Fermat's Library | Thymic health consequences in adults annotated/explained version.

This landmark 2026 article argues that the thymus, long thought to ...

> "The thymus is a specialized immune organ responsible for maturin...

> "Indeed, given the role of the thymus in maintaining an adaptive ...

> "Our results demonstrate that thymic health varies between indivi...

The striking result here is not just that people with better thymic...

The authors quantify “thymic health” by training a deep learning mo...

What makes this result so striking is that it reframes the adult th...

Nature | www.nature.com | 1

Article

Thymic health consequences in adults

Simon Bernatz

1,2,3,4,18

, Vasco Prudente

1,2,3,18

, Suraj Pai

1,2,3,18

, Asbjørn K. Attermann

1,5,6,7

Yumeng Cao

, Jiachen Chen

, Asya Lyass

, Borek Foldyna

1,10

, Leonard Nürnberg

1,2,3

Keno Bressem

1,2,11,12

, Christopher Abbosh

, Charles Swanton

11,12,13

, Mariam Jamal-Hanjani

11,13,14

Michael T. Lu

1,10

, Joanne M. Murabito

15,16,17

, Kathryn L. Lunetta

8,15,19

, Nicolai J. Birkbak

5,6,7,19

Hugo J. W. L. Aerts

1,2,3,10,19

✉

The thymus is essential for establishing Tcell diversity early in life, but undergoes

profound involution with age and has therefore traditionally been regarded as largely

nonfunctional in adults

1,2

. Here we propose that preserving thymic functionality is

integral to adult health and longevity. We developed a deep learning framework to

quantify thymic health from routine radiographic images and evaluated its association

with longevity and risk of major age-associated diseases in two large prospective

cohorts of asymptomatic adults: the National Lung Screening Trial (n = 25,031) and

the Framingham Heart Study (n = 2,581). In both cohorts, thymic health varied markedly

across the population. In the National Lung Screening Trial, higher thymic health was

consistently associated with lower all-cause mortality, reduced lung cancer incidence

and lower cardiovascular mortality over 12 years of follow-up after adjustment for age,

sex, smoking and comorbidities. In the independent Framingham Heart Study cohort,

higher thymic health was signicantly associated with reduced cardiovascular

mortality, independent of age, sex and smoking. Thymic health was further linked to

systemic inammation and metabolic dysregulation, and associated with modiable

lifestyle factors including smoking, obesity and physical activity. Together, these

ndings reposition the thymus as a central regulator of immune-mediated ageing and

disease susceptibility in adulthood, highlighting its potential as a target for preventive

and regenerative strategies to promote healthy ageing and longevity.

The thymus is a specialized immune organ responsible for maturing

Tcells, thereby producing a diverse Tcell repertoire crucial for mount-

ing an adaptive immune response

1,2

. The thymus itself decays with age

and eventually transforms entirely into adipose tissue through a pro-

cess known as thymic involution

. While the absence of a functioning

thymus in children is associated with profound immunodeficiency

the consequences of thymic decay in adulthood are more subtle

5,6

Indeed, it was long believed that once the thymus generates a suffi-

ciently diverse Tcell repertoire in childhood, the Tcell repertoire could

be peripherally maintained to support an adaptive immune response

against a diverse array of pathogens

2,7

. For this reason, the thymus has

long been considered largely nonfunctional in adults.

However, a growing body of evidence challenges this notion

2,6,8–17

. In

a recent landmark study by Kooshesh etal.

investigating the impact

of thymectomy on long-term health, the authors found that adults

who had their thymus removed experienced adverse health conse-

quences across multiple diseases and outcomes, where penetrance can

be decades after thymectomy

. While the consequences of thymectomy

are impactful, only a small fraction of the population is exposed to this

procedure, whereas individual and lifestyle-dependent differences in

thymic decay affect everyone.

Although the impact of thymic function is increasingly recognized in

ageing and across a wide range of clinical settings, thymic decay across

the population has been incompletely explored

2,13

. However, there is

increasing evidence that the rate of thymic involution varies among

individuals

18–20

. Indeed, given the role of the thymus in maintaining

an adaptive immune response, the individualized rate of thymic decay

may be a major driver of age-associated diseases, such as cardiovascular

disease and cancer

In this study, we investigated the impact of thymic functionality,

here called thymic health, in adults. For this purpose, we analysed

two prospectively collected cohorts of 27,612 individuals enrolled

in the Framingham Heart Study (FHS) and National Lung Screening

Trial (NLST). We developed a deep learning system to automatically

https://doi.org/10.1038/s41586-026-10242-y

Received: 13 January 2025

Accepted: 5 February 2026

Published online: xx xx xxxx

Open access

Check for updates

Artiicial Intelligence in Medicine (AIM) Program, Mass General Brigham & Harvard Medical School, Boston, MA, USA.

Department of Radiation Oncology, Brigham and Women’s Hospital and

Dana-Farber Cancer Institute, Harvard Medical School, Boston, MA, USA.

Radiology and Nuclear Medicine, GROW & CARIM, Maastricht University, Maastricht, The Netherlands.

Department

of Radiology and Nuclear Medicine, Goethe University, Frankfurt am Main, Germany.

Department of Molecular Medicine, Aarhus University Hospital, Aarhus, Denmark.

Department of Clinical

Medicine, Aarhus University, Aarhus, Denmark.

Bioinformatics Research Center, Aarhus University, Aarhus, Denmark.

Department of Biostatistics, Boston University School of Public Health,

Boston, MA, USA.

Department of Mathematics and Statistics, Boston University, Boston, MA, USA.

Cardiovascular Imaging Research Center, Massachusetts General Hospital, Harvard Medical

School, Boston, MA, USA.

Cancer Research UK Lung Cancer Centre of Excellence, University College London Cancer Institute, London, UK.

Cancer Evolution and Genome Instability Laboratory,

The Francis Crick Institute, London, UK.

Department of Medical Oncology, University College London Hospitals NHS Foundation Trust, London, UK.

Cancer Metastasis Laboratory, University

College London Cancer Institute, London, UK.

Framingham Heart Study, National Heart, Lung and Blood Institute and Boston University Chobanian and Avedisian School of Medicine, Framingham,

MA, USA.

Department of Medicine, Boston University Chobanian and Avedisian School of Medicine, Boston, MA, USA.

Section of General Internal Medicine, Boston Medical Center, Boston,

MA, USA.

These authors contributed equally: Simon Bernatz, Vasco Prudente, Suraj Pai.

These authors jointly supervised this work: Kathryn L. Lunetta, Nicolai J. Birkbak, Hugo J. W. L. Aerts.

✉

e-mail: haerts@bwh.harvard.edu

2 | Nature | www.nature.com

Article

quantify thymic health on computed tomography (CT) scans. Our

results demonstrate that thymic health varies between individuals

and is impacted by sex, age and lifestyle habits. Notably, we show that

individuals with low thymic health, that is, lost thymic functionality,

have a shorter lifespan and an increased risk of cancer and cardiovas-

cular diseases. These findings strongly suggest that thymic health is

crucial for long-term health and lifespan.

Quantification of thymic health

For quantification of thymic health, we developed a deep learning

system using an independent dataset of 5,674 individuals to determine

compositional radiographic characteristics of the thymus as a proxy for

its functionality (Fig.1, Methods and Supplementary Fig.9). The system

takes a CT scan as input and provides the automatic continuous thymic

health estimate as output. We applied the system to prospectively col-

lected data from a total of 27,612 individuals from two cohorts, including

2,581 participants in the FHS and 25,031 participants in the NLST (Fig.1).

Clinical characteristics are shown in Extended Data Table1. For outcome

analyses, participants were categorized as low, average or high thymic

health based on the bottom 25%, middle 50% and top 25% of the popula-

tion. These thresholds were supported by cut point iterations (Extended

Data Fig.1). Individuals with low thymic health are assumed to have lost

most of their thymic functionality, while individuals with average to

high thymic health have preserved their thymic functionality through-

out ageing to an increasing degree. As expected

21,22

, thymic health was

higher in female than male participantsand significantly declined

with age (Fig.2a,b, Extended Data Fig.2a,b and Supplementary Fig.2).

Furthermore, thymic health was lower in individuals with higher body

mass index (BMI) (Fig.2b, Extended Data Fig.2b and Supplementary

Fig.2).

Thymic health and risk of death

To investigate the associations of thymic health with clinical outcomes,

we assessed all-cause mortality throughout a 12-year follow-up period

using Kaplan–Meier and Cox proportional hazards analysis. For the

NLST, participants with higher thymic health showed lower mortal-

ity than those with low thymic health (high versus low thymic health:

Kaplan–Meier mortality estimate at 12 years 13.4% versus 25.5%; hazard

ratio (HR) 0.49; 95% confidence interval (CI) 0.45–0.53; Fig.2c). The

association was preserved in a Cox analysis adjusted for smoking sta-

tus and pack-years, and stratified by sex and 5-year age bins (Fig.2d).

To verify that these results were not confounded by age, we repeated

the analysis using narrower 3-year age bins (Supplementary Fig.3a)

and, in a separate model, used age as the time scale to implicitly adjust

for the effect of ageing (Fig.2e). To further evaluate the potential

impact of comorbidities, we extensively adjusted for clinical variables

and known diseases in a Cox model (Fig.2f), and further repeated

the analysis in a healthier sub-cohort where individuals with prior

occurrence of cancer or major comorbidities were excluded (Fig.2g).

Throughout all analyses, thymic health remained significantly associ-

ated with outcomes after controlling for the potential confounding

effects of either age or comorbidities (all type III P < 0.001), firmly

indicating a prognostic value of thymic health beyond clinical vari-

ables. Finally, excluding thymic health from multivariate models did

Training data

for pipeline development

Evaluation data

to analyse clinical

impact

FHS

n = 5,674

n = 27,612

2,581

25,031

ECG-gated

thoracic CT scans

Outcome:

Cardiovascular events

Death

NLST

Outcome:

Lung cancer incidence

Death

Non-ECG-gated

thoracic CT scans

Low

High Average

Low

Thymus

location (3D)

Thymus centre

[x | y | z]

Cropping

Thymus CoM

Thymus localizer

CNN model

Feature extraction

SSL foundation model

Prediction

Thymic health score

Output

0–100

Feature vector

4,096 features

Fig. 1 | Overview of study design. a, Illustration of thymic health, that is, an

imaging-based proxy of thymic functionality and three representative examples

of individuals with high, average or low thymic health. The thymus bed is

outlined in orange. b, The model was developed on 5,674 CT scans and validated

on 27,612 independent CT scans from the FHS and NLST. c, Illustration of the

deep learning pipeline, which takes a CT scan as input and outputs a continuous

quantification of thymic health after automatically localizing and quantifying

the thymus on the basis of self-supervised learning(SSL).CNN, convolutional

neural network; CoM, centre of mass;ECG, electrocardiogram. Illustrations

in a and c created in BioRender; Birkbak, N. https://biorender.com/bd3dmmr

(2026).

Nature | www.nature.com | 3

not markedly alter the HR estimates of the remaining covariates,

indicating that thymic health provides independent and potentially

complementary prognostic information (Extended Data Fig.3a,e).

