Anxiety in the Medically Ill: A Systematic Review of the Literature

Background Although anxiety is highly represented in the medically ill and its occurrence has relevant clinical implications, it often remains undetected and not properly treated. This systematic review aimed to report on anxiety, either symptom or disorder, in patients who suffer from a medical illness. Methods English-language papers reporting on anxiety in medically ill adults were evaluated. PubMed, PsycINFO, Web of Science, and Cochrane databases were systematically searched from inception to June 2021. Search term was “anxiety” combined using the Boolean “AND” operator with “medically ill/chronic illness/illness/disorder/disease.” Risk of bias was assessed via the Joanna Briggs Institute (JBI) Critical Appraisal Tools—Checklist for Prevalence Studies. The PRISMA guidelines were followed. Results Of 100,848 citations reviewed, 329 studies met inclusion criteria. Moderate or severe anxious symptoms were common among patients with cardiovascular, respiratory, central nervous system, gastrointestinal, genitourinary, endocrine, musculoskeletal system or connective tissue, dermatological diseases, cancer, AIDS and COVID-19 infections. The most common anxiety disorder was generalized anxiety disorder, observed among patients with cardiovascular, respiratory, central nervous system, dermatologic diseases, cancer, primary aldosteronism, amenorrhea, and COVID-19 infection. Panic disorder was described for cardiovascular, respiratory, dermatology diseases. Social anxiety was found for cardiovascular, respiratory, rheumatoid diseases. Specific phobias were relatively common in irritable bowel syndrome, gastroesophageal reflux, end-stage renal disease. Conclusion Anxiety is a major challenge in medical settings. Recognition and proper assessment of anxiety in patients who suffer from a medical illness is necessary for an appropriate management. Future reviews are warranted in order also to clarify the causal and temporal relationship between anxiety and organic illness.


INTRODUCTION
Anxiety is a feeling characterized by anguish, sense of threat, and fear. When it is explained by a real and objective trigger, it is considered physiological. When there are no objective reasons of being in such status, anxiety becomes pathological (1). For instance, when an organic disease occurs, anxiety can be a normal psychological reaction (2) but it can also flourish and evolve into a symptom with a pathological meaning or into a mental disorder. Anxiety, indeed, is highly represented in the medically ill (3), with generalized anxiety disorder as the most prevalent disorder (10.3%) in primary care settings (4). The pathways between anxiety and physical illness co-occurrence are not fully understood. Several possible mechanisms and synergies exist. Among them, environmental and genetic factors (5) have been proposed as being able to favor such co-occurrence as well as individual vulnerability (6) and socio-economic status (7).
In addition, anxiety may influence how the patient experiences the pathological process of own medical illness and his interaction with others (8,9), including medical and nursing staff (10). In particular anxiety in association with a chronic medical illness worsens the quality of life (11), affects social functioning (12), increases medical burden (13). Anxiety has a negative impact on compliance (14), resulting in exacerbation of illness (15) and high health care utilization and costs (16). Anxiety increases the susceptibility to illness leading to illness progression, rehospitalization, and mortality (17)(18)(19)(20). Among chronic illness patients, anxiety negatively affects emotional stability resulting in depressive symptoms, suicidal ideation, and social isolation (21,22). Several studies reported the strong association between anxiety and somatization (23)(24)(25). In medically ill patients, anxiety amplifies physical symptoms, leading to useless (if not dangerous from a physical or psychological point of view) and inappropriate invasive tests/procedures to investigate hypothetical, but never confirmed, organic explanations (26). The treatment of anxiety, whether pharmacological or psychological, was found to favorably affect the outcome of a number of organic diseases (27,28).
Unfortunately, anxiety often remains unacknowledged, unrecognized, untreated in medically ill patients (29,30). Such a phenomenon is the result of several converging factors. First, the differentiation of anxiety worthy of clinical attention is hindered by the widespread occurrence of non-clinically relevant anxious symptoms in medical settings (31). Secondly, when an anxiety disorder is associated with a medical illness, there is a tendency to regard it as a physiological psychological reaction, secondary to the distress of the medical illness or to the patient's awareness of its consequences (32). However, in the clinical realm it is evident that not all medically ill patients have anxiety or an anxiety disorder (33). In addition, the expression of emotional distress is often disregarded or even discouraged by clinicians, and patients' needs are satisfied if they refer to the body rather than to the psychological sphere (34). Finally, the use of anxiolytics, particularly benzodiazepines, is widespread, especially during hospitalizations (35), often without a real indication. On the contrary the prescription seems justified by the organizational limits of hospital, being the staff able to handle only a limited number of requests.
Although anxiety is highly represented in the medically ill and its occurrence has relevant clinical implications, no systematic reviews on its prevalence and rate seem currently available. To fill this gap, the aim of the present systematic review was to report on anxiety, either symptom or disorder, in patients who suffer from a medical illness.

