Biological and Clinical Markers in Panic Disorder

Article information

Psychiatry Investig. 2019;16(1):27-36
Publication date (electronic) : 2018 September 5
doi :
1Department of Health Sciences, University of Florence, Florence, Italy
2Maastricht University Medical Center, Department of Psychiatry & Psychology, School for Mental Health & Neuroscience, Maastricht, the Netherlands
Correspondence: Fiammetta Cosci, MD, PhD Department of Health Sciences, University of Florence, via di San salvi, 12 50135 Florence, Italy Tel: +39 (0)55 2757066, Fax: +39 (0)55 6236047 E-mail:
Received 2018 June 28; Revised 2018 July 23; Accepted 2018 July 26.



Classifying mental disorders on the basis of objective makers might clarify their aetiology, help in making the diagnosis, identify “at risk” individuals, determine the severity of mental illness, and predict the course of the disorder. This study aims to review biological and clinical markers of panic disorder (PD).


A computerized search was carried out in PubMed and Science Direct using the key words: “marker/biomarker/clinical marker/neurobiology/staging” combined using Boolean AND operator with “panic.” In addition, the reference lists from existing reviews and from the articles retrieved were inspected. Only English language papers published in peer-reviewed journals were included.


Structural changes in the amygdala, hippocampus, cerebral blood level in the left occipital cortex, serotonin 5-TH and noradrenergic systems activation, aberrant respiratory regulation, hearth rate variability, blood cells and peripheral blood stem cells, hypothalamic–pituitary–adrenal axis dysregulation were identified as potential candidate biomarkers of PD. Staging was identified as clinical marker of PD. According to the staging model, PD is described as follows: prodromal phase (stage 1); acute phase (stage 2); panic attacks (stage 3); chronic phase (stage 4).


The clinical utility, sensitivity, specificity, and the predictive value of biomarkers for PD is still questionable. The staging model of PD might be a valid susceptibility, diagnostic, prognostic, and predictive marker of PD. A possible longitudinal model of biological and clinical markers of PD is proposed.


In the last decade increasing efforts have been made to classify mental disorders on the basis of objective makers [1-4] claiming that markers can clarify the aetiology of psychiatric diseases, confirm a diagnosis, identify “at risk” individuals, determine the severity of mental illness, predict the course of the disorder [1,2,5]. Some authors also suggested that the use of markers might lead to personalized psychiatric treatments approach and inform about the type, timing, and course of interventions to be used as well as monitoring the clinical response to them [1,2,5]. Of course, markers must have at least a moderate level of sensitivity, specificity, and predictive value [2] to be useful. In this vein, markers were classified as: 1) susceptibility/risk marker, i.e., the marker indicates the potential for developing a disease in an individual who, from a clinical standpoint, does not have that disease or medical condition clinically relevant, yet; 2) diagnostic marker, i.e. a marker used to identify individuals with the disease or to define a subset of the disease; 3) prognostic markers, i.e., markers used to identify the likelihood of a clinical event, disease recurrence or progression in patients who have the disease or medical condition of interest; 4) predictive markers, i.e., markers used to identify individuals who are more likely than those without the marker to experience a favourable or unfavourable effect from specific intervention or exposure; it provides a forecast of the potential for a patient to respond to one or more specific treatments [6-8].


A computerized search was carried out in PubMed and Science Direct. Search terms were: “marker/biomarker/clinical marker/neurobiology/staging” combined using Boolean AND operator with “panic.” In addition, the reference lists from existing reviews and from the articles retrieved were inspected. Only English language papers published in peer-reviewed journals were included.


Biological markers in panic disorder

A biomarker is defined as a characteristic that is objectively measured and evaluated as an indicator of normal biologic processes, pathologic processes, or biological responses to a therapeutic interventions [3,4]. We will here illustrate the biological markers proposed for panic disorder (PD) on the basis of the literature.

Structural or activity changes in brain regions

According to Bandelow et al. [9,10], structural changes in the volume of amygdala, hippocampus, parahippocampal gyri, and brainstem nuclei are candidate biomarkers of panic disorder. Volume reduction or increase in these areas as well as increased activation of amygdala and hippocampus in response to fearful stimuli have been reported in PD patients but not in controls (p value ranging from 0.002–0.005) [9,10]. Some evidences suggested a greater activation in PD patients than in controls at the level of the insular cortices, of the left inferior frontal gyrus, of the dorsomedial prefrontal cortex, and of the left caudatum [9,10]. A negative correlation between left amygdala volume and anxiety was found in PD patients (r=-0.54; p=0.02) [11].

The cerebral blood flow in the left occipital cortex and the serotonin (5-TH) and noradrenaline (NE) system activation were also proposed as putative biomarkers [9,10]. Compared to controls, patients with PD showed a higher cerebral blood flow (mean±SD 1.35±0.01 vs. 1.24±0.02, p<0.05) [12] and lower 5-TH plasma levels [9,10] in addition a significant association between a hyperactivation of the NE system, anxiety, and somatic symptoms was found in PD patients but not in patients with other psychiatric disorders (e.g., generalized anxiety disorder, obsessive compulsive disorder, depression and schizophrenia) [9,10,13]. These findings were consistent with Gorman’s neuroanatomical hypothesis of PD [14].

Respiratory patterns

Several theories sharing the hypothesis of a causal relationship between aberrant respiratory regulation and panic have been developed [15-19]. Both respiratory symptoms during panic attacks and behavioral and respiratory hypersensitivity to hypercapnic gas mixture inhalation have been found in PD patients [15-19]. In these subjects, hyperventilation may be chronic [16]. In addition, hyperventilation (e.g., higher baseline mean minute ventilation, lower end-tidal partial pressure of CO2, and lower venous pCO2), higher variability of mean minute ventilation, higher respiration rate, higher tidal volume, higher rate of sighs and apnoea in respiratory patterns were found in PD patients but not in healthy controls (effect size Hedges’ g ranging from -0.73 to 0.39, p<0.01). Grassi et al. [17] verified whether these respiratory abnormalities are specific to PD or are also present in anxiety disorders other than PD. Lower baseline mean et-pCO2, indicating a condition of hyperventilation (effect size Hedges’ g ranging from -0.28 to -0.56, p<0.01) and higher respiration rate (effect size Hedges’ g ranging from 0.25 to 0.47, p<0.01) were found in PD patients but not in those with social phobia or generalized anxiety disorder [17]. Thus, the aberrant respiratory pattern (i.e., hyperventilation and higher respiration rate) seems to be a biomarker specific for PD.

