In This Issue
Spring Bridge on Technologies for Aging
March 15, 2019 Volume 49 Issue 1

Supporting Precision Aging: Engineering Health and Lifespan Planning for All of Us

Tuesday, March 12, 2019

Author: Eric Dishman

“The aim of precision medicine is to find ways to make health care more tailored to each person based on their individual differences.” (All of Us Research Program 2018)

“Healthy Ageing [is] the process of developing and maintaining the functional ability that enables well-being in older age.” (WHO 2015, p. 28)

Precision medicine—or precision health—is about delivering care that -precisely addresses the unique biology, needs, and circumstances of an individual. The move to the broader term health recognizes that data streams far more diverse than one’s DNA sequence—and interventions far broader than targeted drug therapies—are required to bring sustainable health and longevity to everyone.

Thanks to dramatic improvements in image analytics, computing, large-scale databases, and cloud capabilities, precision medicine provides next-generation sequencing and molecular diagnostics that deepen understanding of one’s unique biology. These technologies make it possible to uncover an individual’s genetic risk for an increasing number of diseases and to determine pharmacogenetic-based treatments knowing which drugs may help or harm that individual (Collins and Varmus 2015).

The same tools are enabling scientific breakthroughs based on “-multi-omic” analysis of an individual. For example, metabolomics yields knowledge of how a specific individual metabolizes drugs, food, and even environmental chemicals (the “exposome”). Proteomics evaluates the proteins expressed in that individual, and microbiomics produces data about the complex relationship between gut bacteria, the brain, and the individual’s entire biological system. These emerging fields of precision medicine hold great promise for more effectively treating, preventing, or even curing many age-related and other conditions (Fiocchi 2014; O’Toole and Flemer 2017).

Meanwhile, the emergence and growth of consumer apps and electronic devices such as fitness wearables, smartphones, watches, telehealth devices, and “smart home” systems provide the potential to support health through new data streams that provide a view of not only a patient’s biology but also behavior, psychology, and attitudes. The stage is being set for healthcare researchers and providers to develop a more holistic and complex understanding of health at both individual and population levels.

Supporting Health over the Lifespan

Overall health involves more than longevity or the absence of pain and disease to “live long and prosper,” in the immortal words of Mr. Spock from Star Trek. Healthy aging encompasses a broad view of multiple factors -needed to promote the World Health Organization’s goal of “well-being in older age” (WHO 2015, p. 28).

Gerontologists have long known, for example, that a person’s function and independence characterize her progression through life stages more than age alone. This thinking is integrated into the WHO (2015) public health framework for Healthy Ageing that considers the importance of an individual’s intrinsic capacity and functional ability in his or her well-being over time.

What Is Precision Aging?

The mashup of the fields and tools of precision health and healthy aging provides fertile ground for rethinking care for people of all ages. Precision aging, then, is about being able to individualize and optimize a person’s care over the lifespan based on deep, data-informed understanding of her or his unique biological, behavioral, environmental, and socioeconomic circumstances.

The platform for precision aging integrates technological breakthroughs from many industries and scientific fields to produce unprecedented real-world, often real-time data. Understanding this lifespan -thumbprint—or “lifeprint”—can transform aging research and how someone is cared for and prepared for across all stages of life. Simplistic “real age” calculators (e.g., Sharecare or BBC iWonder) that try to assess an individual’s functional capacity and overall health status abound on the internet today, but they pale in comparison to what is possible in an era of precision aging.

Lifespan Planning

Within the next decade or two, it seems reasonable to imagine that most individuals will be able to leverage personalized, predictive models with their care teams to do lifespan planning from an early age, just as some people do today with financial planning. Perhaps life-span planning becomes the norm as part of every annual physical. Or perhaps even the frequency of physicals is customized based on an individual’s health status, “omic” profile, life goals, recent exposures, and continuous functional assessments from everyday devices.

A real-time, risk-adjusted lifespan plan that draws on many data streams may enable clinicians (perhaps “lifespan coaches”?) to recommend a well-balanced portfolio of actions and interventions across life stages with assistive technologies that help an individual to achieve them. The same data enable the research, the plan, and adherence to the plan. The plans will differ significantly from person to person to bolster overall health, function, and resilience.

