The Lucky Years: How to Thrive in the Brave New World of Health (14 page)

Some of the other exciting research I’m involved with entails overlaying health records from millions of people with variables such as the weather and news events. Think about examining, for example, what happens to kids who were born during the week Hurricane Sandy swept over the eastern seaboard of the United States. That one environmental impact alone could have had health consequences.

Stephen J. Elledge is a professor of genetics at Harvard Medical School and Brigham and Women’s Hospital in Boston. His research is leading to tools for tracking patterns of disease in different populations. His work is ultimately helping us to understand the differences between the young and the elderly, as well as people from various parts of the world. A test he’s recently developed, for example, could be used to find out whether viruses, or the body’s immune response to them, have a hand in chronic diseases, including cancer. This test, called VirScan, requires just a drop of blood and can broadcast nearly every virus a person has been exposed to throughout life, past and present. First reported about in the journal
Science
in 2015, VirScan can currently identify more than a thousand strains of viruses from 206 species, which reflects the entire human “virome,” or all the viruses known to infect humans, from the common cold to HIV.
¹
The test works by detecting antibodies, which are the body’s defense mechanisms against an invader. These are highly specific proteins the immune system manufactures to combat germs such as viruses. And once you are exposed to a virus and make an immune response to it, the antibodies stay around and provide a “record” that you were exposed to that virus.

The application of this type of test will be staggering. We will be able to generate all kinds of data from documenting people’s exposures to disease. Some have compared this technology to the development of the electron microscope, which allowed us to have more resolution at the micro level and “see” things previously invisible. One application, for example, will be mapping out historical and current patterns of disease
and seeing how certain diseases are affected by the type of antibodies a person has. We’ve long suspected that viruses may contribute to chronic ailments such as heart disease, asthma, and autoimmune diseases, the latter of which are characterized by a glitch in the immune system whereby it produces antibodies that mistake a person’s own cells for foreign invaders and attack them.

There’s a lot we still don’t know about the relationship between, say, overcoming a flu bug and being diagnosed with type 1 diabetes or multiple sclerosis later on. But no such viruses or antibodies have ever been identified in terms of these diseases, and the research in this realm has been tricky. To look for them, we’ve had to single out suspect viruses and test for them individually. But with a test like VirScan, we can look at the big picture and establish Big Data to find correlations between certain viral infections and future risk of diseases. For example, an infection from virus
X
leads to higher risk for illness
Y
later in life. Or maybe those antibodies you develop as a result of that infection will
protect
you from certain ailments. The test can thus be used to shed light on an individual’s risk factors and inform future health decisions. The technology might even help answer questions about cancer, which operates differently in different people. Perhaps the reason lies in which antibodies a person has, and when they developed, which in turn affects an individual’s response to treatment and drugs such as chemotherapy.

The best part about this test is that it’s inexpensive (about $25) and can be done in a matter of hours. This is one-stop shopping that should become part of your annual checkup so you can be as proactive as possible. And its utility will go up exponentially the more the test is deployed and the larger the database becomes.

Data mining for disease patterns won’t always have to rely on bodily fluids. There will be lots of opportunities to find connections in our complex biology from even the most uncomplicated observations. Case in point: In 2014, it was discovered that artificial sweeteners wreak havoc on the body, disrupting its microbial inhabitants,
²
known collectively as the microbiome (a subject I’ll cover in the next chapter). That, in turn, can affect metabolism and blood sugar balance. Although it was
exploring the microbiome that finally gave us this insight, we could have known years ago about the relationship between drinking diet soda and increased risk for diabetes had another type of database been already in place to collect the information. Such a huge finding could have come from simply knowing what people buy and consume (lots of products made with artificial sweeteners) and their health profiles (lots of insulin resistance and diabetes).

For another example, consider the 2015 headline that probably caused some panic in people who take heartburn drugs that use proton pump inhibitors (PPIs) such as esomeprazole (Nexium), omeprazole (Prilosec), and lansoprazole (Prevacid). It said that these drugs, many of which are now available over the counter, can increase your risk of a heart attack up to 21 percent regardless of whether or not you have a preexisting heart condition. The beauty of this conclusion is that it came from the simplest of data mining. Researchers at Houston Methodist and Stanford University analyzed 16 million clinical documents from some 2.9 million patients.
³
That’s how they traced a connection between people who were prescribed common antacids with PPIs and the chance of having a heart attack. The research was conducted after a report came out in 2013 in the journal
Circulation
that showed how PPIs could potentially lead to long-term cardiovascular disease on a molecular level: they change the lining of blood vessels.
⁴
Interestingly, patients who were prescribed another type of antacid that contains what are called H2 blockers had no increased likelihood of heart attack.

Antacids that contain PPIs are most often prescribed for digestive system issues, including gastroesophageal reflux disease (GERD). According to the FDA, 1 in 14 Americans has taken proton pump inhibitors, among the most commonly used drugs in the United States. So the results of this study are alarming, but keep in mind that they reflect an association—not causation. The people who take PPIs and experience heart trouble could have other health challenges that compound the problem, such as obesity, high blood pressure, and underlying genetic risk factors. Nevertheless, culling this kind of information from Big Data
is part of our brave new world of medicine. It gives us more knowledge to consider in taking care of ourselves.

