Iterated Insights

Jared Edward Reser, PhD with GPT 6 Pro

Abstract

I propose that stress-induced changes in cancer protection can help identify biological maintenance activities whose immediate costs compete with their longer-term benefits. Under some ecological conditions, reducing expenditure on future somatic integrity could favor immediate survival or other components of reproductive success. When such regulation persists in comparatively safe environments, however, it may compromise protection that would be valuable to preserve. This perspective integrates evolutionary theories of maintenance allocation with research on neuroendocrine regulation of DNA-damage responses and antitumor immunity. Its central hypothesis is that the pattern of protection suppressed during stress contains information about the costs, timing, and organization of biological maintenance. I distinguish the time required to perform maintenance from the timescale of its benefits and its tolerance for postponement. I further propose that some protective failures involve incomplete maintenance sequences, in which initial containment is preserved but subsequent clearance or restoration is impaired. A comparative research program should measure baseline protection, susceptibility to stress-related suppression, and recovery after exposure. Some long-lived lineages may preserve protection through resistance to suppression, rather than exclusively through stronger baseline defenses. Identifying these regulatory differences could guide targeted restoration. Demonstrating therapeutic recovery would not, by itself, establish adaptive resource allocation, which requires additional evidence of costs and compensating benefits.

Keywords: evolutionary oncology; somatic maintenance; stress; glucocorticoids; DNA repair; antitumor immunity; life-history trade-offs; cancer prevention

1. An unfinished question about cancer protection

I began writing about this idea in 2005, but the literature connecting stress, gene regulation, cancer, and evolutionary biology was so dense that I never finished the article. Artificial intelligence now makes it possible for me to revisit the notes, connect findings across disciplines, and develop the argument into a more coherent research proposal. My archived notes asked whether cancer prevention might receive less investment when resources were scarce and stress was high. They also proposed maintaining the activity of genes involved in repair. 

The question remains worth pursuing: When stress reduces cancer protection, what protective work is being omitted, what does that work cost, and what could be gained by restoring it?

The body’s protection against cancer involves multiple operations, including maintaining DNA integrity, responding appropriately to damaged cells, and coordinating immune recognition and elimination. Experimental studies have identified points at which stress-related signaling can impair these operations, sometimes with successful prevention or reversal of the impairment through targeted intervention (Feng et al., 2012; Yang et al., 2019). 

I use housekeeping to refer to this broader protective work, not to the narrower molecular category of constitutively expressed housekeeping genes. The proposed framework concerns the regulation of functions, not simply whether particular transcripts increase or decrease.

The evolutionary foundations are substantial. Kirkwood’s account of aging connected somatic maintenance with resource allocation. Boddy and colleagues subsequently modeled trade-offs between cancer defense and reproductive competitiveness. Jacqueline and colleagues explicitly proposed that understanding evolutionary constraints on anticancer adaptations could identify opportunities for intervention (Kirkwood, 1977; Boddy et al., 2015; Jacqueline et al., 2017). 

The contribution proposed here is more specific: stress responses may provide an experimental means of identifying the temporal priorities of cancer protection. By examining which functions are reduced, preserved, redirected, and restored, we may learn where biological investment becomes vulnerable and where medical investment could be productive.

2. Cancer defense as an investment across time

An evolutionary explanation of reduced protection should begin with relative costs and benefits. The most expensive function is not necessarily the first one that can be reduced. An expensive activity that prevents immediate death may remain indispensable. A less expensive activity whose benefits emerge much later may be more vulnerable to postponement.

Consider two hypothetical protective operations with similar current costs. One prevents tissue failure tomorrow. The other primarily reduces the probability of malignant transformation years later. Under conditions that favor immediate survival over delayed gains, reducing the second activity would be the more plausible allocation change.

This reasoning leads to the central prediction:

Among protective functions that can be reduced without immediate catastrophe, stress-related suppression should be greater when their current opportunity costs are high and a larger proportion of their benefits is delayed.

