Chronic Psychological Stress and Cancer

1. Introduction

Psychological stress is experienced cognitively and emotionally but is implemented biologically through coordinated neural, endocrine, autonomic and immune responses. Acute stress responses can be adaptive. When activation is sustained or repeatedly triggered, signalling through the hypothalamic-pituitary-adrenal (HPA) axis and sympathetic nervous system (SNS) can alter immune-cell trafficking, inflammatory regulation, vascular biology and tissue homeostasis [1,3,5,6,7,8,17,21]. These pathways have generated longstanding interest in whether chronic psychological stress contributes to cancer initiation, progression or clinical outcome.

Three questions need to be separated. First, does psychological stress independently increase the probability that cancer will arise? Second, can stress-related physiology modify an established tumour or its microenvironment? Third, is psychological distress associated with recurrence, treatment response or mortality after diagnosis? These questions require different evidence. Cancer incidence is principally addressed by prospective epidemiology; biological modification is most directly tested in mechanistic and experimental models; and prognosis requires well-controlled clinical cohorts and trials.

Mechanistic evidence is substantial. A landmark ovarian-cancer model demonstrated increased tissue catecholamines, tumour burden, invasive growth and angiogenesis under chronic behavioural stress through beta2-adrenergic/cAMP-PKA signalling, VEGF and matrix metalloproteinases [23]. In breast cancer, sympathetic activation produced a marked increase in distant metastasis through beta-adrenergic recruitment of tumour-associated macrophages, while propranolol inhibited this effect [24]. Other studies have implicated host bone-marrow stromal responses [25],

myeloid-derived suppressor cells [20], pre-metastatic niche formation [29], glucocorticoid-receptor pathways [37,43], EMT and STAT3 signalling [35], and tumour-specific adrenergic pathways in gastric and oral cancer [32,33,39].

Human evidence is more heterogeneous and does not justify the statement that psychological stress generally causes cancer. More recent high-level analyses emphasise confounding by smoking, alcohol, adiposity and other behavioural or socioeconomic factors when cancer incidence is considered [2,15,16,45]. This review therefore integrates biological and clinical evidence while maintaining clear boundaries between experimental causality, biological plausibility and causal inference in humans.

2. Methods: Targeted Narrative Literature Search and Verification

This article was developed as a targeted narrative review rather than a systematic review. The approach was informed by established narrative-review quality principles: a clearly stated rationale and objective, transparent description of the search process, appropriate weighting of evidence, critical rather than purely descriptive synthesis, and attention to clinically meaningful outcomes.

A targeted PubMed/MEDLINE search was conducted and iteratively updated through 25 August 2026. Search concepts included combinations of: “chronic psychological stress”, stress, cancer, tumour/tumor progression, tumour/tumor microenvironment, metastasis, HPA axis, sympathetic nervous system, beta-adrenergic, catecholamine, glucocorticoid receptor, immunity, inflammation, immunotherapy, depression, anxiety, cancer incidence, cancer survival and cancer mortality. Citation chaining from directly relevant reviews and landmark primary studies was used to identify mechanistically important older work. The purpose was to assemble a focused and representative evidence base across mechanisms and human outcomes rather than to perform exhaustive PRISMA-style screening.

The final working corpus contains 46 publications from 2006 to 2026, including broad and specialised reviews, experimental studies, prospective and record-linkage human cohorts, systematic reviews, meta-analyses, a rapid review of reviews and an umbrella review. Evidence was weighted by design: experimental studies support causal mechanisms within model systems, while human cohort studies and quantitative evidence syntheses were given greater weight for incidence, survival and mortality.

As a separate editorial-integrity step, every final reference was checked against its PubMed record or linked publisher record for title, journal, publication status and editorial notices. Two sources from the earlier working corpus were removed after this audit because PubMed records showed Expressions of Concern: “Psychological stress on cancer progression and immunosenescence” and “Dopamine blocks stress-mediated ovarian carcinoma growth”. They were replaced with an unflagged 2023 review of stress and cancer pathogenesis [6] and a 2012 PLoS Biology study of sympathetic regulation of breast-cancer bone metastasis [25]. No source carrying a known Expression of Concern is used as evidentiary support in this verified version.

This is still a narrative review: no formal risk-of-bias instrument was applied across all heterogeneous study designs and no new pooled effect estimate was calculated. The search strategy and selection logic are therefore described transparently but should not be interpreted as a systematic-review protocol.

3. Neuroendocrine Pathways and Adrenergic Signalling

The HPA axis and SNS are the principal physiological routes through which sustained stress can influence peripheral tissues. HPA activation increases glucocorticoid exposure, whereas sympathetic activation increases catecholaminergic signalling. Reviews in the final corpus describe glucocorticoid and adrenergic receptors on malignant, immune, stromal and endothelial cells, providing multiple routes by which systemic stress physiology may interact with tumour biology [1,3,6,7,8,11,12,13,14,17,18,21,43].

