The Impact of Attachment Insecurity on Pain and Pain Behaviors in Experimental Pain

Article Information and Overview

• Title: The impact of attachment insecurity on pain and pain behaviors in experimental pain
• Journal: Journal of Psychosomatic Research 111 (2018) 127–132
• Authors: Tonny Elmose Andersen, Sophie Lykkegaard Ravn, Claus Manniche, and Søren O'Neill.
• Institutions:
  • Department of Psychology, University of Southern Denmark, Odense, Denmark.
  • The Specialized Hospital for Polio and Accident Victims, Roedovre, Denmark.
  • Spine Center of Southern Denmark, Middelfart, Denmark.
• Core Objective: To assess how dimensions of insecure attachment (anxiety and avoidance) impact pain behaviors during laboratory-induced pain, specifically using a cold pressor procedure.

Fundamental Concepts and Definitions

• Pain Perception vs. Pain Behavior: These are distinct phenomena. Pain perception is the subjective experience, while pain behaviors are the physical enactments of that experience.
• Pain Behaviors: Defined as "specific body movements enacted during the experience of pain." They serve two primary functions:
  • Protective: Movements that shield the body from further harm.
  • Communicative: Expressions intended to elicit empathy or caring behaviors from observers.
• Attachment Orientations: Complex interpersonal schemas of self and others that influence threat perception, emotion regulation, and responses to stressors like pain.
  • Attachment Anxiety: Characterized by worry over the availability of others and their positive regard. It is associated with hyper-activating strategies such as catastrophizing and hypervigilance.
  • Attachment Avoidance: Characterized by discomfort with closeness and dependence on others. It is associated with under-activating strategies such as underestimating threats.
• Fearful Attachment: A combination of high anxiety and high avoidance, characterized by a disorganized shift between approach and avoidance behaviors.

Theoretical Frameworks and Background

• Gate Control Theory: Proposed over $50$ years ago, stating that painful stimuli are extensively modulated by the central and peripheral nervous systems, making psychosocial factors critical to the pain experience.
• Heuristic Model (Meredith et al.): Suggests attachment insecurity acts as a diathesis for chronic pain. It proposes that attachment moderates the relationship between pain appraisals, emotional responses, and outcomes.
• Kolb's Comparison (1982): Was the first to directly compare pain behaviors (like complaining) to attachment behaviors, viewing them as a "cry for security."
• Interpersonal Context: Pain behaviors can lead to positive responses (social support) or negative responses (burdening others), leading individuals to either suppress or exaggerate expressions based on perceived social consequences.

Methodology and Study Design

• Participants:
  • $N = 60$ patients with low back pain (LBP).
  • Age range: $18-65$ years (Mean age $= 52.8$, $SD = 8.6$).
  • Gender: $45\%$ female.
  • Recruited from a spine center in the Region of Southern Denmark (catchment area of $1.2$ million people).
  • Average chronic LBP duration: $3.97$ years ($SD = 6.16$).
• Assessment Tools:
  • Revised Adult Attachment Scale (RAAS): 18-item self-report. Measured Avoidance ($\alpha = 0.66$) and Anxiety ($\alpha = 0.81$) on a 5-point Likert scale.
  • Visual Analogue Pain Scales (VAS): $100\,mm$ paper scale to measure pain intensity.
  • Cold Pressor Test (CPT):
    • Equipment: $25$-l water tub refrigerated at $0-2\,^\circ C$.
    • Procedure: Submerge non-clenched hand up to the wrist (circulating water at $10\,l/min$, $0.5\,bar$) for $2$ minutes or until unbearable.
• Recording and Coding:
  • Participants were video recorded during CPT and $30\,s$ post-immersion.
  • Five female psychology student assistants conducted the tests, specifically trained to be "friendly but neutral" and avoid small talk.
  • Coders (blind to attachment scores) used a manual adapted from Keefe and Block.
  • Pain Communication: Combined facial expressions (grimacing, wincing) and verbal/para-verbal expressions (grunts, sighs).
  • Pain Protection: Body movements like bending forward, neck arching, hand clenching, and rocking.
  • Inter-rater Reliability: Intra-class correlation coefficient ($ICC$) was $0.78$ for protection and $0.64$ for communication.