This is a hypothesis about the organization of regulation. It does not imply that cells consciously forecast death or that the body calculates how many weeks it has left. The proposed mechanisms would be responses to signals that had some relationship to relevant conditions during evolutionary history. Whether those signals remain informative in a particular environment is a separate question.

Costs should also be measured broadly. Candidate costs include metabolic expenditure, materials used to manufacture protective machinery, time spent repairing rather than proliferating, replacement of cells eliminated for safety, and collateral consequences of immune activation. These costs need not move together, and a change that reduces one may increase another.

There is now experimental evidence that reproductive investment can interact with tumor control. In Hydra oligactis, Stepanskyy and colleagues modulated reproductive effort through food availability and found that greater reproductive effort was associated with more tumor development, reduced remission, and progression toward advanced stages. Resource conditions affected some of these relationships. This supports investigating allocation trade-offs, although it does not establish the molecular mechanisms or their applicability to mammalian stress responses (Stepanskyy et al., 2026). 

The relationship between external mortality and maintenance nevertheless requires care. Formal models show that increased extrinsic mortality can favor faster life histories, have no effect, or favor slower life histories, depending on demographic conditions and how population density affects survival and reproduction. High mortality cannot simply be inserted into an argument as automatic evidence for reduced maintenance (de Vries et al., 2023). 

Accordingly, the present framework does not require a universal rule that short-lived animals invest little in cancer defense. It proposes that ecological conditions can shape the regulation of particular defenses and asks whether identifiable stress responses reflect those priorities.

3. Stress must be separated into its biological components

The relevant exposure is not a single quantity called stress. Psychological adversity, persistent sympathetic activation, glucocorticoid exposure, nutrient limitation, and local tumor-generated signals must be distinguished.

Duration matters as well. Dhabhar and McEwen demonstrated that acute and chronic stress could have opposite effects on particular cell-mediated immune responses. Nutrient limitation can also increase some maintenance activities: glucose starvation activates autophagy through AMPK-dependent regulation of ULK1. These findings argue against a model in which all adversity uniformly reduces housekeeping (Dhabhar & McEwen, 1997; Kim et al., 2011). 

The prediction should therefore concern a pattern of selective allocation, not a global decline. Some functions may be increased to address an immediate threat, others reduced, and still others altered in ways that unintentionally favor cancer.

It is equally important to distinguish cancer initiation from tumor progression and treatment response. A 2026 individual-participant-data meta-analysis of 22 cohorts, including up to 421,799 participants, found no association between the psychosocial factors examined and overall cancer incidence. The measures included general distress, bereavement, perceived social support, relationship status, and neuroticism (van Tuijl et al., 2026). 

Such findings constrain broad claims that psychological distress is a major general cause of cancer. They do not resolve whether a particular hormonal pathway impairs a defined protective function or whether interrupting that pathway improves treatment. The framework developed here does not attribute cancer to an individual’s failure to cope, remain optimistic, or avoid worry.

4. Molecular examples identify where protection can be interrupted

4.1 Restricting immune-cell metabolic activation

Qiao and colleagues found that β2-adrenergic receptor stimulation interfered with the metabolic reprogramming of activated mouse CD8 T cells. The cells showed reduced GLUT1 expression, glucose uptake, glycolysis, and mitochondrial respiratory activity. The receptor dependence of these effects was supported by pharmacological and genetic experiments (Qiao et al., 2019). 

This is particularly relevant to an allocation framework because it connects a stress-responsive receptor with access to fuel and the metabolic activity required for immune function.

The potential intervention point is not merely the availability of nutrients outside the cell. It is the signaling system governing their use. Providing more food would not necessarily overcome a receptor-mediated restriction on metabolic activation.

However, reduced expenditure by an immune cell does not demonstrate a useful saving for the organism. Establishing an adaptive allocation mechanism would require showing what resources were spared, where they became available, and what compensating benefit resulted.