These findings were supported in the FHS with similar effect size and

direction (high versus low thymic health: 3.9% versus 14.5% Kaplan–

Meier mortality estimate at 12 years; HR 0.24; 95% CI 0.16–0.38;

average versus low thymic health: 9.6% versus 14.5%; HR 0.63; 95%

CI 0.48–0.83; Extended Data Fig.2c) (type III P < 0.001), although

statistical significance was not reached after multivariate modelling

in this smaller cohort (Extended Data Fig.2d) (type III P = 0.254). Con-

tinuous and detailed cutoff analyses testing a wide range of thresh-

olds confirmed these results, showing gradually improving survival

HR average: 0.69 (95% CI 0.65−0.74)

HR high: 0.49 (95% CI 0.45−0.53)

n = 25,031

100

0 3 6 9 12

Years since randomization

Survival (%)

Age (years)

Thymic health (%)

BMISex

6,259 6,157 5,976 5,454 4,612

12,515 12,220 11,717 10,399 8,567

6,257 6,032 5,651 4,940 3,919Low

Average

High

At risk

100

0 3 6 9 12

Thymic

health (%)

Sex

Age

(years)

Smoking

BMI

NLST (n = 25,031)

*** *** *** *** *** *** *** ***

High

Average

Low

Thymic health

Low

Average

High

Sex

Male

Female

Age

Older median

Younger median

Smoking status

Non-smoker

Current smoker

Thymic health

Low

Average

High

Number

6,257

12,515

6,259

0.50 0.75 1.00 1.50

HR HR

Reference

0.76 (0.71; 0.81)

0.67 (0.61; 0.73)

95% CI

100

= 1.94 × 10

–22

P = 2.76 × 10

–64

BMI

Underweight

Normal

Overweight

Obese

Variable

Thymic health average

Thymic health low

Thymic health high

BMI

Current smoker

Pack-years

Diabetes

Hypertension

Adult asthma

Asbestosis

Bronchiectasis

Childhood asthma

Chronic bronchitis

COPD

Emphysema

Fibrosis of the lung

Heart disease

Pneumonia

Sarcoidosis

Silicosis

Stroke

Tuberculosis

0.5 1 2

HR HR P value

Reference

1.47 × 10

–09

0.83 (0.78; 0.89)

0.77 (0.70; 0.84)

0.99 (0.98; 1.00)

1.89 (1.78; 2.01)

1.01 (1.01; 1.01)

1.67 (1.53; 1.82)

1.22 (1.14; 1.30)

1.11 (0.99; 1.26)

1.19 (0.94; 1.50)

1.21 (1.04; 1.40)

1.07 (0.91; 1.26)

0.97 (0.88; 1.07)

1.45 (1.30; 1.63)

1.64 (1.49; 1.79)

1.34 (0.84; 2.13)

1.39 (1.29; 1.50)

1.13 (1.05; 1.21)

0.84 (0.37; 1.88)

1.03 (0.48; 2.21)

1.42 (1.24; 1.63)

1.12 (0.85; 1.47)

95% CI

Variable

Thymic health average

Thymic health low

Thymic health high

BMI

Current smoker

Pack-years

0.5 1 2

HR HR

P value

Reference

2.19 × 10

–11

0.2179

2.19 × 10

–65

0.77 (0.71; 0.84)

0.71 (0.64; 0.80)

0.99 (0.99; 1.00)

1.93 (1.79; 2.08)

1.01 (1.01; 1.01)

2.37 × 10

–37

95% CI

0.5 1 2

0.77 (0.72; 0.83)

0.71 (0.65; 0.78)

1.89 (1.78; 2.01)

1.01 (1.01; 1.01)

Variable

Thymic health average

Thymic health low

Thymic health high

Current smoker

Pack-years

HR HR P value

Reference

1.40 × 10

–17

8.59 × 10

–98

1.10 × 10

–46

95% CI

Male

Female

50–59

60–64

65–69

70–75

Under-

weight

Over-

weight

Obese

Normal

100

0.0103

1.25 × 10

–90

5.15 × 10

–34

9.19 × 10

–33

5.39 × 10

–10

0.0835

0.1469

0.0121

0.4041

0.5626

1.10 × 10

–10

5.33 × 10

–26

0.2190

1.62 × 10

–17

0.0007

0.6726

0.9403

3.10 × 10

–7

0.4222

Fig. 2 | Association of thymic health with long-term mortality. a, Overview

of the NLST. Data are sorted by ascending thymic health where each column

represents one patient. Thymic health is categorized into low, average and high

based on the bottom 25% (blue), middle 50% (orange) and top 25% (red) of the

population. The fractional left split of the NLST represents patients who were

defined as having low thymic health through automatic quality control.

b, Associations between thymic health and sex (n = 25,031) in the NLST,

across age groups in years (n = 25,031) and body mass categories (n = 24,948,

missingness n = 83). ***P < 2 × 10

−16

; NS, notsignificant with P = 0.2505. c, Kaplan–

Meier plots for overall survival outcomes in the NLST across thymic health

categories. The inset in the plot shows the same data on an expanded y axis.

Unadjusted HRs are shown onthe bottom left. d, HR for thymic health categories

adjusted for pack-years and smoking status and stratified by sex and age

binned at 5 years. e–g, HRs of all-cause death for participants in the NLST,

using continuous age as the time scale to account for potential residual

confounding by age (n = 25,027; missingness n = 4) adjusted for pack-years

and smoking status and stratified by sex (e); adjusted for the shown clinical

and epidemiological covariates and stratified by sex and age binned at 5 years

(n = 24,597; missingness n = 434) (f); adjusted for the shown clinical and

epidemiological covariates and stratified by sex and age binned at 5 years

in the subgroup of the NLST in which participants with a history of cancer

(n = 575), childhood or adult asthma, diabetes, asbestosis, bronchiectasis, lung

fibrosis, sarcoidosis, silicosis or tuberculosis (n = 5,500) were excluded from

the analysis (n = 18,565; missingness n = 54) (g).COPD, chronic obstructive

pulmonary disease.

4 | Nature | www.nature.com

Article

independent of sex and age with increasing thymic health (Extended

Data Fig.1).

Lung cancer incidence and mortality

We investigated the associations of thymic health with lung cancer

mortality and lung cancer incidence in the NLST, the main endpoints

of the trial. Participants with high and average thymic health were

less likely to develop lung cancer than participants with low thymic

health (high versus low thymic health: 3.4% versus 5.3% Kaplan–Meier

incidence estimate at 6 years; HR 0.64; 95% CI 0.53–0.76; average ver-

sus low thymic health: 4.1% versus 5.3%; HR 0.78; 95% CI 0.67–0.90;

Fig.3a). These associations were preserved in sex and age-stratified

analyses with adjustments for smoking status and pack-years (type III

P = 0.036) (Fig.3b). Again, these associations were consistent across

further analyses, extensively controlled for the confounding effects of

age and comorbidities (Extended Data Figs.3b,f, 4a,d and 5a and Sup-

plementary Fig.3b) (all type III P < 0.05). Furthermore, participants with

higher thymic health had lower lung cancer mortality risks relative to

participants with low thymic health (high versus low thymic health: 1.1%

versus 2.0% Kaplan–Meier mortality estimate at 6 years; HR 0.52; 95%

CI 0.44–0.63; average versus low thymic health: 1.5% versus 2.0%; HR

0.70; 95% CI 0.61–0.80;HR reported through 12 years; Fig.3c). Again,

these associations were preserved across the respective multivariate

models (all type III P < 0.001) (Fig.3d, Extended Data Figs.3c,g, 4b,e

and 5b and Supplementary Fig.3c). We further stratified the cohort into

current and former smokers. For lung cancer incidence, the associa-

tions with thymic health showed similar trends in both groups but did

not reach statistical significance (current smokers type III P = 0.051;

former smokers type III P = 0.25). For lung cancer mortality, the asso-

ciations were preserved in both current (type III P = 0.009) and former

smokers (type III P = 0.003) (Supplementary Fig.4). While the NLST

was designed to screen for lung cancer, pan-cancer mortality was also

recorded (Supplementary Table1). Again, we observed that participants

with higher thymic health had lower cancer-specific mortality risks rela-

tive to participants with low thymic health, which were also preserved

in multivariate models (all type III P < 0.02) (Extended Data Figs.5c

and 6a,b and Supplementary Fig.3d).

Cardiovascular mortality and incidence

We investigated the associations of thymic health with cardiovascular-

specific mortality in the NLST, where participants with high and aver-

age thymic health had lower cardiovascular-specific mortality risks

than participants with low thymic health (high versus low thymic

health: 2.9% versus 7.5% cardiovascular-specific mortality estimates

at 12 years; HR 0.37; 95% CI 0.31–0.44; average versus low thymic

health: 4.4% versus 7.5%; HR 0.57; 95% CI 0.50–0.65) (Fig.4a and Sup-

plementary Table2). Again, these associations were preserved in

sex and age-stratified analyses with adjustments for smoking status

and pack-years (Fig.4b) and also consistent across further sensitiv-

ity analyses, extensively controlled for the confounding effects of

age and comorbidities (Extended Data Figs.3d,h, 4c,f and 5d and

Supplementary Fig.3e) (all type III P < 0.001). Similar results were

obtained in the independent FHS. Participants with higher thymic

health had lower cardiovascular-specific mortality relative to those

with low thymic health (high versus low thymic health: 0.3% versus

3.9% cardiovascular-specific mortality estimate at 12 years; HR 0.08;

95% CI 0.02–0.34, average versus low thymic health: 1.5% versus 3.9%;

HR 0.38; 95% CI 0.20–0.70; Fig.4c). These associations also were pre-

served in multivariate models (type III P = 0.021) (Fig.4d). Further,

participants with higher thymic health had lower cumulative inci-

dence of cardiovascular-specific diseases relative to participants

with low thymic health (high versus low thymic health: 5% versus

16.7% Kaplan–Meier estimate for cardiovascular disease at 12 years;

HR 0.26; 95% CI 0.16–0.40, average versus low thymic health: 10.8%

versus 16.7%; HR 0.65; 95% CI 0.49–0.87; Fig.4e). However, these

latter associations were partly attenuated in sex and age-stratified

HR average: 0.78 (95% CI 0.67–0.90)

HR high: 0.64 (95% CI 0.53–0.76)

n = 23,164

0 1 2 3 4 5 6

Years since randomization

Cumulative lung cancer incidence (%)

5,792 5,748 5,711 5,692 5,652 5,582

11,502 11,407 11,262 11,178 11,102 10,926

5,866

11,633

5,665 5,566 5,504 5,437 5,379 5,307 5,201Low

Average

High

At risk

HR average: 0.70 (95% CI 0.61–0.80)

HR high: 0.52 (95% CI 0.44–0.63)

n = 25,029

0 3 6 9 12

Years since randomization

Cumulative lung cancer mortality (%)

6,258 6,156 5,976 5,454 4,612

12,515 12,220 11,717 10,399 8,567

6,256 6,031 5,650 4,939 3,919

Low

Average

High

At risk

0 3 6 9

Thymic health

Low

Average

High

Number

5,665

11,633

5,866

0.50 0.75 1.00 1.50

HR HR

Reference

0.85

0.80

95% CI

(0.74; 0.98)

(0.66; 0.97)

Thymic health

Low

Average

High

Number

6,256

12,515

6,258

0.50 0.75 1.00 1.50

HR HR

Reference

0.76

0.69

95% CI

(0.66; 0.88)

(0.57; 0.83)

High

Average

Low

b d

P = 3.05 × 10

–6

P = 0.0360

P = 1.61 × 10

–12

P = 5.00 × 10

–5

0 1 2 3 4 5 6

Fig. 3 | Association of thymic health with long-term risk of lung cancer and

lung cancer-specific mortality. a, Percentage of individuals who did develop

lung cancer. b, HR of new lung cancer stratified by sex and age and adjusted

for pack-years and smoking status. c, Lung cancer-specific mortality. d, HR of

death from lung cancer stratified by sex and age and adjusted for pack-years

and smoking status. The insets in the inverted Kaplan–Meier plots show the

same data on an expanded y axis. a–d, Cox proportional hazards regression

was used to estimate HRs. In the forest plots, the centre of each box represents

the estimated HR, and the whiskers denote the corresponding 95% CI; shaded

box size is for visualization only and does not encode statistical weight. The

overall contribution of thymic health to uni- or multivariable models was

evaluated using likelihood ratio tests (χ² tests) comparing full models with

nested models excluding thymic health (type III test, two-sided) without

adjustments for multiple comparisons.

