# Auditory Verbal Hallucinations in Borderline Personality Disorder
### Cognitive, phenomenological and neurophysiological factors

###### 08/09/2020

---

### What is Borderline Personality Disoder?

##### "Borderline" - Patients exits in a 'grey area' between neurosis and psychosis (Stern, 1939; Kernberg, 1967)

##### ICD-10: "Emotionally unstable personality disorder"

#### DSM: "Some develop psychotic like symptoms (e.g.hallucinations, body distortions, hypnagogic phenemenon) during times of stress"

----

### Auditory Verbal Hallucinations (AVH)

##### 50-90% of BPD patients report hearing voices that other people don't hear (Kingdon et al, 2010; Yee et al, 2005)
##### ~40% of patients at Sussex Voices Clinic have BPD diagnosis

##### Hearing voices in BPD is sig risk factor for suicide plans, attempts and hospitalisation (Miller et al, 1993; Slotema et al, 2016)

##### However, concept of 'pseudo'-hallucination has fostered idea that patients' AVH experiences may be malingered.

---

### Aim - Better categorize the phenomonological, cognitive and neurophsyiological factors associated with AVH.

###### 1. Determine the role of AVH location in effecting anxiety
###### 2. Determine whether response bias (SDT) is associated with AVH, as in other groups
###### 3. Determine the spatial pattern of activation associated with AVH (BOLD)
###### 4. Determine patterns of functional connectivity that are associated with symptoms

---

# Methods

#### 3 Phases:
##### 1) Clinical assessment (questionaires - AVH Symptom Severity, Beliefs about Voices etc) (n = 48)

----

##### 2) Behavioural assessment (Signal Detection Task)
###### - Response bias


----

##### 3) Neuroimaging (fMRI) (n = 29)
###### - Measured thoughts; quality and quantity of AVH; and brain activity (BOLD) and functional connectivity (rs-sbfc)

#### If answer to 'I heard my voice/s' = Yes:

'How distressing did your voice/s feel?'
'How loud were your voice/s?'
'What proportion of the time were you hearing your voice/s?' [0% / 100%]
'Where did your voice/s sound like they were coming from?' [Inside Head / Outside Head]
...
'How anxious do you feel?'

---

# Results

----

### Signal Detection Task

###### - Response bias correlated with AVH symptom severity (r = .653, p = 0.001, n = 21) and persecutory beliefs (r = -.637, p = 0.006, n = 17)

----

## Factors related to anxiety



###### - Anxiety related to AVH Intensity (r = .352 , p = 0.004, n = 66) not AVH location (r = -.109, p = 0.382, r = 66)

----

# Task - fMRI Analysis

----

# Rest - fMRI Analysis


---

# Discussion

----

## Signal Detection

- Response Bias significantly related to AVH symptom severity, and Persecutory beliefs

- May represent an 'externalizing bias' in source monitering.
- This may be a trait-like risk factor that increases vulnerability to hallucinatons across different hallucinating groups (Brookwell et al, 2013)

----

## Anxiety

- Hallucination intensity, NOT percieved location, modulated anxiety.

- 'Pseudohallucination': a pseudoconcept (Zwaard & Polak, 2001)

----

## Task based fMRI


- Auditory cortex activity suggests the experience of an AVH is neurally consistent with actually hearing something

----

## Functional Connectivity at Rest

- Found FC between ACC <-> Precuneus & rA1 <-> Cuneus increased as AVH lenth increased.
- Suggests a pattern of coupling that may be important for the the maintainence of hallucinations (not just genesis)

----

5. Coupling between ACC and insula may serve as mechanism that is protective against AVH

---

## To summarize:

###### Cognitive Factors: Externalizing Bias
###### Phenomenological Factors: AVH intensity, not location
###### Neurophysiological Factors: Primary Auditory Cortex

#### - The evidence here substantiates and legitamizes the experience of AVH for those with BPD
#### - It reframes the experience from one that is seen as 'psuedo', to one that is comparible in terms of severity to hallucinations in schizophrenia

---

# Any Questions?

---

To calculate this bias measure, for each participant we calculated their hit and false alarm rates. The hit rate is the probability of reporting “voice” when a voice was present. The false alarm rate is the probability of reporting “voice” when a voice was absent. Then, c is given by -0.5*(Z(hit rate) + Z(false alarm rate)), where z is the standard z-score.