4.2 Interrupting the coordination of antitumor immunity

Yang and colleagues identified a glucocorticoid-dependent increase in TSC22D3 in dendritic cells following social-defeat stress in mice. This change impaired interferon-related responses and therapy-induced antitumor immunity. Blocking glucocorticoid signaling or deleting Tsc22d3 specifically in dendritic cells reversed the detrimental effect on tumor control (Yang et al., 2019). 

The relevant failure concerns coordination: the ability to organize an effective protective response after receiving information about a threat.

This also illustrates why searching only for suppressed genes would miss important targets. In this case, an inhibitor increased, producing a reduction in downstream protection. A functional restoration program must identify excessive inhibitory activity as well as insufficient expression of protective machinery.

4.3 Weakening damage-responsive quality control

Feng and colleagues demonstrated a pathway through which glucocorticoids increased SGK1, promoted MDM2 activity, and reduced p53 function. Chronic restraint stress accelerated radiation-induced tumor development in mice carrying one functional copy of the p53 gene. Complementary cellular experiments examined how the pathway could be interrupted (Feng et al., 2012). 

This supplies a mechanism by which neuroendocrine signaling can interfere with decisions to stop proliferation or eliminate damaged cells. It does not establish that ordinary psychological stress alone initiates cancer in an otherwise unmodified human population.

The therapeutic distinction is also important. Preventing inappropriate suppression of functional p53 is different from restoring a tumor’s missing or defective p53 machinery. Regulation can sometimes be corrected without rebuilding a pathway, but that opportunity depends on what remains intact.

4.4 Preserving the ability of immune cells to kill

Chakraborty and colleagues engineered CAR-NK cells lacking NR3C1, the glucocorticoid-receptor gene. These cells retained antitumor activity under cortisol exposure and improved tumor control in mouse models. The study also examined cortisol-rich tumor environments, including local hormone production and conversion (Chakraborty et al., 2026). 

This demonstrates a strategy for protecting a selected therapeutic cell population from an inhibitory signal. It does not require psychological stress to be the source of that signal.

The broader implication is that immune-cell presence and immune-cell performance must be measured separately. A protective population can remain present while its capacity to execute a response is compromised.

4.5 Redirecting activity rather than simply reducing it

He and colleagues found that chronic stress promoted neutrophil extracellular traps and lung-microenvironment changes that increased metastasis in mice. Neutrophil-specific glucocorticoid-receptor deletion or treatment that degraded these extracellular traps prevented the additional stress-associated metastasis (He et al., 2024). 

This is a boundary case for the cost-cutting hypothesis. The harmful process involves increased or redirected activity, not simply reduced expenditure.

It follows that a useful research program must distinguish underinvestment, inappropriate allocation, and pathological overactivation. Calling all three “reduced immunity” would obscure the mechanisms and could suggest the wrong intervention.

5. Classifying maintenance by the timing of its benefits

The distinction between short-term and long-term housekeeping should be developed as a continuum. It should not become a fixed division between two lists of genes.

A biochemical operation that takes minutes can have consequences lasting decades. Conversely, a function associated with longevity may also be essential for surviving the next day. The relevant question is not how slowly an operation occurs, but how its benefits and costs are distributed over time.

I propose three primary temporal dimensions:

Dimension

Operational question

Relevance to intervention

Execution time

How long does the protective operation take to perform?

Determines how rapidly an intervention can alter function.

Benefit horizon

When does performing the operation now prevent harm or preserve capability?

Identifies immediate and delayed returns from maintaining protection.

Tolerance of postponement

How long can the operation be delayed before protection becomes less effective or impossible?

Distinguishes safely deferrable maintenance from time-critical prevention.

Two additional properties should be measured alongside them: the cost of performing the operation and the reversibility of its suppression.

