Anxiety in Chronic Medical Illness: Cancer, HIV, Epilepsy, and Cardiac Disease — Prevalence, Neurobiological Mechanisms, and Evidence-Based Management

Anxiety disorders are among the most prevalent and underdiagnosed psychiatric comorbidities in patients with chronic medical illness. While a degree of worry following a serious diagnosis is universal and expected, a substantial proportion of medically ill patients develop clinically significant anxiety that meets criteria for DSM-5-TR anxiety disorders — including generalized anxiety disorder (GAD), panic disorder, illness anxiety disorder, adjustment disorder with anxiety, and specific phobias related to medical procedures or disease progression. These are not merely psychological reactions to stress; they involve identifiable neurobiological perturbations driven by the disease process itself, its treatment, and the bidirectional interplay between peripheral inflammation and central nervous system function.

The clinical consequences of untreated anxiety in chronic illness are profound. Anxiety independently predicts poorer treatment adherence, increased healthcare utilization, accelerated disease progression, diminished quality of life, and in several conditions — notably cardiac disease — increased mortality. A meta-analysis by Roest and colleagues (2010) found that anxiety was associated with a 26% increased risk of incident coronary heart disease and a 48% increased risk of cardiac death, independent of depressive symptoms and traditional cardiovascular risk factors. Similar patterns emerge across oncology, HIV medicine, and neurology.

Despite this evidence, anxiety in medically ill populations remains systematically underrecognized. Clinicians often attribute symptoms such as tachycardia, dyspnea, fatigue, and cognitive difficulty to the underlying medical condition rather than considering a comorbid anxiety disorder. This diagnostic overshadowing results in significant treatment gaps. Estimates suggest that fewer than 30% of cancer patients with clinically significant anxiety receive any form of psychiatric or psychological treatment. This article provides a disease-specific review of anxiety prevalence, mechanisms, diagnostic considerations, and evidence-based management across four major chronic illnesses: cancer, HIV/AIDS, epilepsy, and cardiac disease.

Cancer

Anxiety disorders are the most common psychiatric comorbidity in oncology populations, frequently exceeding the prevalence of major depression. A landmark meta-analysis by Mitchell and colleagues (2011), encompassing over 10,000 patients across oncological settings, found pooled prevalence rates of 9.8% for DSM-defined anxiety disorders and 19.4% for clinically significant anxiety symptoms (using validated screening instruments with clinical thresholds). However, prevalence varies dramatically by cancer type, stage, and treatment phase. In lung cancer, anxiety prevalence reaches 30–40%, driven in part by dyspnea and its direct induction of panic-like symptoms. In gynecological cancers, prevalence estimates range from 15–33%. Among patients receiving active chemotherapy, rates are typically 20–30%, while in the palliative care setting, prevalence may reach 25–49% depending on measurement instruments. Notably, anxiety is highest in the peri-diagnostic period and in the first year following treatment completion — the so-called "re-entry" phase — when the structure of active treatment is withdrawn.

HIV/AIDS

Anxiety disorders are pervasive in people living with HIV (PLWH), with prevalence estimates substantially exceeding general population rates. A systematic review by Brandt and colleagues (2017) found that approximately 20–30% of PLWH meet criteria for a current anxiety disorder, compared to roughly 7–8% in the general population (NIMH estimates). GAD is the most frequently diagnosed anxiety disorder (prevalence 15–20%), followed by panic disorder (5–11%) and social anxiety disorder (7–14%). PTSD is also highly prevalent, affecting 10–30% of PLWH, though this often reflects pre-existing trauma histories compounded by the stress of diagnosis and HIV-related stigma. HIV-associated neurocognitive disorder (HAND) further complicates the picture, as cognitive symptoms can both mimic and exacerbate anxiety presentations.

Epilepsy

Epilepsy carries one of the highest psychiatric comorbidity burdens of any neurological condition. Anxiety disorders affect an estimated 20–30% of people with epilepsy (PWE), with some studies reporting rates as high as 40% in treatment-resistant epilepsy. The International League Against Epilepsy (ILAE) Commission on Neuropsychiatry has emphasized that anxiety in epilepsy is often atypical, with interictal (between-seizure) anxiety being the most common presentation, but peri-ictal anxiety — occurring as an aura, during a seizure, or in the postictal period — representing a unique phenotype not captured by standard DSM-5-TR criteria. A large population-based study by Tellez-Zenteno and colleagues (2007), using the Canadian Community Health Survey, found that the lifetime prevalence of any anxiety disorder in epilepsy was 22.8%, compared to 11.2% in the general population (adjusted odds ratio approximately 2.4). GAD and agoraphobia are disproportionately represented. Seizure frequency, temporal lobe epilepsy, and polypharmacy are consistently identified as risk factors.