Nature | www.nature.com | 5

analyses with adjustment for smoking status (type III P = 0.051)

(Fig.4f).

Mortality by disease type

To investigate the relevance of thymic health in different disease types,

aside from cancer and cardiovascular disease, we analysed the NLST,

where the cause of death was recorded through a 12-year follow-up.

Across all investigated disease-specific causes of death, participants

with high or average thymic health were less likely to die as compared

with participants with low thymic health (Extended Data Fig.6), and all

associations remained statistically significant in sex and age-stratified

analyses with adjustments for smoking status and pack-years (Fig.4g

and Supplementary Fig.3f) and in multivariate analyses using con-

tinuous age as the time scale to implicitly adjust for the effect of age-

ing (Extended Data Fig.5e–g) (type III P < 0.05 for all). Mortality from

pulmonary disease for high and average thymic health was 61% and

40% lower, respectively, compared to those with low thymic health

(Extended Data Fig.6c and Supplementary Table3). Likewise, mor-

tality from endocrine, nutritional and metabolic diseases, including

HR average: 0.57 (95% CI 0.50–0.65)

HR high: 0.37 (95% CI 0.31–0.44)

0 3 6 9 12

Years since randomization

Cumulative CVD mortality (%)

6,258 6,156 5,976 5,454 4,612

12,515 12,220 11,717 10,399 8,567

6,256 6,031 5,650 4,939 3,919

Low

Average

High

At risk

0 3 6 9

n = 25,029

Thymic health

Low

Average

High

Number

6,256

12,515

6,258

0.30 0.75 1.50

HR HR

Reference

0.63

0.53

95% CI

(0.55; 0.72)

(0.44; 0.64)

Thymic health

Low

Average

High

Low

Average

High

Low

Average

High

Number

6,256

12,515

6,258

6,256

12,515

6,258

6,256

12,515

6,258

HR HR

0.67

0.54

0.70

0.43

0.56

0.61

95% CI

(0.57; 0.80)

(0.43; 0.68)

(0.48; 1.03)

(0.24; 0.79)

(0.38; 0.82)

(0.37; 1.01)

Pulmonary disease mortality

End/Met/Nutr disease mortality

Digestive disease mortality

Reference

P = 1.72 × 10

–14

P = 7.80 × 10

–8

P = 0.0116

P = 0.0141

High

Average

Low

HR average: 0.38 (95% CI 0.20–0.70)

HR high: 0.08 (95% CI 0.02–0.34)

n = 2,581

0 3 6 9 12

Years since examination

Cumulative CVD mortality (%)

646 644 641 617 306

1,290 1,275 1,239 1,179 536

645 630 601 554 254

Low

Average

High

At risk

HR average: 0.65 (95% CI 0.49–0.87)

HR high: 0.26 (95% CI 0.16–0.40)

n = 2,443

0 3 6 9 12

Years since examination

Cumulative CVD incidence (%)

631

623

608 509 131

1,220

1,175

1,109 873 217

592

557

512 405 93

Low

Average

High

At risk

Thymic health

Low

Average

High

Number

592

1,220

631

HR HR

Reference

0.85

0.55

95% CI

(0.63; 1.14)

(0.34; 0.91)

Average

Low

P = 7.84 × 10

–10

0 3 6 9 12

High

Average

Low

0 3 6 9 12

P = 0.051

P = 0.021

P = 7.72 × 10

–6

P = 2.48 × 10

–30

Thymic health

Low

Average

High

Number

645

1,290

646

0.20 0.75 1.50

HR HR

Reference

0.50

0.21

95% CI

(0.26; 0.95)

(0.05; 0.93)

1.00

0.20 0.75 1.501.00

Fig. 4 | Association of thymic health with long-term risk of CVD-specific

mortality, CVD incidence and disease-specific mortalities. a, Percentage of

participants in the NLST who died from CVD. b, HR of death from CVD adjusted

for pack-yearsand smoking status, and stratified by sex and age. c, Percentage

of participants in the FHS who died from CVD. d, HR of death from CVD adjusted

for smoking status and stratified by sex and age. e, Percentage of participants

in the FHS who had a new CVD-specific event, such as myocardial infarction,

congestive heart failure or cerebral embolism. f, HR of a new CVD-specific

event adjusted for smoking status and stratified by sex and age. g, HR of death

from the specified disease groups among participants in the NLST stratified by

sex and age and adjusted for pack-years and smoking status. a–g, Follow-up for

all analyses was 12 years. The insets in the inverted Kaplan–Meier plots show

the same data on an expanded y axis. Cox proportional hazards regression was

used to estimate HRs. In the forest plots, the centre of each box represents the

estimated HR, and the whiskers denote the corresponding 95% CI; arrowheads

indicate that the 95% CI extends beyond the visualized limits; shaded box size

is for visualization only and does not encode statistical weight. The overall

contribution of thymic health to uni- or multivariable models was evaluated

using likelihood ratio tests (χ² tests) comparing full models with nested models

excluding thymic health (type III test, two-sided) without adjustments for

multiple comparisons.End, endocrine; Met, metabolic; Nutr, nutritional.

6 | Nature | www.nature.com

Article

metabolic disorders such as diabetes mellitus, was 68% and 37% lower

for individuals with high and average thymic health, respectively, com-

pared to those with low thymic health (Extended Data Fig.6d and Sup-

plementary Table4). Finally, mortality from diseases of the digestive

system that included liver, gallbladder or pancreatic diseases was 54%

and 47% lower for individuals with high and average thymic health,

respectively, compared to those with low thymic health (Extended

Data Fig.6e and Supplementary Table5). Together, these data provide

evidence that thymic health is prognostic across diverse diseases, indi-

cating disease-agnostic relevance for health.

Impact of metabolic health and lifestyle

Next, we investigated the associations of cholesterol, triglycerides, fast-

ing glucose and blood pressure with thymic health in the FHS (Fig.5).

Across all clinically relevant categories, female participantshad higher

thymic health as compared with maleparticipants, and thymic health

was positively associated with metabolic health (Fig.5a,b). In sex-,

age- and smoking status-adjusted analyses, and across sex strata, we

consistently found that high-density lipoprotein (HDL) had a significant

positive association with thymic health, while the common variables

of metabolic syndrome, including triglyceride levels, fasting glucose

and blood pressure, showed negative associations with thymic health

(Fig.5b and Supplementary Fig.5).

We assessed the association of actionable lifestyle factors, such as

smoking and consumption of alcohol, with thymic health. In sex- and

age-adjusted analysis, both smoking duration (actively smoking years)

and intensity (packs smoked per day) were negatively associated with

thymic health, as also reflected in the strong negative association with

the composite measure pack-years (Fig.5c). The drinking of alcoholic

beverages showed no association with thymic health.

Furthermore, we used the Fried frailty phenotype

to investigate

whether individuals with low thymic health are at increased risk for

future disabilities and reduced quality of life. We found a significant

association between low thymic health and increased frailty scores

(P < 0.001), independent of sex, age and smoking status (Supplemen-

tary Table6). Exploring the individual components of the Fried frailty

index, we found that this signal was particularly driven by slower walk-

ing speed (P = 0.032), lower physical activity index (P = 0.010) and

increased exhaustion (P = 0.021). Together, these data indicate a strong

relationship between metabolic health and lifestyle with thymic health,

and that thymic health may directly be associated with an individual’s

quality of life and long-term risk of disability.

Chronic inflammation and thymic health

As immune response and inflammation are closely related, we explored

whether low thymic health is associated with dysregulated inflam-

matory processes. For this, we investigated the association of blood

inflammatory proteins with thymic health. First, we analysed prot-

eomic blood plasma data obtained from 317 individuals included in FHS

(Supplementary Table7) before the CT scans (mean time difference,

10.4 years). Plasma samples were analysed using the Olink inflamma-

tion panel comprising 92 proteins, with 68 passing quality control.

Count

2,573

2,553

1,180

2,542

2,581

2,580

Beta

<− Thymic health worse | Thymic health better −>

Beta

5.16

−0.42

−4.71

−3.00

−2.85

−2.52

95% CI

(4.08; 6.25)

(–1.43; 0.59)

(–6.13; –3.29)

(–4.02; –1.98)

(–3.93; –1.78)

(–3.55; –1.50)

Cholesterol

Blood pressure

HDL

LDL

Triglycerides

Fasting glucose

Systolic

Diastolic

Variables

and subgroups

P value

<2 × 10

–100

0.4108

1.15 × 10

–10

1.02 × 10

–8

1.92 × 10

–7

1.60 × 10

–6

100

Ideal

<2 × 10

–16

<2 × 10

–16

<2 × 10

–16

7.7 × 10

–15

0.0417

2.1 × 10

–10

0.0011

Border-

line

Low

100

Ideal

Normal

Border-

line

High

100

Ideal

Border-

line

High

100

Normal

Pre-

diabetes

Diabetes

100

Normal

HT-S1

HT-S2

100

Normal

Elevated

HT-S1

HT-S2

HDL

Fasting glucose

Diastolic blood pressure Systolic blood pressure

LDL

Triglycerides

HGF

OSM

FGF21

VEGFA

TNFSF14

IL-6

S100A12

IL-18R1

CCL2

CXCL11

CXCL10

CCL19

IL-18

CCL13

FLT3LG

CXCL6

–0.2 –0.1 0 0.1

Thymic health effect size (s.d. units)

–log

[FDR adjusted P]

25,031

5,545

12,945

–2 –1 0 1 2

<− Thymic health worse | Thymic health better −>

−1.53

−1.19

−0.94

−0.06

(–1.88; –1.18)

(–1.53; –0.85)

(–1.33; –0.55)

(–0.79; 0.67)

(–0.55; 0.42)

Count Beta Beta 95% CI

Smoking

Alcohol

Pack-years

Cigarettes per day

Actively smoking years

Drinks per week

Drinks per day

Variables

and subgroups

P value

<2 × 10

–100

8.73 × 10

–12

2.30 × 10

–6

0.8687

0.7990

Thymic health (%)Thymic health (%)Thymic health (%)

Sex

Male

Female

<2 × 10

–16

<2 × 10

–16

<2 × 10

–16

2 × 10

–16

7.7 × 10

–15

8.6 × 10

–13

<2 × 10

–16

0.0003

<2 × 10

–16

1.4 × 10

–8

0.0013

<2 × 10

–16

7.9 × 10

–8

6420–2–4–6

Fig. 5 | Associations of thymic health with metabolism, lifestyle and

inflammation in the FHS. a, Associations of metabolically relevant variables

with thymic health, stratified by sex with femaleparticipants shown in green

and maleparticipants in purple. b, Respective associations adjusted for sex,

age and smoking status. c, Associations of smoking-related factors and weekly

consumption of alcoholic beverages with thymic health adjusted for sex and

age. d, Association of Olink-based plasma protein levels with thymic health,

adjusted for sex, age and smoking status (n = 317). In a, box plots show the median

(centre line), interquartile range (25th–75th percentiles; box), and whiskers

extending to the minimum and maximum values within 1.5× the interquartile

range. Statistical comparisons between male and femaleparticipants were

performed using two-sided Wilcoxon rank sum tests. Patient counts shown in a

correspond to the number (Count) of patients depicted in the respective rows

of b. In b and c forest plots, the box centres represent the estimated regression

coefficients, and the whiskers their corresponding 95% CI. The shaded box size

is for visualization only and does not encode statistical weight. Statistical

significance of individual coefficients was evaluated using two-sided t-tests.

