This is the first study to investigate the baseline demographic, clinical, and treatment characteristics that predict improved functioning at 12-month follow up among young people receiving early intervention for BPD. Three main findings emerged from this study. First, having at least one caregiver in employment at baseline uniquely predicted better social adjustment at 12 months. Second, each of lower BPD severity, fewer co-occurring personality disorder diagnoses, and fewer interpersonal problems at baseline uniquely predicted fewer interpersonal problems at 12 months. Third, suicide attempts (SA) and non-suicidal self-injury (NSSI) over the previous 12 months did not seem to be influential upon functional outcomes and were not included in the final model.
While better overall social adjustment (including occupational, leisure, social and relationships) at 12 months was uniquely predicted by having at least one caregiver in employment, it is noteworthy that this was not predicted by commonly used selection criteria for clinical service provision, i.e., general indicators of clinical severity, or the level of social adjustment. This supports findings from a large epidemiological study (N = 2281), which found that the lower levels of exposure to parental joblessness during childhood was associated with improved wage outcomes in early adulthood, even after adjusting for demographic and socioeconomic factors [34]. Although these findings are associations, two mechanisms might explain how caregiver employment might lead to improved long-term occupational functioning in their offspring. First, the ‘resource-based framework’ [35] suggests that parental employment increases a caregiver’s ability to gain access to resources that might be beneficial to the child’s development, including food, housing, and the capacity to invest in early educational environments conducive to growth and learning. In turn, this might enhance the formation of skills required for future labour force attachment [36]. Second, the ‘socialisation framework’ [37] suggests that the presence of work role models and positive parental attitudes toward employment might increase a child’s self-efficacy and motivation for employment [38]. The current findings highlight the importance of further research designed to investigate the potential roles of caregiver educational and occupational attainment in treatment outcome.
The baseline characteristics that predict fewer interpersonal problems at 12 months (BPD severity, number of co-occurring PD diagnoses, severity of interpersonal problems) fall under the rubric of indicators of PD severity. The finding that lower BPD severity at baseline uniquely predicted fewer interpersonal problems following treatment is supported by previous epidemiological research showing that children and adolescents from the community with lower levels of BPD severity experienced improved peer relationships at two, four, nine and twenty year follow-up [14–16, 39]. These findings are also supported by a cross-sectional study of adolescents (N = 177; aged 15–18 years) receiving outpatient treatment for BPD, which found that poorer peer relationships were predicted by a diagnosis of a PD (BPD or other PD), when adjusting for disruptive behaviour disorders, mood disorders, substance use disorders and anxiety disorders [40].
The finding that SA and NSSI over the previous 12 months do not seem to be influential in 12-month functional outcomes reinforces the case to reconsider entry criteria for early intervention treatment programs. In many clinical settings, recent SA and NSSI garner much of the short-term focus of care and are highly influential in clinical decision-making, especially with regard to offering treatment [41]. Evidence supports SA and NSSI as a population level transdiagnostic marker of psychopathology and suicide risk among young people [42]. However, the current finding adds weight to the argument that, among clinical populations, where SA and NSSI are ubiquitous [99% lifetime history among the MOBY sample; 43], and have limited utility as a test for later suicide because of their modest sensitivity and low positive predictive value [44, 45], SA and NSSI should not exert disproportionate influence upon who is selected for treatment.
Strengths and Limitations
This is the first study to investigate predictors of functional outcomes among young people receiving early intervention for BPD. Strengths of this study include the intentional recruitment of a ‘real-world’ sample of young people, typical of frontline clinical services, who had never received a diagnosis of or specialised treatment for BPD. Also, the methodological process of eliminating variables prior to the final regression analysis, along with the relatively large sample size, meant that a large number of predictor variables could be examined, without compromising the power of the study. Despite the large number of variables measured, there remains a substantial amount of unexplained variance, suggesting that there might be important variables not usually measured in clinical trials for personality disorder that might influence functional outcomes.
Limitations include that the method for defining social disadvantage rank was based upon postcode, rather than individual or household variables. It is possible that the relationship between having at least one caregiver in employment and better functional outcomes might be explained by other individual or household indicators of social disadvantage. It is also possible that participants with poor outcomes are under-represented at the 12-month time point, as the presence and number of mental disorders at follow-up, rather than baseline, has been found to be significantly associated with follow-up contact difficulty in longitudinal studies among young people [46].