News

By Dr. David Gorstein, Partner & Population Health Practice Leader

WHOOP and Oura command astronomical valuations on the promise that they’ll become the operating system for health and wellness data—integrated with healthcare in ways that are insightful and actionable.

From personal experience, I know that they can’t even make that true in the narrow, well‑defined use case of training data.

Warning shot across the bow (today’s press release from Training Peaks):

“Health metrics from Oura Ring now integrate directly into TrainingPeaks, giving you a full-circle view of your training and life. Connect your Oura Ring to see sleep, HRV, stress scores and more populate in your Health Insights dashboard.”

So far, most of these partnerships are effectively brand events: “X now integrates with Y” but integration depth is extremely thin. It’s just data appearing on another surface, vs a true synthesis layer that changes recommendations, training plans, or clinical decisions.

Training is hard enough. Tracking it shouldn’t be.

Right now there’s a structural gap in endurance + healthspan tech:

• Training platforms (e.g., TrainingPeaks) own the structured work: power files, TSS/IF, planned sessions, coach workflows.
• Wearables (e.g., WHOOP, Oura) own the continuous physiology: sleep, HRV, strain, illness, meds, travel, labs.

The real “operating system” would simply fuse those two. That’s where the value is.

If a device can’t even ingest power and basic workout metadata from where serious athletes and coaches already live, it’s not an OS. It’s a very good sensor feeding someone else’s OS.

That has two big consequences:

• Coaching: Without structured training data, any “AI coach” is guessing about intent. It sees the strain, but not the plan.
• Strategy: If TrainingPeaks (or someone else) pulls in wearables’ health data while wearables don’t pull in structured work, the platform becomes the hub and the devices become interchangeable inputs.

Taking it to the next level of insight for Whoop or Oura will require:

• Ingesting TrainingPeaks power + planned/completed workouts and fusing that with strain and healthspan.
• Understanding why serious coached athletes already treat TP as the OS and wearables as inputs.
• Defining what has to be true for “OS of performance and health” to be more than a pitch‑deck slogan.

To justify their valuations and investor expectations, WHOOP and Oura are already expanding beyond elite performance into healthspan, chronic conditions, preventive care, labs, and eventually deeper healthcare workflows, but If you can’t tie simple power‑based training load to autonomic response today, you’re not going to be the layer that turns blood + omic data into “next best action” tomorrow. You’ll be another point solution in someone else’s care stack.

Who will emerge?

Every company wants to be the system of record / OS, so everyone gatekeeps their data. Strava talked about being the OS for the athlete years ago (wanting to pull in sleep data from Oura/WHOOP), but no one let them so that vision never came about. Strava’s own API access is also heavily limited. WHOOP has historically been selective with partnerships because they believe data is their primary asset. All players want the context, but nobody really wants to give up control of the user graph or the data layer.

WHOOP is trying to get deeper insight and give better recommendations using their AI coach, which is actually pretty cool. To know more about the person: what they’re training for, what their plan is, what their health concerns are, etc, is derived from asking the user directly. It’s not optimal for companies to collect that information manually (AI chat) or build native workout features (ie, Strength trainer in WHOOP, but there are MUCH better workout apps)

That means the “full-context OS” probably does not happen through normal integrations. It likely only happens through acquisition, otherwise, the incentives are too misaligned. I think it’s not just a product strategy problem but also has to do with the current competitive dynamics in this market and the need to justify these crazy valuations.  But so far the integrations and new insights remain half-baked.

If you’re building in this space and see it differently, I’d genuinely like to hear why.

By Dr. David Gorstein, Partner & Population Health Practice Leader

Recent analyses put the “longevity market” at over $1T in total addressable value. That opportunity is typically framed around demand for:

• $10K–$50K “longevity retreats”
• Extensive multi-omic panels
• Stacks of peptides and nootropics
• Unvalidated “biological age” tests used for serial monitoring
• Concierge “longevity” clinics that over-test and over-image as a business model

This is a compelling story for investors. It is far less compelling if the goal is to extend healthspan in a way that would meet even minimal value-based care standards.

The uncomfortable truth is that the core pillars of healthspan are already well studied, actionable, and essentially free.

• Regular physical activity (at least 150–300 minutes per week of moderate intensity,   plus at least 2 days per week of strength training)
• Sleep regularity and sufficiency
• Tobacco cessation and minimal alcohol
• Vaccination, blood pressure and lipid control, and basic metabolic risk management
• Social connection, fall prevention, and fracture prevention in older adults

These “boring” levers are exactly the ones that move mortality curves, disability curves, and dementia curves the most. They are not scarce, and they are not high-margin.

If your goals are like mine:

• Can I ride a bike at 70?
• Can I climb stairs without stopping?
• Am I still myself cognitively?
• Can I chase my grandchildren without worrying about my balance?

then most of the tools you need already sit in primary care, public health, and your own behavior, not in a boutique longevity stack.