Under the present hypothesis, especially informative functions would combine substantial current costs, limited immediate consequences from temporary reduction, and large cumulative consequences when reduction persists. These properties must be established experimentally rather than inferred from a pathway’s association with aging.

Incomplete maintenance sequences

Some protective failures may concern the completion of a sequence rather than the absence of its first step.

Cellular senescence provides a useful example. In experiments by Ovadya and colleagues, impaired perforin-mediated immune surveillance increased the accumulation of senescent cells and was accompanied by inflammation, tissue dysfunction, and reduced lifespan. Removing senescent cells improved several outcomes in the experimental models (Ovadya et al., 2018). 

This suggests a sequence worth examining: arrest potentially dangerous proliferation, clear appropriate arrested cells, and restore tissue function.

I propose that stress may sometimes preserve an initial containment response while disproportionately impairing subsequent clearance or reconstruction. The senescent-cell study establishes the importance of clearance; it does not demonstrate this particular stress-dependent sequence.

The proposed maintenance-completion deficit would explain how a tissue could remain functional in the near term while accumulating unresolved problems. It also creates a distinct intervention question: should treatment amplify the initial response, or ensure that the later work gets completed?

I use maintenance debt descriptively for the accumulating consequences of unfinished protective work. It should not imply that all resulting damage can later be repaid.

6. Cancer-defense resilience may be an evolved property

Comparative studies have identified meaningful differences in somatic protection. Cagan and colleagues found that annual mutation rates in intestinal crypts were inversely associated with lifespan across 16 mammalian species. Tian and colleagues linked longevity across rodent species to more efficient double-strand-break repair and differences in SIRT6 activity. Firsanov and colleagues identified improved DNA-repair properties in bowhead-whale cells and showed that CIRBP could enhance repair when expressed in human cells (Cagan et al., 2022; Tian et al., 2019; Firsanov et al., 2025). 

These findings support comparative investigation of protective mechanisms. They do not show that long-lived species necessarily spend more energy on protection. Greater efficiency may improve protection without proportionately greater expenditure. Nor do these studies, by themselves, establish resistance to stress-related suppression.

The additional hypothesis proposed here is:

Some lineages may achieve durable cancer protection partly by making particular defenses less susceptible to neuroendocrine suppression, or by restoring those defenses more effectively afterward.

Two species could have similar baseline repair performance yet respond differently to a hormonal challenge. One might preserve function while the other becomes impaired. After exposure ends, their recovery could also differ.

These differences would be missed by a resting measurement. I therefore propose measuring cancer-defense resilience as a profile comprising baseline function, susceptibility during a defined challenge, and the speed and completeness of recovery.

This should not be treated as a single universal score. A species may preserve one repair pathway while reducing another protective function. The rodent SIRT6 study itself found that different repair processes related to different ecological variables: nucleotide excision repair tracked sunlight exposure rather than longevity (Tian et al., 2019). 

Comparisons must match relevant cell types and account for developmental stage, physiological age, shared ancestry, and receptor engagement. Equal hormone concentrations across species do not necessarily produce equal biological exposures. Observed cancer frequency alone would also be an inadequate measure because animals differ in their opportunity to live long enough for cancer to be detected.

A particularly promising discovery strategy would identify defenses that are impaired by stress-related signaling in a susceptible system but preserved in a resistant one. The regulatory difference could then be tested for transferability.

7. A research program must demonstrate both protection and cost

The strongest evidence for this framework would connect regulation, function, expenditure, and outcomes. Gene-expression surveys are useful for discovery, but they cannot establish the proposed trade-off by themselves.

Separate the signal from the actual resource constraint

Experiments should vary neuroendocrine exposure and resource availability independently. A factorial design could compare adequate and limited resources, with and without a defined glucocorticoid or adrenergic challenge, alongside pathway-specific interventions.

