To be continued

Metabolic Risk Stratification in Hospital Practices II

A review and business-model analysis for value-based hospital growth

6.1 The p-tau217 evidence base

 Plasma phosphorylated tau at position 217 (p-tau217) has emerged as the best-validated of these markers. In a large multi-center study published in Nature Medicine in 2025, plasma p-tau217 detected amyloid-related Alzheimer's pathology with areas under the receiver-operating-characteristic curve of 0.93 to 0.96, with accuracy of 89 to 91 percent in secondary (specialist) care settings and 85 percent even in primary care settings. Head-to-head comparisons against cerebrospinal fluid and PET-confirmed diagnoses have repeatedly found p-tau217 to outperform other blood markers such as p-tau181, with reported diagnostic accuracy (area under the curve) as high as 0.95 in some cohorts. A large Swedish community cohort study published in Nature Medicine in 2025 (the SNAC-K study, following more than 2,100 adults aged 60 and older for up to 16 years) found that higher p-tau217, alongside neurofilament light chain and glial fibrillary acidic protein, was associated with increased long-term risk of all-cause dementia and Alzheimer's disease specifically, with p-tau217 showing the strongest 10-year discriminative ability for Alzheimer's disease among the markers tested.

6.2 What these tests can and cannot do: a regulatory reality check

It is essential for any hospital program built on this science to state its limits accurately. As of mid-2026, the U.S. Food and Drug Administration had cleared two plasma tests for specific, narrow clinical uses in symptomatic adults being evaluated by a specialist: a Lumipulse-platform p-tau217/beta-amyloid ratio test, and an Elecsys p-tau181 test intended for initial rule-out before specialist referral. Neither test is authorized, validated, or intended as a population-wide screening tool for asymptomatic or cognitively healthy adults, and neither can tell a healthy individual with certainty whether dementia will eventually develop. A pooled analysis of more than 2,700 cognitively unimpaired adults presented at the 2026 Alzheimer’s Association International Conference found that each one-standard-deviation increase in baseline plasma p-tau217 was associated with a 38 percent higher risk of progressing to mild cognitive impairment or dementia over subsequent years — a genuinely useful prognostic signal for research and risk stratification, but explicitly not, in the authors’ own framing, a basis for treating an abnormal result in a healthy person as a diagnosis, since an abnormal biomarker result in a currently asymptomatic person cannot yet be paired with a treatment proven to prevent the predicted decline. This distinction — between a validated prognostic risk marker useful for stratification within a broader management program, and a diagnostic test that establishes or rules out a disease — is the same distinction the original Chinese-language strategy document rightly insisted on for cancer-related metabolic markers such as the TyG index, and it must be applied with equal rigor to any Alzheimer's-related biomarker a hospital chooses to offer. Amyloid pathology can also be present for twenty to thirty years before dementia symptoms appear, particularly in APOE ε4 carriers, meaning a positive result in a young or middle-aged adult communicates elevated long-term risk on a timescale unsuited to short-term alarm or to any specific clinical promise.

6.3 Combining metabolic and neurodegenerative risk markers: the rationale for integration

The scientific case for integrating a neuro-metabolic risk panel into a hospital's existing insulin-resistance and metabolic-assessment infrastructure, rather than building it as a separate service line, rests on three points. First, the same population already being evaluated for HOMA-IR, the TyG index, and cardiometabolic risk factors is, on the epidemiological evidence reviewed above, a population at elevated background risk for cognitive decline, so the marginal cost of adding a cognitive and neuro-biomarker assessment to an existing visit is low relative to building a freestanding memory clinic. Second, the interventions with the best current evidence for reducing dementia risk in at-risk adults — structured multidomain lifestyle programs combining exercise, nutrition, cognitive and social engagement, and management of cardiovascular and metabolic risk factors, following the model established by the Finnish FINGER trial and since replicated and adapted in the U.S. POINTER trial and in Chinese cohorts including the Hangzhou HERITAGE study and the rural MIND-China initiative — are largely the same interventions already required for metabolic risk reduction, so the two service lines share a delivery infrastructure of nutritionists, exercise physiologists, and structured follow-up rather than requiring a separate one. Third, an emerging 2026 international trial (PROTECT-Cog, led by the Alzheimer's Association) is specifically testing whether combining a FINGER-style multidomain lifestyle program with a metabolism-targeting drug class (GLP-1 receptor agonists) produces a larger reduction in cognitive-decline risk than either approach alone — a direction of travel that reinforces, rather than undermines, the case for treating metabolic and cognitive risk as a single clinical problem rather than two separate ones.

 7. A New Business Model: The Metabolic and Neuro-Metabolic Function Assessment Center

 7.1 Positioning

 The proposed service should be named and positioned carefully. “Metabolic and Neuro-Metabolic Function Assessment Center” or “Insulin Resistance and Brain-Metabolic Health Center” are both defensible names; “Cancer Prevention Center” or “Dementia Prevention Center” are not, and should be avoided regardless of commercial appeal, because neither cancer nor Alzheimer's disease prevention can be promised on the basis of current evidence. The center should be positioned as offering metabolic and cognitive-risk assessment, individualized risk stratification, and evidence-based lifestyle intervention for adults with prediabetes, obesity, fatty liver disease, dyslipidemia, hypertension, polycystic ovary syndrome, a family history of dementia or elevated background dementia risk factors, and health-conscious individuals seeking proactive management — explicitly as a complement to, not a substitute for, standard endocrinology, cardiology, neurology, and oncology care.