Quantitative Results

• Correlations:
  • Pain Intensity and Pain Communication: $r = 0.48$ ($p < 0.001$).
  • Pain Intensity and Time (immersion duration): $r = -0.48$ ($p < 0.001$).
  • Pain Intensity and Pain Protection: $r = 0.23$ ($p = 0.08$, non-significant).
  • Unexpected Result: Neither attachment anxiety ($p = 0.48$) nor attachment avoidance ($p = 0.49$) correlated directly with pain intensity or pain behaviors in the simple correlation matrix.
• Moderation Analysis (Attachment Anxiety):
  • Attachment anxiety significantly moderated the relationship between pain intensity and pain communication ($b = -0.26$, $95\%\,CI [-0.505, -0.020]$, $p = 0.03$).
  • Attachment anxiety significantly moderated the relationship between pain intensity and pain protection ($b = -0.29$, $95\%\,CI [-0.552, -0.034]$, $p = 0.03$).
  • The model accounted for $31\%$ of the variance in pain communication and $14\%$ in pain protection.
• Moderation Analysis (Attachment Avoidance):
  • Attachment avoidance was neither a significant predictor nor a moderator in any model.

Regression Slope Analysis for Attachment Anxiety

• Low Attachment Anxiety:
  • Communication: Significant positive relationship with pain ($b = 3.012$, $p = 0.008$).
  • Protection: Significant positive relationship with pain ($b = 1.968$, $p = 0.03$).
  • Interpretation: Concordance between internal pain and outward behavior is highest in securely attached individuals.
• Mean Attachment Anxiety:
  • Communication: Significant positive relationship ($b = 1.945$, $p = 0.001$).
  • Protection: Non-significant positive relationship ($b = 0.775$, $p = 0.13$).
• High Attachment Anxiety:
  • Communication: Non-significant positive relationship ($b = 0.877$, $p = 0.18$).
  • Protection: Non-significant negative relationship ($b = -0.419$, $p = 0.44$).
  • Interpretation: Individuals with high attachment anxiety appear to decouple their pain behaviors from their actual pain intensity, often downplaying communication.

Discussion and Clinical Implications

• Departure from Theory: Theoretically, attachment anxiety is associated with hyper-activation (over-reporting). However, this study found high-anxiety individuals downplayed behavior (lower concordance).
• Potential Mechanisms:
  • Fear of Rejection: Participants with high anxiety may have a history of rejection when expressing pain. In the presence of a neutral/unknown observer, they may inhibit behaviors to avoid social exclusion or negative responses.
  • Emotional Numbing: Fear of rejection may result in emotional numbing, affecting the overt display of pain.
  • Contextual Shaping: Pain behaviors are shaped by context. If behaviors are perceived as serving no useful function or being counterproductive in a specific setting (e.g., a formal medical experiment), they are inhibited.
• Clinical Significance: If patients with high attachment anxiety suppress pain communication, healthcare providers may underestimate their pain level. This is problematic because pain communication is necessary to elicit the empathy and support essential for health-promoting behaviors.

Study Limitations

• Sample Size: The sample ($N = 60$) was relatively small, limiting statistical power.
• Chronic vs. Acute: Participants were chronic low back pain patients; their responses to experimental cold pressor pain may differ from pain-free individuals used in previous studies.
• Lack of Priming: The study did not actively manipulate or "prime" the attachment system (e.g., through social exclusion paradigms).
• Environment: The interaction with unknown laboratory assistants was kept neutral, which may have influenced the behavioral inhibition observed in anxious participants.
• Attachment Levels: Generally, levels of attachment insecurity in the sample were quite low.
• Alternative Variables: Factors such as pain-catastrophizing, trait anxiety, and fear-avoidance beliefs were not adjusted for in the primary models.