Cardiac Disease

Anxiety disorders are common across the spectrum of cardiovascular disease, including coronary artery disease (CAD), heart failure (HF), and arrhythmias. In patients post-myocardial infarction (MI), prevalence of clinically significant anxiety ranges from 20–30%, with approximately 10–15% meeting criteria for a specific DSM-5-TR anxiety disorder. GAD and panic disorder are the most frequent presentations. In heart failure populations, anxiety prevalence ranges from 13–70% depending on measurement, with a meta-analytic estimate of approximately 32% for clinically significant symptoms (Easton et al., 2016). Among patients with implantable cardioverter-defibrillators (ICDs), 25–50% develop significant anxiety, often characterized by hypervigilance regarding device shocks and avoidance behaviors resembling PTSD. Notably, anxiety in cardiac populations is frequently confounded with cardiac symptoms — chest pain, palpitations, dyspnea, diaphoresis — creating a diagnostic overlap that challenges both cardiologists and psychiatrists.

The neurobiological basis of anxiety in chronic medical illness extends well beyond psychological stress responses. Multiple convergent mechanisms operate across disease states, while certain pathways are specific to individual conditions.

Systemic Inflammation and Cytokine-Mediated Neuropsychiatric Effects

A central mechanism shared across cancer, HIV, cardiac disease, and to a lesser extent epilepsy is systemic inflammation and its effects on brain circuitry. Chronic medical illness elevates pro-inflammatory cytokines — particularly interleukin-1β (IL-1β), interleukin-6 (IL-6), tumor necrosis factor-alpha (TNF-α), and interferon-gamma (IFN-γ). These cytokines cross the blood-brain barrier via circumventricular organs and active transport, and are also produced centrally by activated microglia. In the brain, they exert several anxiogenic effects: (1) activation of the hypothalamic-pituitary-adrenal (HPA) axis, leading to sustained cortisol elevation and glucocorticoid receptor resistance; (2) upregulation of indoleamine 2,3-dioxygenase (IDO), which diverts tryptophan metabolism from serotonin synthesis toward kynurenine pathway metabolites — including quinolinic acid, an NMDA receptor agonist with excitotoxic and anxiogenic properties; (3) reduction of brain-derived neurotrophic factor (BDNF) in the hippocampus and prefrontal cortex; and (4) altered glutamate-GABA balance in the amygdala and anterior cingulate cortex.

HPA Axis Dysregulation

Chronic HPA axis activation is documented in cancer (particularly during chemotherapy and in advanced disease), HIV infection, and heart failure. Elevated cortisol impairs hippocampal negative feedback regulation, produces dendritic retraction in medial prefrontal cortex (mPFC) neurons, and enhances amygdala reactivity — a pattern that directly promotes anxiety. In cardiac disease, cortisol dysregulation is bidirectional: anxiety drives sympathetic hyperactivation and HPA axis overactivity, which in turn promotes arrhythmogenesis, endothelial dysfunction, and platelet aggregation.

Cancer-Specific Mechanisms

Beyond inflammation, cancer-related anxiety involves chemotherapy-induced neurotoxicity (particularly agents crossing the blood-brain barrier such as methotrexate, cytarabine, and certain targeted therapies), direct paraneoplastic effects on the CNS, and treatment-related hormonal disruptions. Anti-estrogen therapies (tamoxifen, aromatase inhibitors) alter serotonergic function and are associated with increased anxiety and mood symptoms. Corticosteroid administration — ubiquitous in oncology — directly activates mineralocorticoid and glucocorticoid receptors in the amygdala and hippocampus, producing dose-dependent anxiety and insomnia.

HIV-Specific Mechanisms

HIV is a neurotropic virus that enters the CNS early in infection, predominantly infecting microglia and perivascular macrophages. Even with effective antiretroviral therapy (ART), low-level viral replication and chronic neuroinflammation persist in the CNS compartment. This produces disruption of fronto-striatal-limbic circuitry, with neuroimaging studies demonstrating reduced volume and functional connectivity in the anterior cingulate cortex, insula, and amygdala — regions central to anxiety processing. Dopaminergic dysfunction in HIV, particularly in basal ganglia circuits, may contribute to both anxiety and apathy. Additionally, certain antiretroviral agents (notably efavirenz) are directly associated with neuropsychiatric adverse effects including anxiety, vivid dreams, and psychosis.

