Jonas Leddington

The Heart of the K-Shaped Economy: Cardiovascular and Metabolic Health in a Divided America

Archive

Researched and written in conjunction with Claude. All ideas mine. Future writing will be all my own words, unless otherwise noted.

A Stelo CGM subscription costs $99 a month. It tells more affluent biohackers how their morning oatmeal affects their glucose curve. (Admittedly, I’ve used one to do exactly that.) Meanwhile, a man in rural Mississippi with undiagnosed diabetes is massively more likely to lose a limb to amputation than to ever wear a continuous glucose monitor. The consumer health industry’s most celebrated innovations are optimizing the already-well. The people dying are not in the addressable market.

The Thesis

The populations carrying the greatest disease burden — lower-income Americans, Black and Hispanic communities, rural residents, the working class caught between Medicaid and disposable income — remain largely outside the aperture of consumer health. They are served, when they are served at all, by an overstretched clinical system built for acute intervention, not prevention, behavior change, or sustained wellness.

This is not just a moral failure. It is a market failure. And it points to a different kind of opportunity: consumer health deployed not as a market category for the affluent, but as a modality — a way of reaching people through consumer channels, community structures, and scalable models that the traditional healthcare system cannot.

This is the first in a series mapping where America’s health outcomes actually diverge, where the existing system breaks down, and where a different approach to consumer health intervention could have the greatest impact. We start with cardiovascular and metabolic disease — the nation’s largest killer and its starkest arena of disparity.

The Burden Is Not Where the Market Is

Cardiovascular disease kills roughly 700,000 Americans a year — about 1 in 5 deaths. Type 2 diabetes affects 37 million, with another 96 million prediabetic. These are the conditions most amenable to behavioral and lifestyle intervention, and the ones where disparities are widest. The data, by population:

Race and Ethnicity

Black Americans die of heart disease at rates 30-40% higher than white Americans. Age-adjusted heart disease mortality runs approximately 220-240 per 100,000 for Black adults versus ~170 for white adults — a gap that has persisted for decades despite overall declines. Black women’s heart disease mortality exceeds that of white men in some analyses, disrupting the assumption that CVD is primarily a male disease. (CDC WONDER/NVSS; AHA Circulation 2023 Statistical Update.)

Hypertension — the key upstream driver — affects 56% of Black adults, compared to ~48% of white adults and ~47% of Hispanic adults. More critically, control rates are lowest among Black and Hispanic adults. The gap doesn’t just exist at diagnosis. It widens at every subsequent stage — treatment, medication adherence, blood pressure at target. (NHANES; Ostchega et al., NCHS Data Brief No. 364; AHA 2023 Update.)

Diabetes prevalence is 12.1% among Black adults and 11.8% among Hispanic adults, versus 7.4% among white adults. But prevalence understates the disparity: diabetes-related mortality is roughly double for Black Americans, reflecting gaps in disease management, complication rates, and access to endocrinology. (CDC National Diabetes Statistics Report, 2022.)

American Indian and Alaska Native populations carry the highest diabetes burden of any racial group — prevalence 2-3x the white rate — and are chronically invisible in both research and product design. (IHS data; CDC NDSR.)

Gender

Men die of heart disease at younger ages and higher rates — roughly 270 per 100,000 versus ~165 for women, age-adjusted. But the gendered story is more complicated than “men’s disease.” (CDC WONDER/NVSS.)

Women are systematically underdiagnosed and undertreated for CVD. Women presenting with heart attack symptoms are less likely to receive timely intervention, less likely to be prescribed statins, and more likely to die of a first cardiac event. The awareness gap has narrowed since the AHA’s “Go Red for Women” campaign, but remains significant among lower-income women and women of color. (Mehta et al., Circulation, 2016; Mosca et al., Circulation, 2013.)

Metabolic syndrome — the precursor cluster — has its highest prevalence among Hispanic and Black women, a population almost entirely absent from consumer metabolic health marketing. (Aguilar et al., JAMA, 2015.)

Income

Income may be the single most powerful predictor of cardiovascular and metabolic health — more consistent in its gradient than race once you control for confounders, though race and income are deeply entangled in America.

Men in the top 1% of income live 14.6 years longer than men in the bottom 1%. For women, the gap is 10.1 years. CVD is the single largest contributor. Critically, this gap widened between 2001 and 2014, and low-income life expectancy varied by 4-5 years depending on geography. (Chetty et al., JAMA, 2016.)

Diabetes prevalence roughly doubles from the top to the bottom income quintile, and the gradient is not just a poverty threshold effect — it operates continuously across quintiles. (NHANES; CDC NDSR; Beckles & Chou, Preventing Chronic Disease, 2016.)

Hypertension control degrades at every income step down. Lower-income adults are less likely to be aware of their hypertension, less likely to be treated if aware, and less likely to be controlled if treated. The failure compounds at each stage. (AHRQ National Healthcare Quality and Disparities Report.)

The income quintile picture matters for market design:

  • Bottom quintile: Substantially served by Medicaid and FQHCs, but these systems focus on acute and chronic management, not prevention or behavior change.
  • Second and third quintiles (~$30K-$100K household income): Too wealthy for Medicaid in most states, often carrying high-deductible employer plans, with limited discretionary health spend. This is the population most likely to fall through the gap between public health infrastructure and consumer wellness markets.
  • Fourth quintile (~$100K-$160K): Where consumer health engagement begins to rise — but offerings remain poorly tailored, marketed as aspirational rather than functional.
  • Top quintile: Well served. Increasingly over-served, with diminishing marginal returns on each new optimization product.

