Researchers must show that the putative neural target of the intervention is engaged by the stimulation regimen. That is, the targeted brain network responsible for the cognitive function of interest must change in response to the intervention.
2. Researchers must use within-experiment control and comparison conditions to validate the effects of stimulation on cognition and brain function. This effectively guards against nonspecific factors.
3. The experimental design and statistical analysis methods must guard against nonspecific or placebo effects and statistical false-positive results.

In addition to standard considerations regarding factors such as sample size, statistical power, and experimental design, researchers must avoid post hoc data analysis decisions. An example of a problematic post hoc decision is splitting research subjects into groups based on positive versus negative treatment responses. This practice is a form of "p hacking" that amplifies the likelihood of false-positive results (Head et al., 2015).

In young adults, researchers found that stimulation increased functional connectivity among cortical-hippocampal network regions and improved associative memory performance (Wang et al., 2014). Thus, the putative neural target — the hippocampal network — was engaged by stimulation. This finding satisfied the first requirement for validation.

In a follow-up experiment using spatial memory testing, researchers measured subjects' general ability to remember the location of an object (success) and also the precision of that memory. Previous findings have indicated that the hippocampus is more critical for memory precision than it is for general success (Kolarik et al., 2017).

The researchers found selective effects of stimulation on precision as well as on its neural correlates, with no effects on success (Nilakantan et al., 2017). The selectivity of the effect on precision but not general success provides a within-experiment control. This control guards against possible nonspecific factors of stimulation and meets the second requirement for validation.

Furthermore, stimulation affected the specific aspect of cognition that researchers hypothesized would respond to stimulation based on previous neuroanatomical findings. This showed further evidence for target engagement and further satisfied the first requirement for validation.

By applying these rigorous methods in older adults, researchers found that stimulation selectively improved recollection memory and increased activity within the hippocampal network. This validated the effects of stimulation on the specific behavioral and neural hallmarks of age-related memory decline (Nilakantan et al., 2019).

Regarding the third requirement for validation, the appropriate design and analysis of any experiment related to cognition is of key importance. All researchers are conscious of the current focus on rigor and reproducibility as it relates to factors such as sample size and statistical corrections for multiple comparisons. However, there are additional factors to consider in analyzing experiments utilizing complex interventions.

For instance, experiments must include appropriate control conditions, including "positive" and "negative" intervention controls. In these experiments, researchers deliver full-intensity stimulation to regions that are not within the hippocampal-cortical network, such as the motor cortex, to guard against nonspecific effects of stimulation.

Furthermore, it is a common practice to divide study participants into intervention "responder" and "nonresponder" groups after the fact and then look for differences between these groups. This practice is problematic because nonspecific differences almost inevitably exist between such groupings. Therefore, the likelihood of falsely identifying effects related to intervention is artificially inflated.

Finally, in studies of brain stimulation, it is important to incorporate adequate measurement of pain, discomfort, and other subjective states that could amplify nonspecific or placebo effects. By following these principles, researchers can increase the likelihood that any effects they observe in their studies are valid, reproducible, and meaningful.

New Neuron Growth Throughout Life

The adult brain continues to generate new neurons throughout life in the dentate gyrus region of the hippocampus. This process is called neurogenesis. However, neurogenesis does decline with age (Altman and Das, 1965; Eriksson et al., 1998; Spalding et al., 2013).

Recent evidence suggests that adult-born neurons play a critical role in decreasing interference between similar memories. These neurons may support pattern separation, which is the ability to distinguish between similar but distinct experiences (Clelland et al., 2009; Knierim and Neunuebel, 2016; McAvoy et al., 2016; Nakashiba et al., 2012; Sahay et al., 2011).

Understanding the mechanisms that regulate adult neurogenesis may identify strategies to promote continued neurogenesis into older age. This could represent an attractive strategy to mitigate cognitive decline.

Genetic studies modulated levels of neurogenesis in the dentate gyrus of adult mice. These studies showed that changes in neurogenesis influenced the ability of mice to distinguish between similar contextual representations (McAvoy et al., 2016). In particular, aged mice that were manipulated to produce more adult-born neurons were better able to discriminate between 2 similar contexts compared with aged controls (McAvoy et al., 2016; McAvoy and Sahay, 2017).