In a–c no adjustment for multiple comparisons was applied. In d, effect size,

in s.d. units, is plotted against false discovery rate (FDR) of the linear regression

association of rank-based inverse-normal transformed Olink inflammatory

proteins with thymic health. The FDRs were computed across 68 proteins using

the two-sided regression P values using the Benjamini and Hochberg method

for multiple comparisons. Proteins with FDR < 0.1 (corresponding to −log

[FDR] > 1) were considered statistically significant. HT-S1, hypertension stage 1;

HT-S2, hypertension stage 2; LDL, low-density lipoprotein.

Nature | www.nature.com | 7

Of these, 16 out of 68 (24%) proteins were negatively associated with

thymic health (Fig.5d), including major mediators and regulators of

inflammation, such as vascular endothelial growth factorA (VEGFA),

interleukin-6 (IL-6), IL-18, hepatocyte growth factor(HGF), oncostatin

M(OSM) and C-X-C motif chemokine ligand (CXCL) family members

10 and 11. These results show that individuals with higher systemic

levels of inflammation had lower thymic health, indicating an interplay

between inflammatory processes and thymic health.

To investigate this further, we assessed whether chronic inflam-

mation lasting for a prolonged period of time correlated with thymic

health. For this, we collected longitudinal C-reactive protein (CRP)

measures over 5 to 10 years for 1,156 participants in the FHS (Supple-

mentary Table8). We found that participants in the FHS with chronic

systemic inflammation, quantified by consistently high CRP levels of

greater than or equal to 3 mg l

−1

(140 out of 1,156, 12.1%) over multiple

longitudinal blood measurements, had substantially lower thymic

health independent of sex, age and smoking status(P = 0.0012).

Taken together, these data highlight that inflammation and thymic

health are closely related; one potentially driving and potentiating the

other, with associated negative health consequences.

Thymic health model stability

To assess model stability, we performed stability investigations

that demonstrated excellent test–retest stability and robustness to

inter-reader input variations (Extended Data Fig.7). Furthermore, to

investigate which anatomical regions the model relies on for making

its quantifications, we performed activation mapping, which demon-

strated specific attention of the model to the thymic bed, while giving

minor attention to adjacent structures, indicating knowledge of the

anatomical context (Extended Data Fig.8). Taken together, these results

demonstrate that the performance of the thymic health model is robust

against input variations, and that it captures contextual information

directly from the anatomical region of the thymus. Extensive details

regarding model development and testing can be found in Supplemen-

tary MethodsS1 and S2, Supplementary Figs.9–17 and Supplementary

Tables9–13.

Discussion

Our results show that thymic decay in adults is highly individualized

and that loss of thymic health increases mortality and disease inci-

dence, including cancer and cardiovascular diseases. These findings

were investigated in two independent prospectively collected clinical

studies of asymptomatic adults, the NLST and the FHS. Among more

than 25,000 participants in the NLST, higher thymic health was con-

sistently associated with a significantly lower risk of mortality by any

cause, lung cancer and cardiovascular diseases, independent of sex,

age, smoking, prior diseases or cancer history. Similar results were

found in the independent FHS, where participants with high thymic

health had a significantly reduced risk of death from cardiovascular

disease, a primary endpoint of the study, independent of sex, age and

smoking. Considering the prevailing perception of the limited role of

the thymus in adults, our findings may change our understanding of

human health, emphasizing the complex yet critical role of the immune

system in long-term well-being and longevity.

We found that thymic health was associated with critical health con-

sequences. Presumed healthy participants in the NLST with high thymic

health had an approximately 50% reduction in the risk of death, were

36% less likely to develop lung cancer and nearly 50% less likely to die

from lung cancer, as compared to participants with low thymic health.

Further, the significant impact of high thymic health on cardiovascular

disease(CVD) was consistent across participants in the NLST and FHS,

with risk reductions in CVD mortality ranging from 63% to 92%. Given

the critical role of the thymus in generating a diverse Tcell repertoire

these results support the broad impact of a sustained adaptive immune

system to combat disease and promote longevity.

Our results challenge the established notion that the cessation of

thymic output in ageing adults is inconsequential as it is naturally

replaced by peripheral expansion of Tcells

. Rather, our results suggest

that loss of thymic tissue in adults may forecast higher risks of disease

and death. The main function of the thymus is to generate a diverse

Tcell repertoire, which provides adaptive immunity throughout life

While the relevance and abundance of the Tcell repertoire at a young

age are welldocumented

, our results indicate that the thymus retains

a continued role in Tcell production throughout adulthood and that

the pattern of decline of thymic function in adults is associated with

poorer health outcomes.

A recent landmark study demonstrated an association between

thymectomy and reduced lifespan and increased risk of cancer, among

othereffects

. While thymectomy is rare, we demonstrate that the

thymic decay is highly individualized even in presumed healthy adults,

indicating that thymic function can also be substantially reduced in

individuals who did not have their thymus surgically removed. Our

work therefore impacts the wider population and aligns with the

well-established age-related decline in immune system function

6,25,26

This is also consistent with previous work modelling declining

Tcell output as a major risk factor for age-related increase of cancer

development

Lifestyle and metabolic health measures, such as smoking, physical

activity or HDL levels, showed strong associations with thymic health.

Likewise, in longitudinal blood-based evaluations, we found individu-

als with chronic inflammation, a hallmark of immunosenescence and

commonly associated with chronic stress, carbohydrate-rich diet and

obesity, had lower thymic health.

Among presumed healthy individuals from the FHS, lower thymic

health was indeed associated with pro-inflammatory modifications of

blood plasma protein levels, consistent with the presence of chronic

inflammation. The pro-inflammatory pattern included increased levels

of cytokines IL-6, IL-18 and OSM, as well as several CXCL chemokines,

all of known relevance in systemic inflammatory diseases such as ath-

erosclerosis, age-associated diseases such as arthritis, and cancer

Our findings are further supported by studies in which thymic invo-

lution was associated with immunosenescence and inflammation,

contributing to illnesses such as metabolic or cardiovascular disease

Metabolic syndrome affects more than a third of all adults in the USA,

with continuously increasing prevalence

. Our results demonstrate

significant associations between metabolic and thymic health. These

findings are consistent with those of recent studies in which fatty degen-

eration of the thymus was associated with obesity

18–20

and smoked

pack-years

19,20

. Together, these findings suggest a profound impact of

actionable lifestyle choices on thymic health and may further clarify

why healthy behaviour improves well-being and lifespan.

In addition, the global increase in early-onset cancers might be linked

to an accelerated rate of thymic decay, potentially driven by factors

such as smoking, minimal physical activity and overall unhealthy life-

styles, leading to more inflammation in the body. Indeed, thymic health

could refine disease and cancer screening strategies, especially for

high-risk individuals.

Here we examined the impact of thymic health on lung cancer inci-

dence separately in current and former smokers. Associations were

primarily driven by current smokers, likely reflecting both higher

statistical power due to higher event rates and the greater biologi-

cal relevance of thymic function under ongoing tobacco exposure

and thereby continuously increasing neoantigen load, making thymic

health particularly relevant for immune surveillance in this setting

Previous studies have attempted to quantify thymic characteristics

using imaging techniques, using conventional approaches such as esti-

mating the proportion of fatty degeneration or measuring attenuation

density

18–20

. However, these studies were limited in scope, focusing

8 | Nature | www.nature.com

Article

on the exploration of associations between thymic features and basic

clinical and epidemiological measures. Previous studies did not find

any associations between thymic imaging characteristics and outcomes

such as survival or disease incidence. These previous investigations

reported presumed residual thymic tissue in only a minor fraction of the

examined population ranging from 41%

to 26%

19,20

, that is, the studies

estimated a fully fatty degenerated thymus in approximately 60–75%

of individuals

. We find these results contradictory to biological asso-

ciations of sustained Tcell output in visually fully fatty degenerated

thymic glands

, and we could corroborate that our measure of thymic

health had preserved impact on outcomes across individuals who had

no appreciable visual thymic tissue according to independent visual

scoring, indicating insufficient thymic quantification using a visual

scoring system. Our deep learning model found favourable thymic

functionalityamong individuals with average and high thymic health,

representing 75% of participants, consistent across both independent

study populations and paralleled by substantial health benefits in these

participants. Indeed, our findings are supported by Kooshesh etal.

who found substantial negative health consequences after thymectomy

in adults, indicating relevant thymic activity throughout life across

the population.

Participants included in this study were prospectively enrolled in the

FHS and NLST, covering a wide age range for both sexes. They are, how-

ever, predominantly white, and further validation in moreethnically

diverse populations is warranted. While we observed an association

between thymic health and overall survival in both cohorts, this did

not remain significant following thorough multivariate adjustment in

the smaller FHS. This further indicates that long-term outcomes may

be influenced by dynamic lifestyle choices in this population that ben-

efits from regular medical examinations and follow-up, and supports

the notion that active engagement in health-promoting behaviours

may potentially attenuate the long-term negative impact of unhealthy

characteristics associated with low thymic health

. Taken together, our

results provide evidence of multifaceted associations between thymic

health and clinically relevant health consequences across clinical, epi-

demiological and biological characteristics and strongly suggest the

critical relevance of the thymus for adult health.

Before the thymic health deep learning model developed here can

be applied in a clinical setting, it is essential to prove generalizabil

ity. Development and application of the thymic health model were

performed in fully independent datasets, with high robustness as

demonstrated by test–retest stability. Given substantial differences

in acquisition protocols, scanners and population characteristics

between theFHS and theNLST, the thymic health analyses were con-

ducted using population-specific thresholding, and no universal cut-

offs can be assumed. Future studies with international and external

validation are needed to explicitly address batch and scanner vari-

ability and enable direct cross-cohort comparisons and generalizable

thresholds.