Urban vs. Rural

Heart disease mortality is 20-25% higher in rural counties than urban ones — and the gap is widening, as urban CVD mortality declines faster. (Moy et al., MMWR, 2017; Garcia et al., Circulation, 2021.)

Rural diabetes prevalence runs roughly 12% versus 9% urban, with rural patients less likely to receive guideline-concordant care, less likely to see an endocrinologist, and more likely to experience preventable complications including amputations. (CDC; Rural Health Information Hub.)

The rural South — the “Stroke Belt” — has age-adjusted stroke death rates 30-40% above the national average, driven by poorly managed hypertension in a region with severe provider shortages. (Howard et al., Stroke; CDC WONDER.)

For intervention design, rural dynamics are structurally different:

  • Provider deserts mean any model dependent on clinical touchpoints will fail to reach these populations at scale.
  • Trust runs through local institutions — churches, employers, community organizations — not through apps or telehealth platforms.
  • The rural health workforce shortage is accelerating, which means the gap will widen without new models.

Where the Breakdown Actually Lives

This is the question that matters for anyone trying to build something: *where in the chain from awareness to outcomes does the system fail, and for whom?

For cardiovascular and metabolic disease, the primary breakdown is not awareness. Most Americans know that heart disease, diabetes, and obesity are dangerous. The failures are downstream:

Engagement is the first fracture. Lower-income and rural populations are far less likely to have a regular primary care relationship where risk is identified and managed proactively. Consumer products that could theoretically fill this gap — blood pressure monitors, CGMs, health coaching — are priced for and marketed to people who already have doctors. The engagement failure is economic, but it is also cultural: existing products don’t speak to these populations’ identities, contexts, or daily realities. A CGM marketed through biohacker podcasts will never reach a second-quintile construction worker with prediabetes. But that worker’s risk is real, and unaddressed.

Adherence is the compounding failure. Even when conditions are diagnosed and treatment initiated, medication adherence for hypertension and diabetes drops sharply as income drops. Cost is a factor, but so are regimen complexity, competing life demands, lack of social support structures, and institutional distrust. This is where isolated clinical encounters fail and where sustained, community-embedded models — built around accountability, identity, and daily rhythm rather than episodic appointments — may have their greatest leverage.

Outcomes even given adherence show residual gaps, likely reflecting accumulated physiological stress (the “weathering” hypothesis, particularly for Black Americans), environmental exposures, and care quality differences. This layer is least addressable by consumer intervention alone — but matters for honest framing of what’s achievable and what requires systemic change.

What This Means

The consumer health industry has built sophisticated, beautifully designed, venture-funded products for a population whose outcomes are already trending in the right direction. The bottom three income quintiles, Black and Hispanic Americans, rural communities, and the vast working and middle class caught between Medicaid and Equinox — these populations carry the overwhelming cardiovascular and metabolic burden, and they are essentially unaddressed by consumer health.

The opportunity is not to make existing consumer health cheaper. A discounted CGM is still a product designed for someone who self-identifies as a health optimizer. The opportunity is to build entirely different models — rooted in community, layered into daily life, designed around the engagement and adherence failures the data reveals — and to deploy consumer health as a modality for reaching people the clinical system cannot sustain.

The hardest part is not the data. The data is clear. The hardest part is building something that works in the second and third quintiles — where willingness to pay is constrained, engagement models have to earn their way in through trust and relevance, and the competitive set is not Whoop or Noom but inertia, distrust, and the daily grind of getting by.

Next: Mental Health and Substance Use — where the crisis is most acute, the consumer market is most confused, and the gap between what’s offered and what’s needed may be even wider.

Sources

  • CDC WONDER / National Vital Statistics System — mortality by demographics and geography
  • NHANES (National Health and Nutrition Examination Survey) — prevalence and biometric data
  • CDC National Diabetes Statistics Report, 2022
  • AHA Heart Disease and Stroke Statistics — 2023 Update (Circulation)
  • Chetty et al., “The Association Between Income and Life Expectancy in the United States, 2001-2014,” JAMA, 2016
  • Ostchega et al., NCHS Data Brief No. 364 — hypertension prevalence and control
  • AHRQ National Healthcare Quality and Disparities Report
  • Moy et al., “Leading Causes of Death in Nonmetropolitan and Metropolitan Areas,” MMWR, 2017
  • Garcia et al., “Cardiovascular Disease in Rural Populations,” Circulation, 2021
  • Aguilar et al., “Prevalence of the Metabolic Syndrome in the United States,” JAMA, 2015
  • Mehta et al., “Acute Myocardial Infarction in Women,” Circulation, 2016
  • Mosca et al., “Awareness and Prevention of CVD in Women,” Circulation, 2013
  • Beckles & Chou, “Diabetes — United States, 2006-2016,” Preventing Chronic Disease, 2016

*Specific statistics represent synthesis of recent data from these sources, typically reflecting 2019-2023 estimates. Ranges are used rather than false-precision point estimates where I am synthesizing across datasets. COVID-era disruptions may have shifted some baselines.