SWITCHED ON
The daily technology series nobody asked for but everyone needed
Measured to Death: Wearables and the Quantified Self
Millions of healthy people are now wearing continuous glucose monitors. The data they generate is real. Whether it is making anyone healthier, or just considerably more anxious, is a separate question entirely.
There is a meaningful difference between information that helps you make a decision and information that simply makes you aware of something you cannot meaningfully act on. A great deal of the wearable technology industry depends on most consumers never quite working out which category their data falls into.
Yesterday's episode on technology and democracy was heavy going — disinformation at scale, deepfakes in elections, the polarisation research, and the structural business model tension that no platform policy has resolved. Today we are deliberately lighter in subject matter and, I promise, no less rigorous in treatment. Wearable technology and the quantified self: the smart rings, continuous glucose monitors, recovery trackers, and biometric dashboards that have moved from elite athlete training rooms into millions of ordinary bathrooms and bedside tables. What this data actually tells you. What it does not. And whether the relationship between measurement and wellbeing is as straightforward as the marketing implies.
01 — From Step Counters to Continuous Everything
The wearable technology category has expanded dramatically from its origins in basic pedometers and the first Fitbits, which tracked steps and not much else. The current generation of consumer wearables — the Oura Ring, the Whoop band, Apple Watch's increasingly sophisticated health sensor suite, Garmin's athletic-focused devices — track heart rate variability, sleep stages, blood oxygen saturation, skin temperature, respiratory rate, and derive composite metrics like "readiness" and "recovery" scores that did not exist as consumer concepts a decade ago.
The most significant recent expansion has been continuous glucose monitoring moving from a medical device for diabetics into a consumer wellness product for the metabolically healthy. Companies including Levels, Lingo, and Signos sell CGM-based programs to non-diabetic consumers, promising insight into how specific foods affect blood sugar and, by extension, energy, weight, and long-term metabolic health. The underlying sensor technology — a small filament inserted under the skin that continuously measures interstitial glucose — is genuinely impressive medical engineering, originally developed and refined for the urgent and well-established purpose of helping diabetics avoid dangerous blood sugar swings.
The sensor is the same. The use case has changed from managing a diagnosed medical condition to generating a real-time anxiety feed about whether your lunch was a personal failure. This is not a technical criticism of the device. It is an observation about what happens when medical-grade monitoring is repurposed for a population that does not have the condition it was designed to manage.
02 — What the Data Actually Means for Healthy People
The scientific question underlying the consumer CGM boom is whether glucose variability in non-diabetic, metabolically healthy individuals is a meaningful signal that should inform dietary choices, or whether it is largely noise that reflects the normal and harmless physiological variation that healthy bodies exhibit. The honest answer, based on the available research, leans considerably more toward the latter than the marketing of consumer CGM companies suggests.
Postprandial glucose spikes — the rise in blood sugar after eating — vary substantially between individuals eating identical meals, and this variation is influenced by factors including sleep quality, stress, exercise timing, and the order in which food is eaten, none of which necessarily indicates underlying metabolic dysfunction. Research on continuous glucose monitoring in non-diabetic populations has found significant glucose variability that does not correlate clearly with metabolic disease markers in people without diabetes risk factors. This does not mean glucose data is meaningless for healthy people — there is legitimate research interest in glycaemic variability as an early marker for prediabetes risk in specific populations — but it means the leap from "your CGM showed a spike after eating white rice" to "you must permanently avoid white rice for metabolic health" is not well-supported by current evidence for someone without diabetes risk factors.
Sleep tracking faces a related and well-documented problem: the devices are reasonably good at detecting whether you are asleep or awake, modestly good at detecting sleep duration, and considerably less reliable at the sleep stage breakdown — REM, deep, light — that generates the most engagement-driving detail in the app interface. Validation studies comparing consumer wearables against polysomnography, the clinical gold standard for sleep measurement, consistently find meaningful discrepancies in stage classification, even as overall sleep-wake detection has improved. The granular sleep score that arrives each morning carries more apparent precision than the underlying measurement actually supports.