Our results connect thymic health with longevity and lower disease

incidence. Our results suggest thatinflammationpotentially drives

an accelerated rate of thymic decay. This finding may provide new

opportunities for preventive strategies that aim to reduce thymic

decay and potentially even reverse it. These can include the use of

anti-inflammatory and anti-obesity drugs and open new avenues of

drug development that aim to improve long-term immune health.

Furthermore, even practical approaches, such as lifestyle changes

including exercise and sleep, as well as healthy food choices and sup-

plement intake, are likely to notably impact thymic health.

While this study investigated the role of thymic health in healthy

individuals, the state of the immune system could also have an impor-

tant role in individuals with disease. This is particularly relevant for

treatments that rely on triggering an immune response, such as immu-

notherapies used to treat patients with cancer. However, it may also be

relevant for other diseases: for example, during the recent COVID-19

pandemic, the response to the virus varied significantly between sex

and age groups, with older men particularly affected

Our analysis reveals previously unrecognized, possible negative

consequences of reduced thymic health. While thymic health declines

with age, we also find considerable variation in thymic health within

age groups, indicating that the rate of decay varies considerably

between individuals. When we analysed blood inflammatory proteins,

we found an association between increased levels of inflammatory

proteins and thymic decay. Although this analysis was limited by a

relatively long interval of 10.4 years between the CT scans used for

thymic health assessment and blood samples used to assess protein

content, these results were supported by an orthogonal assessment

of CRP levels, supporting that chronic inflammation may be a likely

driver of poor thymic health. It is also likely that a genetic component

exists, predisposing certain individuals to increased or decreased

rates of thymic decay. A recent large-scale study highlighted sub-

stantial individual variability in immune resilience and its link to

long-term health outcomes

. Gene expression profiles associated

with immune competence and low inflammation were linked to lon-

gevity, while pro-inflammatory signatures correlated with poorer

outcomes. These findings align with our observations that thymic

health varies widely across individuals and is negatively associated

with chronic inflammation. Awareness of a genetic predisposition

may lead to preventive measures and increased surveillance, and this

should be elucidated in future studies that match single nucleotide

polymorphisms with rates of thymic decay. Other limitations of this

study include the older age and heavy smoking status of individuals in

the NLST cohort relative to the FHS cohort, both likely to affect thymic

health.

An important implication of our study is that the retrospective obser-

vational design does not allow conclusions about causality. It is possible

that lower thymic health contributes to adverse outcomes by weakening

immune resilience, but it is also possible that pathological processes

leading to reduced health status, mortality and disease drive thymic

decline. Clarifying the direction and nature of these associations will

require future prospective and mechanistic studies and will be essential

to determine whether thymic health can serve as a target for preven-

tion or intervention strategies. Recent large-scale initiatives, such as

the Advanced Research Projects Agency for Health-funded Thymus

Rejuvenation programme, underscore growing interest in developing

regenerative approaches to restore thymic function, further highlight-

ing the high clinical relevance of this work

In summary, this study underscores the highly personalized nature

of thymic health and emphasizes the previously unrecognized possible

critical role of maintaining thymic health to preserve an agile, adap-

tive immune response that will accommodate long-term well-being

and longevity. Today, thymic assessments do not have an established

clinical standard, and the thymus is not examined in routine clinical

care. The extent to which the adult appearance of the thymus is associ-

ated with health and whether actionable lifestyle or risk factors may

be harnessed to improve thymic health was unknown. By analysing

27,612 individuals, our results provide evidence that thymic health is

directly associated with critical outcomes and diseases and may be

directly targetable by various approaches, such as smoking reduction

and weight loss in overweight and obese individuals.

Our results underline the relevance of the thymus throughout life.

Clinical investigations of preventive or regenerative strategies will

be vital to help us understand how to harness the thymic potential to

improve population health

34–38

Online content

Any methods, additional references, Nature Portfolio reporting summa-

ries, source data, extended data, supplementary information, acknowl-

edgements, peer review information; details of author contributions

Nature | www.nature.com | 9

and competing interests; and statements of data and code availability

are available at https://doi.org/10.1038/s41586-026-10242-y.

1. Goldrath, A. W. & Bevan, M. J. Selecting and maintaining a diverse T-cell repertoire.

Nature 402, 6–13 (1999).

2. Miller, J. F. A. P. The function of the thymus and its impact on modern medicine. Science

369, eaba2429 (2020).

3. Palmer, D. B. The effect of age on thymic function. Front. Immunol. 4, 316 (2013).

4. Collins, C., Sharpe, E., Silber, A., Kulke, S. & Hsieh, E. W. Y. Congenital athymia: genetic

etiologies, clinical manifestations, diagnosis, and treatment. J. Clin. Immunol. 41, 881–895

(2021).

5. Rezzani, R., Nardo, L., Favero, G., Peroni, M. & Rodella, L. F. Thymus and aging:

morphological, radiological, and functional overview. Age 36, 313–351 (2014).

6. Lynch, H. E. etal. Thymic involution and immune reconstitution. Trends Immunol. 30,

366–373 (2009).

7. Kumar, B. V., Connors, T. J. & Farber, D. L. Human Tcell development, localization, and

function throughout life. Immunity 48, 202–213 (2018).

8. Kooshesh, K. A., Foy, B. H., Sykes, D. B., Gustafsson, K. & Scadden, D. T. Health

consequences of thymus removal in adults. N. Engl. J. Med. 389, 406–417 (2023).

9. Murray, J. M. etal. Naive Tcells are maintained by thymic output in early ages but by

proliferation without phenotypic change after age twenty. Immunol. Cell Biol. 81, 487–495

(2003).

10. Williams, K. M., Hakim, F. T. & Gress, R. E. Tcell immune reconstitution following

lymphodepletion. Semin. Immunol. 19, 318–330 (2007).

11. Weng, N.-P. Aging of the immune system: how much can the adaptive immune system

adapt?Immunity 24, 495–499 (2006).

12. Thomas, R., Wang, W. & Su, D.-M. Contributions of age-related thymic involution to

immunosenescence and inlammaging. Immun. Ageing 17, 2 (2020).

13. Palmer, S., Albergante, L., Blackburn, C. C. & Newman, T. J. Thymic involution and rising

disease incidence with age. Proc. Natl Acad. Sci. USA 115, 1883–1888 (2018).

14. Vadasz, Z., Haj, T., Kessel, A. & Toubi, E. Age-related autoimmunity. BMC Med. 11, 94

(2013).

15. Nunes-Alves, C., Nobrega, C., Behar, S. M. & Correia-Neves, M. Tolerance has its limits:

how the thymus copes with infection. Trends Immunol. 34, 502–510 (2013).

16. Douek, D. C. etal. Changes in thymic function with age and during the treatment of HIV

infection. Nature 396, 690–695 (1998).

17. Wang, W., Thomas, R., Sizova, O. & Su, D.-M. Thymic function associated with cancer

development, relapse, and antitumor immunity—a mini-review. Front. Immunol. 11, 773

(2020).

18. Sandstedt, M. etal. Complete fatty degeneration of thymus associates with male sex,

obesity and loss of circulating naïve CD8+ Tcells in a Swedish middle-aged population.

Immun. Ageing 20, 45 (2023).

19. Araki, T. etal. Normal thymus in adults: appearance on CT and associations with age, sex,

BMI and smoking. Eur. Radiol. 26, 15–24 (2016).

20. Ackman, J. B. etal. Sex difference in normal thymic appearance in adults 20–30 years of

age. Radiology 268, 245–253 (2013).

21. Pido-Lopez, J., Imami, N. & Aspinall, R. Both age and gender affect thymic output: more

recent thymic migrants in females than males as they age: age and gender affect thymic

output. Clin. Exp. Immunol. 125, 409–413 (2001).

22. Heng, T. S. P. etal. Effects of castration on thymocyte development in two different

models of thymic involution. J. Immunol. 175, 2982–2993 (2005).

23. Fried, L. P. etal. Frailty in older adults: evidence for a phenotype. J. Gerontol. A 56,

M146–M157 (2001).

24. Klein, L., Kyewski, B., Allen, P. M. & Hogquist, K. A. Positive and negative selection of the

Tcell repertoire: what thymocytes see (and don’t see). Nat. Rev. Immunol. 14, 377–391

(2014).

25. Britanova, O. V. etal. Age-related decrease in TCR repertoire diversity measured with

deep and normalized sequence proiling. J. Immunol. 192, 2689–2698 (2014).

26. Mittelbrunn, M. & Kroemer, G. Hallmarks of Tcell aging. Nat. Immunol. 22, 687–698 (2021).

27. West, N. R., Owens, B. M. J. & Hegazy, A. N. The oncostatin M-stromal cell axis in health

and disease. Scand. J. Immunol. 88, e12694 (2018).

28. Moore, J. X., Chaudhary, N. & Akinyemiju, T. Metabolic syndrome prevalence by race/

ethnicity and sex in the United States, National Health and Nutrition Examination Survey,

1988–2012. Prev. Chronic Dis. 14, E24 (2017).

29. Alexandrov, L. B. etal. Mutational signatures associated with tobacco smoking in human

cancer. Science 354, 618–622 (2016).

30. Allaire, S. H. etal. Evidence for decline in disability and improved health among persons

aged 55 to 70 years: the Framingham Heart Study. Am. J. Public Health 89, 1678–1683

(1999).

31. Cummings, M. J. etal. Epidemiology, clinical course, and outcomes of critically ill adults

with COVID-19 in New York City: a prospective cohort study. Lancet 395, 1763–1770

(2020).

32. Ahuja, S. K. etal. Immune resilience despite inlammatory stress promotes longevity and

favorable health outcomes including resistance to infection. Nat. Commun. 14, 3286

(2023).

33. Jenkins, A. ARPA-H funds project to restore immune system function. ARPA-H https://

arpa-h.gov/news-and-events/arpa-h-funds-project-restore-immune-system-function

(28 September 2023).

34. Duggal, N. A., Pollock, R. D., Lazarus, N. R., Harridge, S. & Lord, J. M. Major features of

immunesenescence, including reduced thymic output, are ameliorated by high levels of

physical activity in adulthood. Aging Cell 17, e12750 (2018).

35. Duah, M. etal. Thymus degeneration and regeneration. Front. Immunol. 12, 706244

(2021).

36. Chaudhry, M. S., Velardi, E., Dudakov, J. A. & van den Brink, M. R. M. Thymus: the next (re)

generation. Immunol. Rev. 271, 56 (2016).

37. Fahy, G. M. etal. Reversal of epigenetic aging and immunosenescent trends in humans.

Aging Cell 18, e13028 (2019).

38. Buono, M., Thézénas, M.-L., Ceroni, A., Fischer, R. & Nerlov, C. Bi-directional signaling by

membrane-bound KitL induces proliferation and coordinates thymic endothelial cell and

thymocyte expansion. Nat. Commun. 9, 4685 (2018).

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in

published maps and institutional afiliations.

Open Access This article is licensed under a Creative Commons Attribution-

NonCommercial-NoDerivatives 4.0 International License, which permits any

non-commercial use, sharing, distribution and reproduction in any medium or

format, as long as you give appropriate credit to the original author(s) and the source, provide

a link to the Creative Commons licence, and indicate if you modiied the licensed material.

You do not have permission under this licence to share adapted material derived from this

article or parts of it. The images or other third party material in this article are included in the

article’s Creative Commons licence, unless indicated otherwise in a credit line to the material.