The resulting patterns would help distinguish different explanations. Protection that fails only under genuine shortage may reflect a hard resource limitation. Suppression that persists despite adequate resources may reflect anticipatory regulation. Failure that remains after both exposure and resource limitation are removed may indicate persistent reprogramming, missed developmental or activation windows, or accumulated damage.

None of these patterns alone proves adaptation. Each identifies a different mechanism requiring its own explanation.

Measure protective performance

Depending on the system, functional endpoints should include repair fidelity, mutation accumulation, damage-responsive arrest, appropriate apoptosis, antigen presentation, tumor-cell killing, and clearance of abnormal cells.

Measurements of cell viability, proliferation, cell-cycle state, and exposure must accompany these assays. An apparent improvement in repair could otherwise reflect changes in which cells survive or divide. Increased expression of repair-related genes should not be equated with more accurate repair.

Measure the proposed saving

The cost-cutting hypothesis requires evidence that reducing protection changes a relevant expenditure. Candidate measurements include substrate uptake, metabolic flux, biosynthetic demand, protein turnover, and replacement of eliminated cells.

The harder question concerns the destination of any saving. An immune cell consuming less glucose does not establish that another tissue benefits. Whole-organism studies would need to determine whether the reduction improves immediate performance, survival under a defined challenge, or another relevant component of fitness.

A reduction that produces no detectable saving or compensating benefit may still be medically important, but it would not support the proposed allocation explanation.

Distinguish preventing suppression from reversing it

Interventions should be administered before exposure, during established suppression, and after the exposure ends. Restoring a protective function while the stress-related signal remains present is particularly informative because it isolates the downstream mechanism from the many consequences of removing the stressor.

Later restoration must be evaluated separately. Recovering a repair pathway does not necessarily remove mutations already fixed in surviving cells. Recovering immune function may not reverse every consequence of prior tumor escape. The timing of intervention could therefore be as important as the identity of the target.

Falsifiable predictions

The framework yields several testable predictions:

Prediction

Evidence that would support it

Evidence that would weaken it

Suppression reflects delayed-benefit trade-offs.

Reduced protection yields measurable current savings or benefits alongside later harm.

Suppression produces no relevant saving or compensating advantage.

Some failures involve incomplete maintenance.

Later clearance or restoration is disproportionately impaired while initial containment remains effective.

All stages change together, or later failure is explained entirely by earlier damage.

Protective resilience adds information beyond baseline strength.

Challenge-and-recovery measurements predict outcomes after accounting for resting function.

Dynamic measurements add no predictive value.

Comparative resistance can guide restoration.

Transferring a causal regulatory difference preserves protection and improves relevant outcomes.

Transfer changes molecular markers without improving protection, or helps malignant cells instead.

An evolutionary explanation could fail while a therapeutic intervention succeeds. Preserving that distinction is essential to making the framework useful rather than unfalsifiable.

8. Targeted restoration should preserve the right functions in the right cells

The therapeutic objective is to recover an appropriate protective function, not to maximize every repair, survival, or immune pathway continuously.

Three goals must remain distinct. Prevention concerns protecting normal tissues against malignant transformation. Restoration of host control concerns recognizing and eliminating malignant cells or reversing environments that favor their spread. Treatment sensitization concerns preventing cancer cells from surviving therapy.

A defense can have different implications in these settings. Preserving the integrity of a normal cell is not equivalent to helping a malignant cell survive treatment. Interventions should therefore be tested in normal tissue and tumor cells separately, including under the treatments with which they might be combined.

Clinical evidence already shows that glucocorticoid signaling can be therapeutically actionable. In the phase III ROSELLA trial, relacorilant combined with nab-paclitaxel improved median overall survival in previously treated platinum-resistant ovarian cancer from 11.9 to 16.0 months, with a hazard ratio for death of 0.65. This demonstrates treatment benefit in a defined setting, not proof that psychological stress initiated the cancers or that all benefit resulted from restored host immunity (Lorusso et al., 2026). 