Let me give you one more example, one that’s already in action and changing the lives of millions. Just a few years ago, the University of Nottingham developed an app called MyBabyFace with the help of the Bill & Melinda Gates Foundation. The problem: global infant mortality remains high at 22 per 1,000 live births—especially in countries where high-tech ultrasounds and skilled physicians are not readily available. The majority of these deaths occur because of bad math; it’s not easy calculating an accurate age for babies being born at an unknown gestation. Consequently, premature babies are not identified in a timely fashion, and we often miss the opportunity for simple, low-cost interventions that help avoid the complications related to being born premature, such as hypothermia. The MyBabyFace app leverages the power of crowdsourcing. Parents across the world upload a photo of their baby’s feet, face, and ears as well as the gestational age. The current purpose of the app is simply to collect data, and the hope is to develop a database to make it easier to judge a child’s time in the womb. Another app, Neogest, uses the depth of foot wrinkles and the roundness of the eyes as clues for gauging a baby’s degree of prematurity.

What makes a data-mining tool like Google so brilliant is that it’s a constantly evolving, giant cataloging system—a way of organizing all the information on the internet. In the computer world, we call a data structure that holds enormous amounts of data coded in a way that allows one to quickly search and retrieve information a hash table. But Google does even better than, say, the hash table used by a massive library to arrange, store, and track books. Every time somebody searches the site, Google gets better at returning results, and the site grows ever stronger and more powerful. This kind of power will come into the health realm in the Lucky Years; there will be a hash table
for genomics, proteomics, environmental factors (like living next to a freeway), lifestyle habits (like eating Paleo or smoking), as well as medical conditions (like being diabetic and allergic to shellfish). Every day I treat patients similar to the ones I saw the prior week; yet I am not improving the care I give or my recommendations based on knowledge culled from previous patients, because the system is not set up for that. But it will be soon. The power of the “Google” data approach is that my care going forward will improve with every patient I treat; the database will have increasingly more information upon which to base decisions, so that my patients and I can make better ones. A colossal amount of data is being collected: it’s been estimated that as much data is now generated in just two days of 2015 as was generated from the beginning of civilization until 2003.

Here’s a good question: Who will host and manage all of this data, some of which is highly sensitive? A plan needs to be put forward to establish a new nonprofit entity that will host all of this data and keep it secure and anonymous. This data is a major world resource and needs to be safeguarded and kept free of biases that government and for-profit companies can bring, and which can result in abusing the database for purposes of discrimination or, worse, blackmail. We need only one host so that the database can be large enough to yield answers.

My clinic, for example, and eleven others across the country have teamed up with the IBM supercomputer known as Watson, which is being trained to analyze the genetic data from a patient’s cancer and search the scientific literature to help guide treatments for this patient. As the artificial-intelligence, “cognitive” computer gets more information and learns about matching patients to treatments, Watson helps us realize the goal of truly personalized medicine. The
Washington Post
summarized the hope perfectly: “And best of all, Watson will continue to learn on the job as it hunts for appropriate oncology treatments. That means that Watson will gain in value and knowledge over time, based on previous interactions with medical practitioners. The more that participating institutions use Watson to assist clinicians in identifying
cancer-causing mutations, the more that Watson’s rationale and insights will improve.”
⁵

At one time, we were afraid to have our financial info online or use a computer to transfer money and store our financial data. Now we barely think twice about doing it. There are tools and assurances built into the system. The same needs to and will happen with health data. And we’ll be able to move from doctor to doctor without repeating tests or worrying about where our records are and whether they are safe. In cases of emergency, we’ll be able to quickly and effortlessly access our entire health record to optimize our care—especially if we are in dire straits and an ER physician needs to know, for instance, if we’re allergic to a medication he’s about to administer.

When a group of banks created a member-owned association in 1966 to revolutionize how payments could be processed without relying on traditional checks and cash, MasterCard financial services corporation was born. But think about what this company, and other banking companies, have had to overcome to make their product work in the market. They’ve had to establish rules for authorization; standardize the billing process; establish rules for international currency exchange; put in place guidelines and procedures to minimize fraud and other abuses; and handle marketing, security, and legal aspects to running a worldwide organization. And now they have to keep up with the increasingly hackable digital world where thieves, embezzlers, and pirates are lurking invisibly everywhere. But the financial services corporations nonetheless still work, and we consumers and debtors continue to trust and rely on them. Just think about how much more efficient the financial services industry is today—and how much easier your life is—thanks to credit cards, digital payments, and being able to view all of your money data online at the click of a mouse.

It’s hard to imagine what life would be like without this system. In ten to twenty years, we will be saying the same thing about access to our health data and the data of anonymous others that adds context to our own. You check your messages today like clockwork, probably, and
tomorrow you’ll be similarly checking things like your blood sugar levels or heart rate using the tools available to you through handy portable devices including your phone. In the Lucky Years, your smartphone will be an incredibly powerful diary of your life that can inform your health decisions, sometimes at a moment’s notice. Ultimately, this facilitates the management of disease in real time rather than around scheduled doctor’s visits, allowing for quicker access to help or quicker adjustment of treatments to resolve the problem. One area in particular where this is sorely needed today is the realm of mental health. And help is on the way.

Several phone apps in production promise the ability to diagnose clinical depression with exceptional accuracy. They’ll even be able to detect days when you’re just feeling moody or extra anxious but aren’t necessarily depressed. Indeed, your phone will be able to turn into a virtual personal therapist and tell you when you seem gloomy by measuring vocal tone, tracking how much you text, and using facial recognition technology to gauge levels of stress. It will even be able to give you a pep talk to calm your anxiety. Other apps in production promise to detect signs of other mental disorders including forms of dementia. This is also possible by evaluating speech patterns and spotting vocal differences between adults aging normally and those with signs of degenerative brain conditions.