Epilepsy-Specific Mechanisms

Anxiety in epilepsy has a unique neurobiological substrate. Temporal lobe structures — the amygdala and hippocampus — are both seizure-generating zones and core anxiety-processing regions. Kindling models demonstrate that repeated seizure activity in the amygdala progressively lowers the threshold for both seizures and anxiety-like behavior. GABAergic dysfunction is central to both conditions: reduced GABA-A receptor density and altered subunit expression (particularly α2 and α3 subunits involved in anxiolytic benzodiazepine effects) are documented in epilepsy tissue. Serotonergic deficits are also well-established, with PET studies showing reduced 5-HT1A receptor binding in temporal lobe structures of patients with comorbid epilepsy and anxiety. The forced normalization phenomenon — where anxiety or psychosis emerges as seizure frequency is reduced by medication — further illustrates the shared circuitry.

Cardiac Disease-Specific Mechanisms

In cardiovascular disease, autonomic nervous system dysregulation is paramount. Reduced heart rate variability (HRV) — a marker of vagal tone and parasympathetic function — is consistently found in anxious cardiac patients and independently predicts arrhythmias and cardiac mortality. The insular cortex, which processes interoceptive signals from the cardiovascular system, shows altered activity in patients with cardiac anxiety, contributing to amplified symptom perception and catastrophic interpretation of benign cardiac sensations. Shared genetic vulnerability via the serotonin transporter gene (5-HTTLPR) short allele has been associated with both anxiety traits and increased cardiovascular risk, though effect sizes are modest and not reliably replicated across all populations.

Diagnosing anxiety disorders in chronically ill patients presents unique challenges that frequently lead to both underdiagnosis and overdiagnosis. Several factors contribute to diagnostic complexity.

Somatic Symptom Overlap

The cardinal somatic symptoms of anxiety — tachycardia, dyspnea, fatigue, diaphoresis, tremor, gastrointestinal distress, and cognitive impairment — overlap extensively with symptoms of the underlying medical condition and its treatment. In cardiac disease, distinguishing panic attacks from angina or arrhythmia requires careful evaluation; approximately 25–50% of patients presenting to emergency departments with chest pain and no acute cardiac event have panic disorder. In cancer patients receiving chemotherapy, nausea, fatigue, and cognitive fog ("chemobrain") may mask or mimic anxiety. In epilepsy, peri-ictal anxiety symptoms must be differentiated from interictal anxiety disorders — the former is a seizure manifestation, while the latter requires independent psychiatric treatment.

Diagnostic Approaches and Screening Instruments

Standard screening instruments require contextual interpretation in medically ill populations. The Hospital Anxiety and Depression Scale (HADS), specifically developed for medical settings, excludes somatic items and has been validated across cancer, cardiac, and HIV populations. A HADS-Anxiety subscale score ≥8 demonstrates sensitivity of 80–90% and specificity of approximately 75–80% for clinically significant anxiety in these populations. The Generalized Anxiety Disorder 7-Item Scale (GAD-7) is widely used in primary care and has growing validation in medical populations, with a cutoff of ≥10 indicating moderate anxiety. Disease-specific instruments also exist: the Cardiac Anxiety Questionnaire (CAQ) for cardiac patients, the Neurological Disorders Depression Inventory for Epilepsy (NDDI-E) which, while depression-focused, often captures comorbid anxiety, and the HIV/AIDS-Targeted Quality of Life (HAT-QoL) instrument.

Differential Diagnosis Pitfalls

Clinicians must consider medication-induced anxiety, which is common across all four conditions. In oncology, corticosteroids, antiemetics (metoclopramide, prochlorperazine via akathisia), and certain targeted therapies can produce de novo anxiety. In HIV, efavirenz is well-documented for neuropsychiatric effects; integrase inhibitors, particularly dolutegravir, have also been associated with anxiety and insomnia in a subset of patients. In epilepsy, levetiracetam — one of the most prescribed antiepileptic drugs — produces behavioral side effects including anxiety and irritability in 10–15% of patients, while brivaracetam may have a more favorable neuropsychiatric profile. Substance withdrawal (benzodiazepines, alcohol, opioids) must always be considered, as substance use disorders are overrepresented in all four populations. Thyroid dysfunction, common in cancer patients who have received neck radiation or immunotherapy, can mimic GAD. Pheochromocytoma, though rare, should be considered when episodic anxiety is accompanied by hypertension.