A Different Map: Evidence vs. Scalability

Instead of organizing the longevity market by price point or “innovation,” it is more useful to classify interventions along two axes:

Evidence: strength of data for improving meaningful outcomes (mortality, morbidity, function, quality of life)

Scalability: feasibility to deploy at population scale at acceptable cost and risk

This yields four quadrants.

1. High-Value, Population-Level

Definition: Strong evidence; cheap; scalable; low harm.

Examples:

• Physical activity prescriptions with resistance training
• Sleep regularity and duration programs
• Tobacco cessation and alcohol reduction
• Vaccination; blood pressure and lipid control; simple metabolic risk management
• Social connection initiatives; fall and fracture prevention

These are the true “blockbusters” of healthspan. They are exactly what a rational value-based system would buy first, at scale.

They are always mentioned in longevity marketing, but then the pitch quickly pivots to the shiny new objects with little or no evidence to support them.

2. High-Value, Targeted / High-Risk

Definition: Strong evidence in clearly defined, higher-risk populations; often pharmacologic or procedure-based.

Examples:

• SGLT2 inhibitors and GLP-1 agonists for established cardiometabolic disease or obesity
• Antihypertensives and statins in high-risk groups
• Proven secondary-prevention interventions (post-MI, stroke, high-risk heart failure, osteoporosis with prior fragility fracture)

These belong in a longevity conversation, but as precision tools, not wellness accessories.

3. Low-Value, Scalable

Definition: Easy to sell and scale; weak or indirect evidence for hard outcomes; often marketed via biomarkers and engagement metrics instead of morbidity and mortality.

Illustrative examples:

• Serial, unvalidated “biologic age” tests treated as primary endpoints
• Consumer multi-omic panels with no clear treatment pathways
• General-population continuous monitoring with no structured, evidence-based intervention behind the data

These offerings may generate insight, motivation, or a sense of control for some individuals. As core longevity infrastructure, however, they have a poor evidence-to-cost ratio and are easy vehicles for over-claiming.

4. High-Uncertainty / Experimental

Definition: Biologically plausible and often exciting, but with limited or early evidence for clinically meaningful endpoints.

Illustrative examples:

• Many peptides, stacks of nootropics, and “longevity cocktails”
• Intensive multi-omic plus imaging packages in otherwise healthy people
• High-priced longevity retreats and concierge clinics that bundle speculative interventions with a clinical veneer

These are not inherently illegitimate. They are research questions or luxury goods, not standard of care. The problem arises when they are sold and reimbursed as if they occupy the high-value, population-level quadrant.

Exhibit: The Longevity Intervention Matrix

Axes:

Vertical: Evidence for meaningful outcomes

Horizontal: Scalability at acceptable cost and risk

HIGH VALUE longevity interventions are the upper right quadrant

The key takeaway for systems, payers, and employers is simple:

• Start in the high-value, population-level quadrant as the default; these give the highest health return per dollar.
• Use the targeted, high-risk quadrant selectively, for clearly defined groups, with outcome and cost accountability.
• Treat the low-value, scalable quadrant at best as an engagement layer, not core healthspan infrastructure.
• Treat the experimental quadrant explicitly as research or luxury, not standard care

Recommendations for Systems, Payers, and Employers

To align “longevity” with genuine value-based care, stakeholders should:

1. Reallocate investment

Prioritize infrastructure and programs that increase movement, strength, sleep regularity, cardiometabolic control, and fall prevention across the population.

Fund digital and in-person interventions that scale these basics, not boutique offerings.

2. Risk-stratify higher-cost interventions

Reserve advanced pharmacology, imaging, and intensive monitoring for clearly defined high-risk cohorts, with accountability for total cost and net benefit, not just service volume.

3. Reframe boutique longevity as research or luxury, not standard care

Encourage transparent labeling of speculative or high-priced interventions as experimental or elective.

Avoid embedding them in benefit designs as if they were evidence-based essentials.

4. Measure what matters

Track outcomes that people actually care about and understand: mobility, independence, cognitive function, hospitalizations, and days at home, alongside mortality.

Evaluate any “longevity program” against those outcomes and total cost, not against proprietary biomarkers alone.

5. Define success in human terms

Anchor programs on the outcomes people actually care about in later life: mobility, cognition, independence, and the ability to participate in relationships and roles—not just nicer dashboards.

Conclusion

“Longevity” is a useful public-facing language only if it ultimately describes more healthy, independent years for more people per dollar, not just a new total addressable market for the wellness and biotech industries.

High-value longevity care already exists, and it looks far more like safe sidewalks, resistance training, blood pressure control, vaccination, and boring sleep routines than it does like $25,000 retreats or multi-omic dashboards. The task now is to fund and scale the former while being honest about the latter.