Researchers are now working to better understand the mechanistic underpinnings of these findings. They are also working to identify molecules that could safely produce a similar therapeutic effect, which may improve cognition in aging humans.

Discussion and Key Takeaways

Animal models of human disease are just that — models. However, they can yield important insights into disease processes that would be difficult to discover in studies of people. Therefore, animal models represent important systems in which biology can be manipulated and studied in ways and over life course intervals that are impractical in humans.

Not all animal models are equally relevant for investigating human disease processes. One particularly overlooked issue is the nature of the environment in which studies are performed. For example, rodent studies are typically conducted in environments deficient in cognitive stimulation. Simple environmental enrichment is well-known to affect neuroanatomy and physiology, cognitive performance, as well as multiple disease processes (Mustroph et al.).

What This Means for Patients

This research has several important implications for patients and their families who are concerned about cognitive aging.

• Inflammation matters: Chronic inflammation may contribute to cognitive decline. This finding connects cognitive aging to other health conditions such as cardiovascular disease, obesity, diabetes, and cancer.
• Epigenetic changes are reversible in principle: The discovery that HDAC3 acts as a "brake" on memory formation in aging mice suggests that it may be possible to release this brake with targeted treatments.
• Blood-borne factors can influence the brain: The finding that young blood can rejuvenate older brains — and that old blood can impair young brains — opens new avenues for treatment.
• Brain stimulation shows promise but needs rigorous validation: Noninvasive brain stimulation techniques may improve memory, but researchers must follow strict standards to prove these effects are real and not placebo.
• New neurons continue to form in adult brains: The birth of new neurons in the hippocampus declines with age but can potentially be promoted to improve memory discrimination.

These findings are all from animal studies or early-stage human research. They do not yet translate directly into treatments that patients can access today. However, they represent important steps toward developing new interventions.

Limitations and Open Questions

The research presented in this review has several important limitations that patients should understand.

• Animal models are not humans: Even the best animal models cannot fully replicate human cognitive aging. Rodent studies in particular have limited applicability to higher-order human cognitive domains.
• Strain-specific effects can mislead: The example of the C57BL/6 mouse shows that genetic differences between animal strains can produce misleading results.
• Brain stimulation studies face unique challenges: It is difficult to blind participants and researchers in brain stimulation studies. This difficulty makes it hard to separate true treatment effects from placebo effects.
• Post hoc analyses can inflate false-positive results: Dividing participants into "responders" and "nonresponders" after the fact can create the illusion of treatment effects that do not really exist.
• Environmental factors are often overlooked: Many animal studies are conducted in environments that lack cognitive stimulation, which may affect the results.

Many questions remain unanswered. How do early-life and mid-life experiences shape cognitive reserve? How does the brain retain plasticity in later life? Can pro-youthful factors be safely harnessed to rejuvenate the aging brain? These questions require further research.

Recommendations and Future Directions

Based on the research presented, the authors suggest several priorities for future work.

1. Develop better animal models: Domestic dogs represent a promising model for studying cognitive aging because they share many cognitive and social features with humans. The Dog Aging Study at the University of Washington aims to enroll 10,000 dogs in a long-term study.
2. Study epigenetic mechanisms: Understanding how HDAC3 and other epigenetic regulators control memory formation could lead to new treatments for age-related memory loss.
3. Investigate blood-borne factors: Identifying "pro-youthful" factors in young blood that can rejuvenate neural regeneration and restore cognitive function is a key goal.
4. Apply rigorous standards to brain stimulation research: Researchers must demonstrate target engagement, use within-experiment controls, and guard against false-positive results.
5. Promote adult neurogenesis: Finding safe ways to maintain or increase the birth of new neurons in the aging brain could help mitigate cognitive decline.
6. Consider the whole person: Research on cognition must be interpreted in the context of the whole individual. Studies of cognitive function will need to be performed and validated in older adults.

The authors emphasize that delaying or preventing dementia will require a better understanding of the biological and environmental factors that contribute to cognitive aging. Studies in animal models can provide important insights into the factors that contribute to the accumulation of reserve and ways in which it is preserved or depleted.

A better understanding of the molecular processes that underlie these elements would enhance and guide not only research but also treatment approaches. These treatment approaches may include noninvasive brain stimulation and drug treatments to promote youthfulness or combat the aging process.