The Role of DSM-5-TR Diagnostic Criteria

DSM-5-TR criteria for GAD, panic disorder, and other anxiety disorders include the specification that symptoms must not be attributable to the physiological effects of a substance or another medical condition. This creates a practical dilemma: when does "anxiety about cancer" become GAD? The clinical distinction rests on (1) pervasive worry extending beyond the medical condition itself, (2) functional impairment disproportionate to the medical situation, (3) presence of the full syndromal criteria including cognitive symptoms (difficulty concentrating, mind going blank) and physical symptoms not explained by the illness, and (4) persistence beyond what would be expected for an adjustment reaction (typically >6 months). When anxiety is clearly a direct physiological consequence of the medical condition or treatment, DSM-5-TR provides the diagnosis of anxiety disorder due to another medical condition, which requires evidence of a causal pathophysiological mechanism.

Pharmacological management of anxiety in chronic medical illness requires careful attention to drug interactions, organ function, cardiac safety, seizure threshold effects, and the specific neurobiological vulnerabilities of each condition.

SSRIs and SNRIs: First-Line Pharmacotherapy

Selective serotonin reuptake inhibitors (SSRIs) and serotonin-norepinephrine reuptake inhibitors (SNRIs) remain the first-line pharmacological treatment for anxiety disorders across all four medical populations. In cancer, a Cochrane systematic review (Ostuzzi et al., 2018) found moderate-quality evidence supporting the efficacy of SSRIs for anxiety and depression in cancer patients, though the evidence base is smaller than for depression alone. Sertraline and escitalopram are generally preferred due to their relatively favorable drug interaction profiles — a critical consideration given the extensive CYP450-mediated metabolism of many oncology agents. Fluoxetine and paroxetine are potent CYP2D6 inhibitors and should be avoided in patients receiving tamoxifen, as they block conversion to the active metabolite endoxifen. In cancer populations, expected response rates to SSRIs for anxiety symptoms are approximately 40–60%, comparable to but slightly lower than rates in primary psychiatric populations, likely reflecting the ongoing medical stressors.

In HIV, SSRIs have demonstrated efficacy in multiple randomized controlled trials. Sertraline was shown to be effective for major depression and comorbid anxiety in a landmark RCT by Elliott and colleagues (1998). Drug interactions with antiretrovirals are manageable but require attention: protease inhibitors (ritonavir, cobicistat) inhibit CYP3A4 and can increase levels of certain SSRIs and SNRIs. Escitalopram and sertraline are generally recommended as first-line options. Citalopram carries QTc prolongation risk at higher doses and requires monitoring.

In epilepsy, SSRIs are safe and do not lower seizure threshold at therapeutic doses — a finding supported by multiple studies and emphasized by the ILAE. Citalopram, sertraline, and escitalopram are preferred. The ILAE Task Force on Psychiatric Comorbidities has stated that the seizure risk with SSRIs is negligible and that undertreating anxiety and depression is a far greater clinical concern than any theoretical proconvulsant effect. Response rates for anxiety in PWE appear comparable to general psychiatric populations (50–65%), though high-quality RCT data specific to this population remain limited.

In cardiac disease, sertraline was established as safe and modestly effective in the landmark SADHART trial (Sertraline Antidepressant Heart Attack Randomized Trial, Glassman et al., 2002), which demonstrated cardiac safety post-MI. While SADHART focused on depression, secondary analyses showed improvements in anxiety measures. The ENRICHD trial found that CBT with adjunctive SSRI use reduced depression but did not significantly reduce cardiac events, though post-hoc analyses suggested potential benefit. Escitalopram is also well-supported. SNRIs (venlafaxine, duloxetine) may increase blood pressure and require monitoring in cardiac patients, particularly those with heart failure or hypertension.

Benzodiazepines: Role and Limitations

Benzodiazepines provide rapid anxiolysis and are sometimes appropriate for acute anxiety management — particularly in palliative care settings, procedural anxiety, and acute panic. However, their use in chronic medical illness requires extreme caution. In epilepsy, benzodiazepines are both antiepileptic and anxiolytic, creating a complex therapeutic landscape; however, tolerance develops rapidly to their anxiolytic effects, and withdrawal can precipitate seizures. In cardiac disease, benzodiazepines may reduce sympathetic activation acutely but do not improve long-term cardiac outcomes. In HIV, cognitive side effects compound existing HAND vulnerability, and benzodiazepines metabolized via CYP3A4 (midazolam, triazolam) are contraindicated with protease inhibitors due to profound interaction. In cancer, short-acting agents (lorazepam, alprazolam) are commonly used for anticipatory nausea and procedural anxiety, with appropriate time-limited use. The general recommendation across guidelines is to limit benzodiazepine use to short-term, as-needed protocols and to prioritize SSRIs/SNRIs for sustained treatment.