The distinction between regulation and indiscriminate activation also has experimental support. García-Cao and colleagues generated mice with additional normally regulated copies of p53. These animals had enhanced damage responses and greater tumor resistance without evidence of premature aging in the study (García-Cao et al., 2002). 

The design principle is therefore to preserve useful responsiveness while preventing inappropriate suppression. Potential approaches include cell-specific receptor modification, local delivery, downstream blockade of a particular inhibitor, and time-limited interventions during vulnerable periods.

My original notes considered epigenetic regulation as a route to maintaining repair functions. The contemporary version of that idea should be correspondingly selective. Broadly changing methylation or acetylation is not an adequate therapeutic specification. The target should be a defined regulatory change with a demonstrated functional consequence.

These proposals concern experimental and clinical development. They do not justify self-directed hormone blockade, immune stimulation, or attempts to increase repair indiscriminately.

9. Environmental mismatch changes the question of what is worth preserving

For people living in relatively safe and resource-secure conditions, protection with delayed benefits may be worth maintaining more consistently than some evolved regulatory responses permit. The mismatch hypothesis is that physiological signals associated with immediate danger can continue influencing maintenance even when they poorly represent the individual’s actual long-term prospects.

This argument does not require claiming that starvation, infection, or premature death have disappeared. Nor does it assume that resource abundance eliminates all trade-offs. It asks whether particular regulatory responses have become unnecessarily restrictive under particular circumstances.

The distinction between available resources and their regulated use is central. An organism can possess food while a protective cell remains unable to make the appropriate metabolic transition. Conversely, nutrient limitation can stimulate recycling through autophagy, so more nutrients do not uniformly imply more maintenance. 

Medicine could potentially improve on an evolved compromise in two ways. It could supply a missing resource, or it could alter a regulatory decision that no longer serves the patient’s circumstances. Determining which problem is present would prevent ineffective or harmful attempts simply to increase biological activity.

Developmental persistence is another question worth investigating. Some regulatory settings might outlast the conditions that induced them. Longitudinal work could ask whether early exposures predict later susceptibility and recovery of specific defenses, while separating persistent regulation from accumulated damage. Such findings would require careful interpretation and would not justify assigning responsibility to parents or individuals.

The larger objective is to distinguish protection that is unavailable from protection that is available but insufficiently deployed.

10. Conclusion

Stress-related reductions in cancer protection may provide a way to investigate how biological maintenance is prioritized across time. The most informative changes may occur where a protective function has substantial current costs, can be reduced without immediate failure, and produces benefits that accumulate over a longer future.

The evidence already identifies specific interruptions in immune-cell metabolism, immune coordination, damage responses, and cytotoxic activity. What remains to be established is which interruptions represent adaptive allocation, which are harmful side effects, and which reflect exploitation of host regulation by tumors.

I propose that the field examine these functions not only at baseline but also during and after defined challenges. The resulting profiles could reveal differences in suppression, incomplete maintenance, and recovery that resting measurements miss. Comparative studies could then ask whether long-lived or resistant systems preserve protection through distinct regulatory mechanisms.

Finding where protection is reduced can help identify where restoration is possible. Establishing the cost, timing, and consequences of that reduction can help determine when restoration is worthwhile.

The goal is to discover which protective work remains unfinished, which parts can still be completed, and how medicine can secure long-term benefits that existing regulation does not reliably preserve.

References

Boddy, A. M., Kokko, H., Breden, F., Wilkinson, G. S., & Aktipis, C. A. (2015). Cancer susceptibility and reproductive trade-offs: A model of the evolution of cancer defences. Philosophical Transactions of the Royal Society B, 370, 20140220. doi: 10.1098/rstb.2014.0220.

Cagan, A., et al. (2022). Somatic mutation rates scale with lifespan across mammals. Nature, 604, 517–524. doi: 10.1038/s41586-022-04618-z.

