Diagnosing Separation Anxiety Disorder in Adults

“Adult-onset” adult separation anxiety disorder includes those individuals whose separation anxiety symptoms first appeared in adulthood, without a documented history of related symptoms during childhood, whereas “childhood-onset” adult separation anxiety disorder connotes those individuals whose symptoms had a childhood onset and which persisted into adulthood.Reference Bögels, Knappe and Clark ¹ A large proportion of adult patients with anxiety disorders have problems that can be traced back to childhood experiences and environments, but which only become manifest in clinical symptoms during adult life, and separation anxiety disorder is considered in the same way as other anxiety conditions. Removing the age of onset stipulation in the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) aligns separation anxiety disorder with all other anxiety disorders and so removes the need for separate adult and juvenile anxiety groups.Reference Manicavasagar, Marnane and Pini ²

In the DSM-5, ³ separation anxiety disorder is characterized by the presence of developmentally inappropriate and excessive fear or anxiety concerning separation from attachment figures. To meet the diagnostic threshold, the fear, anxiety, or avoidance of separation must have been present for at least 4 weeks in a child or adolescent, and typically for at least 6 months in an adult. Diagnosis also requires the presence of at least 3 of 8 features, including recurrent excessive distress when anticipating or experiencing separation, and persistent and excessive worry about losing major attachment figures; this can be problematic, as it seems difficult to distinguish fear and anxiety concerning separation from attachment figures (the characteristic feature) from excessive distress when anticipating or experiencing separation, or from excessive worry about losing attachment figures (2 of the supporting features). Furthermore, another 2 of the supporting features include a persistent reluctance or refusal to go out, away from home, to school or to work because of fear of separation, and a persistent fear or reluctance about being alone or without major attachment figures at home or in other settings, which seems a rather nuanced distinction. In addition, the criterion of “persistent and excessive worry about experiencing an untoward event that causes separation from an attachment figure” is hard to distinguish from the persistent and excessive worrying about family members, which is common in patients with generalized anxiety disorder (GAD). Also, the features of troubled sleep while being away from attachment figures and repeated nightmares about separation seem difficult to delineate from posttraumatic sleep disturbances in patients whose trauma involved significant threats to family members. The capacity of diagnostic tools to delineate between separation anxiety disorder and panic disorder has been validated,Reference Silove and Marnane ⁴ although panic disorder itself is important in the differential diagnosis of separation anxiety disorder.

But these problems are not restricted to DSM-5 criteria: similar problems are seen in the diagnostic criteria for separation anxiety disorder in childhood included within ICD-10. ⁵ At least 3 of 8 features must be present, including persistent and excessive worry about possible harm befalling major attachment figures, and persistent and excessive worry that some untoward event will separate the child from an attachment figure. But if the worrisome and anticipated untoward event involves harm, should both features count? It also seems difficult to distinguish anxiety and worry about harm potentially befalling an attachment figure (the first listed feature) from distress (for example manifest by anxiety, crying, and misery) in anticipation of separation from an attachment figure (another listed feature). The ICD-10 criteria stipulate that “the criteria for generalized anxiety disorder of childhood are not met,” implying that when they are present, GAD should be diagnosed in preference to separation anxiety disorder. But, again, it could be hard to distinguish 2 other features relating to sleep disturbance, namely insomnia marked by persistent reluctance or refusal to go to sleep without being near an attachment figure, from the presence of repeated nightmares involving themes of separation.

In separation anxiety disorder in childhood, parents are the major attachment figures, and common separation behaviors include crying, tantrums, and school refusal. By contrast, adults with separation anxiety have strong fears that their attachment figures (usually a spouse or child) will come to harm, and so make strenuous efforts to remain in close contact with them, despite troublesome impracticalities.Reference Pini, Abelli and Shear ⁶ Childhood separation anxiety can be the first step along a pathway of a mental disorder that may manifest either as separation anxiety or as other forms of mental disorder in young adults.Reference Lewinsohn, Holm-Denoma, Small, Seeley and Joiner ⁷ Early separation anxiety symptoms and behaviors may act as a “general vulnerability factor,” conferring an increased risk for anxiety and mood disorders.Reference Silove, Alonso and Bromet ^{8 ,} Reference Kossowsky, Pfaltz, Schneider, Taeymans, Locher and Gaab ⁹

