What if we studied a population more motivated to change their environment than the young and healthy? If we want exposomic data to be actionable, we should ask research questions that impact patients. Even a subtle difference in how we frame a research question can change how the results can be clinically translated. To do that, we must address what the environment does to an existing disease.
Exposome Perspectives Blog by Robert O. Wright, MD, MPH
“They teach you there’s a boundary line to music. But, man, there’s no boundary line to art.” Charlie Parker
Why This Isn’t Just Another ‘I’m an Expert, Trust Me’ Essay
If you ever want to make something cool seem dull, just define and explain it—ask any musician. I googled the definition of the blues and got this: “Blues is distinguished by the use of the flattened third, fifth, and seventh of the associated major scale. Blues shuffles or walking bass reinforce the trance-like rhythm and call-and-response, and they form a repetitive effect called a groove.”
Forgive them Muddy Waters, they know not what they do. To provide context to this post about clinical translational research, I had coronary artery bypass surgery three months ago. I thought I’d get that bit of information out of the way. It is also a flagrant attempt at garnering empathy—I’m telling you that I have been a patient—and now I will strategically add that I am also a physician and public health practitioner who has experienced the biomedical enterprise at multiple levels. It’s akin to that news feed article that starts out: “I’m a nutritionist, this is what you should eat” or “I’m an economist, this is how we fix the federal deficit.”
But I’m not the usual expert with decades of experience in a single field. I’m a public health scientist, a physician and a patient – my opinion doesn’t come from one field of study and my life actually has been impacted by the problem of integrating public health with medicine. When I say “this is how we integrate exposomics into the clinical research enterprise,” it comes from looking at the barriers to doing this from different angles. On the surface, this may seem ho-hum, but it is actually a call for a radical shift in perspectives. Because of my cross-training and life experiences, I’ve looked at environment from many different sides: public health researcher, clinical practitioner, and patient. I’ve seen where the lines blur and where they demarcate. There are a lot of misconceptions about the differences and overlaps among clinical trials, precision medicine, research translation, and public health. Most scientists even assume they are intertwined, and it took me about 30 years (and major surgery) to get enough expertise and experience in these areas to take a stab at describing their boundaries. Despite it all, I still get things wrong, and sometimes it’s just the illusions I recall.
What Medical School Didn’t Teach Me About Public Health and Vice Verse
I remember, as a young medical student, having lunch in the cafeteria of the public health school, which had cheaper food than the medical school and was only two blocks away. I asked my schoolmates what the difference was between public health and medicine. None of us three hungry medical students knew. So I just shrugged my shoulders and bit into a hamburger. Graduation, pediatric residency, and two clinical fellowships followed, and if I were asked, I still couldn’t define public health, something I find is still common among young physicians. Truth be told, people seldom ask that question, and at age 35, I entered public health school and started to learn how they differ.
Now, speaking as a public health scientist and physician, here’s my succinct description of the difference: Public health focuses on preventing diseases before they occur, while medicine is about treating individuals after diseases occur—whether through cures or by preventing further progression.
These differences are far greater than they sound.
For the last 25 years, I told myself I lived in the public health and clinical medicine worlds simultaneously, but that is incorrect. I spent time in each world sequentially in short bursts. In the morning, I’d come to my office in the Public Health School, then I would leave my research team to work a shift in the Pediatric Emergency Room. The physical proximity of the school to the ER was the only actual connection, as these worlds rarely overlapped. It was like stepping through the closet of a C.S. Lewis novel into Narnia. I sutured lacerations, but I didn’t design playgrounds to be safer. I treated asthma attacks with nebulizers and steroids, but I didn’t ask my patients about the neighborhood where they lived or check its air quality. When my physician colleagues asked about my research, my descriptions of studying toxic environmental chemicals on brain development mostly drew polite nods and rarely a follow-up question. The next day public health colleagues might tell me about their research on preventing lead poisoning, sometimes telling me that I should use it clinically. Then I would nod politely and almost never ask a follow-up question.
Public Health Prevention Won’t Fix a Blocked Artery
Probably the biggest surprise I can offer is that public health research can seem “irrelevant” to clinical practice. That sentence will likely get a very negative reaction from some, but if you indulge me, I will explain. The flow and rhythm of clinical care is hard to describe. Research papers don’t capture it. Clinical decisions are often highly contextual to the individual, while public health professionals think about populations—context gets “averaged” out. Often, they get frustrated that clinicians don’t incorporate their research or recommendations into clinical practice.
Since I just had surgery (i.e., feel a little sorry for me), let’s use smoking and coronary artery disease as an example. If I present to the Emergency Room with chest pain as a 61-year-old man, the staff will immediately assume it is a heart attack, test me for it, and treat me based on symptom severity before it is too late. If I had been a smoker, if I were overweight, if I drank alcohol or had high cholesterol—none of those factors would have changed the treatment protocol that day. Nor will they play a role in the subsequent decision to place a stent or refer for coronary artery bypass surgery. All those public health prevention variables were useful as risk factors to prevent heart disease, but once I have a heart attack, they play almost no role in how the medical field will treat me.