Hearth rate variability, blood cells, peripheral blood stem cells

The hearth rate variability (HRV) (i.e., the extent to which the interval between beats varies with time) was also assumed as a candidate biomarker for PD [9,20]. HRV is a core feature of cardiovascular diseases [21] which was found to be positively correlated with panic attacks across clinical and non-clinical samples [9,22,23]. The HRV was considered a possible core feature of PD [9,10,22,23] on the basis of studies reporting an increased risk of cardiovascular diseases in PD patients than in subjects without psychiatric disorders [24] and a prevalence of panic disorder ranging from 4% to 12.5% in cardiac outpatients. HRV is measured using the time domain (i.e., differences between adjacent beat intervals) and frequency domain (i.e., measures based on power spectral analysis, which allows detection of lower and high frequency oscillation) [9,10,20]. A meta-analysis [20] reported that time and frequency domains were significantly lower in patients with PD than in healthy controls, beyond potential confounding effects of medication use and medical and psychiatric co-morbidity (effect size Hedges’ g ranging from -0.69 to -0.29, p<0.05). Chalmers et al. [20] reported similar findings comparing patients with anxiety disorders other than PD (i.e., post traumatic stress disorder, generalized anxiety disorder, social anxiety) with healthy controls; they showed that HRV is an aspecific marker across anxiety disorders [9,10,20].

Recent findings [25] observed increased platelet distribution width (PDW), red cell distribution width (RDW), and mean platelet volume (MPV) in PD patients if compared to healthy controls. These results emphasized the role of blood cells and peripheral blood stem cells as possible biomarkers of PD [25-27] in accordance with the theory of the inflammatory origin of PD [26].

Hypothalamic–pituitary–adrenal axis dysregulation

It has been assumed that panic attacks might be a result of a disturbance in stress response regulation by the strong activation of the hypothalamic–pituitary–adrenal (HPA) axis [28]. The HPA axis is the major endocrine system which regulates the physiological response to stress and as a result drive how an organism might adapt its own behavior or environment in order to copy with that stress [29-31]. When the stress is persistent, the negative feedback system, which dampens the HPA axis activation, is impaired inducing chronic cortisol release [29,32] which may provoke flatter circadian variation and heightened daily cortisol secretion [29-32].

Studies addressing the role of HPA axis activation in panic disorder used cortisol secretion as index of the HPA functioning during panic attacks or compared PD patients with controls. These studies provided inconsistent results. Some reported a higher cortisol secretion during panic attacks compared to the values obtained in the same individuals at comparable times on panic-free days [28,33]. Other studies reported unchanged or only marginal HPA axis activation during spontaneous attacks [28]. In addition, while some evidences reported higher cortisol secretion in PD patients when compared to controls, other studies showed comparable levels of cortisol between PD patients and healthy controls [9,10,28,32,34]. As pointed out by Bandelow et al. [35], it is not clear whether the dysfunctions of the HPA axis are a potential cause of PD or a consequence of permanent stress induced by recurrent panic attacks.

These heterogeneous findings might be explained by several elements; first, at the beginning of the panic attack the HPA axis activation might be due to an arousal in reaction to novelty cues and/or to anxiety anticipatory about further attack and/or the avoidance of places where having an attack is embarrassing develop [32,34,36,37]; second, after the acute phase of a panic attack, the HPA axis might normalize due to a successful habituation to the repeated experiences of panic [32,34,38]. However, the role of cortisol secretion as biomarker of PD is not fully confirmed.

Classification of putative biomarkers of PD

As previously reported, markers might be categorized as susceptibility/risk markers; diagnostic, prognostic, and predictive markers [6-8]. Given that the biomarkers illustrated were found in PD patients and not in controls [9,10,16,17,20,25], they might be considered susceptibility and/or diagnostic biomarkers, however longitudinal studies are warranted to confirm this. Also the HPA axis dysregulation and the heart rate variability might be considered prognostic biomarkers given that higher cortisol secretion was found to predict poorer long-term outcome in PD patients at 2–4 year follow-up (p<0.003) [39,40] and lower HRV in PD patients might increase vulnerability to cardiovascular diseases [41]. Predictive biomarkers of PD are still unclear due to the inconsistency of results. Grambal et al. [42] suggested that increased pre-treatment activation of the dorsolateral prefrontal cortex, right parietal cortex, left frontal eye field, orbito-frontal cortex, and left amygdala predict poor outcome in cognitive behaviour therapy; Bandelow et al. [9] and Fisher et al. [43] reported that neither brain regions structural changes nor basal cortisol concentrations predict responses to psychological treatments. In Table 1 we reported a proposal for classifying PD biomarkers.

Putative makers of panic disorder classified as: susceptibility/risk marker; diagnostic marker, prognostic marker, predictive marker [6-8]

The biomarker crisis

A body of research on biomarkers considered single brain region/circuit or a specific neurotransmitter without ravelling one-to-one relationships with a specific disease [4]. Given that mental disorders are multidimensional in their description, multifactorial in their origins, and involve non-linear interactions in their development [44], it is unlikely that a single biomarker might explain the multifaceted nature of a specific psychiatry disease [2,45]. Multimodal approach where the diagnosis and/or the course of diseases are explained by a combination of different biomarkers could be more reliable [2,4,45]. Moreover, it is likely that different biomarkers are associated with a cluster of symptoms rather than to a specific diagnosis [2,4,45]. A further hurdle is that some biomarkers, as for example structural changes in the amygdala and/or hippocampus, although they showed acceptable reliability [9,10], can be used in clinical practice rarely for practical and economic reasons [45]. To facilitate the use of biomarkers on a broad-based scale, we would need simple and cost-effective biomarkers [2,45] such as for instance urine or saliva cortisol, heart rate variability, blood cells, peripheral blood stem cells, or respiratory pattern [9,10,16,17,20,25]. However, although the biomarkers mentioned for PD showed high sensibility in distinguishing PD from the healthy condition, they did not show enough specificity in distinguishing PD from other psychiatric disorders [9,10]. Basically, the identification of biomarkers is based on observations that the specific biomarker is detected in PD patients and not in healthy controls or in patients with other psychiatric disorders [9,10,45]. This might lead to interpretative bias given that, due to an overlap in pathophysiological findings among psychiatric disorders, biomarkers might be shared by different psychiatry disorders. Most of the candidate biomarkers described above (e.g., structural brain morphology, lower 5-HT plasma, norepinephrine, hypo/hyper secretion of cortisol, instability of the cortical arousal system, HRV, PWD and RDW) differentiate PD patients form healthy controls but not PD patients from patients having other psychiatric disorders, such as anxiety disorders, schizophrenia, mood disorders [9,10,46-58].

