If you have diabetes, you know the drill. Every two weeks, you’re shelling out $75 for a sensor that may or may not fall off before it expires. You’ve pricked your finger thousands of times. You’ve been promised a ‘painless’ revolution for a decade. It’s not coming.
I’ve spent the last year talking to the physicists, engineers, and clinicians who actually build these devices. And what I’ve learned is this: The problem isn’t engineering. It’s physics. And physics doesn’t care about your Kickstarter campaign.
Non-invasive glucose monitoring—using a watch or a patch to read your blood sugar through the skin—sounds like magic. But magic is just a word we use when we don’t understand the limits of reality. Here’s the reality: glucose is a small molecule that lives in your blood and interstitial fluid. To measure it accurately, you need direct access to that fluid. Light can’t do it reliably. Sound can’t do it. Sweat won’t work because it’s not the same as blood.
Every major tech company has tried. Apple has a secret team. Google tried. Samsung tried. They all hit the same wall: light scattering in the skin creates noise that drowns out the glucose signal. You can’t filter it out. You can’t algorithm it away. You can’t ‘just use AI’ to fix it. It’s like trying to hear a whisper in a hurricane.
But here’s the part that makes me angry: we keep pouring billions into this dead end. The entire non-invasive industry has raised over $3 billion in the last decade. That money could have made existing continuous glucose monitors (CGMs) radically cheaper, smaller, and more reliable. Instead, diabetics are still paying out of pocket for sensors that break when you sneeze.
I’ve seen this firsthand. A friend with Type 1 diabetes spent $600 last month on sensors. One fell off in the shower. Another gave false lows all night. She still pricks her finger eight times a day. The promise of non-invasive is a lie that keeps us from demanding better from the technology we already have.
Let me be clear: I’m not saying we should stop innovating. I’m saying we should stop pretending a watch can replace a needle. The physics is clear. The money is wasted. The real win is a $5 CGM that lasts a month, sticks to your skin, and doesn’t need a prescription. That’s possible. That’s worth fighting for.
So next time someone tells you ‘the next Apple Watch will track your blood sugar,’ ask them one question: ‘How?’ If they can’t answer, they’re selling you a dream. And you’ve been buying dreams long enough.
FAQ
Q: But what about the Apple Watch rumors? They have patents and a secret team.
A: Patents are not products. Apple has been working on non-invasive glucose for over a decade, and every credible leak suggests they're still far from a reliable, medical-grade reading. The physics of light scattering in the skin is a fundamental barrier that no amount of engineering has yet overcome. Even if Apple releases something, it will likely be a wellness indicator, not a diagnostic tool.
Q: So what should a diabetic do right now instead of waiting for a non-invasive device?
A: Invest in the best existing CGMs like Dexcom G7 or Freestyle Libre 3. Push for insurance coverage that makes them affordable. Advocate for research into cheaper, longer-lasting disposable sensors rather than non-invasive tech. The incremental improvements in current CGMs (smaller, more accurate, longer wear time) are far more likely to improve your life than a moonshot that may never work.
Q: Isn't it possible that a breakthrough in materials science or AI could solve the problem?
A: It's possible, but unlikely within the next decade. The fundamental issue is that glucose concentration in interstitial fluid lags behind blood glucose by 5-15 minutes, and the signal-to-noise ratio through skin is terrible. Even with perfect AI, you can't extract a signal that isn't there. The most promising research is into microneedle patches or fluid extraction, which are still minimally invasive. 'Non-invasive' as a pure optical or transdermal measurement is probably a dead end.