These plans would need to start early. It is becoming clear that what happens to each person in utero and -early childhood establishes very long-term health and functional trajectories for better and for worse. For example, in an article on “The Obstetric Origins of Health for a Lifetime,” Barker and Thornburg (2013, p. 511) write: “There is now clear evidence that the pace and pathways of early growth and development are major risk -factors for the development of a range of chronic diseases including coronary heart disease and type 2 diabetes.”

The variability of longevity, health, function, freedom, and vitality of roughly same-age individuals runs the gamut from lives of fun and fitness to frailty and futility. What data are needed to help scientists better understand the complex combination of factors that cause some 80-year-olds to be bed-bound or dependent on a loved one for basic daily activities and others to still be running races and fully active in the workforce? Moreover, as the basic understanding of etiology becomes more nuanced and holistic, what are the diverse vectors of health that come together to predict and prepare an individual for a healthy, high-quality life?

A Portfolio Framework for Precision Aging

As with a good financial plan, a lifespan plan requires a diversified portfolio of investments (of one’s time, actions, and money) that target the unique risks an individual may face in many different domains of health. Based on almost two decades of global fieldwork, surveys, and design sessions with people over age 65 and their family/friends/professional care providers, my Intel lab created an “Opportunity Map” to drive innovations in “aging in place” and independent living (Plowman et al. 2009). It guided our technology development and in-home pilots by focusing on one opportunity area at a time in great depth. Across our studies it became clear that the presence and perception of health or frailty are highly variable, complex, and multifactorial for individuals, so I have revised the map here as a framework for a precision aging approach.

 Figure 1

Seven overlapping vectors, in different and dynamic combinations for each individual, help define a port-folio of overall health and quality of life at any point in time (figure 1). An individual may have full function and capacity (represented by a full piece of the pie in that vector) or varying needs and gaps in function. Four foundational vectors—akin to Maslow’s (1943) “basic” hierarchy of needs—play significant roles in total health status: Safety and Comfort, Physical and Bodily Health,[1] Cognitive and Mental Health, and Connection and Community. The other three -vectors—Choice of Time, Choice of Place, Meaning and Purpose—-represent higher-order aspects of personal autonomy that enhance quality of life.

Illustrative Scenarios

A few scenarios based loosely on three families from Intel’s fieldwork help to tease out the complex, interrelated dynamics of these seven vectors. In our observations, two individuals with seemingly similar foundations of safety, physical and cognitive health, and connection may vary greatly in health and quality of life. And it is rarely the gaps in a single vector that wipe out whole-person health if there is a balanced portfolio of function and capacity in the others. However, slow, unnoticed declines in numerous vectors may feel like a sudden, catastrophic event to those caught unaware.

“Joanne” is a former software engineer living with her husband Bernard in Arizona, forced to “retire young,” as she put it, at age 57 due to the onset of mild cognitive impairment. Most of her memory issues (struggles with numbers, sequences of activities, and names of familiars) have remained stable for 20 years. She also has a long personal and family history of diabetes that is well managed and a congenital heart condition that has caused her no problems so far. A move to Arizona was initially good for her allergies, but not anymore since transplants from all over brought their houseplants, too.

 Figure 2

At Time A (figure 2), Joanne is 79 with a healthy, fulfilled life thanks to what she calls the “heroic efforts” of Bernard, who is 10 years younger and working two full-time jobs as an engineer and caregiver for Joanne. He helps her with her meds, food prep, daily schedule, and transportation to see former work and tennis friends. She finds joy in teaching young children from her church how to read and lives for her short, solitary neighborhood walks.

Bernard is becoming worried because she forgets to lock the front door and close the garage that fronts a major street and sidewalk. Things have been stolen out of the garage. Although he tracks her short walks with a location app on her smartphone, he is concerned because she has gotten lost and confused recently about how to call him for help.

For Physical & Bodily Health and Cognitive & -Mental Health, where Joanne has gaps, her husband’s compensation for those means the rest of her health vectors are fulfilled. Even her Safety & Comfort has reduced only slightly because of Bernard’s “free taxi service,” as he referred to it.