Other Pharmacological Options

Buspirone, a 5-HT1A partial agonist, is an option for GAD in medical populations. It lacks sedation, cognitive impairment, and drug interaction concerns seen with benzodiazepines. However, its efficacy is modest (NNT ≈ 7–8 for GAD) and onset is delayed (2–4 weeks), limiting its clinical utility for acute anxiety. Mirtazapine is useful in cancer patients with comorbid anxiety, insomnia, and anorexia due to its antihistaminic, anti-serotonergic (5-HT2A/2C, 5-HT3), and appetite-stimulating properties. Pregabalin, an α2δ voltage-gated calcium channel ligand, has demonstrated efficacy for GAD (NNT ≈ 5–7) and may be particularly useful in patients with comorbid neuropathic pain — common in cancer, HIV, and epilepsy populations. In epilepsy, pregabalin carries additional benefit as an adjunctive antiseizure medication, though it can cause sedation and weight gain. Gabapentin has a weaker evidence base for anxiety but is used clinically in similar contexts. Hydroxyzine, an H1 antihistamine, has modest evidence for GAD and may be appropriate for patients in whom SSRIs are contraindicated, but carries anticholinergic burden.

Psychotherapy is a cornerstone of anxiety management in chronic medical illness, with several modalities demonstrating efficacy in disease-specific populations.

Cognitive-Behavioral Therapy (CBT)

CBT has the strongest evidence base for anxiety disorders in medically ill populations. A meta-analysis by Tatrow and Montgomery (2006) found that CBT produced significant reductions in anxiety among cancer patients, with a pooled effect size (Cohen's d) of approximately 0.42 for anxiety outcomes — a moderate effect. CBT adaptations for medical populations typically emphasize: cognitive restructuring of health-related catastrophic cognitions; behavioral activation to counter illness-related avoidance and deconditioning; graded exposure to feared medical situations (scans, blood draws, hospital environments); and relaxation training including diaphragmatic breathing and progressive muscle relaxation. In cardiac disease, a meta-analysis by Whalley and colleagues (2011) for the Cochrane Collaboration found that psychological interventions (predominantly CBT-based) reduced anxiety symptoms in CHD patients with a small-to-moderate effect, though effects on cardiac events were inconsistent. In epilepsy, adapted CBT protocols targeting seizure fear, anticipatory anxiety, and social avoidance have shown promise in small RCTs, with a systematic review by Gandy and colleagues (2013) supporting efficacy. In HIV, CBT has demonstrated efficacy for anxiety with effect sizes comparable to primary psychiatric populations (d = 0.40–0.60) in multiple trials.

Acceptance and Commitment Therapy (ACT)

ACT is increasingly applied in chronic illness populations, with a theoretical emphasis on psychological flexibility — the ability to remain engaged with valued activities despite the presence of anxiety and disease-related distress. ACT may be particularly well-suited to chronic illness contexts where anxiety cannot be entirely eliminated because the threat (disease progression, mortality) is real. A growing evidence base supports ACT for anxiety in cancer (Graham et al., 2016), with preliminary trials showing improvements in anxiety and quality of life. Effect sizes appear comparable to CBT (d = 0.40–0.55), though head-to-head comparisons in medical populations are limited.

Mindfulness-Based Stress Reduction (MBSR) and Mindfulness-Based Cognitive Therapy (MBCT)

MBSR, originally developed by Kabat-Zinn for chronic pain and medical illness populations, has substantial evidence in cancer. A meta-analysis by Piet and colleagues (2012) found that MBSR produced significant reductions in anxiety in cancer patients with an effect size of approximately 0.37. MBCT, which combines mindfulness with cognitive therapy elements, has shown similar effects. These interventions appear to reduce anxiety through improved interoceptive awareness, reduced rumination, and enhanced prefrontal regulation of amygdala activity — mechanisms supported by neuroimaging studies. In cardiac rehabilitation populations, mindfulness-based interventions have shown modest but significant anxiety reduction.