This brings me to exposomics—a field that arose from public health. The purely public health perspective of exposomics prevents it from being incorporated into clinical medicine. By that, I mean exposomics is mainly studied as something we can modify to prevent illness. Public health prevention requires understanding what causes a disease since you can’t prevent a disease if you don’t know what causes it. For example, smoking causes lung cancer, alcohol causes liver failure, or, in my work, toxic chemicals impair brain development. The goal is preventing the outcome, not treating it. Your primary care physician may counsel you to stop smoking, but by the time you’ve seen the thoracic surgeon, it’s too late. Prevention is the domain of the healthy, and clinical management is for those with disease. Clinical management requires understanding how to reverse or mitigate the effects of a disease; it’s not about preventing it. Understanding the cause of the disease even seems like it doesn’t help. Let’s take my left anterior descending artery as an example. My coronary artery disease likely arose from a combination of diet in my younger life and my genetics. But when I presented to the thoracic surgeon, no amount of eating vegetables was going to save me from surgery. Likewise, knowing whether pesticides cause Parkinson’s Disease may offer ways to prevent the disease but doesn’t address how to treat it. In Parkinson’s Disease, neurons have died, and there is no bringing them back to life. So if my coronary arteries are beyond “prevention” and exposomics is only about prevention, what is the value of exposomics to me now?
This brings me to today’s word “actionable.” Clinical researchers have created the catchy phrase “bench to the bedside” to describe different layers of the “translational research” process from the research lab to the clinic. The foundation of precision medicine is to refine medical treatment using genomic data to be specific to the individual. First, that approach only works if genes are the major cause of the disease, and they rarely are. Further, even genomic research primarily looks for the cause of the disease, and precision medicine cannot apply that research very easily. Perhaps a shortcoming of both exposomics and genomics is the failure to identify actionable information for clinicians and patients.
For example, this is how clinicians think: if you tell me my patient has a genetic variant that triples his risk of getting Alzheimer’s Disease, even though he doesn’t have it now and may never get it, how will that change my management? I have already told him to eat healthy and exercise. Will it change the likelihood he will practice good health behaviors? What if he asks what that means, and I tell him, instead of 1 in 50, your risk is now 3 in 50? Is that enough incentive to get him to change his behavior? Saying the risk is triple doesn’t mean the same thing for all diseases, either. For rare diseases like many cancers, a tripling of lifetime risk may mean the difference between 1 in 10,000 to 3 in 10,000. Is that actionable to a 25-year-old? We know how much smoking and alcohol impair health. How successful is advice to stop smoking and drinking to get a healthy person to quit? What is enough risk increase in the future to motivate someone to change behavior now when they are healthy?
As the author Ursula K. LeGuinn, once said, “There are no right answers to the wrong questions.”
How Does the Exposome Affect Patients with Disease?
So, what is the right question? What if we studied a population more motivated to change their environment than the young and healthy? If we want exposomic data to be actionable, we should ask research questions that impact patients. Even a subtle difference in how we frame a research question can change how the results can be clinically translated. To do that, we must address what the environment does to an existing disease.
If we expand exposomics to include the impact of environment on disease progression or disease treatment, we may be able to get the attention of patients and physicians. Right now, we are not doing that. Here’s an example: there is no treatment for childhood lead poisoning that reduces its toxic impact on the developing brain—this is why we focus on prevention. But lead exposure doesn’t just happen to healthy children. What if we measured lead exposure in children with a neurologic disease, such as autism? Children with autism have high rates of pica, which refers to eating nonfood substances. This places them at higher risk of lead exposure. If lead is neurotoxic, it is likely to have negative effects on a child with autism, whose brain is likely more vulnerable to lead. This is different than proposing that autism is caused by lead poisoning. That would be a public health prevention study and would require measuring lead exposure before the diagnosis of autism to establish cause and effect. The study I am proposing is about assessing lead poisoning after the diagnosis to see if it makes autism symptoms worse, and by extension, whether reducing lead exposure improves those symptoms. Disease causation does not matter—only the impact on the patient matters. If lead is present in the blood, the research is actionable.
Here’s another example: there is a body of literature that shows that obesogenic chemicals (i.e. microplastics or perfluoryl alkyls chemicals) cause type 2 diabetes. But what is the effect of these chemicals on someone who already has type 2 diabetes? If someone already has diabetes and is being treated with insulin, what happens if they get exposed to chemical obesogens? Is it possible that exposure to obesogenic chemicals underlies the recent change to a higher insulin dose in a diabetic patient? People with diabetes live in the same environments as the rest of us, and they are being exposed to chemicals in ultraprocessed food. Nonetheless, we don’t teach physicians to think about how these chemicals may impact clinical care of diabetic patients.
There are even genetic examples. Patients with Cystic Fibrosis (CF) develop chronic lung disease. Therefore, perhaps air pollution exacerbates Cystic Fibrosis? Would HEPA filters to clean their indoor air help? CF Patients also have high rates of diabetes—might endocrine disrupting chemicals impact their diabetes? All of these examples seem logical and are actionable. All can lead to environmental interventions that may improve patient outcomes, but they are rarely considered in the clinical setting or by clinical research. We need to get studies into the medical literature and plausible interventions so that testing can become actionable.
The Art of Medicine, The Science of the Environment
When I talk about this issue, I sometimes get pushback that I am advocating for taking research funding away from public health prevention. I don’t think clinical and public health research are mutually exclusive. We can still search for the causes of diseases and work to prevent them, but at the same time, let’s not forget about people who have chronic diseases today. Prevention is irrelevant to them, they are thinking about their situation now. They live in the world as the healthy and the same environmental exposures impact them as the rest of us. They live in the same world, breathe the same air, use the same products, and there are things we can do now to help them cope with the effects of the exposome on their diseases. We don’t have to be restricted to studying the exposome’s impact on disease prevention alone. As clinicians, we are taught that medicine combines science and art. Each patient is contextual. Few adhere to the textbook descriptions of signs and symptoms and only a subset respond to the treatments we prescribe. Perhaps if we look at exposomics from both sides—clinical and public health—we can start to truly understand its role on human health.
As Joni Mitchell once noted about clouds.
I’ve looked at clouds from both sides now
From up and down and still somehow
It’s cloud illusions I recall
I really don’t know clouds at all
Joni Mitchell