But just 18 months later, the portfolio and her overall health and vitality are in decline because of Bernard’s massive stroke and subsequent need for home care. Their adult children who live hours away by plane insisted on moving Joanne and Bernard into an assisted living facility.

 Figure 3

Although Joanne’s chronic disease management (Physical & Bodily Health) improved thanks to nursing care and her cognition remained stable, her overall health plummeted (figure 3). She hates where they live; the stroke and move disconnected her from -Bernard and her old neighbors and friends (-Connection & Community has gone almost to zero). Choice of Time is now half because she is forced into scheduled activities she detests. Choice of Place is almost nil because the facility staff limit her movement on campus, and she has lost most of her -Meaning & -Purpose because her “church kids” no longer come for tutoring.

Figure 4 

In conrast, “Carter,” age 80, appears similar to Joanne since he also lives in an assisted care facility, with memory loss from slow-moving dementia and several chronic diseases, including arthritis (figure 4). But he is “loving my life here just fine.” He has lived alone since childhood, so moving into the facility 10 years ago was a boost to his social network. His Safety & Comfort is lower because of occasional falls at night when he forgets to grab his cane by the bedside before heading to the bathroom. But by day he is quite content and socially active, attending yard sales and eating out weekly with a new-found “friend with wheels!” who lives in the independent homes on campus.

 Figure 5

A third scenario, with “Emma,” age 77, a retired school teacher who still lives in the farmhouse she and her husband built 50 years ago, shows someone who is “healthy as an ox,” with no physical ailments or disease (figure 5). She can still drive the 24 miles to the nearest grocery store, even in the snow. On summer days when the humidity is not too bad, she uses a push mower on the small patch of lawn in front of the house, but points out that “the tractor Aaron used for the fields just rusts since I never learned to drive it.”

But, in contrast to her physical health, Emma suffers from lifelong, mostly untreated depression, made significantly worse by the loss of Aaron 5 years earlier. Her life orbited around him, and with the closest neighbor 6 miles away (“they are strangers to me,” she says), she now has almost no Connection & Community. She is distrustful of strangers and increasingly fearful someone will break in and kill her at night. Her Choice of Time, Place, and Meaning are all low because she has nowhere to go amid the farm fields. She leads a mostly sedentary life with the companionship of 24/7 television noise.

 Figure 6

These three stories illustrate some of the common challenges—with unique and different impacts for -individuals—that our social science team observed repeatedly in the families we studied (figure 6). While it is not easy today to predict whether and how a person will face any of these challenges, what might be possible in the near future?

The Future for Precision Aging

The platform for doing precision aging research and, eventually, lifespan prediction, planning, and intervention is becoming widely available and affordable. The growing ubiquity of smartphones, homes, and fitness wearables with increasing diagnostic capabilities enables real-time assessment of biological, behavioral, cognitive, and social contexts (see Dodge and Estrin in this issue). These same technologies afford individualized interventions in just the right way at just the right time.

The Internet of Things and cloud computing are establishing the architecture and infrastructure for precision aging assistants—from robotic companions to social and memory support bots to self-driving cars. Progress and significant global investment in artificial intelligence, deep learning, and multidimensional analytics at the enormous scale of omic and other molecular diagnostic datasets mean the pace and power of precision health discoveries for people of all ages is growing exponentially.

So for a future Joanne, perhaps there will be options to optimize for a healthy lifespan even before she is born. Her parents may opt for prenatal whole genome sequencing that reveals her risk of eventually facing a particular kind of dementia based on her variants and key proteins in her blood. Her Arizona allergies never become a problem since the same analysis may show a set of pollens and plants she needs to be exposed to in early childhood to build resistance.

In her early 20s, subtle changes in her speech and decision-making patterns—unconscious to her and -others—may show the earliest signs of her dementia risk, triggering a series of cognitive games that may help. If her predementia continues to progress into her 40s and 50s, an individualized diet, meditation, and medication regimen may slow that progression, with the help of a behavior change coach designed for her psychology, attitudes, and preferences to keep her on track.