Beta-adrenergic signalling is the most extensively demonstrated experimental pathway. In ovarian cancer, chronic behavioural stress increased catecholamines, tumour burden, invasive growth and vascularisation through ADRB2-dependent cAMP-PKA signalling and increased VEGF, MMP2 and MMP9 [23]. In breast cancer, stress-induced neuroendocrine activation had little effect on primary-tumour growth but produced a large increase in distant metastasis; macrophage recruitment and beta-adrenergic signalling were central, and propranolol inhibited tumour spread [24].

Stress physiology can also act on the metastatic host environment. In a breast-cancer bone-metastasis model, sympathetic activation increased RANKL expression in bone-marrow osteoblasts and promoted tumour-cell colonisation of bone; propranolol and interference with RANK signalling reduced this effect [25]. In pancreatic cancer, chronic stress increased tumour growth and reduced survival in an experimental model, with beta-adrenergic blockade antagonising the effect [27]. Gastric and oral-cancer studies likewise support tumour-specific beta2-adrenergic pathways affecting proliferation, invasion or metastasis [32,33,39].

Not all relevant signalling is explained by circulating epinephrine. Experimental work has shown that chronic stress can enhance metastatic biology even when circulating epinephrine is not required, pointing toward local sympathetic norepinephrine and tissue-specific neural regulation [30]. A 2025 Neuron study further identified a corticoamygdala circuit through which social interaction reduced intratumour sympathetic activity and suppressed breast-cancer progression in mice [42]. These findings support an emerging cancer-neuroscience model rather than a single-hormone explanation.

4. Glucocorticoid Signalling and Treatment Response

Glucocorticoid signalling provides a second major pathway. The glucocorticoid receptor can regulate apoptosis, metabolism, immune surveillance and treatment-response programmes, with effects varying by tumour context [6,8,17,21,43]. A 2026 Cancer Research review describes accumulating evidence that GR signalling may support tumour survival or resistance to several anticancer therapies in defined settings [43].

Primary studies add mechanistic detail. Stress-hormone-mediated acceleration of breast-cancer metastasis has been linked to inducible nitric oxide synthase-related signalling, with inhibition of nitric oxide synthase attenuating metastatic effects in an experimental model [31]. In cervical cancer, GR overexpression was associated with adverse progression and cisplatin resistance, with p38 MAP kinase implicated mechanistically [37]. These findings support investigation of GR biology but do not establish that psychological stress itself is the dominant clinical source of glucocorticoid exposure in patients receiving cancer treatment.

Stress may also influence immunotherapy through neuroendocrine-immune pathways. A 2026 review describes mechanisms by which SNS and HPA activation may impair local and systemic antitumour immunity relevant to checkpoint therapy [4]. This area remains translational: the mechanistic rationale is substantial, but clinical efficacy data for stress-targeted strategies alongside modern immunotherapy are not yet sufficient for routine oncological recommendations.

5. Immune Dysregulation and the Tumour Microenvironment

The immune system is a major interface between stress physiology and cancer. Chronic catecholamine and glucocorticoid signalling can influence leukocyte trafficking and function, and reviews in the corpus describe altered natural-killer and cytotoxic T-cell activity, dendritic-cell function, macrophage polarisation, inflammatory mediators and expansion or recruitment of immunosuppressive populations [3,7,10,14,17].

Primary studies illustrate these effects. Chronic stress promoted breast-carcinoma metastasis through accumulation of myeloid-derived suppressor cells and beta-adrenergic activation of an IL-6/STAT3 pathway [20]. Different stress paradigms produced different patterns of adaptive immune regulation and tumour angiogenesis in a breast-cancer model, illustrating that stress effects are not biologically uniform [28].

A 2026 gastric-cancer liver-metastasis study extended this framework to stress-associated hepatic immune remodelling. Chronic stress increased glucocorticoid exposure, neutrophil recruitment and neutrophil extracellular trap formation in experimental models; the publication also reported a prospective clinical cohort in which higher stress burden was associated with higher cortisol and NET-related markers and shorter progression-free survival [44]. This is a promising translational example, but replication across independent cohorts and tumour types is necessary.

The broader neuro-immune oncology literature emphasises bidirectional interactions: neural activity can influence immune-cell states and tumour biology, while cancer, pain, inflammation and treatment can themselves alter neural and endocrine stress systems [10,14]. Consequently, cross-sectional associations between distress, cortisol, immune markers and tumour outcomes cannot by themselves determine causal direction.

