EVIDENCE OF AGE BIAS ON THE ALTERNATIVE MODEL OF PERSONALITY DISORDERS
Lisa Stone-Bury, Alyssa Premovich, Daniel Segal- Life-span and Life-course Studies
- Health Professions (miscellaneous)
- Health (social science)
Abstract
Introduction
Previous research established substantial age bias on personality disorder (PD) diagnostic criteria for older adults. The Alternative Model of Personality Disorders (AMPD) is a new model of PDs proposed in DSM-5’s Section III that measures PDs dimensionally along two criteria: personality functioning with four domains (measured by the Levels of Personality Functioning Scale; LPFS) and pathological personality traits with five domains (measured by the Personality Inventory for DSM-5; PID-5). This study examined age bias on the AMPD in a cross-sectional design. Method: Older (n = 200) and younger adults (n = 213) completed the LPFS-Self-Report (LPFS-SR) and the PID-5-Brief Form (PID-5-BF).
Results
Differential Item Functioning (DIF) analyses were computed on items of the LPFS-SR and PID-5-BF. An item exhibits DIF if two age groups with similar levels of an underlying scale have different probabilities of endorsing an item. Overall, 18 of 80 items on the LPFS-SR demonstrated large DIF. The Empathy (50% of items) and Intimacy (25% of items) domains were most aged biased. For the PID-5-BF, 10 of 25 items showed large DIF. All five domains demonstrated large age bias, with Psychoticism (100% of items) most impacted. Across both DIF models, items showed evidence of being biased for (disproportionately high endorsement of an item) or against (disproportionately low endorsement) older adults.
Discussion
Findings indicate meaningful age bias on both AMPD diagnostic criteria, indicating that older adults are at-risk of being under- and over-diagnosed with PDs under the AMPD. Revision of the AMPD is warranted to achieve age neutrality.