Registration
This review protocol was registered in the "International Prospective Register of Systematic Reviews" (PROSPERO) in 2021, under the registration number: CRD42021296741, and available at: https://www.crd.york.ac.uk/prospero/display_ record.php?ID=CRD42021296741, and not published elsewhere.

Eligibility Criteria
Eligible articles included English-language papers published in peer-reviewed journals reporting data on anxiety in medically ill adults. Anxiety disorder had to be diagnosed according to the Diagnostic and Statistical Manual of mental disorders (DSM-III, -III R, -IV, -IVTR or -5th edition) or the International Classification of Diseases (ICD-9, -10, -11). Anxious symptoms had to be assessed via standardized rating scales.
Additional inclusion criteria were: age of at least 18 years, a sample of at least 10 subjects (as it was already used in 36 in order to guarantee and minimum representativeness of results). Studies with different designs were included (i.e., cross-sectional, longitudinal study, observational, case-control studies).
Exclusion criteria were: (a) patients with multiple organic diagnoses, (b) not original data, (c) non-clinical samples (d) results on anxiety aggregated with results on depression or other psychological features. Treatment outcome studies were not included, being off topic for the present review.

Information Sources and Search Strategy
The following electronic databases were systematically searched from inception to June 2021: PubMed, PsycINFO, Web Of Science, and Cochrane. In addition, a manual search of reference lists from relevant reviews was done. Search term was "anxiety" combined using the Boolean "AND" operator with "medically ill/chronic illness/illness/disorder/disease."

Selection and Data Collection Process, Data Items
Titles and abstracts were screened by two authors (S.R. and G.M.). Articles potentially relevant were retrieved and the authors independently assessed each in full. Any disagreement was resolved by consensus. Risk of bias (quality of the studies) was assessed via the Joanna Briggs Institute (JBI) Critical Appraisal Tools-Checklist for Prevalence Studies (36). It consists of 9 questions and the scoring system is: "yes" scores 1, "no" or "not clear" or "not applicable" score 0. The JBI Checklist-Total score is the sum of the items.

Effect Measures
The search strings, the list of relevant reviews, the data coding, and the quality criteria are available on request to the corresponding author. No missing data were found. The methods described fulfilled the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (37).

RESULTS
The search provided a total of 100,848 citations. After reviewing the abstracts to exclude those which clearly did not meet the criteria, 1,289 remained. Of these, 960 were excluded, not meeting the inclusion criteria ( Figure 1 shows the flow diagram of the search). A total of 329 studies were identified for inclusion, the following parameters were extracted: country/city where data were collected, study design, organic disease diagnosis, sample size, instrument used to assess anxiety, eventual psychopathological manifestations co-occurring with anxiety (for details see Supplementary Table S1). Among them, 7 studies reached the maximum JBI score, 68 had a score of 8, 244 studies had a score of 7, and 10 studies had a score of 6, which means that at least 6 out of 9 criteria were fulfilled. Results will be here qualitatively presented based on the apparatus involved in the organic illness (see also Tables 1, 2). Symptoms severity levels were taken directly from respective studies and may refer to the use of different cut-offs for the same scale of assessment. No meta-analysis was conducted due to the methodological . About 35% of atrial fibrillation patients showed high levels of anxiety, in particular women (49%). Patients having "bad" relations with nursing staff had 6.58 times higher possibility to experience anxiety than patients having "very good" relations (p  Among patients with angiographically normal or near normal coronary arteries, 34% met criteria for current PD and 11% presented specific phobia (Beitman et al., 1989 * ).
Clinically significant anxiety was also found in 31.8% of patients with incidental pulmonary nodules (Li et al., 2020b * ) and in 12% of those with interstitial lung diseases (Holland et al., 2014 * ).