Disease-Specific Psychotherapeutic Considerations

In cardiac disease, psychoeducation about the difference between panic and cardiac symptoms is a critical therapeutic component. Cardiac rehabilitation programs that incorporate psychological components show better anxiety outcomes than exercise alone. In epilepsy, psychoeducation about seizure triggers, the development of seizure action plans, and systematic desensitization to seizure fear are important components. In HIV, interventions addressing HIV-related stigma, disclosure anxiety, and trauma histories are essential; trauma-focused CBT or prolonged exposure may be necessary for comorbid PTSD. In palliative cancer care, meaning-centered psychotherapy (developed by Breitbart and colleagues) and dignity therapy address existential anxiety, which is poorly captured by standard anxiety disorder diagnoses but causes profound suffering.

Direct head-to-head comparisons of pharmacotherapy versus psychotherapy for anxiety in specific chronic illness populations are sparse. Most evidence derives from extrapolation of general anxiety disorder literature combined with disease-specific trial data.

In the general anxiety literature, a network meta-analysis by Bandelow and colleagues (2015) found that both CBT and pharmacotherapy (SSRIs/SNRIs) were superior to placebo for GAD, with comparable effect sizes. However, pharmacotherapy showed faster onset of effect, while CBT demonstrated more durable benefits after treatment discontinuation. This general pattern likely holds in medical populations, with several disease-specific considerations influencing treatment selection.

When to Prioritize Pharmacotherapy

Pharmacotherapy may be preferred when anxiety is severe (GAD-7 ≥15), impairing medical treatment adherence, or accompanied by significant comorbid depression. In patients with limited access to specialized psychotherapy — common in rural areas and resource-limited settings — pharmacotherapy is more immediately accessible. In cardiac disease, SSRIs may confer secondary cardiovascular benefit through antiplatelet effects and autonomic modulation, though this remains debated. In epilepsy, pregabalin offers the advantage of dual antiseizure and anxiolytic properties.

When to Prioritize Psychotherapy

Psychotherapy is preferred when patients have mild-to-moderate anxiety, prefer non-pharmacological approaches, have concerns about drug interactions or polypharmacy (common in all four populations), or when anxiety is predominantly characterized by avoidance behaviors that respond well to exposure-based interventions. In patients with cardiac anxiety and ICD-related fear, specialized CBT protocols have demonstrated efficacy superior to usual care, with NNT estimates of 3–5 for clinically meaningful anxiety reduction in some trials. In cancer survivors with fear of recurrence — a highly prevalent and functionally impairing anxiety presentation — fear of cancer recurrence (FCR)-specific interventions based on metacognitive and acceptance-based frameworks show effect sizes of d = 0.40–0.70.

Combined Treatment

Combined pharmacotherapy and psychotherapy is standard for moderate-to-severe anxiety in these populations, mirroring evidence from the general anxiety literature. The combination likely provides the broadest benefit, addressing both the neurobiological and cognitive-behavioral dimensions of anxiety. However, evidence from the STAR*D trial framework and analogous studies highlights that treatment response in the context of medical comorbidity is often slower and requires more treatment steps than in primary psychiatric populations.

Understanding prognostic factors is essential for clinical decision-making and patient counseling in the management of anxiety comorbid with chronic illness.

Predictors of Poor Outcome

Several factors consistently predict poorer anxiety outcomes across medical populations: prior psychiatric history (particularly pre-existing anxiety or mood disorders) is one of the strongest predictors of persistent anxiety following medical diagnosis; disease severity and progression, including advanced cancer stage, AIDS-defining illness, refractory epilepsy, and NYHA Class III-IV heart failure, predict more severe and treatment-resistant anxiety; social isolation and low social support are robust predictors across all conditions, with a meta-analysis in cancer showing that perceived social support has a correlation of r = −0.30 to −0.40 with anxiety severity; pain, present in up to 60% of cancer patients and 30–50% of HIV and epilepsy patients, independently predicts anxiety severity and treatment resistance; and polypharmacy and medication-induced neuropsychiatric effects create treatment-emergent anxiety that may be misattributed to the primary anxiety disorder.

Predictors of Good Outcome

Modifiable predictors of favorable outcome include: early identification and treatment initiation — anxiety that is recognized and treated within the first three months of medical diagnosis has better outcomes than chronic, untreated presentations; treatment adherence to both psychiatric and medical regimens; physical activity, which has anxiolytic effects mediated through BDNF upregulation, HPA axis normalization, and vagal tone improvement — cardiac rehabilitation programs that include exercise show significant anxiety benefit; stable disease, with well-controlled HIV viral load, seizure freedom, and cardiac stability each predicting reduced anxiety over time; and psychological flexibility and active coping strategies, which predict better adjustment across conditions.

