The Thymus Is Not Dead — It Predicts How Long You Live

For decades, the thymus was dismissed as a vestigial organ in adults. This Nature 2026 study proves otherwise: thymic health, quantified by deep learning from routine chest CT, independently predicts all-cause mortality, cancer, and cardiovascular death across two large prospective cohorts.

27,612
Total participants
(NLST + FHS)
0.49
Hazard ratio
High vs Low (NLST)
12 yr
Maximum
follow-up
51%
Mortality reduction
High vs Low thymic health
−36%
Lung cancer
incidence reduction
−48%
Lung cancer
mortality reduction
The Thymic Health Paradigm Shift OLD PARADIGM Thymus = "done" after childhood T cell repertoire maintained peripherally Involution = normal, irrelevant to health NEW PARADIGM Thymus = active longevity regulator Thymic health predicts mortality Involution rate = modifiable, individual THYMECTOMY EVIDENCE Kooshesh et al. NEJM 2023 Thymectomy → ↑ cancer risk ↑ autoimmune disease ↑ all-cause mortality DEEP LEARNING QUANT Bernatz et al. Nature 2026 CT-based thymic health score n=27,612 prospective cohorts HR 0.49 (high vs low) CLINICAL IMPLICATIONS Risk stratification from CT Thymic regeneration therapies Lifestyle intervention targets Preventive immunoaging REGENERATIVE Growth hormone + DHEA Thymic transplant FOXN1 gene therapy KGF/IL-7 cytokines

Fig. 1 — The paradigm shift: from dismissing the adult thymus to recognizing it as a central immune-aging regulator and longevity predictor.

🫀
Immune Aging Clock
The thymus acts as a biological clock for immune system aging. As it involutes, T cell diversity shrinks, cancer surveillance weakens, and systemic inflammation rises — all key drivers of age-related disease.
🧠
Deep Learning Pipeline
A self-supervised CNN automatically localizes the thymus bed on CT, then quantifies compositional characteristics (functional tissue vs adipose) to output a continuous thymic health score — no manual annotation needed.
🎯
Modifiable Target
Thymic health is linked to modifiable factors — smoking cessation, weight management, and physical activity can preserve thymic function. This opens the door to lifestyle and pharmacological interventions targeting immune aging.

The Thymus — Your Forgotten Longevity Organ

Located behind the sternum, the thymus trains T cells — the adaptive immune system's soldiers. But unlike other organs, it actively self-destructs with age in a process called involution.

Thymic Anatomy & Involution YOUNG THYMUS (Active) Cortex Medulla ~40-50g at puberty Dense with thymocytes Active T cell production Involution ~3% / year AGED THYMUS (Involuted) residual Fat ~5-15g in elderly Mostly adipose tissue Minimal T cell output Bone Marrow Progenitors Thymus Selection & maturation Naïve T cells CD4+ / CD8+ (RTE) Periphery Immune surveillance Defense Cancer, infection, etc.

Fig. 2 — Thymic anatomy, involution process, and T cell maturation pipeline. The thymus shrinks ~3% per year, replacing functional tissue with fat.

🔬
What the Thymus Does

Positive selection: Thymocytes that bind self-MHC survive (cortex).

Negative selection: Thymocytes that react too strongly to self-antigens are eliminated (medulla).

Output: Naïve CD4+ helper T cells and CD8+ cytotoxic T cells with diverse TCR repertoire.

Recent Thymic Emigrants (RTEs): Fresh T cells with TREC markers — measurable indicators of ongoing thymic output even in adults.

⏳
Involution Consequences

Shrinking TCR diversity: Reduced ability to recognize novel pathogens and neoantigens.

Oligoclonal expansion: Memory T cells expand to fill the gap, but with limited diversity.

Immunosenescence: Accumulation of senescent T cells, chronic low-grade inflammation.

Cancer immune evasion: Weakened immune surveillance allows nascent tumors to escape detection.

Thymic Mass & T Cell Output Across the Lifespan

Study Design & Deep Learning Framework

A self-supervised deep learning system trained on 5,674 CT scans, validated on 27,612 independent scans from two landmark prospective cohorts.

Deep Learning Pipeline for Thymic Health Quantification 1. CT SCAN INPUT Routine chest CT Low-dose or standard Non-contrast or gated NLST / FHS / ECG-gated 2. AUTO-LOCALIZE Self-supervised learning Centre-of-mass detection Thymus bed extraction CNN + CoM algorithm 3. QUANTIFY Compositional analysis Functional tissue vs fat Radiographic density Continuous score output 4. RISK STRATIFY Low (bottom 25%) Average (middle 50%) High (top 25%) Cox PH + KM analysis NLST — National Lung Screening Trial n = 25,031 Ages 55–74 · Smokers/former · 12yr follow-up FHS — Framingham Heart Study n = 2,581 Community-based · ECG-gated CT · Independent validation

Fig. 3 — Deep learning pipeline: automatic thymus localization → compositional quantification → risk stratification, applied to two independent cohorts.