The lack of availability of biomarkers with high specificity has been widely discussed [45,59,60]. As reported by Kapur et al. [60], psychiatry seems to be in a Catch-22 given that the contemporary diagnostic system was not designed to facilitate biological differentiation. In addition, current diagnostic definitions of psychiatric disorders based on symptoms encompass very heterogeneous populations and are thus likely to yield spurious results when exploring biological correlates of mental disturbances [59].

To be also noted that biomarkers are influenced by environmental and lifestyle factors such as stress, physical activity, comorbidity, psychotropic medications [45]. Psychotropic drug treatments, particularly after long-term use, might cause or precipitate adverse effects on the course, characteristics, and responsiveness of an illness that do not necessarily subside with discontinuation of the drug or of modifying responsiveness to subsequent treatments [61,62]. Such vulnerabilities are subsumed under the rubric of iatrogenic comorbidity [61,62].

Psychometric and clinimetric approach

The psychometric approach aims to develop instruments that measure a single construct using multi-items [63], given that, its customary goal is to achieve a unidimensional construct in which the relatively homogeneous components all measure essentially the same phenomenon [63,64]. Methodological difficulties in applying psychometric principles to diagnostic testing and to detect changes related to the course of illness and/or to difference between pre- and post-treatment have been outlined [64]. Wright and Feinstein [65] provide an explanation for the disappointing performance that multi-item scales may present. In psychometrics, the homogeneity of components is considered as the most important requirement for a rating scale; however, the same properties that give a scale a high score for homogeneity, as the redundant nature of the items of a scale, may obscure its ability to detect change decreasing its sensitivity [65]. A high correlation between scales is also regarded as evidence that the two scales measure the same factor, but a high correlation does not indicate similar sensitivity: a common content of two scales may ensure a high positive correlation between them, but the items they do not share may be important in determining their sensitivity [64,66]. On the basis of the psychometric approach all items of a scale have same weight [66]. Although multi-item scales might be valid and reliable, it might show lack sensitivity [66,67].

Fava et al. [66] suggested that psychometrics has now become an obstacle to the progress of clinical research in psychiatry and clinical psychology and that the more clinically oriented clinimetrics could offer a valid alternative. Clinimetrics’ is the term introduced by Feinstein in the early 1980s [68] to indicate a domain concerned with the measurement of clinical issues that do not find room in customary clinical taxonomy. Such issues include types, severity and sequence of symptoms; rate of progression in illness (staging); severity of comorbidity; problems of functional capacity; reasons for medical decisions (e.g. treatment choices), and many other aspects of daily life, such as well-being and distress [68]. Differently from the psychometric approach, in clinimetrics homogeneity of components is not needed and single items may be weighed in different ways: what matters is the capacity of an index to discriminate between different groups of subjects and to reflect changes in experimental settings [64,66]. In addition, the clinimetric approach is directed at the development of instruments to measure multiple constructs with a single index that might be divided into ailment-oriented indexes (refer to specific diseases, states and clinical manifestations) and general indexes (refer to general health and functional states that are not distinctive for a particular disease or condition) [64,66]. The clinimetric analysis proposed convergent, discriminant, and incremental validity as important features that determine the sensibility of a clinical measurement [69-71].

The current diagnostic systems, the Diagnostic and Statistical Manual of Mental Disorders (DSM) [72-76] and the International Classification of Diseases (ICD) [77-80], are mostly influenced by psychometric models according to which the severity of the disorder is determined by the number of symptoms rather than to their intensity or quality [66,81,82]. Thus, all symptoms have an equivalent value and load in determining the severity of a psychiatric disorder, unlike in clinical medicine where major and minor symptoms are often differentiated (e.g., the Jones criteria for rheumatic fever) [68]. This might bias the diagnostic process given that psychological symptoms that do not reach the diagnostic threshold might be not considered [83] although they might affect the quality of life of the patient and lead to pathophysiological and therapeutic implications [67]. In addition, different patients could meet the diagnostic criteria for the same disorder but presenting different scores on a rating scale which reflect different symptom severities, perceptions, and illness attitudes, that in turn might affect the clinical course of the disorder [67]. Differently form the psychometric models, clinimetrics assumed that different symptoms have different weigh [67]. Moreover, clinimetrics approach focuses on symptoms sequence, rate of progression of illness, and comorbidity, and on problems of functional life, such as well-being [64,66,67].

