The Unreasonable Effectiveness of Lifestyle

It wasn’t quite like the dream where you’re standing in front of an audience and suddenly realize you’re not wearing pants, but at a recent Harvard Business School function, I felt a flush of affinity.

The occasion was an annual event for students in the Harvard life sciences MS/MBA program, designed to recognize the hard work of students and faculty, and to connect trainees with program advisors like me. Before dinner, advisors were invited to stand and briefly introduce ourselves to students.

The experience was something else. One woman was leading something like her fifth life science company, as I recall; another was a former Brigham infectious disease physician who pivoted to life science investing and just retired after 30 years. A third was a highly effective social impact investor who trained as a lawyer, served in the Navy where he was an officer on a guided missile frigate, and was a former managing director at Bain Capital. You get the idea.

And then it was my turn.

I explained how after a quarter century in biopharma and healthtech, I’m now focused on … improving health by cultivating agency, in the context of a capacious vision of health that also embraces physiology, purpose and connection.

Which was when I felt like I forgot my pants.

The audience seemed to stare at me with looks suggesting a combination of confusion and pity. They may have wondered if I had stumbled into the wrong event — a thought I considered as well. I passed the mic to the next advisor (I think a life science VC with a MBA from Harvard), and the event continued.

But it made me wonder, as I know many have before me, why we tend to view efforts to treat the sick so differently — so much more seriously — than efforts to keep us healthy in the first place.

The issue isn’t, or shouldn’t be, the high esteem in which we hold new medical products — such as the therapeutics I’ve spent much of my career trying to discover and develop.

As a patient, when you are seriously sick, your entire world collapses down to your illness, and there’s nothing you care about more than trying to get better. In these moments of urgency, we’re grateful — beyond grateful — for the availability of powerful diagnostic tools that help delineate our disease, and effective medicines that, if we’re fortunate, can be deployed to treat it.

But start to talk about lifestyle, and you’re immediately shunted somewhere else — akin, perhaps, to the “lower stakes” tables in Swingers, or into the company of Jugdish and Sidney in Animal House.

More to the point, I can just imagine a medical colleague muttering, “I didn’t spend years of miserable training memorizing the Krebs cycle and the coagulation cascade — and drop tens of thousands of dollars in tuition — just to tell my patients to move more, drink less, and eat better.”

And yet: this is just what some of the strongest data suggest.

The Unreasonable Effectiveness of Lifestyle Interventions

In a 1959 lecture at New York University, physicist Eugene Wigner introduced the phrase “the unreasonable effectiveness of mathematics.” We might borrow the framing to describe the unreasonable effectiveness of lifestyle interventions, given their outsized impact on human health.

I’ve previously outlined an evidence-based model for flourishing health built around three pillars: physiology (movement, sleep, nutrition, prevention), connection and purpose, and agency. It’s the third pillar I find most underappreciated — and potentially tractable. But all three matter, and the data confirm it.

The impact of lifestyle on keeping us healthy far exceeds the impact of most medicines –which is perhaps why the 2019 ACC/AHA prevention guidelines explicitly state that the most important way to prevent cardiovascular disease is to promote a healthy lifestyle throughout life. A compelling 2018 publication in Circulation found that adults who maintained five healthy behaviors — not smoking, healthy body weight, moderate physical activity, avoiding heavy alcohol use, and a healthy diet — reduced their risk of major disease by a staggering 74%, translating to 12–14 extra years of life compared with people who had none of these habits.

Research has also reinforced the importance of sleep. I found a recent discussion on the always-informative Future of Everything podcast hosted by Stanford’s Russ Altman particularly valuable: the perspective offered by guest Jamie Zeitzer — a circadian physiologist at Stanford — was surprisingly pragmatic and down-to-earth, mercifully lacking the dogmatism often associated with the topic.

For a vivid illustration of what agency actually looks like — and why it matters — consider this memorable LinkedIn post from WHOOP founder Will Ahmed, whose platform is now valued at over $10B, exceeding most local biotechs. Ahmed himself clearly possesses remarkable agency, which was evidently an animating force in his leadership of the company. The irony is that WHOOP itself, like other fitness platforms including Peloton and Oura, seems especially focused on enhancing performance through metric optimization rather than deliberately cultivating agency in its users — a gap that thoughtfully deployed AI could effectively address (as we’ll discuss).

