With the help of Arthur Conan Doyle’s famous detective and the board game Clue, Dr. Robert Wright explains the differences between exposomics and traditional environmental health research – and the importance of integrating them
Exposome Perspectives Blog by Robert O. Wright, MD, MPH
I have been asked the question how exposomics, which is the study of the totality of environmental exposures across the lifespan, differs from traditional environmental health research. After all, haven’t we always been interested in all parts of our environment? Well, yes, but the exposome is not just adding a new name to the same ideas, rather it is philosophically different.
Let’s start with linguistics; because we use the suffix “ome” we conjure the notion that the difference is size. According to the dictionary – “ome” in biology refers to all constituents considered collectively. The exposome is the totality of all constituents of the environment.
The problem, however, with our focus on size is that size is not actually the most important difference between exposomics and environmental health research. It is the process that matters. Exposomics vs environmental health is about a difference in scientific approach and rapidity of discovering new information. Perhaps the most surprising thing is that exposomics needs environmental health approaches.
Discovery vs. Hypothesis
Another way that exposomics is different is because it is about “discovery” and not about testing hypotheses. Testing hypotheses means starting with a theory then gathering data to prove whether your theory was correct. Discovery means gathering data first and then developing a theory to form a hypothesis. This may seem logical, but until very recently, environmental scientists were trained only to test hypotheses. By starting with a hypothesis, you avoid the temptation of trying to make the data fit (i.e. post hoc modifying your hypothesis to be consistent with your findings).
Exposomics is a logical, rigorous scientific process of discovery and validation of data and theories and ultimately, hypothesis testing. Discovery research and hypothesis testing research should be integrated, and therefore exposomics and traditional environmental health research should be integrated.
Perhaps an analogy will help. Good scientific questions are like a good crime novel. In both cases, there are mysteries that we are compelled to solve. In a mystery novel, the mystery is often a crime. We search for who committed the crime by discovering how and why they did it. Like scientists, detectives use evidence to find and verify the truth. So to solve our Exposomic true-crime analogy, and explain its novelty, let’s discuss a classic book character and a famous board game.
Playing Clue with Sherlock Holmes
Let’s pretend we are playing the popular board game Clue with Sir Arthur Conan Doyle’s detective Sherlock Holmes. There are 6 characters in the game and a victim named Mr. Boddy. We know in advance that one of the 6 is the killer.
Imagine Sherlock Holmes weaving his way through Mr. Boddy’s mansion and being shown the Clue cards held by different players as he moves through rooms. After a few turns, Sherlock pieces together bits of information and begins to believe Colonel Mustard committed the murder with the pistol in the dining room. He came to this conclusion because he’d previously been shown cards that Miss Scarlett and Mrs. White were innocent, and he noted that no one could produce the pistol card. He himself holds the cards for the candlestick, Professor Plum, and lead pipe and he surmises that you, the player to his left, hold the card for Mr. Green and the knife based on what you have shown others who made guesses about them. Based on the data he collected, his hypothesis is that Colonel Mustard is the killer. This is the equivalent of discovery research.
If this were a real murder mystery and not a board game, he might base his theory on data collected from interviews. Perhaps Colonel Mustard owed large sums of money to Mr. Boddy, the victim, and this desperation led to murder. In this case, Sherlock would use hypothesis testing to try to solve the mystery. He might look into Colonel Mustard’s finances or check his bank statements. He might search the victim’s desk for any evidence of conflict with Colonel Mustard over money. He might ask the Colonel’s daughter where he was the night of the murder or go to the local gun store to see if the Colonel has made a recent purchase. While completely valid, a weakness of this approach is that Colonel Mustard may not be the killer. The approach may clear the good Colonel, and no one will be falsely accused (a “false positive” as we say in research and medicine”), but in the process, time has been wasted. There are 5 other potential killers—Mrs. Peacock, Mr. Green, Professor Plum, etc. While Sherlock is gathering evidence for the Colonel, he is ignoring their possible roles in the murder. If the killer is Mr. Green, then there may even be an awkward moment when Sherlock looks through the cards in the answer envelope to learn he is wrong. But this is what hypothesis testing is all about, and sometimes scientists are wrong. If Sherlock has the correct hypothesis, he will find the killer, but if he is wrong, the true killer may get away.
Prior to the advent of discovery science, there was a tendency to be singularly deductive in science and focus on hypothesis testing for a very small number of risk factors. Instead of searching for new clues, we simply took what we already knew and reapplied it in a slightly different fashion. This is why there are tens of thousands of studies that focus on a handful of environmental factors, such as smoking, lead poisoning, phthalates, and air pollution, even though we know there are millions of important environmental factors in our lives. This ad hoc approach of sticking to what we already understand slows down scientific progress and limits our understanding of how environment impacts health.
Compared to exposomics, traditional environmental health by itself is incremental and slow. While a bit simplified, the game of Clue intuitively shows that data collection and inductive reasoning leads to deductions that in turn cause us to test hypotheses. This is how exposomics and environmental health can work together. We can be inductive by collecting evidence without a hypothesis (exposomics), analyze that evidence to try to piece together what is happening, weed out extraneous information, and deduce a hypothesis that we can rigorously test (environmental health).
The mystery continues – read part II