If dementia progresses into her 70s and 80s, there is a better chance that safety monitoring, smart transportation, and digital assistants can supplement the help of her partner so that her church tutoring and neighborhood walks are still something to live for. Self-driving cars may be a new normal, enabling her to safely get out into the world.

For Carter, perhaps his slow-progressing dementia never forms because the stress and inflammation of his youth that would trigger the “ministrokes” across decades are prevented. An understanding of how his body metabolizes certain foods, drugs, and chemicals keeps him away from the careers and places with toxins he is susceptible to. His doctor shifts his diet and exercise over the years, with adjustments enabled by a customized do-it-yourself blood analyzing patch. While his form of arthritis is still incurable, a nonaddictive pain medication sustains his golf game and even an occasional run well into his 80s. As his fall risk increases in his early 100s, his personal health system identifies times of day when his gait and stamina are good enough to share a walk with his friends.

For Emma and her isolation and loneliness, an early diagnosis of a specific depression yields a pharmaco-genetic solution that calms her mood swings and bouts of paranoia. At an early age, she, her family, and her friends know the risks of her self-imposed isolation, and that collective mindfulness prompts others to reach out at just the right moment. While rural living may still be her preference or only option for affordable housing, Emma’s assistive apps keep her socially engaged, cognitively stimulated, and purposeful.

These vignettes are not meant to be a utopic celebration in which technology magically solves all human health challenges. They describe ways in which science and technology are enabling more nuanced understanding of health and aging processes that can be brought to bear at an individual level. They also describe the kinds of support systems that may be needed across one’s life-span. The opportunity to do lifespan planning across the many vectors of overall health may improve longevity and quality of life even during the current age wave that is increasing the number of people affected by frailty, disability, and age-related disease.

A Personal Perspective: Persistent Engineering Questions and Challenges

In 1993, as a junior social scientist and programming geek freshly arrived at Paul Allen’s think tank, I was asked to do fieldwork in places where there was no computing. I started studying nursing homes and families dealing with dementia.

As part of a project called “Elderspace,” we engineered mock-ups—some real, most fake—of ways to improve nursing homes: wearable sensors to track an elder’s steps and balance, video tracking to alert the staff to falls, voice analytics to detect cognitive decline and depression, wireless tablets and earbuds for computing everywhere, and wall-sized telehealth stations for remote chronic care. Almost nothing worked; we were too early in the maturity curve of these technologies to be effective (Dishman 2003). We later realized our engineering energies could have been spent on a better question: how to use computing and new technologies to bypass nursing homes altogether?

As I write this perspective 25 years later, having -funded around 200 Intel university grants and started several companies to support consumer health and aging in place, I have seen great promise but few sustainable outcomes. Many of the technologies have been too early, incomplete, or expensive.

Now at the National Institutes of Health to lead the All of Us Research Program, I see that it is technological déjà vu all over again. All of Us is a life stage, life-span, big data study, with eventually 1 million pregnant -women, newborns, kids, young adults, and people of all ages sharing a wide range of health information and data—clinical, genetic, behavioral, social, environ-mental, lifestyle—so that scientists can better understand the complex interactions of health described in the seven vectors above.

 Figure 7

While much closer to enabling precision aging for everyone, the current state of the underlying technologies and infrastructure is still not ready for prime time, in my opinion. The list of technical challenges just to do the research is daunting (figure 7). And integration of these capabilities into everyday lives and homes so people can manage their portfolio of health needs will require tackling systems -challenges—-technological, political, economic, and social—in at least four areas.

Data Sharing at Scale

The persistence of data silos and dataset protectionism is a threat to discovery and cures. As most health plans and medical centers rush to brand themselves as the cutting edge of precision medicine, they often hoard patient data for only their own researchers to make discoveries, publish the first papers, and garner the first patents. As tech firms large and small rush to sell their black-boxed data on consumers for product targeting and marketing, progress in precision aging suffers from the lack of large-scale datasets that could power faster discoveries for both common and rare diseases.

All of Us is an open resource, but even at 1 million people, it will not have the statistical power needed to predict and individualize care. The field needs sample sizes in the billions, not millions. While we are building data linkages to other large research programs and to nonmedical datasets (e.g., weather, pollution), the lack of interoperability and, in some cases, adequate security in systems we are connecting to must be overcome.