Appreciating the Burden of Separation Anxiety Disorder in Adults

Separation anxiety symptoms are not uncommon, across the lifespan: a prospective longitudinal study in adolescents and young adults (aged 14–24 years) found a 7.8% “lifetime prevalence” of separation anxiety symptoms,Reference Brückl, Wittchen, Höfler, Pfister, Scheider and Lieb ¹⁰ and an opportunity sample in first-semester college students found that 21% met or exceeded the symptom threshold on a checklist of separation anxiety symptoms.Reference Seligman and Wuyek ¹¹ This threshold was exceeded by 3.5% of elderly (60 years or older) primary care patientsReference Wijeratne and Manicavasagar ¹² ; supra-threshold levels of separation anxiety were present in 24.5% of women attending a hospital ante-natal clinicReference Eapen, Silove, Johnston, Apler and Rees ¹³ and in 35% of primiparous women with unsettled infants admitted to a parent/child residential unit.Reference Kohlhoff, Barnett and Eapen ¹⁴ But the presence of separation anxiety symptoms does not necessarily equate to the presence of separation anxiety disorder, which is characterized by significant personal distress and associated impairment.

The findings of the large (n=5692) U.S. National Co-morbidity Survey Replication (NCS-R) epidemiological study, in which the estimated lifetime prevalence was 6.6%, provided substantial data on the existence of separation anxiety in adults. Although 36.1% of cases with an onset of separation anxiety symptoms in childhood persisted into adult life, the majority (77.5%) of “cases” in adults had a reported onset in adulthood. The prevalence was higher in females than in males, this difference being more marked in the sub-group of individuals in which separation anxiety started in childhood.Reference Shear, Jin, Ruscio, Walters and Kessler ¹⁵ Adult separation anxiety disorder markedly impairs social and work function: more than 40% of respondents classified as having the condition reported severe impairment in at least one life domain, most commonly in the domains of social and personal life, with impairment being more marked in those with comorbid conditions, which were present in more than half the cases in the NCS-R study.Reference Shear, Jin, Ruscio, Walters and Kessler ¹⁵ Silove et alReference Silove, Alonso and Bromet ⁸ have re-examined prevalence rates of separation anxiety disorder in the general population (n=38,993 adults) in 18 countries using data from the World Health Organization (WHO) World Mental Health Surveys. Analyses focused on prevalence, age at onset, comorbidity, predictors of onset and persistence, and separation anxiety–related role impairment. In this analysis, the lifetime separation anxiety disorder prevalence averaged 4.8% across countries, and 43.1% of lifetime onsets occurred after the age of 18 years. Rates of comorbidity with other psychiatric disorders were similar to those found in previous large-scale epidemiological studies. Significant predictors of lifetime separation anxiety disorder included female gender, retrospectively reported childhood adversities, and lifetime traumatic events. These predictors were largely comparable for separation anxiety disorder onsets in childhood, adolescence, and adulthood and across country income groups.

The extensive comorbidity of separation anxiety disorder has been confirmed in a number of clinical samples. Separation anxiety symptoms, of both childhood-onset and adult-onset, are common in patients with bipolar disorder, particularly in those patients with an early onset of bipolar disorderReference Pini, Abelli and Mauri ¹⁶ and in outpatients with eating disorders.Reference Troisi, Massaroni and Cuzzolaro ¹⁷ Separation anxiety symptoms are common in outpatients with “complicated grief”Reference Dell’Osso, Carmassi and Corsi ¹⁸ and posttraumatic stress disorder,Reference Dell’Osso, Carmassi and Musetti ¹⁹ and separation anxiety disorder was diagnosed in 56% of outpatients with affective disorders and complicated grief.Reference Pini, Gesi and Abelli ²⁰ Also, features of prolonged grief, depression, and separation anxiety disorder were found to correlate in an opportunity sample of bereaved individuals.Reference Boelen ²¹

Observational studies in patients with obsessive-compulsive disorder have found a point prevalence of separation anxiety disorder of 27.2%Reference Franz, Rateke and Hartmann ²² and a lifetime prevalence of 17%.Reference Mroczkowski, Goes and Riddle ²³ The relationship between separation anxiety and “obsessivity” (as an abnormal trait rather than a diagnosis of OCD) is intriguing. Unfortunately, the few data in the literature arise from a focus on the relationship with insecure attachment, with the observation of an association between obsessiveness and high scores on the attachment fearful dimension style.Reference Wiltgen, Adler and Smith ²⁴ However, it is a common impression among clinicians working with adult individuals with separation anxiety that, in some cases, they have excessive interpersonal control over significant figures (a core feature of obsessive personality), characterized by over-thinking or rumination on attachment figures, with reticence toward unplanned changes in routine movements and trips. Future research in this direction is recommended to clarify the extent to which the trait of obsessiveness impacts on separation anxiety (and vice versa).