6. Angiogenesis, EMT and Metastatic Niche Formation

Metastasis requires invasion, dissemination, survival in circulation and successful colonisation of distant tissues. Stress-related pathways have been linked experimentally to several of these steps. The ovarian-cancer model identified VEGF-dependent angiogenesis as a major downstream consequence of beta-adrenergic signalling [23]. Breast-cancer studies subsequently demonstrated macrophage-dependent metastatic programming [24], host bone-marrow remodelling [25], pre-metastatic lung niche formation [29], and stress-related EMT/STAT3 signalling [35].

EMT and related invasion programmes are plausible points of convergence between stress signalling and malignant-cell behaviour. Chronic stress promoted EMT-mediated metastasis through a miR-337-3p/STAT3 pathway in breast cancer [35], while beta2-adrenergic/PlexinA1 signalling has been implicated in gastric-cancer progression [39]. A 2024 pharmacological study reported reduced stress-associated breast-cancer metastasis after targeting beta2-adrenergic signalling with baicalin [40]. These intervention studies strengthen mechanistic inference within their experimental systems but do not yet establish clinically validated anticancer therapies.

Pre-metastatic niche biology is particularly relevant because it focuses on systemic host tissues rather than only the primary tumour. Chronic stress enhanced macrophage recruitment and lung metastatic colonisation in a breast-cancer model [29], while the 2026 gastric-cancer study implicated glucocorticoid-driven NET formation in the hepatic metastatic environment [44]. Together, these observations support the idea that stress-related physiology may modify the receptivity of distant tissues to metastatic cells.

7. Human Evidence: Cancer Incidence

Human evidence is the critical test of claims that psychological stress causes cancer. A 2020 systematic review and meta-analysis of cohort studies reported modest associations between depression or anxiety and cancer incidence and mortality [45]. Earlier pooled work also identified associations between psychological distress and mortality from selected cancer sites [46]. However, these findings are difficult to interpret causally because psychological distress is correlated with smoking, alcohol use, body mass index, physical inactivity, socioeconomic disadvantage and healthcare utilisation.

The 2023 individual-participant-data meta-analysis by van Tuijl and colleagues provides a stronger test of incidence because harmonised individual-level adjustment was possible across 18 cohorts. Depression and anxiety were not associated with overall, breast, prostate, colorectal or alcohol-related cancer incidence. Associations with lung and smoking-related cancers were substantially attenuated after adjustment for smoking, alcohol use and body mass index [16].

A 2023 rapid review of reviews found heterogeneous conclusions across exposure definitions and cancer sites [15]. The 2026 umbrella review in the final corpus similarly reported no significant pooled association between stress-related exposures and overall cancer incidence while identifying a more consistent association with cancer mortality [2]. Taken together, the higher-level human evidence does not support a broad statement that chronic psychological stress, anxiety or depression independently causes cancer.

This conclusion is compatible with experimental findings. A pathway may be biologically real and clinically relevant in selected tumour-host contexts without producing a large independent population-level incidence effect. Cancer initiation is multifactorial, and stress-related physiology is better conceptualised as a potential modifier than as a universal carcinogenic exposure.

8. Human Evidence: Prognosis, Survival and Mortality

Associations are more consistent for outcomes after cancer has developed. A US record-linkage analysis evaluated psychological distress in relation to cancer mortality and found evidence of elevated risk in more distressed groups, while highlighting the importance of demographic, socioeconomic and behavioural covariates [36]. A systematic review of solid-tumour patients likewise found that psychological distress was frequently associated with poorer survival, although tumour type, distress measurement and covariate adjustment varied substantially [38].

The 2026 umbrella review reported a pooled association between stress-related exposures and cancer mortality in contrast with its null pooled finding for overall cancer incidence [2]. A separate meta-analysis of 65 cohort studies reported higher cancer-specific mortality among patients with depression across colorectal, breast, lung and prostate cancer, with substantial between-study heterogeneity [41]. These data support the clinical importance of depression and distress after diagnosis, but they do not establish that neuroendocrine stress biology is the only or primary mechanism.

Behavioural and clinical pathways are plausible contributors. Depression and anxiety can affect sleep, smoking, alcohol use, physical activity, adherence, treatment completion, social support and healthcare engagement; advanced disease can itself cause distress, creating reverse causation. Mechanistic studies provide biologically plausible additional pathways, but the most defensible interpretation is that psychological distress is a meaningful prognostic correlate and possible biological modifier rather than a universally proven independent determinant of survival.

9. Clinical Implications and Psychological Interventions

Psychological distress should be recognised and treated in oncology regardless of whether stress reduction changes tumour biology. Distress, anxiety and depression affect quality of life, sleep, functioning, relationships and treatment engagement. Reviews of psychological intervention in cancer care describe established psychosocial benefits and possible effects on neuroendocrine or immune pathways, while emphasising that evidence for direct improvements in tumour control or survival is less certain [19].

