Exposome Perspectives Blog by Robert O. Wright, MD, MPH
“Regular maps have few surprises: their contour lines reveal where the Andes are and are reasonably clear. More precious, though, are the unpublished maps we make ourselves, of our city, our place, our daily world, our life; those maps of our private world we use every day.” – Alexander McCall Smith, Love over Scotland (abridged)
“If we are something, we are our past aren’t we?” – Jorge Luis Borges
Linking Geography and Biology
Most journeys go smoother when you use a map. It is more than a cliché to say life is a journey, and metaphors about life, journeys, and maps abound. We all have our own personal map of our life’s journey—which once drawn—reflects our personal exposome. Perhaps most importantly, our personal exposome map links our past to our present, as our health is more a reflection of our past exposome than our exposome today. The real challenge of exposomics isn’t measuring “everything”, that’s the easier part. The most difficult part is measuring the past—those personal maps of our private selves going back to childhood, which explain how and why we evolved to be who we are.
Geography and biology have been intertwined for centuries. We evolved as a species in response to Earth’s environment. In fact, all of biology is about responding to the environment, whether it is synaptic pruning in a toddler, digesting food, attracting a mate, planning your workday, or estimating your needs in retirement; you are using your memory of the past to plan your future actions within the context of your present environment. This is how “nature and nurture” interact—in a series of exposure-response loops influenced by the memory of how we reacted to our past exposures. The exposome does not begin inside our cells, but instead lives outside us and sends signals from the outside in. To estimate your exposome, simply look around you to see where you are and try to remember where you’ve been. The exposome is our past and present food, water, and air, the buildings in which we live and work, and the social connections of our families, friends, and coworkers. If there is one commonality to all the many components of the exposome, it is planet earth—which conceptually is a giant map.
Mapping Middle Earth
The value of maps in telling life stories has a long history. Readers of The Lord of the Rings or The Hobbit no doubt combed over the books’ detailed maps of the terrain and waters of Middle Earth. There is even an interactive website that lets you plot the journeys of different characters, such as Frodo and Sam in the main story or Bilbo in The Hobbit. Did you know that nearly 70 years separate the events of The Hobbit—Bilbo Baggins’ sojourn to the Lonely Mountain and back—and the tale of Frodo’s quest to destroy the Ring of Power? Yet, one map can recreate both these journeys. From that map, we can tell that Bilbo likely experienced wildfire smoke exposure when Smaug the Dragon set fire to Lake Town. Likewise, Frodo and Sam inhaled ambient metals in Mount Doom. Even if the smelting had stopped, there had to be leftover dust that was stirred up when they fought Gollum. We can even attach dates to these events, including the amount of green space the Hobbits experienced while living in the Shire.
Given all this, would it surprise you if I told you that geography predicts your health better than your DNA? Google it and you may be surprised.
You may even be surprised to learn that Dr. Francis Collins, the geneticist, once said this.
“Your likelihood of having a certain life span depends heavily on the ZIP code where you were born, and that is a reflection of all of the…environmental exposures, socioeconomics, social determinants of health, et cetera.”
– Francis Collins, NPR Interview, 2021
The Proxy is not a Problem: ZIP Codes and Biomarkers
A common caveat when discussing ZIP code and health is that ZIP code isn’t the cause, but is a proxy for the social determinants of health—factors such as socioeconomic status, and access to resources, among others. I guess your reaction to hearing that depends on how you define “cause” and whether you believe that correlation by itself is meaningless unless the correlation is also the direct cause. Given that ZIP code predicts health, why would thinking this is just correlation prevent us from using ZIP code clinically? It’s the prediction part that matters. If the prediction is reproducible, then the information is valuable. Netflix doesn’t care why you pick a certain show to watch, but it does try very hard to predict what show you will watch. Causation to Netflix is less important than prediction. So don’t let anyone talk you out of the importance of your address in predicting your health. With the right tools, we can use that information and figure out the causes separately, which likely include culture, diet, socioeconomic status, access to resources, and levels of stress, pollution, and good nutrition.
In other words, the reason ZIP code predicts health is that it is a proxy for differences in our exposome. A biologist may argue that this means ZIP code is less objective than a blood-based biomarker—i.e., assays that indicate a biological state of health or disease. That’s not a unique issue. Many biomarkers are also proxies for something else that is the cause. For example, a biomarker measured in the blood is just a proxy of the disease if that disease (e.g., dementia) is found in a different tissue, like the brain. Biomarkers may or may not correlate with what is happening in the actual tissue of interest, or with what happened yesterday, or with what will happen tomorrow. We may think that molecular biomarkers are more objective than questionnaires or maps, but they have the same caveats. Molecular biomarkers also reflect only a moment in time in the blood and often tell us little about our past. Much medical research is dedicated to molecular biomarker discovery, but health information is not limited to assays. Maps can give us information similar to biomarkers, they also can give information unattainable from a biomarker—like when something happened and the context in which it happened.