Clinical markers in panic disorder

From cross-sectional diagnostic systems to the staging model

Even thought the current diagnostic systems [72-80] showed acceptable reliability, their clinical utility remains elusive [84-86]. In a large epidemiological survey run on a sample of 1,764 users of ICD or DSM, it was evident that 1,123 subjects (64%) rated those diagnostic systems as with low utility in selecting a treatment and assessing probable prognosis [84]. This might be partially explained by the several reasons. First, the taxonomy in psychiatry derived from the traditional method of clinical medicine which provides operating specifications for making a clinical decision about the existence of a specific disease [87]. As a consequence, the diagnostic reasoning process ends with the identification of a disorder [88] while, in reality, the clinical judgment should go through a series of ‘transfer stations’ where potential connections between presenting symptoms and pathophysiological processes are drawn and are amenable to longitudinal verification and modification as long as therapeutic goals are achieved [81,89-91]. Second, the diagnostic systems are mostly influenced by psychometric models [81-83]; as a result: 1) patterns of symptoms, severity of illness, effects of comorbid conditions, timing of phenomena, rate of progression of illness (staging), responses to previous treatments, and other clinical distinctions that might demarcate major prognostic and therapeutic differences among patients who otherwise might be deceptively similar since they share the same psychiatric diagnosis are poorly taken into account; 2) the target of therapy tends to become syndromes resulting from a certain number of symptoms, that could be of mild intensity and of doubtful impact on quality of life, instead of symptoms that might be incapacitating for the patient [81]. Within a sample of patients with major depressive episode, the number of symptoms (either including all depressive symptoms or selecting only those relevant to a DSM diagnosis) does not correlate with the illness severity which was instead strongly correlated with certain symptoms than other ones [92]. Exclusive reliance on diagnostic criteria impoverishes the clinical process and does not reflect the complex thinking that underlies decisions in psychiatric practice [89]. The mental disorders are not static, sharply defined illnesses with separate aetiologies and courses, but rather syndromes that overlap and develop in stages [86,93].

Moving from a cross-sectional nosography to the longitudinal view of the development of psychiatric disorders, staging was proposed as a strategy to improve the diagnostic and treatment process in psychiatry and clinical psychology [67,68,86].

The use of staging in psychiatry

Staging [86,94] allows to define the extent of progression of a disorder at a particular point in time and where a person is currently located along the continuum of the course of the illness; define prodromes (e.g. early symptoms and signs that differ from the acute clinical phase) and residual symptoms (e.g. persistent symptoms and signs despite apparent remission or recovery); characterize a psychiatric disorder development according to different stages [94]. In 2013, Cosci and Fava [86] systematically reviewed the literature to synthesize the different models of staging available in psychiatry and clinical psychology across different diagnostic categories such as schizophrenia, unipolar depression, bipolar disorder, panic disorder, substance use disorders, anorexia and bulimia nervosa.

As stage 1 they found the prodromal phase, defined as the time interval between the onset of prodromal symptoms and the onset of the characteristic manifestations of the fully developed illness [86]. After the acute phase (stage 2), it might be difficult to assess whether partial or full remission has occurred; attenuated symptoms, the so called residual symptoms, might be observed at stage 3; they are due to partial persistence of the disorder or an aggravation of a pre-existing abnormal personality trait [86]. Stage 4 represents chronicity of the psychiatric disorder [86].

Staging as clinical marker of panic disorder

In 1993, Fava and Kellner [94] described a staging model for PD with agoraphobia based on the fact that, in a substantial proportion of patients, agoraphobia, hypochondriacal fears and beliefs, and generalized anxiety precede the first panic attack. However, given that for some patients, the first panic attack can occur without conspicuous prodromal symptoms, while anticipatory anxiety, phobic avoidance, and hypochondriasis may develop subsequently [95], Sheehan and Sheehan [96] proposed different staging process: stage 1 characterized by panic attacks with limited symptoms (subpanic); stage 2 is panic; stage 3 is hypochondriasis; stage 4 is single phobia; stage 5 is social phobia; stage 6 is agoraphobia, and stage 7 is depression.

In 2008, Fava et al. [97] proposed an updated version of the staging model of PD [94] which was described as follows: stage 1 defined by pre-agoraphobia with predisposing factors, such as health anxiety and anxiety sensitivity, genetic vulnerability, premorbid personality, hypochondriacal fears and beliefs and impaired psychological well-being; the relative weight of these factors may vary from patient to patient and lead to subtle avoidance patterns and to the stage 2, which is agoraphobia; stage 3 is characterized by the occurrence of panic attack; health anxiety may turn into hypochondriasis and/or disease phobia and/or thanatophobia; demoralization and/or major depression may occur; stage 4 in which the duration of PD with agoraphobia might predispose to the development of other psychiatric complications, as depression; agoraphobia may become more severe and hypochondriacal fears and beliefs may be accentuated.

As regard to the sub-clinical symptoms, the most common prodromal symptoms in PD were depressed mood, illness phobia, distress and avoidance of closed spaces, excessive worries, negative affectivity, anxiety sensitivity and health anxiety or fear of disease; whereas the more prevalent residual symptoms were generalized anxiety, somatic anxiety, health anxiety, low self-esteem, agoraphobia, hypochondriasis, reduced psychological and physical well-being, limited symptoms of panic attacks, anticipatory anxiety and depression [86,95,98,99]. Within the framework of the Diagnostic Criteria for Psychosomatic Research [100], the following subclinical symptoms were proposed given that all of them were found to be associated with PD and agoraphobia [99]: disease phobia, somatization, irritable mood. Disease phobia might be part of the hypochondriacal syndrome, yet they may also occur independently; disease phobia differs from hypochondriasis for three characteristics: fears concern a specific disease and are unlikely to be shifted to another disease or organ system; fears tend to manifest themselves in attacks rather than in constant worries as in hypochondriasis; it often results in the avoidance of internal and external illness-related stimuli, while hypochondriasis involves reassurance-seeking or checking behaviors [100]. Somatization is conceptualized as a clustering of somatic symptoms involving different organ systems probably due to an enhanced general sensitivity to pain and discomfort [100]. Irritable mood refers to the concept of irritability that might be part of psychiatric syndromes; it is always unpleasant for the individual and its overt manifestation lacks a cathartic effect [100].

In 2013, the staging model was updated [86] as follows (Table 2): the prodromal phase (stage 1) was defined by the presence of subclinical symptoms of agoraphobia, social phobia, generalized anxiety disorder, hypochondriasis; these symptom being stable during the acute phase (stage 2); panic attack start with subsequent worsening of anxiety and hypochondriacal symptoms and possible co-occurrence of demoralization and major depression (stage 3); PD might endure, in persistent or attenuated form, with agoraphobia and/or social phobia and/or generalized anxiety disorder and/or hypochondriasis, which in turn might predispose to the development of other psychiatric complications as depression (chronic phase, stage 4).