7.2 A tiered testing menu

Testing should be organized in tiers of increasing depth, each with a stated clinical rationale, an explicit statement of what the result does and does not mean, and a defined follow-up pathway, to avoid both over-testing and the appearance of diagnostic overreach.

Tier Core content What it establishes What it must not claim 1. Basic metabolic screen Fasting glucose, HbA1c, fasting insulin, lipid panel, liver and kidney function, body composition, waist circumference, blood pressure 

2. Insulin-resistance panel Fasting insulin, HOMA-IR, TyG index, waist-to-height ratio, body composition for degree of insulin resistance as a modifiable risk factor.

3. Inflammatory / metabolic-stress panel hs-CRP, uric acid, liver enzymes, selected oxidative-stress markers where clinically indicated chronic low-grade inflammation associated with metabolic dysfunction.

4. Neuro-metabolic risk panel (new) Validated cognitive screening instrument; plasma p-tau217 or an equivalent FDA-cleared / regulator-approved assay where clinically appropriate and consented; APOE genotype only with genetic counselling; correlation with the insulin-resistance panel A prognostic risk signal for Alzheimer's-related pathology, to be interpreted alongside cognitive testing and family history for a diagnosis of Alzheimer's disease or dementia in an asymptomatic person; any guarantee about future cognitive outcome.

5. Deep metabolic profiling Amino-acid panel, organic acids, lactate/pyruvate, selected TCA-cycle intermediates Research-grade or high-end assessment with defined, narrow clinical indications.

6. Integrated risk-stratification report Synthesis of tiers 1–5 with individualized intervention plan A personalized, actionable risk profile and follow-up plan A prognosis or a promise of disease prevention.

The neuro-metabolic panel is the one clear addition this review makes to the original strategy document, and it carries the highest compliance burden of any tier: it should only be offered under a clear informed-consent process, should never be marketed as a stand-alone screening product, should be interpreted by a clinician with training in cognitive assessment (not returned as a raw number to the patient), and should be paired from the outset with referral criteria to neurology or memory-clinic specialists for any result that, combined with cognitive testing, warrants further evaluation.

7.3 Intervention design

 The intervention side of the center should mirror the model already validated in multidomain dementia-prevention trials, because it is also the correct model for insulin-resistance management: individualized medical nutrition therapy delivered by a registered dietitian or nutrition physician, emphasizing energy balance, reduced refined carbohydrate intake, adequate protein, dietary fiber, and a low-glycemic-index eating pattern; a structured exercise prescription combining aerobic and resistance training, designed by rehabilitation or sports-medicine specialists; regular monitoring of weight, waist circumference, body fat percentage, muscle mass, and visceral fat; assessment of sleep quality and psychological stress, both of which independently affect insulin sensitivity and cognitive health; multidisciplinary follow-up spanning endocrinology, nutrition, rehabilitation medicine, cardiology, hepatology, and, where the neuro-metabolic tier is used, neurology or geriatric psychiatry; and digital follow-up through outpatient visits, telephone or online check-ins, wearable devices, or a chronic-disease management platform to sustain adherence, since adherence — not any single test result — is what determines whether the program produces a measurable benefit.

8. Policy Basis and Compliance Guardrails

 The center can reasonably be positioned within several converging policy directions: high-quality development of public hospitals, comprehensive chronic-disease prevention and control, the national weight-management and metabolic-disease action initiatives, extension of health-management services, development of special-needs medical services, upgrading of physical-examination center offerings, multidisciplinary collaborative care, and integration with commercial health insurance and corporate health-management contracts. Several provinces permit medical institutions to offer special-needs services or to apply market-adjusted pricing to individualized services that meet regulatory conditions, but this varies materially by province and must be confirmed against local pricing-bureau, health-security-bureau, and health-commission requirements before launch rather than assumed.

A small number of compliance red lines should be treated as non-negotiable regardless of commercial pressure to relax them: Self-pay items must be clearly separated from, and never conflated or bundled with, insurance-covered items.

No claim may state or imply that the program reverses, cures, or prevents cancer, dementia, or Alzheimer's disease; permissible language is limited to improving modifiable risk factors and supporting risk management. Nutrition and lifestyle interventions must be delivered as professional clinical services, never as a channel for selling unregulated supplements or products of unverified origin. The program must supplement, not substitute for, standard-of-care treatment in endocrinology, cardiology, hepatology, neurology, and oncology. High-end or novel testing (including any neuro-metabolic biomarker) must be offered only with a documented clinical indication, a clear statement of its limitations, and an informed-consent process, never as an undifferentiated up-sell. New chargeable items must complete internal ethics, medical-affairs, and finance review and local price-registration or price-disclosure procedures before launch.

