Debugging Your Body: What Longevity Research Taught Me About Systems Thinking

I spent most of my twenties treating sleep as the variable I could compress. Deadline coming? Take it out of sleep. It's the only budget line with no immediate consequences, which is exactly what makes it dangerous.

Two books rearranged that for me: Matthew Walker's Why We Sleep and Peter Attia's Outlive. Not because either told me something I couldn't have guessed, but because both frame the body as a system with maintenance requirements, failure modes, and accumulating debt - and that's a frame I already know how to think in.

A caveat before any of the rest, because it matters. Walker's book has taken substantial criticism since publication for overstating individual findings, and some of the specific numbers in it don't survive a look at the underlying papers. I still think the central argument holds and it genuinely changed my behavior. But I'd read it as a persuasive case for taking sleep seriously rather than as a reference text, and I'm no longer comfortable repeating its more dramatic statistics without checking them.

The maintenance window

Sleep isn't downtime in the sense of nothing happening. It's when the brain clears metabolic waste, consolidates memory, repairs tissue, and rebalances hormones. Skip it and you're running production without garbage collection: no crash, just gradual degradation across cognition, emotional regulation, and immune function.

The finding that actually changed my behavior wasn't about disease risk. It was that the degradation is invisible from the inside. Chronic partial sleep restriction produces measurable performance decline while subjective alertness recovers to something near baseline - so you feel roughly fine while performing meaningfully worse. You are not a reliable narrator of your own sleep debt.

That's a familiar shape. It's every system that reports healthy while the error rate climbs, because the health check tests the wrong thing.

So sleep stopped being a variable. Seven to eight hours, and the only exceptions are actual emergencies rather than the ordinary kind of deadline I would previously have described as one.

Medicine 2.0 vs Medicine 3.0

Attia's argument is about optimizing for the wrong metric.

Conventional medicine - his "Medicine 2.0" - treats disease once it presents. Symptoms, diagnosis, treatment. The problem is that the diseases that kill most people in wealthy countries take decades to develop, so by the time they're symptomatic you've spent most of the runway during which intervention was cheap.

Medicine 3.0 is prevention-first and thinks in decades rather than appointments. Watch the metrics, catch the drift early, intervene before anything breaks.

The obvious criticism, which Attia mostly acknowledges: this requires time, money, and access that most people don't have, and a lot of the interventions have less evidence behind them than the confidence with which they're delivered. I find the framework useful without believing every recommendation is equally well supported.

The body as a complex system

The parallel that keeps holding up is cascade behavior. Poor sleep elevates cortisol. Elevated cortisol contributes to insulin resistance. Insulin resistance drives metabolic dysfunction, which drives inflammation, which accelerates more or less everything else. One degraded input propagates through subsystems that look unrelated on the org chart.

The other one is compounding. Thirty minutes of walking a day is nothing on any given day, which is precisely why it's easy to skip and why it matters over a decade. Same reason nobody prioritizes the small performance fix that only pays off across a million requests.

And then the debt metaphor, which I use more than the others and trust slightly less: inflammation, insulin resistance, arterial plaque accumulate quietly and get expensive to service later. It's a good intuition pump. It's also the kind of analogy that can talk you into thinking you understand the biology better than you do.

What I actually track

I use a Whoop for sleep duration, efficiency, and stages, and I aim for seven and a half hours or more. The single biggest improvement wasn't a metric though - it was fixing my sleep and wake times, weekends included. Consistency did more than duration.

The tracker also made two things unavoidable that I'd been ignoring. Alcohol wrecks my REM sleep, including amounts small enough that I'd have told you they were fine. And caffeine in the afternoon measurably cuts my deep sleep. I cut both, and I want to be clear that I only did it because I could see it in data I couldn't argue with. Being told this in a book didn't work. Watching it happen to my own numbers did.

Beyond that: heart rate variability as a rough proxy for recovery, mostly used to decide whether to train hard or take it easy. Resting heart rate, which trends down as fitness improves and spikes before I notice I'm getting sick. Zone 2 cardio a few hours a week, which Attia is evangelical about and which is extremely boring. Strength training three times a week, for muscle mass as a metabolic reserve and as protection against falling over when I'm old. Protein at somewhere around 1.6g per kg of body weight, which is a commonly cited target for active adults rather than a settled number. And regular sauna, where the longevity evidence is suggestive rather than conclusive, but which I'd probably keep doing anyway because it's an hour with no screen in it.

I should say that the tracking itself has a cost. There's a version of this where the data becomes the hobby and the anxiety about the numbers does more damage than the thing being measured. I've been closer to that line than I'd like.

Where the engineering metaphor is useful

Observability is the real one. You can't make decisions about a system you can't see, and most people run their health entirely blind until something fails hard enough to force a doctor's visit. Sleep data, HRV, an annual blood panel - it's not exotic, it's just the difference between having metrics and guessing.

Capacity planning is the second. VO2 max and muscle mass are headroom. When something goes wrong - illness, injury, a bad year - you draw down reserves, and how much you have determines whether the system stays up or fails cascading. That reframing did more to get me into the gym consistently than any aesthetic motivation ever did.

Graceful degradation is the goal, and it's worth stating precisely, because "longevity" invites a fantasy about not declining. You will decline. The question is whether it's gradual and survivable or sudden and total.

The part I'm least sure about

There's a failure mode in this whole genre where systems thinking becomes a way to feel in control of something that isn't very controllable. Genetics, luck, and circumstance do enormous amounts of work here, and a well-instrumented person can still get unlucky at 50. I notice a pull toward believing the dashboard protects me. It doesn't. It just means I'll see some things earlier.

What I'm reasonably confident about is small: sleep, movement, and muscle have the best ratio of effort to benefit, and nearly everything else - supplements, protocols, the more exotic end of the biohacking world - is marginal by comparison and much more fun to talk about. I've spent more hours reading about the marginal stuff than the fundamentals deserve. That's a bias worth naming.

What I'm optimizing for

Not lifespan. Healthspan - the years where I'm functional and independent and able to do the things I care about.

The goal at 80 isn't to still be here. It's to be able to pick up a grandkid, get on a plane, play guitar, write something. That's a capacity target, and capacity is built decades before you need it.

Both books are worth reading with different expectations. Walker's is more accessible and will change how you treat sleep this week, provided you read it as an argument rather than a citation. Attia's is denser, more careful about uncertainty, and better as a framework than as a protocol.

Happy to compare notes if you're deep in this: LinkedIn or email.