If material is not included in the article’s Creative Commons licence and your intended use is

not permitted by statutory regulation or exceeds the permitted use, you will need to obtain

permission directly from the copyright holder. To view a copy of this licence, visit http://

creativecommons.org/licenses/by-nc-nd/4.0/.

Article

Methods

Study cohorts

We used two independent prospectively enrolled cohort studies

(n = 27,612), consisting of the FHS (n = 2,581) and the NLST (n = 25,031),

to conduct our retrospective secondary analysis (Supplementary

Fig.6).

The FHS is a longitudinal community-based prospective cohort study.

It started enrolling the original participants in 1948 and has since been

followed by consecutive enrolments of the Offspring cohort (children

of the original participants and their spouses) in the 1970s and the

Third-Generation cohort (children of the Offspring participants) from

2002 to 2005. Eligibility for enrolment in the FHS CT study required an

age of 35 years or older for maleparticipants and 40 years or older for

female participants. All participants had regular follow-up examination

cycles every 4–8 years, and they provided written informed consent for

the CT study and each attended examination. For the current study, we

identified 2,581 participants from the Offspring and Third-Generation

cohorts who had non-contrast-enhanced, non-gated, full-thoracic CT

scans covering the entire thymic bed between 2005 and 2011. Partici-

pants are under ongoing surveillance for cardiovascular disease end-

points and death. An endpoint review committee of senior investigators

using all available data (for example, hospital records) adjudicated the

FHS study endpoints using standardized criteria

. For subanalyses, Off-

spring participants who attended Exam 7 (1998–2001), 8 (2005–2008)

or 9 (2011–2014) for Olink inflammation proteomics panel, longitudinal

CRP and Fried frailty analyses were identified. The measurement of

thymic health occurred an average of 10.4 and 3.75 years after Exams 7

and 8,respectively, and 2 years before Exam 9, in our study population.

The NLST is a longitudinal randomized trial of screening for lung

cancer with the use of low-dose chest CT as compared with chest radi-

ography. It enrolled participants from 2002 to 2004 with screening

examinations from 2002 to 2007. Primary event follow-up was carried

out until 31 December 2009, and extended follow-up for overall survival

and disease-specific survival was registered until 2015. Eligibility for

enrolment required an age between 55 and 74 at randomization and a

smoking history of a minimum of 30 pack-years without quitting smok-

ing for more than 15 years before enrolment. Each participant provided

written informed consent. For the current study, we identified 25,031

participants from the first low-dose CT screening exam (T0), which was

performed soon after the time of randomization in Institutional Review

Board-approved centres of the Lung Screening Study or in centres

of the American College of Radiology Imaging Network, which were

responsible for collecting participant data.

Our retrospective secondary analysis of the FHS and NLST was

reviewed and approved by their corresponding review boards. All par-

ticipants provided written informed consent at study enrolment. Each

participant provided written informed consent during each examina-

tion attended as part of the FHS, a procedure evaluated and approved

by the Institutional Review Board at Boston University Medical Center.

Artificial intelligence-based thymic health assessment

We developed a deep learning system that automatically extracts a

thymic health score ranging from zero (complete thymic decay) to

one (high thymic health) from a given CT scan that covers the thoracic

region. A detailed description of the model development, including

training data and architecture, can be found in Supplementary Meth-

odsS1. In short, first, the system identifies the location of the thymic

bed to generate a centre-of-mass (CoM). Next, a deep learning model

performs a thymic health assessment using a second model that lever-

ages self-supervised learning in which information from unlabelled

imaging data was used to quantify CT characteristics of the thymic

area. Model development was done in a collection of 5,674 thoracic

CT scans (independent of FHS and NLST), covering a wide range of

acquisition settings of patients that were imaged for various diseases,

including cancer and infectious diseases. The system was fine-tuned

against expert image annotations of thymic involution to obtain an

objective measure of remaining thymic soft tissue, hypothesized to

reflect thymic functionality, on the basis of an existing protocol

18–20

FHS and NLST were defined as test sets and remained unseen until the

fully independent and externally developed model was locked. A tech-

nical evaluation of the performance of the end-to-end deep learning

system can be found in Supplementary MethodsS2. An evaluation of

thymic health against the previously proposed manual protocol of thy-

mus scoring

18–20

is depicted in Supplementary Fig.1, and an evaluation

against thymic bed volume is shown in Supplementary Figs.7 and 8.

Assessment of inter-reader stability

To evaluate the stability of the end-to-end system, a sub-sample of

5,081 CT scans was selected, consisting of 2,500 random scans from

the NLST and 2,581 random scans from the FHS. The CoM coordinates

were perturbed by shifting the CoM input in all three directional axes,

adding a random amount of simulated inter-reader noise sampled from

a Gaussian distribution with a mean of 0 and a variance of 16 mm. The

results were bootstrapped for 50 iterations, and the concordance index

was calculated between predicted thymic scores for a pair of trials and

averaged over all trials to assess the robustness of the system’s perfor-

mance against input perturbations. Further details for this assessment

can be found in Supplementary MethodsS2.2.3.

Analysis of test–retest variability

To assess the test–retest variability of the system, the Reference Image

Database to Evaluate Therapy Response dataset was used. This dataset

consists of 31 pairs of CT scans acquired approximately 15 min apart

from patients with non-small cell lung cancer. The intraclass correlation

coefficient and Cohen’s Kappa were computed for both the automatic

quality control and thymic health model, grouped by predicted scores

and categories, respectively. The intraclass correlation coefficient

was used to evaluate the consistency of the automatic quality con-

trol scores, while Cohen’s Kappa was used to measure the agreement

between the thymic health categories assigned by the model across the

test and retest scans. Furthermore, the R

coefficient was calculated

to compare the model’s predictions across the test and retest scans,

attesting to the model’s reliability and reproducibility in the context

of short-term variability inherent in the dataset.

Quality control of the artificial intelligence system

To investigate the associations between thymic health scores and the

CT-based input, that is, the thymic bed area, we used Shapley value

analysis and saliency maps. Shapley value analysis is a game-theoretic

approach that assigns importance scores to each input feature, indicat-

ing its contribution to the model’s output. This method allowed us to

identify the most attributable features for a predicted thymic health

score, providing insights into the model’s decision-making process.

After determining the most attributable features using Shapley value

analysis, we selected the feature with the largest contribution to the

output prediction. Next, we determined regions in the input volume

that affected this feature substantially. We used occlusion sensitivity,

where sliding windows of 7 mm

were occluded across the volume, and

the difference in feature values was observed. Intuitively, windows that

change the feature value the most are the most salient regions. These

visualizations aid in understanding the artificial intelligence system’s

measures, demonstrating a visual representation of the imaging regions

that contributed the most to the model’s predictions.

Statistical analysis

We scaled the raw output of the thymic health classification model, after

supervised learning, that is, the scalar probability corresponding to the

likelihood of the thymus being not ‘fully-fatty degenerated’, to match

the percentile distribution ranks of the examined study populations

to facilitate thymic health interpretation and potential clinical transla-

tion. By doing so, thymic health values range from 0 to 100 and can be

easily interpreted, such as a value of 50 ranking a participant exactly

in the median of the study population distribution. Further, using the

first and third quartiles in each respective study population, thymic

health was ranked into low (less than or equal to 25), average (25–75)

and high (greater than 75).

Summary statistics used mean, median or range for continuous vari-

ables and proportions or percentages for categorical variables. Overall

survival was calculated from the date of the CT examination (FHS) or

from the date of randomization (NLST) until the date of death. For

time-to-event analyses, we used a cutoff at 12 years of follow-up for both

FHS and NLST for all survival and incidence analyses except for lung

cancer incidence, for which data were only available for a follow-up of 6

years after randomization. If participants were alive or did not have an

event after 12 or 6 years of follow-up, respectively, they were censored

at that time. Association analyses used Wilcoxon rank sum test or lin-

ear regression as appropriate. For linear regression analyses, clinical

variables were z-score normalized and treated as predictors for thymic

health. The beta coefficients represent the change in thymic health

per one s.d. increase in each predictor. Time-to-event distributions

were analysed with the Kaplan–Meier estimator and Cox proportional

hazards models as appropriate for univariate and multivariate testing.

The Schoenfeld residuals were assessed for inspection of the propor-

tional hazards assumption. Initial Cox proportional hazards models

used thymic health as predictor, with adjustments for sex and age,

which violated the proportional hazards assumptions. We addressed

this by implementing sex- and age-stratified analyses in which age was

categorized into distinct bins using the age distribution of the studied

population (FHS, younger than 45 years, 45–48 years, 48–51 years, 51–54

years, 54–57 years, 57–60 years, 60–63 years, 63–66 years, 66–69 years,

69–72 years, 72–75 years and 75 years or older; NLST, 55–59 years, 60–64

years, 65–69 years and 70–75 years) and this stratification approach

was used in all sex- and age-adjusted analysis throughout the article if

not described otherwise. Both the Akaike Information Criterion and

Bayesian Information Criterion were inspected, and the model fit and

parsimony were improved using the implemented stratified models.

Unless explicitly stated otherwise, the following apply: box plots

show the median (centre line), interquartile range (25th–75th percen-

tiles; box), and whiskers extending to the minimum and maximum

values within 1.5× the interquartile range. Statistical comparisons

between groups were performed using two-sided Wilcoxon rank sum

tests with no adjustment for multiple comparisons. P values are inter-

preted against a significance level that is defined as less than 0.05. Cox

proportional hazards regressions were used to estimate HRs, and the

corresponding 95% CI are provided for all statistical values. In the for-

est plots, the centre of each box represents the estimated HR, and the

whiskers denote the corresponding 95% CI; arrowheads indicate that

the 95% CI extends beyond the visualized limits; shaded box size is for

visualization only and does not encode statistical weight. The overall

contribution of thymic health to uni- or multivariable models was evalu-

ated using likelihood ratio tests (χ² tests) comparing full models with

nested models excluding thymic health (type III test, two-sided). Statis-

tical significance of individual covariate coefficients was assessed using

two-sided Wald z-tests with no adjustments for multiple comparisons.

There were no missing values in all time-to-event analyses if not

otherwise described. In the case of missing values in the association

analyses, values were assumed to be missing completely at random,

and the analyses handled missing data by their row-wise exclusion.No

statistical methods were used to predetermine sample size. Statistical

analyses were done in R v.4.2.2 (R Project for Statistical Computing).

Inclusion and ethics statement

This retrospective secondary analysis of NLST data was reviewed and

approved by their corresponding review boards and approved under

NLST-367 and NLST-374. FHS analysis was reviewed and approved by the

Boston University Medical Center Internationl Review Board, protocol

number H-32132.

Reporting summary

Further information on research design is available in theNature Port-

folio Reporting Summary linked to this article.

Data availability

NLST data may be requested from the National Cancer Institute (https://

biometry.nci.nih.gov/cdas/nlst/), imaging data from the Imaging Data

Commons portal (https://portal.imaging.datacommons.cancer.gov/)

and the Thymic Health scores for NLST are available at Zenodo (https://

doi.org/10.5281/zenodo.18306999)

. Data from the FHS are available

from the online repositories BioLINCC and dbGap, or from the submis-

sion of a research proposal via FHS ResApp (https://www.framing-

hamheartstudy.org/fhs-for-researchers/research-application/). An

overview of public imaging data collections used for the development

of the deep learning model, which were downloaded from the Imaging

Data Commons (https://portal.imaging.datacommons.cancer.gov/),

can be found in Supplementary Table9.