Holding lifestyle to a higher standard

Part of the problem faced by preemptive medicine is that because patients start out relatively healthy, the stakes seem lower. Correctly diagnosing and treating a dangerous disease has much less room for error than offering lifestyle advice to healthy people — which also means the bar for lifestyle advice is vanishingly low. To the extent this encourages diversity of approaches and marketplace competition, that’s good. But the worry is accountability: since it’s so difficult to connect specific lifestyle interventions with downstream consequences, anyone can argue anything. It’s all but impossible to discern (in a time frame likely to matter) what’s truly effective, and to separate an approach that’s trendy or fun from one more likely to yield durable benefit.

Moreover, most lifestyle-oriented platforms are paid if they engage users, independent of any health benefit those users receive. One perverse consequence is the tendency of platforms to pursue users already relatively engaged in their health — for the simple reason that customer acquisition costs are viewed as lower. Getting a fairly healthy person to opt into a Peloton or a WHOOP is a lot easier than motivating a habitually sedentary couch potato to do the same; yet in the case of physical activity, the health benefit of going from nothing to something is about as large as the benefit of going from something to a whole lot more. The real inflection point occurs when people get off the couch.

Capacious and accountable

A longstanding theme of this column has been my aversion to the pursuit and culture of relentless optimization — a view Brad Stulberg, among others, has expressed with particular eloquence — and my embrace of a more expansive view of health.

This was very much the theme of my recent keynote at Personify Health’s annual Thrive Summit, where I tried to make the case that genuine flourishing isn’t built on metric optimization — it’s built on function, connection and purpose, and agency –- the belief that your choices matter, and that you can create a better tomorrow. Life, I reminded the audience, is richer than any dashboard can capture.

What I find most promising is the possibility of cultivating agency deliberately — and we have a model for what that actually looks like. The Diabetes Prevention Program, which achieved a remarkable 58% reduction in type 2 diabetes incidence through “lifestyle alone,” worked through intensive, individualized coaching: helping participants understand their own goals, set achievable targets, practice self-monitoring as a skill rather than a score, and, crucially, develop strategies for recovering from lapses rather than treating them as failures. The typically overlooked self-regulatory skills you develop in sticking to an exercise routine, for example, if made explicit and deliberately extended to new contexts, can be portable to other domains of life.

Thus, while relentless optimization in the Sisyphean pursuit of perfection seems misguided, staying accountable seems sensible. I weigh myself every day. I have a weekly exercise schedule I generally keep to. I know how much I lift, and a sense of how fast I jog. I don’t use leaderboards on Peloton and other apps, but I choose rides based on the time and anticipated intensity. The comfort of routine feels reassuring and serves as a great foundation to my day.

I suspect that if lifestyle platforms can motivate you to keep exercising — the way I find Peloton instructors can — and make consistent exercise a bit more engaging and fun, that may be a real benefit. 

The proliferating metrics offered by some platforms are trickier: markers of biological age, for instance, remain population-derived and haven’t been validated for use in individuals. But the more interesting question, as Dr. Isaac Kohane often notes, is “compared to what?”

The honest answer is that chronological age (as Dr. Martin Lee eloquently discussed in a recent New England Journal of Medicine) is itself a poor proxy for our actual physiological reserve and potential. What the escalating interesting in new markers of aging gesture toward, however imperfectly, is the animating idea of my recent Wall Street Journal op-ed: that our biological trajectory is not fixed by the number of times we’ve circled the sun, and that through the choices we make — movement, recovery, connection, and the cultivation of agency — we can genuinely elevate our health and enhance our flourishing.

Lifestyle’s AI opportunity

A particular opportunity for lifestyle platforms may involve AI — but perhaps not in the way you think. I’m extremely skeptical that AI is going to churn through the often-dubious data and questionable metrics wearables tend to generate and deliver trenchant insights on the perfect fitness program. But I have been impressed by the quality of dialog AI enables. I’m trying out a base model WHOOP right now, and the AI coaching chat, in beta mode, is surprisingly good.

A decade ago, I wrote about the untapped potential of personalizing behavioral approaches — the idea that behavioral coaching could be far more effective if it were grounded in a deeper, more nuanced understanding of the individual — the kind that only emerges over time. An AI that could get to know you well enough to offer not better exercises, but a more effective approach to motivating and encouraging you specifically — and begin to do so with the sort of attentiveness and continuity that characterized the DPP — could be important and valuable. And if that capability extended beyond exercise into other areas of life, you’d have something more interesting than a fitness app: an engine for agency itself. This, as I see it, is the best-case scenario for where fitness and other “lifestyle” platforms could wind up — and how they might finally make good on their often-heralded, long-awaited promise to transform our health and enlarge our lives.