Central Nervous System Diseases
The rate of anxiety among stroke patients was found to increase significantly between 6 months (17%) and 5 years after the event (29%) (Lincoln et al., 2013 * ). Point prevalence of DSM-III-R anxiety disorders was higher in the stroke survivors than in healthy controls. The most common diagnoses were: PD (24 vs. 8%) and GAD (27 vs. 8%). The prevalence of any anxiety disorder was higher in the stroke group (42 vs

Gastrointestinal Diseases
Anxiety was present in 28% of gastroenterological patients (Alosaimi et al., 2014 * ). Severe anxiety was observed in 27% of patients with chronic digestive system diseases. Patients with digestive system tumors had the highest rate of anxiety (55. The incidence of severe anxiety was significantly higher in the non-erosive reflux disease group (16.5%) than in the reflux esophagitis one (10.4%) (Yang et al., 2015 * ). Among gastroesophageal reflux disease patients, a current clinical anxiety symptomology was found in 20.7% of cases (On et al., 2017 * ). Among gastroesophageal reflux disease patients, anxiety disorders were diagnosed in 30% of cases: 11.6% met the diagnostic criteria for specific phobias, 5.6% for lifetime social phobia, 5.6% for other phobic anxiety disorders, 3.8% for GAD, and 3.8% for agoraphobia with panic attacks ( In patients with 79 different rare diseases, the rate of anxious symptoms was 23% and females presented significantly more severe symptomatology than males (Uhlenbusch et al., 2019 * ).
Patients with systemic hypertension and type 2 diabetes had mild anxiety in 32% of cases, moderate anxiety in 29%, and severe anxiety in 26% of cases (Teixeira et al Among inpatients with amenorrhea, the prevalence of GAD was 23.5% (Fava et al., 1984 *

DISCUSSION
Moderate or severe anxiety occurs particularly among patients with chronic kidney disease, end-stage renal disease, hip pathology, systemic lupus erythematosus patients, hereditary angioedema and chronic urticarial, metastatic breast cancer, bladder cancer. Severe anxiety had the highest rates among patients with chronic illnesses, atrial fibrillation, coronary artery bypass graft, chronic thromboembolic pulmonary hypertension, pulmonary hypertension, chronic rhino sinusitis, asthma, migraine, multiple sclerosis, epilepsy, digestive system tumors, liver cirrhosis, irritable bowel syndrome, obesity, type 2 diabetes, hyperprolactinemia, COVID-19 infection.
The most common anxiety disorder was GAD, which was observed among patients with cardiovascular, respiratory, CNS, dermatologic diseases, and also in cancer, primary aldosteronism, amenorrhea, COVID-19 infection. Patients with cardiovascular, respiratory, or dermatology diseases also presented PD. Social anxiety was described for cardiovascular, respiratory, rheumatoid diseases. Specific phobias were relatively common in irritable bowel syndrome, gastroesophageal reflux, end-stage renal disease.
The present results should be considered as an overview of such clinical phenomenon which is in need of being further explored clarifying, among the others, temporal or causal relationships between anxiety and organic illnesses. In addition, we found studies referring to different levels of severity (i.e., mild/moderate/severe) of anxious symptoms that should be taken into account since the various levels of severity may impact the discomfort and the functioning of patients differently. In addition, such heterogeneity of severity, which was measured via different tools and in some cases also using different cut-offs for the same tool, suggest caution in interpreting the results and in using them to draw conclusions in comparability. Future studies increasing the body of evidence for each level of severity of anxiety in each medical disease are warranted to overcome this limitation. Also research aimed at disentangling between a physiological anxious reaction to the physical illness and a pathological, thus for instance maladaptive, response to the status of being medically ill are needed.
Anxiety represents a major challenge in medical settings, being highly represented either as a symptom or as a disorder (39). Nowadays, anxiety can be properly assessed in the medically ill via clinician-or self-reported measures (40,41). This may provide information on the overall health condition, also according to a longitudinal view of development of disorders (42,43), thus demarcating major prognostic and therapeutic differences among patients who otherwise might seem to be deceptively similar since they share the same diagnoses. It also allows to catch the possible interplay between mental and organic disease, for instance clarifying if there is a primary/secondary relationship (44,45). It can help verifying whether the patient is at risk of developing depression, which often coexists with anxiety (46,47) and worsen its prognosis. It allows to investigate other areas associated to anxiety. Among them, it allows to investigate illness behavior, the ways in which individuals experience, perceive, evaluate, and respond to their health status (48,49). This is a transdiagnostic core characterization (50), with multiple expressions (51,52), providing an explanatory model for clinical phenomena (8). Relevant information can be obtained also assessing mental pain (53,54), which captures a feeling state characterized by emotional pain, emptiness, and internal perturbation (55), sometimes at the core of the suicidal process (56). Finally, a comprehensive assessment may also include evaluating specific positive features, i.e. psychological well-being (57,58), which can be eventually empowered to cope with anxiety (59) and the organic disease (60,61).
Recognition and proper assessment of anxiety represent the necessary steps for its appropriate management. Clinicians boast a large and effective armamentarium to treat anxiety, which include both pharmacological (e.g., benzodiazepines) (62,63) and non-pharmacological (e.g., wellbeing therapy) interventions (59), they need to use it.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author/s.