Separation anxiety disorder or high separation anxiety was found to be associated with the presence of comorbid personality disorders (Clusters B and C) in patients attending an anxiety clinic.Reference Silove, Marnane, Wagner and Manicavasagar ²⁵ Meta-analytic findings indicate that childhood separation anxiety disorder increases the risk of panic disorder in later life.Reference Kossowsky, Pfaltz, Schneider, Taeymans, Locher and Gaab ⁹ Comorbidity of separation anxiety disorder with panic disorder and agoraphobia in adults is common, but the two disorders can be distinguished.Reference Silove and Marnane ⁴ Childhood separation anxiety disorder, adult panic disorder, and hypersensitivity to a 35% carbon dioxide challenge may share a common latent intervening variable.Reference Roberson-Nay, Klein and Klein ^{26 –} Reference Battaglia, Pesenti-Gritti, Medland, Ogliari, Tambs and Spatola ²⁸

The presence of comorbid separation anxiety disorder confers a generally worse prognosis for other conditions. Successful outcomes in patients with social anxiety disorder, panic disorder, or GAD were found to be less likely in the presence of comorbid adult separation anxiety disorder.Reference Kirsten, Greyner, Wagner and Manicavasagar ²⁹ A successful outcome for cognitive behavior therapy (CBT) in patients with panic disorder was 4 times less likely in the presence of comorbid separation anxiety disorder,Reference Aaronson, Shear and Goetz ³⁰ possibly because CBT undertaken for other anxiety disorders does not reduce the intensity of separation anxiety symptoms.Reference Manicavasagar, Silove and Curtis ³¹ A more recent study indicates that patients undergoing CBT for panic disorder, GAD, or social phobia all have worse outcomes if they also have adult separation anxiety disorder.Reference Bögels, Knappe and Clark ¹

Measuring Separation Anxiety Symptoms

A number of diagnostic interviews and rating scales have been developed to facilitate the diagnosis and assessment of severity of adult separation anxiety disorder. The Adult Separation Anxiety Structured Interview (ASA-SI) is derived from Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) criteria for the childhood-onset condition, with items modified for adulthood.Reference Manicavasagar and Silove ³² The Structured Clinical Interview for Separation Anxiety Symptoms (SCI-SAS) assesses childhood-onset and adult-onset conditions, with items based on DSM-IV criteria.Reference Cyranowski, Shear and Rucci ³³ The NCS-R study involved the development of a new module for adult separation anxiety disorder for addition to the Composite International Diagnostic Interview, based on the SCI-SAS. Symptom severity can be assessed through completion of the self-report Adult Separation Anxiety-27 (ASA-27), which contains the same items as the ASA-SI: it has an established threshold for “caseness,”Reference Manicavasagar and Silove ³² adequate internal consistency and test–retest reliability, and, as expected, high concordance with the ASA-SI.Reference Manicavasagar, Silove, Wagner and Drobny ³⁴

Distinguishing Separation Anxiety Disorder from Other Conditions

It seems reasonable to suppose that separation anxiety in adults was previously often construed as a form of panic disorder with or without agoraphobia,Reference Silove and Rees ³⁵ while for many individuals a diagnosis of adult separation anxiety disorder may have been more appropriate, with panic attacks being regarded as a feature of separation anxiety, rather than of panic disorder. Differential diagnosis between separation anxiety disorder and panic disorder rests on establishing that the fear of having a panic attack does not drive separation anxiety, whereas it does in agoraphobia: patients with separation anxiety disorder may experience “panic” that manifests as extreme stress or distress,Reference Silove and Marnane ⁴ but the primary fear is the fear of separation rather than the fear of having a panic attack.Reference Bögels, Knappe and Clark ¹ In addition, patients with panic disorder are preoccupied with their own personal health and safety, whereas individuals with separation anxiety disorder are concerned for the well-being of their attachment figures. The distinction from agoraphobia is potentially troublesome: though agoraphobic situations are defined as places and situations from which escape might be difficult or in which help might not be available to the affected person, conversely in separation anxiety disorder, the primary fear is that something untoward will happen to someone else (loved ones and other major attachment figures).Reference Bögels, Knappe and Clark ¹