Long-Term Trajectory

Longitudinal studies suggest that anxiety in chronic illness follows heterogeneous trajectories. In cancer, Lam and colleagues (2018) identified four distinct anxiety trajectories over the first two years post-diagnosis: low-stable (approximately 40%), recovery (30%), chronic (20%), and late-onset (10%). In cardiac disease, anxiety tends to peak in the peri-event period and decline over 6–12 months in most patients, but approximately 20–25% develop chronic anxiety persisting beyond one year. In epilepsy, anxiety may be relatively stable when seizure control is achieved but fluctuates with breakthrough seizures and medication changes. In HIV, the advent of effective ART and transition to chronic disease management has shifted the anxiety profile from acute mortality-related fear to chronic concerns about stigma, adherence, aging, and neurocognitive decline.

Anxiety in chronic medical illness rarely occurs in isolation. The most common psychiatric comorbidity is major depressive disorder (MDD), with co-occurrence rates of 50–70% across all four conditions — substantially higher than the 40–60% overlap seen in primary psychiatric settings. This comorbidity is bidirectional and synergistic: combined anxiety and depression predict worse medical outcomes than either condition alone.

In cancer, the combination of anxiety and depression is associated with 30–40% greater odds of treatment nonadherence and significantly reduced survival in some tumor types. In cardiac disease, the prognostic impact of combined anxiety-depression on mortality exceeds that of either disorder alone, with some studies suggesting a hazard ratio of 1.5–2.0 for major adverse cardiac events.

PTSD deserves specific attention. In cancer, PTSD prevalence following diagnosis ranges from 5–20%, with higher rates in ICU survivors and those who experienced traumatic medical procedures. In HIV, pre-existing PTSD prevalence is high (20–40%) due to the disproportionate impact of HIV on populations with trauma histories (sexual minorities, people who use drugs, survivors of sexual violence). In cardiac disease, up to 15% of MI survivors and 20–30% of cardiac arrest survivors develop PTSD. In epilepsy, PTSD is less studied but likely underrecognized, particularly in patients with histories of seizure-related injuries or traumatic medical experiences.

Substance use disorders (SUDs) are comorbid with anxiety in 15–30% of medically ill patients. In HIV, alcohol and substance use are among the strongest predictors of anxiety severity and treatment non-response. In epilepsy, alcohol use is a common seizure trigger, creating a dangerous cycle where anxiety drives alcohol use, which lowers seizure threshold, which increases anxiety. In cardiac disease, tobacco dependence — driven in part by anxiety-related self-medication with nicotine — directly worsens cardiovascular risk.

Insomnia co-occurs with anxiety in 40–60% of patients across all four conditions and requires targeted treatment, as it perpetuates both anxiety and medical illness through HPA axis activation, inflammation, and immune dysregulation.

In advanced illness and palliative care settings, anxiety takes on dimensions that standard diagnostic frameworks may inadequately capture. Existential anxiety — characterized by fear of death, loss of meaning, and confrontation with non-being — is highly prevalent but does not map neatly onto DSM-5-TR categories. Breitbart's research on meaning-centered group psychotherapy (MCGP) demonstrated significant reductions in anxiety, hopelessness, and desire for hastened death in advanced cancer patients, with effects persisting at two-month follow-up (Breitbart et al., 2010). Dignity therapy, developed by Chochinov, addresses distress related to loss of purpose and legacy, with preliminary evidence of anxiety reduction in terminally ill patients.

Pharmacologically, benzodiazepines have a more prominent role in palliative care, where concerns about dependence are less relevant than symptom relief. Midazolam is used for acute anxiety crises and terminal agitation. Haloperidol, while primarily an antipsychotic, is used at low doses for anxiety with agitation in palliative settings. Ketamine has emerging evidence for rapid anxiolytic effects in palliative populations, likely mediated through NMDA receptor antagonism and rapid synaptogenesis, though data remain preliminary.

Cultural factors significantly influence the expression, recognition, and treatment-seeking for anxiety in chronic illness. Somatization of anxiety — experiencing and reporting anxiety primarily through physical symptoms — is more common in certain cultural contexts and contributes to diagnostic underrecognition. Stigma surrounding mental health in many cultural communities creates additional barriers to treatment. Language-concordant providers, culturally adapted CBT, and integration of traditional healing practices with evidence-based treatment may improve engagement and outcomes, though the evidence base for culturally adapted interventions in medically ill populations remains limited.