Staging of panic disorder [86]

The phenomenological clinical sequence of PD might be defined as follow: phobic avoidance and hypochondriasis leading to panic, which, in turn, leads to more phobic avoidance and hypochondriasis. The rollback phenomenon might thus favour a decrease in avoidance by exposure, which improves agoraphobia and panic, with eventual disappearance of panic, whereas agoraphobia persists although to a less degree [95]. Prodromal symptoms of PD might tend to become residual symptoms, which, in turn, may progress to prodromal symptoms of relapse [95]. Alternatively, the rollback phenomenon might be characterized by anxiety elicited by bodily sensations, which influences catastrophic beliefs and such beliefs influence avoidant behaviour [95].

According to the staging model, the acute phase of PD represents a “transfer station” from prodromal to residual symptoms [86]. Within the staging framework seems unlikely that panic attack is a pathognomonic features of panic disorder, while it is plausible that it might be a trans-diagnostic risk factor for the development of agoraphobia and/or social phobia and/or generalized anxiety disorder, and/or hypochondriasis [86]. The staging model of PD has been supported by a vast bulk of literature [86] and can be a good clinical marker (Table 2) for PD:

1) It is a “susceptibility/risk marker,” allowing to detect the presence of prodromal symptoms of PD (e.g., stage 1);

2) It is a “diagnostic marker,” allowing to identify individuals at each single stage of the disorder and the longitudinal evolution of PD across different stages;

3) It is a “prognostic marker,” Fava et al. [101] underline the prognostic value of residual symptoms given that they are negatively correlated with psychological well-being [101] and that may represent the onset of prodromes of relapse. Thus, residual symptoms should be a target of therapy [95];

4) It is a “predictive marker,” staging might improve the clinician’s ability to select a proper treatment to prevent the progression to further stages or promote regression to an earlier stage [86]. Stage-specific therapies have been shown to be effective for depression [102]. Furthermore, agoraphobic avoidance was found to be a strong predictor of non-response to pharmacotherapy and poor response to CBT [103,104].

Given that the staging model of PD is specific for panic disorder [86], it is also a specific marker for PD and its use is easily practically and economically viable in clinics.


According to the clinimetric principles [68,87,88], a valid marker must be sensitive, specific, and predictive [2]. Staging [86] has such properties while biomarkers for PD present low specificity and low predictive value [9,10,45,49]. Biomarkers would be more specific if related to specific stages of panic disorder. We here present an attempt to enrich the staging model of panic disorder with stage-specific biological markers (Figure 1). Changes in amygdala volume, lower HRV, HPA-axis dysregulation were observed in generalized anxiety, social anxiety, and agoraphobia [9-11,20,28,32], thus they may be specific for stage 1 and 2. All biomarkers here described were studied in samples of patients satisfying the DSM diagnosis of PD, thus having panic attacks. This suggests that these biomarkers can be specific for stage 3 or 4. HPA-axis dysregulation, aberrant respiratory pattern, and changes in amygdala volume might be involved in stage 4 given that: 1) hyperventilation might be chronic [16,17]; 2) higher cortisol levels predict higher symptoms severity at 2–4 year follow-up; 3) lower amygdala volumes were found in patients with mean (±SD) duration of illness of 5.4 (±6.4) years [11]. However, this proposal has the limitation that each biomarker is apparently specific for more than one stage. This underlines the limitations already illustrated for biomarkers which, taken without a clinical marker, are still aspecific and, as a consequence, with limited clinical utility.

Figure 1.

Putative biomarker stage-specific for panic disorder.