Differential diagnosis between separation anxiety disorder and GAD can also be troublesome, as worrying about something bad happening to loved ones is a feature that is common to the 2 conditions. Approximately half of children and adolescents with separation anxiety disorder worry about harm befalling their parents when not with them,Reference Masi, Mucci, Favilla, Romano and Poli ³⁶ and the majority of children with the diagnosis of separation anxiety disorder also have the diagnosis of comorbid GAD.Reference Verduin and Kendall ³⁷ The distinction between the 2 disorders in adults is often difficult, but largely rests on establishing that the fear and worry of losing loved ones is just one of a wide range of worrisome themes in patients with GAD (in which other common themes include personal health, financial concerns, and interpersonal difficulties), whereas in separation anxiety disorder, the central and often only concern is the fear of and worry about losing a major attachment figure.Reference Bögels, Knappe and Clark ¹

The distinction of separation anxiety disorder from “dependent personality disorder” is also difficult. Dependent personality disorder is characterized by a persistent, pervasive, and indiscriminate tendency to rely excessively on other individuals, particularly when making decisions. By contrast, separation anxiety disorder is characterized by a limited array of concerns about the proximity and safety of major attachment figures.Reference Pini, Abelli and Shear ⁶ Adults with separation anxiety disorder typically perceive no difficulty in caring for themselves. Although frantic efforts to avoid real or imagined abandonment are a feature of “borderline personality disorder,” the distinction between the 2 conditions is helped by identifying the pervasive mood instability and uncertainty with regard to self and the majority of relationships in borderline personality disorder, which is not a typical feature of separation anxiety disorder.Reference Manicavasagar, Marnane and Pini ² Finally, a concern regarding the misbehavior of romantically attached figures is a feature of morbid jealousy, and this condition should be included within the differential diagnosis, particularly when the worries and concerns about attachment figures in patients with separation anxiety disorder have acquired a persistently preoccupying and obsessive quality.

Linking Separation Anxiety Disorder and Adult Attachment Style

According to attachment theory, securely attached children exhibit less anxious behavior toward attachment figures, while insecurely attached children display more symptoms of separation anxiety.Reference Bowlby ³⁸ From this perspective, a link between insecure attachment and separation anxiety revealed that insecure attachment constituted a general risk factor in the development of anxiety. BowlbyReference Bowlby ³⁸ explained that attachment, separation, and reunion responses are learned as infants develop. Anxiety and the fear of abandonment are the driving forces behind attachment formation. Insecure attachment can often result when an attachment relationship is threatened, or the attachment figure is not consistently available. BowlbyReference Bowlby ³⁹ proposed that children’s levels of anxiety might be affected by the way in which they are attached to their caregivers: ambivalently attached children were constantly afraid of being alone and in danger, because their caregivers were unreliable concerning their needs.Reference Bowlby ³⁹

Clinical studies have revealed the association between specific adult attachment styles and depressive disorders.Reference Shaver, Schachner and Mikulincer ⁴⁰ Attachment anxiety, which concerns apprehension over rejection and abandonment, has been associated with increased rates of psychopathology. Attachment anxiety represents 1 of the 2 primary dimensions involved in self-report measures of adult attachment style. The second dimension, namely attachment avoidance, concerns the degree to which a person feels uncomfortable depending on and being emotionally close to others. Hypothesized explanations for the association between attachment anxiety and depression include anxious peoples’ negative models of self (eg, believing they are unlovableReference Bartholomew and Horowitz ⁴¹ ), low self-esteem, self-criticism, and dysfunctional attributions to partners’ behavior that increase the likelihood of separation anxiety.Reference Greenberg ⁴² People high in avoidance generally invest less in relationships, are less upset when they end, and are relatively low in commitment and relationship satisfaction.Reference Mikulincer and Shaver ⁴³