It is important to realize, however, that these processes occur in the context of the human experience. Studies of cognitive aging must consider the complexity and individuality of each person's life.

Frequently Asked Questions

What is cognitive reserve and why does it matter?

Cognitive reserve is the brain's ability to withstand damage and still function well. Resilience is its capacity to adapt and recover. Researchers say progress in delaying dementia will come from understanding how early-life and mid-life experiences build these capacities, and how they interact with accumulating disease-related damage in the aging brain.

Why are dogs being studied to understand human cognitive aging?

Domestic dogs are diurnal like humans, show great physical variability, respond to humans, and can be trained to perform cognitively complex tasks. Dogs also appear to undergo cognitive decline with age, and different breeds age cognitively at different rates. A long-term study aims to enroll 10,000 dogs to study aging.

What is HDAC3 and how might it relate to age-related memory loss?

HDAC3 is an enzyme that acts as a critical negative regulator of long-term memory formation. In studies of aging mice, deleting or inhibiting HDAC3 improved synaptic plasticity and long-term memory, and older mice lacking HDAC3 performed like younger animals. Researchers describe HDAC3 as a molecular brake pad on new memories in the aging brain.

Can young blood rejuvenate an aging brain?

In mouse experiments using shared blood supplies, researchers found immune factors prominent in old blood, including beta-2 microglobulin (B2M), whose levels rise with age in mice and humans. Young mice injected with B2M performed worse on cognitive tests, but the effect was reversible when B2M was stopped. Pro-youthful factors are still being explored.

Does the adult brain still make new neurons?

Yes. The adult brain continues to generate new neurons throughout life in the dentate gyrus region of the hippocampus, a process called neurogenesis, though it declines with age. These adult-born neurons appear to help distinguish between similar but distinct experiences. In aged mice manipulated to produce more adult-born neurons, discrimination between similar contexts improved.

Are noninvasive brain stimulation treatments proven for memory?

Not yet. Noninvasive brain stimulation shows promise, but researchers say strict validation is needed because these interventions are hard to blind and placebo effects are difficult to separate. In older adults, one study found stimulation selectively improved recollection memory and increased activity in the hippocampal network. These are early-stage findings, not treatments patients can access today.

Can I get any of these treatments now?

No. These findings come from animal studies or early-stage human research and do not yet translate directly into treatments patients can access today. They represent steps toward developing new interventions. Delaying or preventing dementia will require understanding both biological mechanisms and each person's individual life experience, with studies validated in older adults.

When should someone worried about age-related cognitive decline seek a second opinion?

Cognitive aging research is still at the animal-model and early human-study stage, so findings on inflammation, epigenetic changes, blood-borne factors, brain stimulation, and adult neurogenesis do not yet translate into treatments available today. A second opinion is reasonable when memory or thinking changes are being managed without a clear explanation of the underlying cause, or when the plan rests on interventions whose effects have not been validated in older adults. Because cognitive function must be interpreted in the context of the whole individual, an independent review of the full picture can help. Diagnostic Detectives Network provides independent expert second opinions.

Source Information

Original article title: Innovative approaches in cognitive aging

Authors: Steven N. Austad, Marcelo A. Wood, Saul A. Villeda, Joel L. Voss, Amar Sahay, Marilyn Albert

Author affiliations: Department of Biology, University of Alabama at Birmingham; Department of Neurobiology and Behavior, University of California, Irvine; Department of Anatomy, University of California, San Francisco; Department of Medical Social Sciences, Feinberg School of Medicine, Northwestern University; Harvard Stem Cell Institute and Department of Psychiatry, Massachusetts General Hospital, Harvard Medical School; Department of Neurology, Johns Hopkins School of Medicine

Publication details: Neurobiol Aging. 2019 November; 83: 150–154. doi:10.1016/j.neurobiolaging.2019.04.013. Published in final edited form. Available in PubMed Central (PMC) as an HHS Public Access author manuscript.

Disclosure: One author (SAV) is on the scientific advisory board for Alkahest, a company developing commercial products targeting geronic blood factors. The other authors report no conflicts of interest.

This patient-friendly article is based on peer-reviewed research.