1. Kalia M, Silva JC. Biomarkers of psychiatric diseases: current status and future prospects. Metabolism 2015;64:S11–S15.
2. Scarr E, Millan MJ, Bahn S, Bertolino A, Turck CW, Kapur S, et al. Biomarkers for psychiatry: the journey from fantasy to fact, a report of the 2013 CINP think tank. Int J Neuropsychopharmacol 2015;18:1–9.
3. Biomarkers Definitions Working Group. Biomarkers and surrogate endpoints. Clin Pharmacol Ther 2001;69:89–95.
4. Venkatasubramanian G, Keshavan MS. Biomarkers in psychiatry-a critique. Ann Neurosci 2016;23:3–5.
5. Macaluso M, Preskorn SH. How biomarkers will change psychiatry: from clinical trials to practice. Part I, Introduction. J Psychiatr Pract 2012;18:118–121.
6. Amur S, LaVange L, Zineh I, Buckman-Garner S, Woodcock J. Biomarker qualification: toward a multiple stakeholder framework for biomarker development, regulatory acceptance, and utilization. Clin Pharacol Ther 2015;98:34–46.
7. Ankeny JS, Labadie B, Luke J, Hsueh E, Messina J, Zager JS. Review of diagnostic, prognostic, and predictive biomarkers in melanoma. Clin Exp Metastasis 2018;35:487–493.
8. FDA-NIH Biomarker Working Group. BEST (Biomarkers, Endpoints, and Other Tools) Resource Silver Spring, MD: Food and Drug Administration (US). Bethesda, MD: National Institutes of Health (US); 2016.
9. Bandelow B, Baldwin D, Abelli M, Altamura C, Dell’Osso B, Domschke K, et al. Biological markers for anxiety disorders, OCD and PTSD–a consensus statement. Part I: neuroimaging and genetics. World J Biol Psychiatry 2016;17:321–365.
10. Bandelow B, Baldwin D, Abelli M, Bolea-Alamanac B, Bourin M, Chamberlain SR, et al. Biological markers for anxiety disorders, OCD and PTSD: a consensus statement. Part II: Neurochemistry, neurophysiology and neurocognition. World J Biol Psychiatry 2017;18:162–214.
11. Hayano F, Nakamura M, Asami T, Uehara K, Yoshida T, Roppongi T, et al. Smaller amygdala is associated with anxiety in patients with panic disorder. Psychiatry Clin Neurosci 2009;63:266–276.
12. De Cristofaro MT, Sessarego A, Pupi A, Biondi F, Faravelli C. Brain perfusion abnormalities in drug-naive, lactate-sensitive panic patients: a SPECT study. Biol Psychiatry 1993;33:505–512.
13. Gecici O, Acar M, Haktanir A, Emul M, Demirel R, YüCEL A, et al. Evaluation of cerebral blood flow volume using color duplex sonography in patients with untreated panic disorder. Psychiatry Clin Neurosci 2005;59:676–682.
14. Gorman JM, Liebowitz MR, Fyer AJ, Stein J. A neuroanatomical hypothesis for panic disorder. Am J Psychiatry 1989;146:148–161.
15. Perna G, Caldirola D, Bellodi L. Panic disorder: from respiration to the homeostatic brain. Acta Neuropsychiatr 2004;16:57–67.
16. Grassi M, Caldirola D, Vanni G, Guerriero G, Piccinni M, Valchera A, et al. Baseline re- spiratory parameters in panic disorder a meta-analysis. J Affect Disord 2013;146:158–173.
17. Grassi M, Caldirola D, Di Chiaro N, Riva A, Daccò S, Pompili M, et al. Are respiratory abnormalities specific for panic disorder? A meta-analysis. Neuropsychobiology 2014;70:52–60.
18. Roth WT. Physiological markers for anxiety: panic disorder and phobias. Int J Psychophysiol 2005;58:190–198.
19. Niccolai V, van Duinen MA, Griez EJ. Respiratory patterns in panic disorder reviewed: a focus on biological challenge tests. Acta Psychiatr Scand 2009;120:167–177.
20. Chalmers JA, Quintana DS, Abbott MJ, Kemp AH. Anxiety disorders are associated with reduced heart rate variability: a meta-analysis. Front Psychiatry 2014;:5–80.
21. Hillebrand S, Gast KB, de Mutsert R, Swenne CA, Jukema JW, Middeldorp S, et al. Heart rate variability and first cardiovascular event in populations without known cardiovascular disease: meta-analysis and dose–response meta-regression. Europace 2013;15:742–749.
22. Davies SJ, Ghahramani P, Jackson PR, Noble TW, Hardy PG, Hippisley-Cox J, et al. Association of panic disorder and panic attacks with hypertension. Am J Med 1999;107:310–316.
23. Davies SJ, Bjerkeset O, Nutt DJ, Lewis G. A U-shaped relationship between systolic blood pressure and panic symptoms: the HUNT study. Psychol Med 2012;42:1969–1976.
24. Gorman JM, Sloan RP. Heart rate variability in depressive and anxiety disorders. Am Heart J 2000;140:S77–S83.
25. Bhad R. Red blood cell and platelet indices: a potential biomarker for panic disorder. J Neurosci Rural Pract 2017;8:164.
26. Ransing RS, Patil B, Grigo O. Mean platelet volume and platelet distribution width level in patients with panic disorder. J Neurosci Rural Pract 2017;8:174–178.
27. Asoglu M, Aslan M, Imre O, Kivrak Y, Akil O, Savik E, et al. Mean platelet volume and red cell distribution width levels in initial evaluation of panic disorder. Neuropsychiatr Dis Treat 2016;12:2435–2438.
28. Ising M, Hohne N, Siebertz A, Parchmann AM, Erhardt A, Keck M. Stress response regulation in panic disorder. Curr Pharm Des 2012;18:5675–5684.
29. Heim C, Shugart M, Craighead WE, Nemeroff CB. Neurobiological and psychiatric consequences of child abuse and neglect. Dev Psychobiol 2010;52:671–690.
30. Danese A, McEwen BS. Adverse childhood experiences, allostasis, allostatic load, and age-related disease. Physiol Behav 2012;106:29–39.
31. Du X, Pang TY. Is dysregulation of the HPA-axis a core pathophysiology mediating co-morbid depression in neurodegenerative diseases? Front Psychiatry 2015;6:32.
32. Faravelli C, Lo Sauro C, Lelli L, Pietrini F, Lazzeretti L, Godini L, et al. The role of life events and HPA axis in anxiety disorders: a review. Curr Pharm Des 2012;18:5663–5674.
33. Bandelow B, Wedekind D, Pauls J, Broocks A, Hajak G, Ruther E. Salivary cortisol in panic attacks. Am J Psychiatry 2000;157:454–456.
34. Jakuszkowiak-Wojten K, Landowski J, Wiglusz MS, Cubała WJ. Cortisol as an indicator of hypothalmic-pitituary-adrenal axis dysregulation in patients with panic disorder: a literature review. Psychiatr Danub 2015;27(Suppl 1):S445–S451.
35. Bandelow B, Domschke K, Baldwin D. Panic Disorder and Agoraphobia Oxford, UK: Oxford Psychiatry Library; 2013.
36. Graeff FG. Anxiety, panic and the hypothalamic-pituitary-adrenal axis. Revista Brasileira de Psiquiatria 2007;29:s3–s6.
37. Graeff FG, Zangrossi H Jr. The hypothalamic-pituitary-adrenal axis in anxiety and panic. Psychol Neurosci 2010;3:3–8.
38. Schommer NC, Hellhammer DH, Kirschbaum C. Dissociation between reactivity of the hypothalamus-pituitary-adrenal axis and the sympathetic-adrenal-medullary system to repeated psychosocial stress. Psychosom Med 2003;65:450–460.
39. Abelson JL, Curtis GC. Hypothalamic-pituitary-adrenal axis activity in panic disorder: prediction of long-term outcome by pretreatment cortisol levels. Am J Psychiatry 1996;153:69–73.
40. Coryell W, Noyes R Jr, Reich J. The prognostic significance of HPA-axis disturbance in panic disorder: a three-year follow-up. Biol Psychiatry 1991;29:96–102.
41. Machado S, Sancassiani F, Paes F, Rocha N, Murillo-Rodriguez E, Nardi AE. Panic disorder and cardiovascular diseases: an overview. Int Rev Psychiatry 2017;29:436–444.
42. Grambal A, Tüdös Z, Hok P, Kamarádová D, Divéky T, Hluštík P, et al. Predictors of poor treatment response to additional CBT in real panic disorder patients: the role of DLPF, orbitofrontal cortex, parietal lobule, frontal eye field and amygdala in PD. Neuro Endocrinol Lett 2015;36:269–281.
43. Fischer S, Cleare AJ. Cortisol as a predictor of psychological therapy response in anxiety disorders-systematic review and meta-analysis. J Anxiety Disord 2017;47:60–68.
44. Cloninger CR. Implications of Comorbidity for the Classification of Mental Disorders: the Need for a Psychobiology of Coherence. In : Maj M, Gaebel W, Lopez-Ibor JJ, Sartorius N, eds. Psychiatric Diagnosis and Classification Chichester: Wiley; 2002. p. 79–105.
45. Boksa P. A way forward for research on biomarkers for psychiatric disorders. J Psychiatry Neurosci JPN 2013;38:75–77.
46. Bassett D. A literature review of heart rate variability in depressive and bipolar disorders. Aust N Z J Psychiatry 2016;50:511–519.
47. Bradley AJ, Dinan TG. A systematic review of hypothalamic-pituitary-adrenal axis function in schizophrenia: implications for mortality. J Psychopharmacol 2010;24(4 Suppl):91–118.
48. Brunoni AR, Kemp AH, Dantas EM, Goulart AC, Nunes MA, Boggio PS, et al. Heart rate variability is a trait marker of major depressive disorder: evidence from the sertraline vs. electric current therapy to treat depression clinical study. Int J Neuropsychopharmacol 2013;16:1937–1949.
49. Delgado PL, Moreno FA. Role of norepinephrine in depression. J Clin Psychiatry 2000;61(Suppl 1):5–12.
50. Faravelli C, Mansueto G, Palmieri S, Lo Sauro C, Rotella F, Pietrini F, et al. Childhood adversity, cortisol levels, and psychosis: a retrospective investigation. J Nerv Ment 2017;205:574–579.
