Against Utopia Weekly #1 – Eugenics Underpins Digital Health Technology

** Issue 1
Eugenics: The Hidden Driver of Healthcare and Digital Health Technology

Welcome to the first edition of the Against Utopia weekly newsletter, where I’ll expound on incomplete thoughts I’ve had, play with them a bit, and hopefully engage you with some new ideas and perceptions on our current situation.

Before we get into it, spread my propaganda to your friends and tell them to sign up for the newsletter here. (

My long form essay series can be found here. (

If you really like me and want to help me spread more propaganda faster, support me on Patreon. (

My first hot take weekly newsletter topic is going to be that digital health, and modern healthcare in general, operates on a principle tantamount to eugenics. Let’s get into it.

So what is digital health? Digital health is the use of information and communication technologies to digitize the delivery of healthcare services, and to provide efficiencies in care, risk, and condition management. In terms of investments, the sector has grown 10x since 2010, reaching $11.7 billion in venture capital funding in 2017. (

For a little over ten years, the digital health sector has been heralded as the next great bringer-of-progress on the frontier of information technology.

Fitbit ( would help us figure out how to move more often, and measure our movement, so we can anticipate and reduce our heart disease risk.

Teladoc ( would help us manage our chronic and acute mental health issues remotely, anywhere, on demand.

iRhythm ( would help us use continuous monitoring technology to predict cardiac health events, and take proactive action to reduce their risk and severity.

Tabula Rasa Healthcare ( would centralize and systematize the mitigation of medication risk, and through this lens, use digital systems to reduce

And so on.

These digital health companies are assuming that the efficiency gains from information technology are there for the taking, and that the underlying medical technology actually works: therefore the problems to be solved are in distribution, form factor, or interoperability, key areas that IT directly improves.

And there’s another pattern here as well – notice how often the word “manage” and “risk” appears on the webpage explainers of these technologies.

If we have the diseases cracked, particularly in the chronic disease realm (heart disease, cancer, diabetes, Alzheimer’s, etc), then it’s merely a matter of scaling the reach of the managerial healthcare state. As soon as we do that, we should watch the risk mitigation sharply reduce mortality and deliver us all the digital health revolution we’ve been waiting for.

However, the digital health revolution will most likely never arrive – almost entirely because it is built on the shady, rickety foundation known as eugenics. In order for us to understand that, we’ll need to go back in time a bit.

In my long form essays, I covered the historical development of depression as a treatable illness, and the many factors that manifested in the medicalization of grieving. One of the key developments in the history of medicine and depression was the methodology of observation, identification, and language standardization that enabled doctors to communicate in objective ways about the mental suffering of their patients.

This professionalization process began in earnest around the turn of the 19th century with the work of a German psychiatrist, Dr. Emil Kraepelin, the founder of scientific psychiatry. Kraepelin was renowned for his scientific approach to managing patient care, his intricate data and reporting methodologies, and his development of objective pattern-seeking methods for comparative psychology. He was also a eugenicist.

I don’t necessarily mean to use eugenicist in this context as a slur, nor do I mean to excuse it. It’s just a simple fact that is crucial to understanding Kraepelin and the way that early 20th century doctors thought about the plight of their patients and the possibilities for their coping and recovery.

Doctors glommed on to Kraepelin’s framing of the problem of mental illness, and this frame was one that spoke deeply to genetic origins. He believed that the amount of mental illness in the population is always fixed, and that this mental illness was of a genetic origin, therefore it is folly for doctors to spend their time trying to figure out how to “cure” mental illness – better they spend time on rigorous scientific methods for identifying those at risk of mental illness in the population, so that their illness can be managed. “Managed” typically meant by a doctor in an asylum, sequestered from the public, and not much more, at least in Kraepelin’s time.

Fast forwarding a bit, it’s not an understatement to say that in 2019, our current medical culture has maintained much of this approach as well. Look no further than genomics and chronic care management. There are scores and scores of genome-wide association studies, where the human (or animal) genome’s variation is broken down into its constituent parts, called single-nucleotide polymorphisms (SNPs), and associations are sought between embodied disease states and SNPs. The important consideration in questioning this work is not to ask whether it should be done at all, but rather to ask:

Even if we did find out as a result of this work, that the majority of chronic illnesses like Alzheimer’s, heart disease, and diabetes have a strong genomic association, what would we actually be able to do if we had that information?

When looking at the medical culture, the answer seems to be that we’d find drugs that address the symptoms of the chronic illnesses, drugs that reduce the risk of carrying these genes in your genome, try to return the values that they are meant to influence to the “normal” range, but never bother to understand the baseline physiological causes or etiology.