Chakraborty, D., et al. (2026). Cortisol-resistant CAR-NK cells overcome steroid-induced immunosuppression in lung cancer. Signal Transduction and Targeted Therapy, 11, 128. doi: 10.1038/s41392-026-02638-z.

de Vries, C., Galipaud, M., & Kokko, H. (2023). Extrinsic mortality and senescence: A guide for the perplexed. Peer Community Journal, 3, e29. doi: 10.24072/pcjournal.253.

Dhabhar, F. S., & McEwen, B. S. (1997). Acute stress enhances while chronic stress suppresses cell-mediated immunity in vivo: A potential role for leukocyte trafficking. Brain, Behavior, and Immunity, 11, 286–306. doi: 10.1006/brbi.1997.0508.

Feng, Z., et al. (2012). Chronic restraint stress attenuates p53 function and promotes tumorigenesis. Proceedings of the National Academy of Sciences, 109, 7013–7018. doi: 10.1073/pnas.1203930109.

Firsanov, D., et al. (2025). Evidence for improved DNA repair in the long-lived bowhead whale. Nature, 648, 717–725. doi: 10.1038/s41586-025-09694-5.

García-Cao, I., et al. (2002). “Super p53” mice exhibit enhanced DNA damage response, are tumor resistant and age normally. The EMBO Journal, 21, 6225–6235. doi: 10.1093/emboj/cdf595.

He, X.-Y., et al. (2024). Chronic stress increases metastasis via neutrophil-mediated changes to the microenvironment. Cancer Cell, 42. doi: 10.1016/j.ccell.2024.01.013.

Jacqueline, C., et al. (2017). Cancer: A disease at the crossroads of trade-offs. Evolutionary Applications, 10, 215–225. doi: 10.1111/eva.12444.

Kim, J., Kundu, M., Viollet, B., & Guan, K.-L. (2011). AMPK and mTOR regulate autophagy through direct phosphorylation of Ulk1. Nature Cell Biology, 13, 132–141. doi: 10.1038/ncb2152.

Kirkwood, T. B. L. (1977). Evolution of ageing. Nature, 270, 301–304. doi: 10.1038/270301a0.

Lorusso, D., et al. (2026). Overall survival with relacorilant and nab-paclitaxel in patients with platinum-resistant ovarian cancer (ROSELLA): A phase 3 randomised controlled trial. The Lancet, 407, 1513–1524. doi: 10.1016/S0140-6736(26)00462-9.

Ovadya, Y., et al. (2018). Impaired immune surveillance accelerates accumulation of senescent cells and aging. Nature Communications, 9, 5435. doi: 10.1038/s41467-018-07825-3.

Qiao, G., et al. (2019). β-Adrenergic signaling blocks murine CD8+ T-cell metabolic reprogramming during activation: A mechanism for immunosuppression by adrenergic stress. Cancer Immunology, Immunotherapy, 68, 11–22. doi: 10.1007/s00262-018-2243-8.

Reser, J. (n.d.). Personal Research Archive: Raw Notes & Observations from 1998–2025. Archived research notes, pp. 623 and 629.

Stepanskyy, N., et al. (2026). Empirical evidence for a trade-off between reproduction and tumour dynamics in the freshwater cnidarian Hydra oligactis. Proceedings of the Royal Society B, 293, 20261116. doi: 10.1098/rspb.2026.1116.

Tian, X., et al. (2019). SIRT6 is responsible for more efficient DNA double-strand break repair in long-lived species. Cell, 177, 622–638.e22. doi: 10.1016/j.cell.2019.03.043.

van Tuijl, L. A., et al. (2026). Psychosocial factors and the risk of cancer: An individual-participant data meta-analysis. Cancer, 132, e70271. doi: 10.1002/cncr.70271.

Yang, H., et al. (2019). Stress–glucocorticoid–TSC22D3 axis compromises therapy-induced antitumor immunity. Nature Medicine, 25, 1428–1441. doi: 10.1038/s41591-019-0566-4.