So if it is widely agreed that place/geography predicts health, then why don’t we treat maps like biomarkers? We already know that our childhood environment predicts our health in adult life. There is a whole field dedicated to this concept called DOHaD, or Developmental Origins of Health and Disease. So why don’t we use that information? Has your physician ever taken an address history? What if he/she did? Then your past exposome could be used to predict your present health. Few biomarkers can capture our past, much less multiple aspects of our past. Yet it is our past exposome that is most predictive of our health today. If we draw blood to measure our current exposome with an assay, we only know what happened today. We won’t know anything about what happened to us years ago, especially when we were children. Maps can fill in that gap.
From Okinawa to Detroit to New York
What would my personal exposome map say? I was born in Okinawa in 1963 and lived across the street from the entrance to the U.S. Kadena Air Force Base for the first 3 years of my life. This means that as an infant/toddler, I may have been exposed to ambient jet fuel and maybe even Agent Orange. I also lived with my maternal grandmother, aunts, uncles, and cousins. My grandmother took me everywhere with her and my mother. I was surrounded by love. I moved to Downriver Detroit in 1966. I also lived in Chicago in my late twenties. Then, suburban Massachusetts for 19 years, and the last 13 years in Westchester, NY. My exposome is radically different today than it was when I was a child.
Here is a cool website that can tell you your life expectancy based on your ZIP code. But here is its shortcoming. If I put in my present ZIP code (10538), the life expectancy is 83 years. The ZIP code where I spent the majority of my childhood (48101) has a life expectancy of 74 years. That’s a big gap, but it may accurately reflect the neighborhoods in which I have lived. My childhood home was burglarized repeatedly. When I was a teenager, there was a fatal police shootout in my next door neighbor’s backyard. Where I live now, the biggest crime is turning right on a red light. There was a lot of pollution in the 1960s and 1970s; it was an era when whole rivers caught fire. My neighborhood was not spared, and I remember falling into the Ecorse River with a bowling ball at age 12, trying to cross the ice as a shortcut home, and a pungent smell of sewage, metals, and solvents that I couldn’t shake as I walked home with icicle pants. I assume my bowling ball is still down there somewhere, if it didn’t dissolve. The metals and solvents were from the nearby automobile factories and refineries that employed most of the neighbors where I lived. I am also not without health problems—even though I’ve exercised regularly for the last 15 years. I’ve had bypass surgery and am pre-diabetic. If my adult and childhood exposomes are competing, my childhood exposome may be winning. In fact, my past exposome map explains some of my present health biomarkers better than my present exposome, which is more the response to my health problems, not the cause. If enough people link health data to life span maps going back to childhood, we will be able to use maps as biomarkers more effectively, because we will better understand the relationships between place and health.
What If Every Health Record Could Link With an Exposome Map?
An exposome map can tell unique stories for each person because the probability of exposure is a lot higher if you once lived near the Ecorse River than if you didn’t. Imagine if we had a record of our country’s or, better yet, Earth’s past and present exposome —to link to our addresses, travels, work, and school. Imagine if it could be linked to our electronic health record (EHR), and then to current exposome measured in an assay, as well as our genome and epigenome. How much better could we predict our future health or our present health if we used these resources in unison and not in isolation? If we had such an exposome map, all my physician would have to do is ask about where I used to live and the years in which I lived there. Then he or she could plop me into the map for all my life’s calendar years and see what else was there with me and add up all the exposures. He/She could even weigh the early life exposures more, as they likely have more impact, which is the concept of DOHaD. Even when I am too old to remember, if we made this part of the clinical visit, there would be a record of where I have lived. I am not suggesting that we replace genomics or exposome assays with geographical maps. I am suggesting that we can combine them in ways that would make genomics and exposomic predictions of health far superior. In fact, I would bet that my genome-exposome map combination would do a better job than my genome or blood exposome alone.
“The future is already here, it’s just not very evenly distributed.” William Gibson
We’ve never tried to map the exposome, and until recently, there was little interest in exposomics in the greater scientific community, but the exposome map’s time may finally have arrived. What’s particularly exciting about maps is that much of the information needed to make our personal exposome maps already exists. We just haven’t had the will or interest to create them, but it may be easier than we think. EPA makes maps of where toxic waste sites are located, as well as air pollution. USGS maps our forests and water, CDC maps our health, and NASA has satellites that map our air, wildfire exposure, and light at night. If we know ZIP code histories or even better—home addresses—we can map health data, air and water quality, exposure biomarkers like blood lead, liver function, or renal function tests to see spatial and temporal patterns and trends. It won’t be straightforward, but it is definitely feasible. All this is attainable if we work to bring all these puzzle pieces together in a relatively short amount of time, if we just develop the will to do it. Everything we need already exists- we just have to bring it all together.
As REM once sang
“Maybe he’s caught in the legend
Maybe he’s caught in the mood
Maybe these maps and legends
Have been misunderstood”