51. Gundogmus İ, Algul A, Karagöz A, Kıyançiçek M. PDW and RDW are new parameters for bipolar episodes and unipolar depression. Psychiatry Clin Psychopharmacol 2018;:1–7.
52. Belvederi Murri M, Prestia D, Mondelli V, Pariante C, Patti S, Olivieri B, et al. The HPA axis in bipolar disorder: systematic review and meta-analysis. Psychoneuroendocrinology 2016;63:327–342.
53. Rajarethinam R, DeQuardo JR, Miedler J, Arndt S, Kirbat R, Brunberg JA, et al. Hippocampus and amygdala in schizophrenia: assessment of the relationship of neuroanatomy to psychopathology. Psychiatry Res 2001;108:79–87.
54. Ressler KJ, Nemeroff CB. Role of norepinephrine in the pathophysiology and treatment of mood disorders. Biol Psychiatry 1999;46:1219–1233.
55. Goodkind M, Eickhoff SB, Oathes DJ, Jiang Y, Chang A, Jones-Hagata LB, et al. Identification of a common neurobiological substrate for mental illness. JAMA Psychiatry 2015;72:305–315.
56. Valkonen‐Korhonen M, Tarvainen MP, Ranta‐Aho P, Karjalainen PA, Partanen J, Karhu J, et al. Heart rate variability in acute psychosis. Psychophysiology 2003;40:716–726.
57. Young AH. Cortisol in mood disorders. Stress 2004;7:205–208.
58. Kim JH, Ann JH, Lee J. Relationship between heart rate variability and the severity of psychotic symptoms in schizophrenia. Acta Neuropsychiatr 2011;23:161–166.
59. Fava GA, Guidi J, Grandi S, Hasler G. The missing link between clinical states and biomarkers in mental disorders. Psychother Psychosom 2014;83:136–141.
60. Kapur S, Phillips AG, Insel TR. Why has it taken so long for biological psychiatry to develop clinical tests and what to do about it? Mol Psychiatry 2012;17:1174–1179.
61. Fava GA, Cosci F, Offidani E, Guidi J. Behavioral toxicity revisited: iatrogenic comorbidity in psychiatric evaluation and treatment. J Clin Psychopharmacol 2016;36:550–553.
62. Tomba E, Guidi J, Fava GA. What psychologists need to know about psychotropic medications. Clin Psychol Psychother 2018;25:181–187.
63. Feinstein AR. Basic biomedical science and the destruction of the pathophysiological bridge from bench to bedside. Am J Med 1999;107:461–467.
64. Fava GA, Rafanelli C, Tomba E. The clinical process in psychiatry: a clinimetric approach. J Clin Psychiatry 2012;73:177–184.
65. Wright JG, Feinsten AR. A comparative contrast of clinimetric and psychometric methods for constructing indexes and rating scales. J Clin Epidemiol 1992;45:1201–1218.
66. Fava GA, Ruini C, Rafanelli C. Psychometric theory is an obstacle to the progress of clinical research. Psychother Psychosom 2004;73:145–148.
67. Tomba E, Bech P. Clinimetrics and clinical psychometrics: macro-and micro-analysis. Psychother Psychosom 2012;81:333–343.
68. Feinstein AR. T. Duckett Jones Memorial Lecture. The Jones criteria and the challenge of clinimetrics. Circulation 1982;66:1–5.
69. Cosci F, Svicher A, Bech P. The family assessment device: a clinimetric analysis. Psychother Psychosom 2016;85:241–243.
70. Timmerby N, Andersen JH, Søndergaard S, Østergaard SD, Bech P. A systematic review of the clinimetric properties of the 6-item version of the Hamilton Depression Rating Scale (HAM-D6). Psychother Psychosom 2017;86:141–149.
71. Timmerby N, Cosci F, Watson M, Csillag C, Schmitt F, Steck B, et al. A confirmative clinimetric analysis of the 36-item Family Assessment Device. Nord J Psychiatry 2018;72:268–272.
72. American Psychiatric Association. Diagnostic and statistical Manual of Mental Disorders 3th Edth ed. Washington, DC: American Psychiatric Association; 1980.
73. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders 3th Ed.-Revth ed. Washington, DC: American Psychiatric Association; 1987.
74. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders (4th Ed.) (DSM-IV) Washington, DC: American Psychiatric Association; 1994.
75. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders (4th Ed-Rev) Washington, DC: American Psychiatric Association; 2000.
76. American Psychiatric Association. Diagnostic and Satistical Mnual of Mntal Dsorders 5th Edth ed. Washington, DC: American Psychiatric Association; 2013.
77. World Health Organization. Manual of the 28. International Statistical Classification of Diseases, Injuries, and Causes of Death. Sixth Revision of the International Lists of Diseases and Causes of Death. Adopted 1948 Geneva: World Health Organization; 1948.
78. World Health Organization. Manual of the 30. International Statistical Classification of Diseases, Injuries, and Causes of Death. Based on the Recommendations of the Seventh Revision Conference, 1955, and Adopted by the Ninth World Health Assembly under the WHO Nomenclature Regulations Geneva: World Health Organization; 1957.
79. World Health Organization. Manual of the 32. International Statistical Classification of Diseases, Injuries, and Causes of Death. Based on Recommendations of the Eighth Revision Conference, 1965, and Adopted by the Nineteenth World Health Assembly Geneva: World Health Organization; 1967.
80. World Health Organization. Manual of the 34. International Classification of Diseases, Injuries, and Causes of Death, Based on the Recommendations of the Ninth Revision Conference, 1975, and Adopted by the Twenty-Ninth World Health Assembly Geneva: World Health Organization; 1977.
81. Fava GA, Tomba E. New modalities of assessment and treatment planning in depression: the sequential approach. CNS Drugs 2010;24:453–465.
82. Bech P. Fifty years with the Hamilton scales for anxiety and depression. A tribute to Max Hamilton. Psychother Psychosom 2009;78:202–211.
83. Faravelli C. Assessment of psychopathology. Psychother Psychosom 2004;73:139–141.
84. First MB, Rebello TJ, Keeley JW, Bhargava R, Dai Y, Kulygina M, et al. Do mental health professionals use diagnostic classifications the way we think they do? A global survey. World Psychiatry 2018;17:187–195.
85. Maj M. Why the clinical utility of diagnostic categories in psychiatry is intrinsically limited and how we can use new approaches to complement them. World Psychiatry 2018;17:121–122.
86. Cosci F, Fava GA. Staging of mental disorders: systematic review. Psychother Psychosom 2013;82:20–34.
87. Feinstein AR. Clinimetrics New Haven: Yale University Press; 1987.
88. Feinstein AR. An analysis of diagnostic reasoning: I. The domains and disorders of clinical macrobiology. Yale J Biol Med 1973;46:212–232.
89. Fava GA. Clinical judgment in psychiatry. Requiem or reveille? Nord J Psychiatry 2013;67:1–10.
90. Feinstein AR. Multi-item ‘instruments’ versus Virginia Apgar’s principles of clinimetrics. Arch Intern Med 1999;159:125–128.
91. Taylor GJ. Affects, trauma, and mechanisms of symptom formation: a tribute to John C. Nemiah, MD (1918-2009). Psychother Psychosom 2010;79:339–349.
92. Faravelli C, Servi P, Arends JA, Strik WK. Number of symptoms, quantification, and qualification of depression. Compr Psychiatry 1996;37:307–315.
93. McGorry PD. The next stage for diagnosis: validity through utility. World Psychiatry 2013;12:213–215.
94. Fava GA, Kellner R. Staging: a neglected dimension in psychiatric classification. Acta Psychiatr Scand 1993;87:225–230.
95. Cosci F. The psychological development of panic disorder: implications for neurobiology and treatment. Revista Brasileira de Psiquiatria 2012;34:9–19.
96. Sheehan DV, Sheehan KH. The classification of anxiety and hysterical states. Part II. Toward a more heuristic classification. J Clin Psychopharmacol 1982;2:386–393.
97. Fava GA, Rafanelli C, Tossani E, Grandi S. Agoraphobia is a disease: a tribute to Sir Martin Roth. Psychother Psychosom 2008;77:133–138.
98. Fava GA, Mangelli L. Subclinical symptoms of panic disorder: new insights into pathophysiology and treatment. Psychother Psychosom 1999;68:281–289.
99. Fava GA, Porcelli P, Rafanelli C, Mangelli L, Grandi S. The spectrum of anxiety disorders in the medically ill. J Clin Psychiatry 2010;71:910–914.
100. Fava GA, Cosci F, Sonino N. Current psychosomatic practice. Psychother Psychosom 2017;86:13–30.
101. Fava GA, Rafanelli C, Ottolini F, Ruini C, Cazzaro M, Grandi S. Psychological well-being and residual symptoms in remitted patients with panic disorder and agoraphobia. J Affect Disord 2001;65:185–190.
102. The role of staging in planning psychotherapeutic interventions in depression. J Clin Pychiatry 2017;78:456–463.
103. Porter E, Chambless DL. A systematic review of predictors and moderators of improvement in cognitive-behavioral therapy for panic disorder and agoraphobia. Clin Psychol Rev 2015;42:179–192.
104. Slaap BR, den Boer JA. The prediction of nonresponse to pharmacotherapy in panic disorder: a review. Depress Anxiety 2001;14:112–122.