Code availability

The software used in the publication is available on GitHub for aca-

demic, non-commercial use in our GitHub (https://github.com/

AIM-Harvard/thymus_health_deeplearning_system.git). Additional

technical details about both the development and evaluation of our

deep learning framework can also be found in the Supplementary

Information. The models’ weights are subject to intellectual property

obligations and cannot be shared publicly, but may be made available

through academic collaboration. For more details, please contact the

corresponding author.

39. Vasan, R. S., Enserro, D. M., Xanthakis, V., Beiser, A. S. & Seshadri, S. Temporal trends in

the remaining lifetime risk of cardiovascular disease among middle-aged adults across

6 decades: the Framingham Study. Circulation 145, 1324–1338 (2022).

40. Prudente, V. C. G. etal. Resources “Thymic health consequences in adults”. Zenodo

https://doi.org/10.5281/zenodo.18306999 (2026).

Acknowledgements S.B., V.P. and S.P. contributed equally as leading authors to this work.

K.L.L., N.J.B. and H.J.W.L.A. acted jointly as a supervisory team. We thank the National Cancer

Institute for collecting and making the data from the NLST accessible, and The Cancer Imaging

Archive and the Imaging Data Commons for making this and other imaging collections used

for developing our deep learning model available on their platforms. H.J.W.L.A. acknowledges

inancial support from NIH (HA: NIH-USA U24CA194354, NIH-USA U01CA190234, NIH-USA

U01CA209414 and NIH-USA R35CA22052; BHK: NIH-USA K08DE030216-01) and the European

Union–European Research Council (HA: 866504). K.L.L., J.M.M., Y.C. and J.C. received inancial

support from NIH-USA R01AG067457). S.B. acknowledges funding from the Deutsche

Forschungsgemeinschaft (DFG,German Research Foundation)—502050303. K.B.

acknowledges funding from Bayern Innovativ, German Federal Ministry of Research,

Technology and Space, Max Kade Foundation and Wilhelm-Sander Foundation. N.J.B.

acknowledges funding from the Lundbeck Foundation (R272-2017-4040), the Novo Nordisk

Foundation (NNF21OC0071483 and NNF23OC0085954) and Savvaerksejer Jeppe Juhl og

Hustru Ovita Juhl Research Stipend. M.J.-H. has received funding from CRUK, NIH National

Cancer Institute, IASLC International Lung Cancer Foundation, Lung Cancer Research

Foundation, Rosetrees Trust, UKI NETs and NIHR. The FHS is funded by a contract from the

National, Heart, Lung, and Blood Institute (75N92019D0031, contract number

75N92019D0031).

Author contributions Conceptualization: S.B., V.P., S.P., H.J.W.L.A. and N.J.B. Methodology:

S.B., V.P., S.P., A.K.A., Y.C., J.C., L.N., M.J.-H., C.A., C.S., K.L.L., J.M.M., H.J.W.L.A. and N.J.B.

Software: V.P. and S.P. Validation: S.B., V.P., S.P., Y.C., J.C., K.B., K.L.L., A.L., H.J.W.L.A. and N.J.B.

Formal analysis: S.B., V.P., S.P., Y.C., J.C., K.L.L. and A.L. Resources: H.J.W.L.A., K.L.L. and J.M.M.

Data curation: S.B., V.P., S.P., Y.C., J.C., K.L.L., J.M.M., B.F., M.T.L. and A.L. Writing—original draft:

S.B., V.P., S.P., H.J.W.L.A. and N.J.B. Writing—review and editing: all authors. Visualization: S.B.,

V.P., S.P., A.K.A., L.N., H.J.W.L.A. and N.J.B. Supervision: H.J.W.L.A., N.J.B., K.L.L. and J.M.M. Project

administration: H.J.W.L.A. and N.J.B. Funding acquisition: S.B., H.J.W.L.A., K.L.L. and J.M.M.

Competing interests M.J.-H. has consulted for Astex Pharmaceutical and Achilles

Therapeutics, is a member of the Achilles Therapeutics Scientiic Advisory Board and Steering

Committee, andhas received speaker honoraria from Pizer, Astex Pharmaceuticals, Oslo

Cancer Cluster, Bristol Myers Squibb and Genentech. M.J.-H. is listed as a co-inventor on a

Article

European patent application relating to methods to detect lung cancer (PCT/US2017/028013),

this patent has been licensed to commercial entities and, under terms of employment, M.J.-H.

is due a share of any revenue generated from such license(s), and is also listed as a co-inventor

on the GB priority patent application (GB2400424.4) with title Treatment and Prevention of

Lung Cancer. C.S. reports receiving support from the Francis Crick Institute and the Royal

Society; has received grants or contracts from AstraZeneca, Boehringer Ingelheim, Bristol

Myers Squibb, Invitae (formerly Archer Dx), Ono Pharmaceuticals, Pizer and Roche-Ventana;

has received consulting fees from Bicycle Therapeutics, Genentech, Medicxi, Metabomed,

Novartis, and is a member of the GRAIL SAB, Relay Therapeutics SAB, China Innovation Center

of Roche (CICoR), SAGA Diagnostics SAB and the Sarah Cannon Research Institute; has

received honoraria from Amgen, AstraZeneca, Bristol Myers Squibb, Illumina, GlaxoSmithKline,

MSD, Roche-Ventana and Pizer; holds patents, planned, issued, or pending, including PCT/

GB2017/053289, PCT/EP2016/059401, PCT/EP2016/071471, PCT/GB2018/052004, PCT/GB2020/

050221, PCT/GB2018/051912, PCT/US2017/28013 and PCT/GB2018/051892; has leadership or

iduciary roles with Cancer Research UK and AACR; holds stock or stock options in Apogen

Biotech, Epic Biosciences, GRAIL and Achilles Therapeutics; andhas other inancial or non-

inancial interests with AstraZeneca and GRAIL Bio UK. M.T.L. reports grants research funding

from the American Heart Association, AstraZeneca, Ionis, Johnson & Johnson Innovation, Kowa

Pharmaceuticals America, MedImmune, National Academy of Medicine, the National Heart,

Lung, and Blood Institute and the Risk Management Foundation of the Harvard Medical

Institutions outside the submitted work. S.B. reports consulting fees from Ambient. H.J.W.L.A.

reports consulting fees and/or stock from Onc.AI, Love Health, Sphera, Health-AI, Ambient,

and AstraZeneca. N.J.B. reports consulting fees from Ambient, is listed as a co-inventor on a

patent application (PCT/GB2020/050221) on methods for cancer prognostication and a patent

on methods for predicting anti-cancer response (US14/466,208). The other authors declare

no competing interests.

Additional information

Supplementary information The online version contains supplementary material available at

https://doi.org/10.1038/s41586-026-10242-y.

Correspondence and requests for materials should be addressed to Hugo J. W. L. Aerts.

Peer review information Nature thanks Bharat Thyagarajan and the other, anonymous,

reviewer(s) for their contribution to the peer review of this work.

Reprints and permissions information is available at http://www.nature.com/reprints.

Extended Data Fig. 1 | Long-term risk of mortality by increasing continuous

thymic health, by increasing thymic health percentiles cut-off thresholds,

and in each thymic health decile. a, Risk of death according to a one standard

deviation increase of the continuous thymic health measure, and b, with

stratification by sex and age. c, Risk of death by increasing thymic health

percentile thresholds in the pooled (FHS and NLST) (upper panel) and

independent FHS (middle panel) and NLST (lower panel) cohorts. d, Risk of

death in each thymic health decile versus the lowest decile, as reference, in the

pooled (upper panel) and independent FHS (middle panel) and NLST (lower

panel) cohorts. a–d, FHS, n = 2,581; NLST, n = 25,031. Cox proportional hazards

regression was used to estimate HRs. In the forest plots, the center of each

box represents the estimated hazard ratio, and the whiskers denote the

corresponding 95% CI; arrowheads indicate that the 95% CI extends beyond the

visualized limits; shaded box size is for visualization only and does not encode

statistical weight. Statistical significance of the continuous or binarized thymic

health covariate coefficients at the different cut points was assessed using

two-sided Wald z-tests without adjustments for multiple comparisons.

CI Confidence Interval, FHS Framingham Heart Study, HR Hazard Ratio, NLST

National Lung Screening Trial, SD standard deviation.

Article

Extended Data Fig. 2 | Association of thymic health with long-term

mortality in the Framingham Heart Study. a, Overview of the Framingham

Heart Study (FHS). Each column represents one patient. The data is sorted by

ascending thymic health, i.e., an imaging-based proxy for thymic functionality,

from left to right. Thymic health is categorized into low, average, and high

based on the bottom 25%, middle 50%, and top 25% of the population. The

fractional left split of the FHS represents the patients who were defined as

having low thymic health through automatic quality control. Thymic health

quantification and basic clinical information are shown. b, Associations

between thymic health and sex in FHS (n = 2,581), across age groups in years

(n = 2,581) and body mass categories (n = 2,577, missingness n = 4) using

pairwise Wilcoxon Rank Sum test. Box plots show the median (center line),

interquartile range (25th–75th percentiles; box), and whiskers extending to

the minimum and maximum values within 1.5 x the interquartile range.

Statistical comparisons between groups were performed using two-sided

Wilcoxon rank-sum tests without adjustment for multiple comparisons.

c, Survival of participants from the FHS without adjustments. d, Hazard ratio

(HR) of death from any cause for FHS participants adjusted for smoking status

and stratified by sex and age binned at 3-years. Follow-up for all analyses was

12 years. The inset in the Kaplan-Meier plot shows the same data on an expanded

y-axis. c,d, Cox proportional hazards regression was used to estimate HRs. In

the forest plots, the center of each box represents the estimated hazard ratio,

and the whiskers denote the corresponding 95% CI; arrowheads indicate

that the 95% CI extends beyond the visualized limits; shaded box size is for

visualization only and does not encode statistical weight. The overall

contribution of thymic health to uni- or multivariable models was evaluated

using likelihood ratio tests (χ² tests) comparing full models with nested models

excluding thymic health (type III test, two-sided) with no adjustments for

multiple comparisons. BMI Body Mass Index, CI Confidence Interval,

FHS Framingham Heart Study, HR Hazard Ratio.