03 — Orthosomnia and the Anxiety the Data Creates
Sleep researchers have documented and named a specific phenomenon arising directly from consumer sleep tracking: orthosomnia, a term coined in 2017 to describe a preoccupation with achieving perfect sleep as measured by a tracking device, to the point that the anxiety about sleep quality itself becomes a cause of poor sleep. Patients have presented to sleep clinics distressed not by how they feel, but by what their wearable's dashboard tells them their sleep score was the night before — a measurement they have limited ability to independently verify and an even more limited ability to directly control.
This pattern extends beyond sleep into the broader quantified self movement. Recovery scores, readiness metrics, and HRV trends create daily verdicts on bodily performance that some users experience as motivating and others experience as a persistent low-grade source of stress — checking a score before deciding whether a day "counts" as a good one, modifying behaviour around optimising a metric rather than around how one actually feels, and in some documented cases developing disordered relationships with exercise and eating organised around hitting specific tracked targets rather than genuine physical signals.
A tool that was supposed to help people listen to their bodies has, for a meaningful subset of users, replaced listening to the body with checking the app — outsourcing a form of self-awareness that humans previously developed through direct experience to a number generated by a sensor of uncertain precision.
This is not a universal experience. Many people use wearable data productively — identifying genuine patterns in their own physiology, catching early warning signs of illness, building sustainable exercise habits through objective feedback. The honest framing is that wearable data is a tool whose effect on wellbeing depends heavily on the user's existing relationship with their body, their tendency toward anxiety or perfectionism, and whether they treat the data as one input among many or as an authoritative verdict on their worth as a healthy person for that day.
04 — The Business Model Underneath the Health Halo
Wearable companies are, with few exceptions, venture-backed or publicly traded companies whose growth depends on continuous subscription revenue and expanding feature sets, and this commercial reality shapes product design in ways that are not always aligned with the most accurate or most calming presentation of health data. Whoop's subscription model, in which the hardware is functionally free but the data and insights require an ongoing monthly payment, depends on users finding the daily dashboard sufficiently compelling — sufficiently worth checking — to justify continued payment. A calm, infrequent, "nothing notable today" presentation of health data is commercially worse than a daily score that fluctuates, generates a small dopamine hit or anxiety spike, and creates a reason to open the app again tomorrow.
This is not a conspiracy theory about deliberately manipulative design. It is a structural observation about what kind of product survives in a subscription-driven market, and it parallels exactly the engagement-optimisation dynamic we examined in social media in Season One. A health metric that genuinely reassured users they were fine and required no further attention would be commercially worse than one that generates ongoing engagement, regardless of which one better serves the user's actual wellbeing.
05 — Using the Data Without Being Used By It
None of this is an argument that wearable technology is worthless or that the underlying sensor engineering is not genuinely impressive — it is, in many cases, remarkable miniaturised medical-grade technology made accessible at consumer prices. The argument is for a specific and limited use of the data: as one input among several, interpreted with appropriate scepticism about precision, and explicitly decoupled from daily self-worth.
The most defensible use cases are longitudinal trend identification — noticing that resting heart rate has crept up over several months, which might warrant a conversation with a doctor — rather than daily score optimisation. Using sleep data to identify gross patterns, like consistently going to bed too late, rather than treating nightly sleep stage percentages as precise clinical measurements. Treating any single day's reading as low-confidence noise rather than a verdict, because the measurement uncertainty in consumer-grade sensors is real and the marketing rarely communicates it honestly.
The quantified self movement promised that more data about your own body would produce better decisions and better outcomes. The more accurate finding, several years into mass adoption, is that more data produces better outcomes for people who already have a secure, low-anxiety relationship with their health, and produces worse outcomes for people who do not — turning an existing tendency toward health anxiety into a continuously fed, algorithmically reinforced loop. The technology is neutral. What it does to you depends substantially on what you bring to it.
Tomorrow we are heading into a frontier that combines almost everything this series has touched — synthetic biology and the engineering of life itself beyond CRISPR's gene editing. Designer microbes, programmable cells, and the companies trying to turn biology into a manufacturing platform. See you then.
Switched On is a daily technology series covering the ideas, systems, and arguments shaping the digital world. Opinionated. Witty. Occasionally wrong. Always worth the argument.