📊
NLST Cohort

25,031 participants from the National Lung Screening Trial

Ages 55–74, current or former heavy smokers (≥30 pack-years)

Low-dose chest CT at baseline

12-year follow-up for mortality, cancer incidence

Primary analysis cohort with extensive clinical covariates

🏥
FHS Cohort

2,581 participants from the Framingham Heart Study

Community-based, broader age range, not restricted to smokers

ECG-gated cardiac CT (different imaging protocol)

Independent validation cohort

Confirmed cardiovascular mortality associations with similar effect size

Key methodological strength: The DL model was developed on an independent dataset of 5,674 CT scans, then applied to both cohorts — ensuring no data leakage. Results were consistent across cohorts despite different imaging protocols (low-dose NLST vs ECG-gated FHS).

Thymic Health Distribution by Category

Thymic Health & Mortality Risk

Higher thymic health = dramatically lower mortality. The association persists after adjustment for age, sex, smoking, BMI, COPD, and major comorbidities.

13.4%
12-yr mortality
HIGH thymic health
19.2%
12-yr mortality
AVERAGE thymic health
25.5%
12-yr mortality
LOW thymic health
HR 0.49
High vs Low
(95% CI: 0.45–0.53)
Simulated Kaplan-Meier Survival by Thymic Health (NLST)
High (top 25%)
Average (middle 50%)
Low (bottom 25%)
Hazard Ratios — All-Cause Mortality
Hazard Ratios Across Analysis Models (High vs Low Thymic Health)
Analysis Model Cohort HR (95% CI) Type III P Adjustments
Unadjusted NLST 0.49 (0.45–0.53) <0.001 None
Smoking-adjusted NLST 0.55 (0.50–0.60) <0.001 Pack-years, smoking status, sex, age bins
Age as timescale NLST 0.57 (0.52–0.62) <0.001 Age (timescale), smoking, sex
Full multivariate NLST 0.60 (0.54–0.66) <0.001 Age, sex, BMI, smoking, COPD, comorbidities
Healthy subgroup NLST 0.58 (0.51–0.66) <0.001 Excluded cancer, asthma, diabetes, etc.
Unadjusted FHS 0.24 (0.16–0.38) <0.001 None
Multivariate FHS 0.55 (0.30–1.01) 0.254 Age, sex, smoking (smaller cohort)

FHS validation: Despite different imaging (ECG-gated cardiac CT), FHS showed even stronger unadjusted effect (HR 0.24, mortality 3.9% vs 14.5%). The multivariate model didn't reach significance due to smaller sample size (n=2,581), but effect direction and magnitude were consistent.

Disease Associations

Thymic health predicts cancer incidence, cancer-specific mortality, and cardiovascular death — linking immune system fitness to the major killers of aging.

🫁
Lung Cancer
Incidence at 6yr (High)3.4%
Incidence at 6yr (Low)5.3%

HR 0.64 (0.53–0.76) for incidence
HR 0.52 (0.44–0.63) for mortality
Preserved in current and former smokers

❤️
Cardiovascular Disease
CV mortality (High)2.8%
CV mortality (Low)5.6%

Consistent across both NLST and FHS cohorts. FHS independently validated CV mortality association after adjusting for age, sex, smoking.

🔥
Systemic Inflammation
Inflammation (High TH)Low
Inflammation (Low TH)High

Low thymic health linked to systemic inflammation and metabolic dysregulation — the hallmarks of immunosenescence and inflammaging.

Thymic Involution → Disease Cascade Thymic Involution Functional tissue → fat ↓ naïve T cell output ↓ TCR Diversity Repertoire contraction Oligoclonal expansion Immunosenescence Senescent T cells ↑ SASP + inflammaging ↓ Surveillance Cancer immune evasion Infection susceptibility Cancer HR 0.64 Lung, pan-cancer CVD Death Significant Both cohorts Metabolic Dysregulation BMI, inflammation All-Cause HR 0.49 12-year follow-up

Fig. 4 — Causal cascade from thymic involution to disease outcomes: reduced T cell diversity → immunosenescence → weakened surveillance → cancer, CVD, and mortality.

Disease-Specific Hazard Ratios (High vs Low Thymic Health)
12-Year Mortality by Outcome Type

Modifiable Factors & Thymic Health

Thymic health is not purely genetic — smoking, obesity, and physical activity significantly influence thymic function. This makes the thymus a modifiable target for healthy aging.

🚬
Smoking

Active smokers had significantly lower thymic health than former smokers.

Pack-years correlation: Higher cumulative smoking → lower thymic health score.

Smoking accelerates thymic involution through oxidative stress, direct toxicity to thymocytes, and chronic inflammation.

⚖️
Obesity / BMI

Higher BMI was significantly associated with lower thymic health.

Mechanism: Adipose tissue invasion of thymus, lipotoxicity to thymic epithelial cells.

Visceral fat produces inflammatory cytokines (TNF-α, IL-6) that accelerate thymic atrophy.

🏃
Physical Activity

Physical activity was associated with preserved thymic health.

Mechanism: Exercise reduces inflammation, improves metabolic health, and may directly stimulate thymic function via IL-7.

