My shopping cart registered an infection four days before my sensors did. The population evidence now says the same thing at scale: the target was never structure. It was trajectory.
Potential for new readouts for chronic inflammation, emerging as a 'master marker' for long-term disease. Could be very interesting as part of your correlation analysis.
Also note the reference therein to Eric Topol's book 'Super Agers', looks to be in your wheelhouse.
Steve, this was a well-written and thought-provoking article. It will take me some time to fully digest. How does self-selection bias affect this? Some people (my wife) are devotees of wearables and data tracking; others (me) have no wearables and could care less. If this results in systematic issue at a population level, it seems like it could be an issue. A similar phenomenon may hold across other parameters, like purchasing, depending on people's behavior (takes me a long time to admit that I need therapy for a cold). At a minimum, seems like it would increase the noise, and possibly lead to false correlates. Anyway, thanks - write that book!
Great comment! I see it as signal, not just bias. People make very different choices about what they measure, when they seek care, and how they respond to illness, even within the same household which often has orthogonal behavior. This seems common when there are strong divisions of labor within a household. Those behavioral differences are themselves informative. The goal is to let those black boxes disentangle biology and behavior at the macroscopic level. It’s fractal - the microscopic often mimics the macroscopic - and those neural nets seem to figure out why. We can just leverage accordingly.
P.S. These mini-articles are, in many ways, me writing a chapter at a time for a future book or two. Hopefully enough people will be reading books by the time I get there…
I was thinking of your article as I read this in the NYT
https://www.nytimes.com/interactive/2026/07/29/magazine/inflammation-chronic-immune-system-health.html?campaign_id=190&emc=edit_ufn_20260801&instance_id=179685&nl=from-the-times®i_id=76257506&segment_id=224041&user_id=0a49ffa1798023044714530743be5e92
Potential for new readouts for chronic inflammation, emerging as a 'master marker' for long-term disease. Could be very interesting as part of your correlation analysis.
Also note the reference therein to Eric Topol's book 'Super Agers', looks to be in your wheelhouse.
Steve, this was a well-written and thought-provoking article. It will take me some time to fully digest. How does self-selection bias affect this? Some people (my wife) are devotees of wearables and data tracking; others (me) have no wearables and could care less. If this results in systematic issue at a population level, it seems like it could be an issue. A similar phenomenon may hold across other parameters, like purchasing, depending on people's behavior (takes me a long time to admit that I need therapy for a cold). At a minimum, seems like it would increase the noise, and possibly lead to false correlates. Anyway, thanks - write that book!
Great comment! I see it as signal, not just bias. People make very different choices about what they measure, when they seek care, and how they respond to illness, even within the same household which often has orthogonal behavior. This seems common when there are strong divisions of labor within a household. Those behavioral differences are themselves informative. The goal is to let those black boxes disentangle biology and behavior at the macroscopic level. It’s fractal - the microscopic often mimics the macroscopic - and those neural nets seem to figure out why. We can just leverage accordingly.
P.S. These mini-articles are, in many ways, me writing a chapter at a time for a future book or two. Hopefully enough people will be reading books by the time I get there…