Article information Continued

Figure 1.

Putative biomarker stage-specific for panic disorder.

Table 1.

Putative makers of panic disorder classified as: susceptibility/risk marker; diagnostic marker, prognostic marker, predictive marker [6-8]

Susceptibility/risk marker Diagnostic marker Prognostic marker Predictive marker
 Structural changes in the amygdala, hippocampus
 Cerebral blood level in the left occipital cortex
 Serotonin 5-TH and noradrenergic systems activation
 Aberrant respiratory regulation
 Hearth rate variability
 Blood cells and peripheral blood stem cells
 Dysregulation hypothalamic-pituitary-adrenal axis/Cortisol secretion
Clinical marker

Table 2.

Staging of panic disorder [86]

Stage 1 Prodromes: agoraphobia and/or social phobia and/or; generalized anxiety disorder and/or hypochondriasis
Stage 2 Acute manifestations of agoraphobia and/or social phobia and/or generalized anxiety disorder and/or hypochondriasis
Stage 3 Panic attacks occur, panic disorder with aggravation of anxiety and/or hypochondriacal symptoms; demoralization and/or major depression
Stage 4 Chronic (attenuated or persistent form): panic disorder and/or agoraphobia and/or social phobia and/or generalized anxiety disorder and/or hypochondriasis. Increased vulnerability to major depression