9. Phased Implementation Roadmap

 1. Pilot 6–12 months Build a minimum viable service loop. Launch a metabolic-assessment outpatient clinic within endocrinology or the health-management center; deploy basic and insulin-resistance panels; train staff; establish informed-consent and reporting templates. 

2. System-building 12–24 months. Finalize test bundles and intervention pathways; integrate nutrition, exercise, sleep, and psychological support; link with the physical-examination center; begin real-world data collection; pilot the neuro-metabolic tier under strict protocol; explore commercial-insurance partnerships. Patient retention; improvement in metabolic indicators; goal-attainment rates for weight, waist circumference, lipids, and glucose; repeat-visit rate are included. 3. Scale-up 24–36 months. Establish an independent, replicable brand, standardize training for physicians, dietitians, and health managers; connect with community health centers and medical consortia; explore corporate wellness and insurance-linked membership models; publish real-world evidence. 

10. Economic Outlook: Scenario-Based, Not Promised

 The original strategy document illustrated potential revenue using a single deterministic projection for a 500-bed tertiary hospital, growing from roughly 22 million yuan in year two to 240 million yuan by year five. That figure is a model output under a specific set of assumptions, not a commitment, and should not be presented to hospital leadership, investors, or regulators as an expected outcome. A more defensible approach uses a three-scenario range, with each scenario tied explicitly to the phase of program maturity rather than to a calendar year, and with revenue always presented net of staffing, reagent and consumable costs, equipment depreciation, facility costs, information-system costs, marketing, and quality-control expenditure — since gross self-pay billing is not equivalent to net contribution margin, and treating it as such is the single most common error in hospital new-service-line planning.

Conservative scenario (pilot phase): limited annual patient volume, basic-tier testing predominant, modest repeat-visit rate. Base-case scenario (system-building phase): stable conversion from the physical-examination center, an established multidisciplinary pathway, and improving repeat-visit and retention rates. Optimistic scenario (scale-up phase): a mature brand, commercial-insurance and corporate-wellness integration, and a growing base of long-term self-pay members. An annual review mechanism should be built in from the outset, both to prevent premature over-expansion on the basis of an early, favorable quarter and to allow early termination or restructuring of the neuro-metabolic tier specifically if the regulatory or evidentiary basis for a given biomarker test changes, which in this fast-moving field is a realistic possibility over a three-year planning horizon.

 11. Conclusion

 Hospitals are navigating a structural transition from volume-based to value-based operation, and the traditional levers of revenue growth — more testing, more medication, more consumables, more beds — are being deliberately compressed by payment reform. Metabolic disease, and insulin resistance specifically as its common upstream mechanism, offers hospitals a genuine, evidence-grounded pathway toward a new, largely self-pay service line, because the affected population is large, the disease burden is long-term, the intervention pathway is standardizable across a multidisciplinary team, and the revenue model does not depend primarily on insurance reimbursement.

This review has extended that case in one specific direction: the same mechanistic and epidemiological logic that connects insulin resistance to diabetes, fatty liver disease, and a subset of cancers also connects it, with substantial and rapidly maturing evidentiary support, to Alzheimer's disease and cognitive decline. The arrival of validated, blood-based Alzheimer's biomarkers such as p-tau217 since 2024 makes it newly feasible, for the first time, to incorporate a defensible neuro-metabolic risk dimension into a hospital's existing metabolic-assessment infrastructure, using the same population, the same multidisciplinary delivery model, and largely the same lifestyle interventions already required for metabolic risk reduction. Whether this model succeeds will not be determined by how compellingly it is packaged, but by three things: whether its central claims are anchored in verifiable, correctly cited evidence rather than in numbers that cannot be traced to a primary source; whether its clinical pathway is standardized, reproducible, and subject to ongoing quality control; and whether its pricing, marketing, and testing practices are conservative enough to withstand regulatory scrutiny as this field — particularly the Alzheimer's biomarker literature — continues to move quickly. Built on those three foundations, a Metabolic and Neuro-Metabolic Function Assessment Center is a credible answer to the question the hospitals are now being forced to ask: not how to do more, but how to do something that is genuinely worth doing, for a population large enough and underserved enough to sustain it.

Note: This article was completed by Dr. Jin Haining and Dr. Xu Bin with AI assistance. Due to the limitations of our professional knowledge, there must be inaccuracies in this paper. Please feel free to point them out and give your corrections. Thank you!

Ref.:

1. Chang J,  Facilitators and barriers to the implementation of DIP payment methodology reform in a public hospital in Guangzhou: a qualitative study based on the CFIR framework. Front Public Health. 2025;13:1569855. 

2. Caixin Global. In depth: China's public hospitals face debt crisis amid health care reforms. Caixin Global, May 9, 2025.

3.Rooney MR, Global and regional prediabetes prevalence: updates for 2024 and projections for 2050. Diabetes Care. 2025;48(11):e142–e144. doi:10.2337/dc25-1640.

支持单位：