Despite growing recognition of anxiety in chronic medical illness, significant evidence gaps remain.

Limitations of Current Evidence

The most critical limitation is the paucity of large, disease-specific randomized controlled trials for anxiety treatment. Most evidence is extrapolated from general anxiety disorder studies or from secondary analyses of depression-focused trials in medical populations. The SADHART and ENRICHD trials in cardiac disease, while landmark, were primarily depression-focused, and their anxiety findings are secondary outcomes. In cancer, the heterogeneity of tumor types, stages, and treatments makes generalizable conclusions difficult. In HIV, the transformation of the disease through ART means that older studies may not reflect contemporary patient experiences. In epilepsy, the ILAE has called for dedicated RCTs of psychiatric interventions in PWE, noting that this population is routinely excluded from psychiatric clinical trials due to seizure risk concerns.

Emerging Research Directions

Psychedelic-assisted therapy is a major frontier. Psilocybin-assisted psychotherapy has shown dramatic anxiolytic effects in cancer-related existential distress. Two landmark trials — by Griffiths and colleagues at Johns Hopkins (2016) and Ross and colleagues at NYU (2016) — demonstrated that a single high-dose psilocybin session, combined with psychotherapy, produced rapid and sustained reductions in anxiety and depression in patients with life-threatening cancer, with approximately 60–80% showing clinically significant improvement at 6-month follow-up and effect sizes exceeding d = 1.0. These results have catalyzed Phase II and III trials.

Digital and remote interventions are expanding access. Internet-delivered CBT (iCBT) has demonstrated efficacy for anxiety in cancer survivors (effect sizes d = 0.30–0.60) and shows promise in HIV and cardiac populations. Smartphone-based ecological momentary interventions allow real-time anxiety management. Virtual reality exposure therapy is being explored for procedure-related anxiety in cancer and cardiac populations.

Biomarker-guided treatment is an aspirational frontier. Inflammatory markers (CRP, IL-6), HPA axis measures (cortisol awakening response), and neuroimaging signatures may eventually allow personalized treatment selection — directing anti-inflammatory strategies toward patients with elevated cytokine-driven anxiety and psychological interventions toward those with primarily cognitive-behavioral presentations. However, this approach remains largely theoretical.

Integrated collaborative care models — embedding psychiatric and psychological services within oncology, cardiology, neurology, and infectious disease clinics — have demonstrated superior outcomes to usual care in depression (IMPACT trial) and are being adapted for anxiety across medical settings. The challenge remains implementation, funding, and workforce availability.

Evidence across all four chronic illness populations supports the following clinical practice principles:

• Universal screening: Routine anxiety screening using validated instruments (HADS-Anxiety subscale ≥8 or GAD-7 ≥10) should be implemented at diagnosis, at treatment transitions, and periodically during follow-up. The American Society of Clinical Oncology (ASCO) recommends distress screening at pivotal medical visits; similar guidelines exist from the American Heart Association for depression screening in cardiac patients and should be extended to anxiety.
• Differentiate anxiety disorders from normal distress: Not all anxiety in the medically ill requires treatment. Key indicators for clinical intervention include functional impairment, interference with medical treatment, persistence beyond expected adjustment periods, and severity meeting syndromal thresholds.
• Address modifiable contributors: Review all medications for anxiogenic potential. Optimize pain management, treat insomnia, and address substance use. Ensure thyroid function is assessed. These interventions alone may substantially reduce anxiety severity.
• Pharmacotherapy selection should be disease-informed: Use SSRIs with favorable interaction profiles (sertraline, escitalopram) in most populations. Avoid paroxetine and fluoxetine with tamoxifen. Monitor QTc with citalopram in cardiac patients. Consider pregabalin in epilepsy and neuropathic pain contexts. Limit benzodiazepines to short-term, specific indications.
• Psychotherapy should be offered as first-line or adjunctive treatment: CBT, ACT, and mindfulness-based interventions have evidence across populations. Adaptations for medical illness contexts improve relevance and engagement. Existential and meaning-centered approaches are appropriate for advanced illness.
• Integrate care: The most effective models embed mental health within medical care teams rather than relying on referral to separate psychiatric services. Collaborative care, co-location of services, and systematic care pathways improve detection and treatment rates.
• Monitor longitudinally: Anxiety trajectories are dynamic in chronic illness. Patients who are resilient early may develop anxiety at later disease stages, during recurrence, or in survivorship. Longitudinal monitoring allows timely intervention.