Endurance athletes in their 60s–70s show thymic output comparable to adults decades younger.

Thymic Health by Sex and Age Group (NLST)
Thymic Health by BMI Category

Sex differences: Females consistently had higher thymic health than males across all age groups (p < 2 × 10⁻¹⁶). This is consistent with known sex dimorphism in thymic biology — estrogen has protective effects on thymic epithelial cells, while testosterone accelerates involution.

Thymic Regeneration Strategies

Emerging interventions targeting thymic rejuvenation — from lifestyle to pharmacological to regenerative medicine.

💊
GH + DHEA + Metformin

TRIIM trial (Fahy 2019): growth hormone + DHEA + metformin reversed ~2.5 years of epigenetic aging and regenerated thymic tissue in 9 men.

Clinical trial
🧬
FOXN1 Gene Therapy

FOXN1 is the master regulator of thymic epithelium. Upregulation in aged mice regenerated thymic architecture and restored T cell output.

Preclinical
💉
IL-7 / KGF Cytokines

IL-7 drives thymocyte proliferation; KGF (FGF7) regenerates thymic epithelium. Both show thymic regeneration in post-transplant patients.

Phase I/II
🔬
Thymus Bioengineering

Decellularized thymus scaffolds + patient TECs could create functional thymic organoids for implantation — promising for DiGeorge syndrome and aging.

Early research

Thymic Health Estimator

Estimate your relative thymic health category based on known risk factors from the study. This is a simplified educational tool — actual thymic health requires CT-based DL quantification.

Risk Factor Inputs
Age55
SexMale
BMI25
Smoking (pack-years)15
Physical Activity (hrs/week)3
Chronic Inflammation LevelLow
62
Estimated Thymic Health Score (0–100)
Average
Risk Profile
Estimated Hazard Ratio
0.72 vs. low thymic health reference

References & Evidence

Peer-reviewed evidence underpinning the thymic health longevity paradigm.

  1. Bernatz, S., Prudente, V., Pai, S. et al. Thymic health consequences in adults. Nature (2026). doi:10.1038/s41586-026-10242-y
  2. Kooshesh, K.A. et al. Health consequences of thymus removal in adults. N. Engl. J. Med. 389, 406–417 (2023).
  3. Palmer, D.B. The effect of age on thymic function. Front. Immunol. 4, 316 (2013).
  4. Goldrath, A.W. & Bevan, M.J. Selecting and maintaining a diverse T-cell repertoire. Nature 402, 255–262 (1999).
  5. Chinn, I.K. et al. Changes in primary lymphoid organs with aging. Semin. Immunol. 24, 309–320 (2012).
  6. Fahy, G.M. et al. Reversal of epigenetic aging and immunosenescent trends in humans. Aging Cell 18, e13028 (2019).
  7. Dunn-Walters, D.K. & Ademokun, A.A. B cell repertoire and ageing. Curr. Opin. Immunol. 22, 514–520 (2010).
  8. Bredenkamp, N. et al. Regeneration of the aged thymus by a single transcription factor. Development 141, 1627–1637 (2014).
  9. Rossi, S.W. et al. Keratinocyte growth factor (KGF) enhances postnatal T-cell development via enhancements in proliferation and function of thymic epithelial cells. Blood 109, 3803–3811 (2007).
  10. Duah, M. et al. Thymus degeneration and regeneration. Front. Immunol. 12, 706244 (2021).
  11. den Braber, I. et al. Maintenance of peripheral naive T cells is sustained by thymus output in mice but not humans. Immunity 36, 288–297 (2012).
  12. Krenger, W., Blazar, B.R. & Holländer, G.A. Thymic T-cell development in allogeneic stem cell transplantation. Blood 117, 6768–6776 (2011).
  13. Coder, B.D. et al. Thymic involution perturbs negative selection leading to autoreactive T cells that induce chronic inflammation. J. Immunol. 194, 5825–5837 (2015).
  14. Hale, J.S. et al. Thymic output in aged mice. Proc. Natl Acad. Sci. USA 103, 17048–17053 (2006).
  15. Sutherland, J.S. et al. Activation of thymic regeneration in mice and humans following androgen blockade. J. Immunol. 175, 2741–2753 (2005).
  16. Mackall, C.L. et al. Age, thymopoiesis, and CD4+ T-lymphocyte regeneration after intensive chemotherapy. N. Engl. J. Med. 332, 143–149 (1995).
  17. Weinstein, Y. et al. Thymus involution and immunity. Annu. Rev. Immunol. 33, 31–53 (2015).
  18. Sempowski, G.D. et al. Leukemia inhibitory factor, oncostatin M, IL-6, and stem cell factor mRNA expression in human thymus increases with age and is associated with thymic atrophy. J. Immunol. 164, 2180–2187 (2000).
  19. Dixit, V.D. Thymic fatness and approaches to enhance thymopoietic fitness in aging. Curr. Opin. Immunol. 22, 521–528 (2010).
  20. Nikolich-Zugich, J. The twilight of immunity: emerging concepts in aging of the immune system. Nat. Immunol. 19, 10–19 (2018).