Extended Data Fig. 3 | Association of clinical variables and known diseases

prior to trial enrollment with health outcomes in NLST. a–d, Association

analyses between the baseline clinical variables body mass index (BMI), smoking

status, and pack-years with a, overall survival, b, lung cancer incidence, c, lung

cancer-specific mortality, and d, cardiovascular disease (CVD)-specific mortality

in NLST, stratified by sex and age. 83 participants were excluded due to missing

BMI values, resulting in 24,948 individuals. Exact cohort sizes after outcome-

specific missingness were: overall survival (n = 24,948; 83 missing), lung cancer

incidence (n = 23,087; 1,944 missing), and lung cancer mortality and CVD

mortality (both n = 24,946; 85 missing). e–h, Association analyses between the

baseline clinical variables BMI, smoking status, and pack-years together with

diagnosed diseases prior to trial enrollment with e, overall survival, f, lung

cancer incidence, g, lung cancer-specific mortality, and h, CVD-specific

mortality in NLST, stratified by sex and age. 434 participants

were excluded due to missing values in BMI and diagnosed diseases, resulting

in 24,597 individuals. Exact cohort sizes after outcome-specific covariate

missingness were: overall survival (n = 24,597; 434 missing), lung cancer

incidence (n = 22,776; 2,255 missing), and lung cancer mortality and CVD

mortality (both n = 24,595; 436 missing). a–h, Cox proportional hazards

regression was used to estimate HRs. In the forest plots, the center of each

box represents the estimated hazard ratio, and the whiskers denote the

corresponding 95% CI; arrowheads indicate that the 95% CI extends beyond the

visualized limits; shaded box size is for visualization only and does not encode

statistical weight. Statistical significance of individual covariate coefficients

was assessed using two-sided Wald z-tests with no adjustments for multiple

comparisons. CI Confidence Interval, HR Hazard Ratio, NLST National Lung

Screening Trial.

Article

Extended Data Fig. 4 | Association of thymic health, clinical variables, and

known diseases prior to trial enrollment with health outcomes in NLST.

a–c, Association analyses between thymic health, the baseline clinical variables

body mass index (BMI), smoking status, and pack-years together with diagnosed

diseases prior to trial enrollment with a, lung cancer incidence, b, lung cancer-

specific mortality, and c, cardiovascular disease (CVD)-specific mortality in

NLST, stratified by sex and age. 434 participants were excluded due to missing

values in BMI and diagnosed diseases, resulting in 24,597 individuals. Exact

cohort sizes after outcome-specific missingness were: overall survival

(n = 24,597; 434 missing), lung cancer incidence (n = 22,776; 2,255 missing),

and lung cancer and CVD mortality (n = 24,595; 436 missing). d–f, Association

analyses between thymic health and the baseline clinical variables BMI, smoking

status, and pack-years, stratified by sex and age excluding participants with

known diseases or cancer prior to trial enrollment with d, lung cancer

incidence, e, lung cancer-specific mortality, and f, CVD-specific mortality

in the subgroup of the NLST in which participants with a history of cancer

(n = 575), childhood or adult asthma, diabetes, asbestosis, bronchiectasis, lung

fibrosis, sarcoidosis, silicosis, or tuberculosis (n = 5,500) were excluded from

the analysis (n = 18,619). Exact cohort sizes after outcome-specific missingness

were: overall survival (n = 18,565; 54 missing), lung cancer incidence (n = 17,280;

1,339 missing), and lung cancer and CVD mortality (n = 18,564; 55 missing). Cox

proportional hazards regression was used to estimate HRs. In the forest plots,

the center of each box represents the estimated hazard ratio, and the whiskers

denote the corresponding 95% CI; arrowheads indicate that the 95% CI extends

beyond the visualized limits; shaded box size is for visualization only and does

not encode statistical weight. The overall contribution of thymic health to

multivariable models was evaluated using likelihood ratio tests (χ² tests)

comparing full models with nested models excluding thymic health (type III

test, two-sided). Statistical significance of individual covariate coefficients

was assessed using two-sided Wald z-tests with no adjustments for multiple

comparisons. CI Confidence Interval, HR Hazard Ratio, NLST National Lung

Screening Trial.

Extended Data Fig. 5 | Association of thymic health with health outcomes

in NLST using continuous age as the time scale. Analyses were adjusted for

smoking status and continuous pack-years and stratified by sex. Modeling age

as the underlying time scale accounts for potential residual confounding by age.

Hazard ratios of a, lung cancer incidence (n = 23,163), b, lung cancer mortality

(n = 25,027), c, malignant neoplasm mortality (n = 25,025), d, cardiovascular-

disease specific mortality (n = 25,025), e, pulmonary-disease specific mortality

(n = 25,025), f, endocrine, nutritional, or metabolic-disease specific mortality

(n = 25,025), and g, digestive-disease specific mortality (n = 25,025). Follow-up

for all analyses was 12 years. Cox proportional hazards regression was used

to estimate HRs. In the forest plots, the center of each box represents the

estimated hazard ratio, and the whiskers denote the corresponding 95%

CI; arrowheads indicate that the 95% CI extends beyond the visualized limits;

shaded box size is for visualization only and does not encode statistical weight.

The overall contribution of thymic health to multivariable models was evaluated

using likelihood ratio tests (χ² tests) comparing full models with nested models

excluding thymic health (type III test, two-sided). Statistical significance of

individual covariate coefficients was assessed using two-sided Wald z-tests

with no adjustments for multiple comparisons. CI Confidence Interval, End/

Met/Nutr endocrine, nutritional, and metabolic diseases, HR Hazard Ratio,

NLST National Lung Screening Trial.

Article

Extended Data Fig. 6 | Impact of thymic health on long-term risk of disease-

specific mortality and clinically relevant cardiovascular disease. Percentage

of individuals who died from a, any malignancy, b, adjusted for sex, age, and

smoking status. Percentage of individuals who died from c, pulmonary disease,

d, endocrine, metabolic, or nutritional disease; or e, digestive disease. Adjusted

analyses of c,d,e, see Fig.4g. Follow-up for all analyses was 12 years. The insets

in the inverted Kaplan-Meier plots show the same data on an expanded y-axis.

a–e, Cox proportional hazards regression was used to estimate HRs. In the

forest plot, the center of each box represents the estimated hazard ratio, and

the whiskers denote the corresponding 95% CI; arrowheads indicate

that the 95% CI extends beyond the visualized limits; shaded box size is

for visualization only and does not encode statistical weight. The overall

contribution of thymic health to uni- or multivariable models was evaluated

using likelihood ratio tests (χ² tests) comparing full models with nested models

excluding thymic health (type III test, two-sided) with no adjustments for

multiple comparisons. CI Confidence Interval, FHS Framingham Heart Study,

HR Hazard Ratio, NLST National Lung Screening Trial.

Extended Data Fig. 7 | Stability assessment through test-retest and input

variability analysis. a, Test-retest stability of the two prediction models

(QC model and 0 vs 1 grading) is analyzed first through intraclass correlation

coefficient for continuous measures (n = 31), followed by Cohen’s kappa for

grade stability. b, Test-retest predictions (n = 31) for both models are displayed

as a scatter plot to show their linearity. c, Sampling distribution for generating

input perturbations for the 50 perturbation trials, where for each scan (patient),

the seed point is perturbed with a drawn sample. Although a 2D distribution is

shown, the perturbations are performed in 3 dimensions. Stability to input

variation/perturbation is shown through the concordance between predicted

scores across different trials for the d, FHS (n = 2,581), and e, NLST (n = 2,500)

cohorts. The box plots show the median (center line), interquartile range (25th–

75th percentiles; box), and whiskers extending to the minimum and maximum

values within 1.5 x the interquartile range. ICC Intraclass Correlation Coefficient,

FHS Framingham Heart Study, NLST National Lung Screening Trial, QC Quality

Control.

Article

Extended Data Fig. 8 | Explainability: Saliency Maps. Representative CT scans

from the FHS cohort overlaid with the generated occlusion-driven saliency

maps using the most important feature, highlighting the regions of the thymus

that the model focuses on when making its predictions. The saliency maps are

displayed on the central slice of the thymus, and the jet color scale is used to

indicate contribution intensity.

Extended Data Table 1 | Study population clinical and epidemiological characteristics

∗ Race data is not available for participants of the Framingham Heart Study (FHS). The FHS is a longitudinal community-based prospective cohort study. It started enrolling the original participants

in 1948 and has since been followed by consecutive enrollments of the Offspring cohort (children of the original participants and their spouses) in the 1970s and the Third-Generation (Gen 3)

cohort (children of the Offspring participants) from 2002-2005. All original participants are of White/European ancestry. Few exceptions might be spouses or adoptees. The degree of missing

data for Race and Smoking status is the difference between the summed values and the total number of participants in each cohort. The degree of missingness of BMI was n=4 for FHS and

n=83 for NLST. Percentages refer to the data after the exclusion of missing values. BMI Body mass index, FHS Framingham Heart Study, NLST National Lung Screening Trial.

≥ ≥

Discussion

The striking result here is not just that people with better thymic health lived longer, but that this association held up even after the authors tried hard to explain it away. They controlled for age, smoking, sex, and major comorbidities in several different ways—including narrower age matching and analyses restricted to healthier participants—and the signal remained strong. In other words, thymic health seems to provide prognostic information about mortality that is not simply a proxy for being younger or less sick. > "Indeed, given the role of the thymus in maintaining an adaptive immune response, the individualized rate of thymic decay may be a major driver of age-associated diseases, such as cardiovascular disease and cancer" This landmark 2026 article argues that the thymus, long thought to become mostly irrelevant in adulthood, remains an important predictor of health and longevity later in life. Using deep learning on routine chest images from two large cohorts, the authors found that adults with healthier-looking thymuses had lower risks of death, lung cancer, and cardiovascular mortality, even after accounting for age, sex, smoking, and other health factors. They also show that thymic health is tied to inflammation, metabolism, and modifiable habits like exercise, obesity, and smoking, suggesting that preserving thymus function could become a new strategy for healthier aging. The authors quantify “thymic health” by training a deep learning model to read chest CT scans and assign each person a continuous score based on the thymus’s radiographic appearance. They then divide the population into low, average, and high thymic health groups using percentile cutoffs, treating the score as an imaging-based proxy for how much thymic function has been preserved with age. The fact that the score declines with age and is higher in women—both expected biologic patterns—supports that the model is measuring something meaningful. > "Our results demonstrate that thymic health varies between individuals and is impacted by sex, age and lifestyle habits. Notably, we show that individuals with low thymic health, that is, lost thymic functionality, have a shorter lifespan and an increased risk of cancer and cardiovascular diseases. These findings strongly suggest that thymic health is crucial for long-term health and lifespan." > "The thymus is a specialized immune organ responsible for maturing Tcells, thereby producing a diverse Tcell repertoire crucial for mounting an adaptive immune response. The thymus itself decays with age and eventually transforms entirely into adipose tissue through a process known as thymic involution. While the absence of a functioning thymus in children is associated with profound immunodeficiency, the consequences of thymic decay in adulthood are more subtle. Indeed, it was long believed that once the thymus generates a sufficiently diverse Tcell repertoire in childhood, the Tcell repertoire could be peripherally maintained to support an adaptive immune response against a diverse array of pathogens. For this reason, the thymus has long been considered largely nonfunctional in adults." What makes this result so striking is that it reframes the adult thymus from a biological relic into something more like a barometer—and perhaps even a driver—of long-term health. The authors argue that thymic decline is not uniform across adults: some people seem to preserve thymic function far better than others, and those differences track strongly with mortality, cancer risk, cardiovascular death, inflammation, and metabolic health. Read this way, the thymus is not just fading quietly in the background of aging; its condition may help explain why lifestyle factors like smoking, inactivity, and obesity so powerfully shape disease risk and lifespan. The broader implication is that immune aging may be more modifiable than we assumed, and that preserving thymic function could become part of how we think about prevention, screening, and healthy longevity.

Comments

Products

Project