We’d find ways to manage the conditions within their reference ranges (e.g. keep blood sugar below x target, keep blood pressure within x/y targets), and we’d assume that if you happen to fall outside of the range, you’re just unlucky, which is another way of saying you have bad genetics. It doesn’t seem to be, at least so far, that we’d want to find ways to engineer genetic interventions for the perceived genetic root causes.

It might seem like I’m putting words in physicians’ mouths, but this is basically what they say by their own admission. Much of the 80s and 90s were spent telling people with clinical depression as measured on the Hamilton depression scale that their experience e.g. is caused by a “chemical imbalance” that is genetic in nature. This language even in the 90s remained unsubstantiated – to this day it is a doubly dubious claim, both from the perspective of a measured chemical imbalance (no one has measured serotonin in the brain, it’s too risky), and from the perspective of genetics (there are extremely weak correlations between some SNPs for serotonin transporters and depression risk).

I’d say it plainly as, the second that a eugenic epistemology takes hold, we stop viewing diseases as things that have root causes and can be cured, and start to view them as things to be managed, with genetic root causes. And since we all have genes, we all have some genetic risk, no matter how little it is.

So it should be no surprise then, that the medical culture has stopped searching for cures for illnesses, and bought wholesale into an ideology of eugenics for disease origination and management.

It may not appear so.

You might say we’re still curing stuff like acute hepatitis. Gilead Sciences released a cure for hepatitis type C a few years ago, and achieved cure rates of ~90% ( . The problem is, cures exhaust the pool of treatable patients, and they end up being bad for business. This might seem like a common trope your reactionary uncle who never believes experts told you at Thanksgiving, and it seemed nutty coming from him.

Allow me to quote Goldman Sachs at length, validating your uncle:

“GILD [Gilead Sciences] is a case in point, where the success of its hepatitis C franchise has gradually exhausted the available pool of treatable patients,” the analyst wrote. “In the case of infectious diseases such as hepatitis C, curing existing patients also decreases the number of carriers able to transmit the virus to new patients, thus the incident pool also declines … Where an incident pool remains stable (eg, in cancer) the potential for a cure poses less risk to the sustainability of a franchise.”

Acute conditions that cause harm or death such as hepatitis C might not have a genetic risk component. And as this analyst at Goldman above says, it’s doubly worse if they’re not carrying the disease, as it can’t be spread, creating new customers for the cure.

Chronic conditions by comparison present endlessly growing, inexhaustible pools of patients limited only by medical perception, language, and standard of care and as a result, there exists a strong incentive to find an ultimate genetic cause for them. Should one be found, or justified into existence, then it can be managed, and suddenly everyone with that SNP is a customer.

So in summary, two key ideas driving eugenics in healthcare technology:
1. the idea that we already have the solutions, they just need to be scaled to manage the health of the population
2. There are no cures anymore, all disease has a genetic profile, and since everyone has genes, everyone is at risk

Let’s take a closer look at healthcare and what we’re actually trying to scale with digital health technology.

Take continuous blood pressure monitoring technologies. Blood pressure (BP) is one of the primary independent risk factors of cardiac illnesses of all kinds, and interventions targeting it are the first line therapies available to reduce overall cardiac illness mortality. It is believed that by intervening in hypertension for patients, we will prevent future heart attacks, and deaths that result from them.

The first line therapies included thiazide-diuretics, calcium channel blockers, ACE inhibitors, and angiotensin receptor blockers. Knowing what they are and how they work is not material to our investigation here – the most important thing to know about them is that, despite being the first line therapies, for all but the most sick hypertensives they are actually useless ( .

A Cochrane meta-analysis, one of the largest and most statistically powerful, showed that the standard of care for hypertensive patients, putting them on a first-line therapy for any BP measurement over 140/90, fails to reduce cardiac events, cardiac mortality, and overall cardiac disease risk.

Despite the utter failure of medications to manage blood pressure risk, the search for genomic causes or associations for blood pressure risk continues ( . It hasn’t stopped a veritable sea of medical devices and digital health technologies from receiving funding to scale the measurement of blood pressure, and the management of the associated medications meant to manage the condition itself.

Lastly, let’s take the case of type 2 diabetes. By now it’s basically a full blown epidemic. Approximately 25.8 million Americans suffer from it, and estimates in 2013 had 113 million Chinese suffering from it, which represents about 10% of the Chinese population.

Diabetes has been linked to increased risk of cardiovascular illness, cancer, and Alzheimer’s, among many other maladies. These maladies all told are the major cost contributors as well as mortality drivers for all healthcare systems the world over. Uncovering the major contributors of disease risk for diabetes and lowering them is and continues to be a major goal of all healthcare state funded research, and private research.

Since the discovery of recombinant insulin, it has been one of the primary therapies for managing glycemic load and limiting the systemic damage of diabetes. The other top line therapies also focus on glycemic control but via pharmacological means. The standard of care here has become metformin.