Separating Adult Separation Anxiety Disorder

Although anxious attachment style and separation anxiety were found to be highly correlated in patients with panic disorder or separation anxiety disorder,Reference Manicavasagar, Silove, Marnane and Wagner ⁴⁴ not all individuals with separation anxiety disorder have an insecure attachment style.Reference Pini, Abelli and Troisi ⁴⁵ The condition is, however, associated with the temperament and character dimensions of high “harm avoidance” and low “self-directedness,”Reference Mertol and Alkin ⁴⁶ but is not especially linked to “intolerance of uncertainty.”Reference Boelen, Reijntjes and Carleton ⁴⁷

There have been few attempts to identify the possible biological basis of separation anxiety symptoms in adults. The role of environmental and genetic factors and the extent to which they interact is unclear.Reference Scaini, Ogliari, Eley, Zavos and Battaglia ⁴⁸ Peripheral-type benzodiazepine binding site receptors, which have a role in the biosynthesis of steroids during stress and anxiety states, were found to have a significantly lower density in the sub-group of panic disorder patients in which coexisting separation anxiety was a prominent feature.Reference Pini, Martini and Abelli ⁴⁹ Separation anxiety disorder appears to be associated with hypersensitivity to inhaled carbon dioxide, with a similar pattern to that in patients with panic disorder, suggesting the 2 diagnoses may have a shared pathophysiological basis.Reference Battaglia, Pesenti-Gritti, Medland, Ogliari, Tambs and Spatola ^{28 ,} Reference Atli, Bayin and Alkin ⁵⁰

Complex perturbations of the hypothalamic neuropeptide oxytocin may be involved in disordered attachment, as it is known to play a part in trust, reducing distress, and helping “read” the mental and emotional states of others. Extensive research using animal models and human volunteers has established a significant role for oxytocin in mother–infant interactions. Peripheral oxytocin levels are potential markers of social responsivity and the capacity for attachment.Reference Milrod, Markowitz and Gerber ⁵¹ However, mutation analysis of the oxytocin gene has shown no consistent disturbance in patients with adult separation anxiety disorder, although single nucleotide and intron 2 molecular variants have been detected,Reference Costa, Pini and Martini ⁵² and the GG genotype single nucleotide polymorphism within the oxytocin receptor gene is linked to high levels of separation anxiety in patients with major depression.Reference Costa, Pini and Gabelloni ⁵³ Attachment anxiety is linked to higher plasma cortisol levels, and further exploration of the links between cortisol and oxytocin may clarify mechanisms underlying the stress response.Reference Gordon, Zagoory-Sharon, Schneiderman, Leckman, Weller and Feldman ⁵⁴

Fear of separation is one of the basic fears of humans; it is readily observable in children, and generally becomes less prominent with maturity. It is presumed that marked separation anxiety represents a psychopathological phenomenon, but diagnostic criteria for separation anxiety disorder would benefit from being conceptually tightened. It has been argued by analogy that extreme versions of normal “death anxiety” could lead to the unnecessary category of “death anxiety disorder,” as fears of severe illness and dying are common across many psychopathological conditions.Reference Starcevic ⁵⁵ But the inclusion of separation anxiety disorder within the DSM-5 group of anxiety disorders diagnosable in adults certainly provides impetus for further research into its distinguishing psychopathological phenotypic features, and also identifying possible endophenotypes, which might help delineate it from frequent comorbid conditions, such as panic disorder, agoraphobia, and social anxiety disorder. Studies, such as the sensitivity of separation anxiety symptoms to carbon dioxide challenge, and also of the potential role of oxytocin in dysfunctional attachment, may provide fruitful lines of enquiry in helping separate this common and burdensome condition from apparently similar conditions.

Disclosures

David S. Baldwin reports “other” from NIHR Clinical Academic Fellow training programme, grants from Research Management Committee of the Faculty of Medicine at the University of Southampton, during the conduct of the study; grants from Pfizer Ltd, grants from H. Lundbeck A/S, outside the submitted work. Robert Gordon reports “other” from NIHR Clinical Academic Fellow training programme and grants from Research Management Committee of the Faculty of Medicine at the University of Southampton during the conduct of the study. Marianna Abelli and Stefano Pini have nothing to disclose.