Borderline Features and Prescription Opioid Misuse in a Substance Use Disorder Treatment Sample

Abstract

Background: An association between borderline personality disorder (BPD) and substance use disorders has been well established. However, very little is known about the relationship between BPD and prescription opioid misuse, specifically. Objectives: The relationship between borderline personality disorder features and prescription opioid misuse was examined in a sample of 208 substance use disorder treatment patients in the outpatient level of care. Results: Controlling for use of alcohol and cannabis, as well as other relevant covariates, we found that BPD features were associated with age of first use of prescription opioids, prescription opioid use disorder symptom count, lifetime use, past 12-month use, problem use, and cravings. Additionally, we found that BPD features were not associated with greater use of medically necessary opioid pain killers as prescribed by a physician; rather the association with BPD was in the greater likelihood of misuse (non-prescribed) of prescription opioid pain killers. The self-harm/impulsivity facet of BPD was most strongly associated with prescription opioid-related variables. Conclusions/Importance: These findings suggest that BPD is related to prescription opioid misuse, above and beyond the tendency to use other drugs of abuse, and that the self-harm impulsivity facet appears to be driving this relationship.

Keywords:

• Prescription opioid misuse
• impulsivity
• self-harm
• borderline personality disorder
• substance use disorder treatment sample

Notes

1  Additionally, we fit a structural equation model in Mplus with manifest BPD features variable predicting the correlated alcohol, cannabis, and prescription opioid use latent variables, while controlling for sex and age. Although the model offered results consistent with those presented here in a more parsimonious analysis, the model fit statistics were not acceptable (x² = 429.18, TLI = .80, CFI = .84, RMSEA = .11 [CI .10-.12], SRMR = .07).

2  We also ran the model with each of the drug use variables simultaneously predicting BPD features, controlling for age and gender. In each of the models, the prescription opioid measures were the strongest predictor of BPD features: NSDUH symptoms, β = .29, p < .001; lifetime use β = .22, p = .001; past 12 mo. use, β = .22, p = .001; problem use, β = .20, p = .006; cravings, β = .17, p = .013.