DOI: 10.1192/j.eurpsy.2026.10501 ISSN: 0924-9338

Diminishing Loss Sensitivity during Risky Decision-Making among Individuals with Gambling Disorder

H. Wei, G. Zhong, J. Liu, Y. Wei, X. Zhang, P. Yang, M. Zhao, J. Du
Show PDF Cite

Introduction

Gambling disorder (GD) poses severe impacts on both individuals and society. Impairment in risky decision-making is a key behavioral characteristic of GD, but the underlying cognitive processes of these deficits remain unclear.

Objectives

This study decomposed the risky decision-making processes of GD with cognitive computational modeling.

Methods

A total of 100 participants with GD and 59 healthy controls (HCs) were recruited to complete psychometric assessments and the Balloon Analog Risk Task. Since GD involved abnormal loss evaluation in decision-making, we developed a novel cognitive model incorporating diminishing loss sensitivity and revealed the processes underlying the risk-taking behaviors with hierarchical Bayesian analysis.

Results

GD participants exhibited stronger loss aversion ( p < 0.001) but faster-diminishing loss sensitivity ( p < 0.001), regardless of severity, which drove the deficits in the overall performance of risky decision-making ( p < 0.001) (Fig1). Overconfident prior belief ( p = 0.002) and higher updating rate ( p < 0.001) were observed among participants with GD (Fig3). Diminishing loss sensitivity was negatively correlated with impulsivity ( p = 0.021), and loss aversion was negatively related to craving for gambling ( p = 0.027) (Fig4).

Table 1.

Demographic and clinical characteristics of participants

Table 1. Long description.

Characteristic
Total, N = 159
HC, n = 59 GD, n = 100 F / H p Effect Size
Age, Years
30.84 ± 7.91
31.71 ± 9.91 30.33 ± 6.45 H 1 = 0.28 0.595 η 2 = 0
Education, Years
14.96 ± 3.25 15.27 ± 4.49 14.77 ± 2.23 H 1 = 2.57 0.109 η 2 = 0.010
BIS Score
82.78 ± 17.64 69.20 ± 16.04 90.79 ± 13.09 F 1, 157 = 85.127 < 0.001 η 2 = 0.352
VAS Score
3.65 ± 3.34

HC: Health controls; GD: Gambling disorder; BIS: Barret Impulsivity Scale; VAS: Visual Analog Scale for gambling craving.
Table 2.

Demographic and clinical characteristics of subgroups of participants.

Table 2. Long description.

Characteristic
Total, N = 159
HC, n = 59 Mild&Moderate, n = 46 Severe, n = 54 F / H p Effect Size
Age, Years
30.84 ± 7.91
31.71 ± 9.91 30.78 ± 5.86 29.94 ± 6.95 H 2 = 1.48 0.476 η 2 = 0
Education, Years
14.96 ± 3.25 15.27 ± 4.49 14.93 ± 1.84 14.63 ± 2.52 H 2 = 2.60 0.273 η 2 = 0.004
BIS Score
82.78 ± 17.64 69.20 ± 16.04 88.50 ± 13.99 92.74 ± 12.07 F 2, 156 = 44.001 < 0.001 η 2 = 0.361
VAS Score
3.07 ± 3.01 4.15 ± 3.55 H 1 = 2.00 0.158 η 2 = 0.010

HC: Health controls; Mild&Moderate: Mild and moderate gambling disorder participants; Severe: Severe gambling disorder participants; BIS: Barret Impulsivity Scale; VAS: Visual Analog Scale for gambling craving.
Image 1:

Image 1: Long description.

Image 2:

Image 2: Long description.

Image 3:

Image 3: Long description.

Conclusions

This research provides novel perspectives on the cognitive processes underlying the risky decision-making of GD patients, highlighting the role of diminishing loss sensitivity during loss evaluation and its clinical implications, which inspires future research on assessment and psychotherapy for GD.

Disclosure of Interest

None Declared

More from our Archive