Patient communication is especially important. Patients should not be told or left to infer that insufficient positivity, unresolved stress or a mental-health condition caused their cancer. Such framing can create guilt and stigma and exceeds the current human evidence. A biologically integrated model of mental and physical health should increase scientific precision and compassion, not relocate responsibility for malignancy onto the patient.

Neuroendocrine pathways remain legitimate translational targets. Beta-adrenergic signalling and GR activity are being investigated as adjunctive targets [11,12,13,43]. Experimental beta-blocker studies repeatedly attenuated stress-associated tumour or metastatic effects [23,24,25,27,32], but prospective tumour-specific clinical trials are required before neuroendocrine modulation can be considered an established anticancer strategy.

10. Why Experimental and Human Findings Can Diverge

Experimental studies maximise control: stress exposure is defined, tumour models are selected, timing is known and pathways can be blocked pharmacologically or genetically. This permits strong causal inference within the model but can also amplify effects that are smaller or more heterogeneous in humans.

Human psychological stress is not a single exposure. Bereavement, social isolation, occupational stress, anxiety, depression and perceived stress differ in duration, severity, behavioural consequences and physiological correlates. Cancers likewise differ in receptor expression, immune phenotype, stage, metastatic pattern and treatment context. A stress-sensitive pathway in one tumour-host system may have little importance in another.

Recent findings reinforce this context dependence. A specific corticoamygdala circuit can modulate intratumour sympathetic activity in mice [42]; beta2-adrenergic/PlexinA1 signalling has been implicated in gastric cancer [39]; and glucocorticoid-driven NET formation may contribute to gastric-cancer liver metastasis [44]. The field is therefore moving away from a single ‘stress hormone causes cancer’ model toward tumour-specific, host-specific and systems-level mechanisms.

11. Limitations

Much of the mechanistic evidence derives from animal models or cell systems. Stress paradigms, tumour models, host immune status, sex, timing and duration differ substantially across studies. Publication bias toward positive mechanistic findings is possible.

Human studies are limited by confounding, reverse causation and heterogeneous measurement. Depression, anxiety, stressful life events and perceived stress are related but not interchangeable. Smoking, alcohol use, adiposity, sleep, physical activity, socioeconomic factors and comorbidity may confound or mediate observed associations.

This review is targeted and narrative rather than systematic. Although every final reference was verified for existence and editorial status and the corpus includes both recent syntheses and landmark primary studies, the search was not exhaustive and no formal risk-of-bias instrument was applied across all study types. Conclusions should therefore be interpreted as an evidence-weighted clinical synthesis rather than a quantitative estimate of effect.

12. Future Research Priorities

Future studies should combine repeated psychological assessment with longitudinal biological measurement. Concurrent measures of perceived stress, depression or anxiety, cortisol or other HPA-axis markers, autonomic activity, inflammatory mediators, immune phenotype and tumour-specific molecular features would allow temporal relationships to be tested more directly.

Prospective oncology cohorts should distinguish cancer incidence, recurrence, progression and mortality and should stratify by tumour type, stage, therapy and molecular or immune phenotype. Where stress-sensitive pathways are strongly implicated, clinical trials could evaluate beta-adrenergic or GR modulation alongside standard therapy using prespecified mechanistic biomarkers and clinically meaningful endpoints.

New systems-level findings also require independent replication. Neural-circuit regulation of tumour sympathetic activity [42] and glucocorticoid-associated NET formation in metastatic niches [44] are compelling, but their generalisability across tumour types and human populations is not yet known.

Psychological-intervention trials should maintain clear therapeutic priorities. Improving distress, functioning and quality of life is an established clinical goal. Biological and oncological endpoints can be studied rigorously where justified, but claims about recurrence or survival should follow prospective evidence rather than precede it.

13. Conclusion

Chronic psychological stress has biologically credible pathways through which it may modify cancer-related processes. HPA-axis and sympathetic activation can influence glucocorticoid and catecholamine signalling, immune regulation, angiogenesis, tumour-cell behaviour, EMT, metastatic niche formation and treatment-response pathways. Experimental causality has been demonstrated in multiple tumour models, particularly for beta-adrenergic mechanisms. Human evidence requires a more restrained interpretation. Current high-level analyses do not establish a general causal association between psychological stress and overall cancer incidence, whereas associations with prognosis and mortality are more consistent but remain vulnerable to confounding and reverse causation. The most defensible conclusion is therefore not that psychological stress ‘causes cancer’, but that chronic stress is a systemic biological state capable of interacting with tumour and host biology in selected contexts. Progress will depend on tumour-specific longitudinal human studies that integrate psychological, neuroendocrine, immune and oncological measures.