After over a half century of intervention, the results are mostly in – insulin therapy does nothing to reduce cardiovascular mortality from diabetes, and doesn’t reduce cardiac events. I won’t mediate, let’s just quote from the source ( :

“Starting insulin therapy early in the course of chronic treatment of patients with type 2 diabetes would imply that there are unique benefits to insulin treatment. As addressed above, there is little evidence to support such a view. Insulin treatment is neither durable in maintaining glycemic control nor is unique in preserving β-cells. Better clinical outcomes than those that occur with other antihyperglycemic regimens have not been shown. The downside of insulin therapy is the need to increase the dose and the regimen complexity with time, the increase in severe hypoglycemia, and the potential increase in mortality as well as the potential increased risk for specific cancers.”

Ok, so insulin doesn’t really work for reducing negative outcomes, but it’s not the standard of care.

How’s the standard of care doing?

Metformin is a safe, easy to manage drug that sensitizes cells to insulin. Truth be told, nobody actually knows how it works, we just know that it gets into cells and something something sugar seems to get into cells more easily, and hyperglycemia is reduced.


A recent meta-analysis ( shows that while metformin reduces the incidence of hyperglycemia, it does fuck-all for mortality outcomes. This is kind of like saying that if you take metformin, you won’t suffer from the symptoms of your uncontrolled blood sugar, but you’re still going to die from the disease on a different (shorter) timetable than you would if you didn’t have it.

I’m simplifying this for the sake of bombast, sure, but just take me up on going into Google Scholar and hunting these studies and meta-analyses down for yourself. You’ll find the difference is not that much in the other direction for the few cases where metformin actually improves quality of life AND outcomes.

So that’s two major areas, heart disease and diabetes, where the top line scientific management methods turn out to be actually useless in reducing mortality.

And if we turn our gaze back to the digital health and healthcare technology sector, we see a spate of technologies built to scale, digitize, and automate the administration of these disease states.

Glooko ( builds remote monitoring technology for diabetics and their doctors to seamlessly integrate information about their blood sugar from their devices, for expert oversight.

Livongo ( integrates body weight, behavior change coaching, and medication management to purportedly engineer better outcomes for its members.

Sano ( is pioneering continuous glucose monitoring, mining data about the interaction of food, blood sugar, and insulin constantly, in order to better manage outcomes for all people, not just diabetics.

You see the problem here, right?

Insulin and various medications that control hyperglycemia are still the first line therapies in these digitized methods. The digital methods that drive these companies above in particular are no different from the things we’ve already been trying for the last half a century, to the tune of, ya know, hundreds of millions of diabetics and a worldwide epidemic of crisis proportions.

This shit didn’t happen overnight! It’s BEEN not working!

And my bet would be that digital information technologies built on top of this eugenic epistemology that merely structure data to manage health conditions are making a strong statement that patients under their care can never be cured. Their health can only be managed, because all they are doing is delivering interventions that do not work to cure anyone, merely to manage conditions within a range as declared by a medical expert.

The only companies that don’t treat their patients with the eugenic gaze seem to be having success reversing diabetes on the other hand. It appears that if you believe patients can be cured, then you’re free to act in the decision space that lets you unearth and scale that cure.

Companies like Omada Health, who coach prediabetics to lose weight and prevent progression to diabetes, or Virta Health, who use low carbohydrate and ketogenic methods to manage behavior change and reverse diabetes, have been experiencing some success on this front, but I’m not here to talk about solutions (yet), I’m just pointing out problems.

All this is not to say that we’re wasting our time trying to build digital health technologies that might radically lower barriers of access to healthcare, cheapen healthcare delivery, or systematize repeatable, computable things such as chronic medication management that humans can’t do as well as algorithms.

It’s just that, in its current manifestation, digital health is tantamount to a bad algorithm of sorts – it’s feeding garbage in, and spewing garbage out.

If my thought experiment here is directionally correct, that the underlying driver of our social organization of science is eugenics, where genetic “risks” define our propensity for ill health, and these risks neglect embodied physiology, then our systems will only ever create garbage.

Cures, better value, and better outcomes, can only ever be a happy accident when padding massive profit margins with a faulty model is the real goal.

And who’s to say the model is faulty if it makes everyone a target with science-y looking things like genome-wide association studies, and then manages to influence its own targets with medicine that never cures anyone?

With eugenics as the driver of our medical epistemology, we may think we’ve got a simple picture finally figured out, but we’ll never tangle with the messy interactions between genetics, environmental inputs, and embodied physiology, leaving many non-genetic possibilities that are more complex but potentially more fruitful unexplored.

If we assume that everyone will get diseases because everyone has genes, and build digital health technologies accordingly, we’ll fashion beautiful digital interfaces that imprison us in our genetics, and leave entire possibilities for our liberation from disease unexplored.

Like what you’re reading? Support me on Patreon! (