Core Framework & Diagram What Are Immune Risk Scores?
July 23, 20261 Min Read What Is Immune Monitoring?
July 23, 2026Compressing your immune state into a number — then using that number to predict your future health risks
—— Immune Risk Scores: from IRP to iAGE, the science and clinical applications of quantified immune state assessment.
I. IRP: the earliest immune 'mortality warning' score
The immune risk score concept's earliest systematic research came from Sweden's OCTO study (1992) and NONA study (1997), led by Bernadette Johansson and Anders Wikby's teams. They performed detailed immunological examination on people aged eighty-five and above, then tracked their two-year and four-year survival rates. They discovered that certain specific immune indicator combinations had stronger predictive power for predicting elderly people's near-term mortality risk than most traditional clinical indicators. This combination was named 'Immune Risk Profile' (IRP): CD4/CD8 T cell ratio inversion (more CD8+ T cells than CD4+ T cells, when normally CD4 should outnumber CD8); elevated CD8+CD28- T cells (terminally differentiated senescent T cells); declining NK cell numbers; CMV (cytomegalovirus) seropositivity; and declining B cell numbers.
Elderly people meeting IRP criteria had significantly higher mortality during follow-up than same-age peers not meeting criteria — even after controlling for age, sex, and underlying conditions. This was the first powerful epidemiological evidence that immune system state can independently predict death risk. IRP's limitations: it mainly predicts short-term (two to four years) mortality risk, mainly applies to extreme old age (eighty-five and above), and its clinical application is limited by requiring flow cytometry to measure T cell subsets, not common in routine checkups.
|
IRP is immune risk score's cornerstone — for the first time systematically proving that 'several immune indicators combined' can predict elderly people's survival probability better than traditional clinical indicators. This opened the entire research direction of quantifying immune state as risk prediction tools. |
2. Inflammatory aging scores: turning 'inflammaging' into a measurable number
After IRP, with the establishment of the 'Inflammaging' concept, researchers began focusing on using comprehensive inflammatory marker levels to quantify a person's 'inflammatory aging burden.' The simplest method is directly measuring several key pro-inflammatory cytokines and using weighted combinations to calculate a comprehensive inflammation index. Several commonly used combinations: IL-6 + CRP (hsCRP) — the most widely used clinically accessible inflammation assessment combination. IL-6 is one of inflammaging's most core markers; hsCRP is the liver's product under IL-6 stimulation, more sensitive than regular CRP. Together, they have more predictive value than either alone. IL-6 + TNF-α + IL-1β — a more comprehensive pro-inflammatory factor combination, but higher testing cost, currently more used in research than routine clinical use. CMV combined with inflammation markers — combining CMV serostatus (positive/negative) with IL-6 levels; some research shows this combination's prediction of elderly people's functional status and mortality rate is better than using either indicator alone.
3. iAGE: using AI to extract 'immune age' from cytokine profiles
In 2021, Stanford University's research team (David Furman et al., published in Nature Aging) developed a new immune aging score — iAGE (Inflammatory Aging Clock). iAGE's development logic represents third-generation immune risk score thinking: not artificially selecting several indicators, but using machine learning to let data tell us which indicator combinations best predict deviation of 'immune age' from actual age. Researchers used: plasma cytokine profiles (concentrations of 100+ cytokines, chemokines, and growth factors) from over 1,000 healthy subjects of different ages; peripheral blood immune cell subset flow cytometry data; and some subjects' DNA methylation clock data (as a 'validation standard'). The ML model extracted from this data a single iAGE score accurately reflecting individual 'immune inflammatory aging degree.'
Key findings: iAGE highly correlates with actual age, but with large individual variation — same sixty-year-old people can have iAGE differing by twenty years; iAGE is a strong independent predictor of multiple chronic disease risks (diabetes, cardiovascular disease, cognitive decline), superior to using any single cytokine alone; CXCL9 (a chemokine mainly induced by IFN-γ, participating in T cell chemotaxis homing) was identified by AI as the single largest contributor to iAGE — an unexpected finding suggesting CXCL9 may be a key driving molecule in inflammaging; lifestyle interventions (exercise, sleep improvement, anti-inflammatory diet) can measurably reduce iAGE score, suggesting iAGE is sensitive to intervention responses, suitable as a tracking indicator of intervention effects.
|
iAGE is one of machine learning's most influential achievements in immune aging assessment: not a human-selected indicator combination, but letting data itself tell us what signals matter most. CXCL9 as the most important contributor's unexpected discovery shows AI's value in biological discovery — it can find key signals outside human prior assumptions. |
4. Commercial immune age testing: questions worth asking
As academic research like iAGE's influence expanded, some commercial health tech companies began productizing similar concepts, offering consumer-purchasable 'immune age testing' services. Commercial products' quality and scientific rigor vary greatly. When evaluating a commercial immune age test, worth asking: Has this score's predictive validity been prospectively validated in independent populations? What specific indicators does it integrate, and how strong is the scientific support for these indicators? How interpretable is the score — can it tell you which aspects have problems, not just a number? Does it provide evidence-based intervention recommendations, rather than coincidentally selling supplements? Currently, immune risk scoring tools with sufficient prospective validation are mainly in academic research (IRP, iAGE, etc.). Commercial products mostly have varying degrees of scientific rigor compromise, requiring critical evaluation.
5. Building your own 'simplified immune risk reference': what can be done now
Waiting for complete commercialized immune age testing isn't a necessary prerequisite. Based on blood tests currently accessible in many regions (including Malaysia), a meaningful 'simplified immune risk reference' can be built: hsCRP (high-sensitivity C-reactive protein) — target: below 1 mg/L. Most comprehensive health checks can include it or can be separately requested. The most practical proxy indicator for inflammaging. Blood count lymphocyte/white blood cell ratio — a simple 'immune reserve' reference indicator. Persistently low lymphocyte proportion suggests the immune system may be in a chronic stress state. NK cell activity (NKCA) — requires specialty or functional medicine institutions, relatively higher testing cost, but has the most direct predictive value for cancer surveillance and infection risk. 25-OH-D (vitamin D) — target: 50–100 nmol/L. A critical hormone affecting comprehensive immune function, deficiency is extremely common in Asian urban populations, one of the most cost-effective immune health assessment indicators.
The combination of these four indicators doesn't equal AI-comprehensive scores like iAGE, but among clinical accessibility, cost, and information value, provides the optimal balance that most people can currently achieve. They aren't just numbers, but repeatable indicators for tracking your immune system's trends — testing once a year, building trends, is the truly meaningful usage pattern.
|
Immune risk scores' most important value is not the absolute number of one test, but trend tracking. Seeing your hsCRP drop from 1.2 to 0.6 over three years — THAT is proof that lifestyle interventions are genuinely working — more reliable than any subjective feeling. |
Frequently Asked Questions