Systems Engineering and Integration

There have been thousands of pilot projects and early products to monitor and support the seven vectors of health and quality of life, but most are point solutions. Oftentimes, they have been tested for short periods with a very small sample. But a few-hundred-person pilot of a medication-prompting system, social support agent, mobility assistant, or fall prevention system provides only anecdotal evidence. Few, if any, of these point products will move the needle on health and quality of life outcomes. Integration into a flexible, standards-based, adaptive system to meet the changing needs of health over time would improve the research, user experience, and outcomes.

Mixed-Discipline Innovation

There is still inadequate funding for and focus on precision aging solutions by governments, foundations, and philanthropists. The kind of systems integration described above is costly and complex. Iterative development and testing require mixed-discipline teams of electrical engineers, computer scientists, software developers, experience designers, usability engineers, bioinformaticians, social scientists, social workers, -clinicians, economists, and many more. These teams can rarely muster the disciplines and dollars to do the kinds of systems development and longitudinal testing that will build the scientific, clinical, and economic case for precision aging solutions at scale.

Reengineering Care and Business Models

The biggest challenge I see in almost every research experiment or startup in the broad space of precision aging is that they rarely do the true experiment. Time and time again, important innovation in technology and infrastructure is not complemented by the reengineer-ing of the care or business models that could transform and sustain health care and aging care.

For example, the pilot of the preventive care solution to stop most heart attacks occurs in a hospital-driven, fee-for-service paradigm that thrives on heart surgeries to pay the bills. The clinical care model and incentives are not realigned to make the best workflows and outcomes happen around new technical systems. If the care models and business models—the workflows and -dollar flows—are not reengineered as well, then precision aging solutions have little chance to gain traction over the next 25 years.

Conclusion

The capabilities to turn precision aging from science fiction to scientific discovery to everyday practice are all around. The emerging technologies and tools to collect and analyze new data types from all seven health vectors increase every month. Breakthroughs in precision medicine for cancer and other conditions appear almost weekly.

With systems engineering and well-designed experiments, the potential to build support systems—-embedded in what must become newly designed systems of care—for a variety of changing health needs across the lifespan is very real. “Live long and prosper” can be more than the mantra of a science fiction character. It can be a rallying cry to make individualized lifespan planning a reality—to enable health and longevity for Joanne, Bernard, Carter, Emma, and all of us.

Acknowledgments

I thank Daozhong Jin and Ashley Armstrong for their help in the preparation of this manuscript.

References

All of Us Research Program. 2018. The costs of imprecision. Online at https://www.joinallofus.org/en/costs-of--imprecision.

Barker DJP, Thornburg KL. 2013. The obstetric origins of health for a lifetime. Clinical Obstetrics and Gynecology 56(3):511–519.

Collins FS, Varmus H. 2015. A new initiative on precision medicine. New England Journal of Medicine 372(9):793–795.

Dishman EJ. 2003. Designing for the new old: Asking, observing, and performing future elders. In: Design Research: Methods and Perspectives, ed Laurel B. Cambridge: MIT Press.

Dodge HH, Estrin D. 2019. Making sense of aging with data big and small. The Bridge 49(1):39–46.

Fiocchi C. 2014. Integrating omics: The future of IBD? Digestive Diseases 32(suppl 1):96–102.

Maslow AH. 1943. A theory of human motivation. Psychological Review 50(4):370–396.

O’Toole PW, Flemer B. 2017. From culture to high throughput sequencing and beyond: A layperson’s guide to the “omics” and diagnostic potential of the microbiome. Gastro-enterology Clinics 46(1):9–17.

Plowman T, Prendergast DK, Roberts SW. 2009. From people to prototypes and products: Ethnographic liquidity and the Intel Global Aging Experience Study. Intel Technology Journal 13(3):20–39.

WHO [World Health Organization]. 2015. World Report on Ageing and Health. Geneva.

[1]  The two terms comprise both physical injury and negative or harmful effects on the body as a system (e.g., disease, pain), to accommodate different associations among readers/cultures internationally.

About the Author:Eric Dishman is director of the All of Us Research Program at the National Institutes of Health.