An exploratory study of the relationship between changes in emotion and cognitive processes and treatment outcome in borderline personality disorder

Abstract

This exploratory study examined specific emotion processes and cognitive problem-solving processes in individuals with borderline personality disorder (BPD), and assessed the relationship of these changes to treatment outcome. Emotion and cognitive problem-solving processes were assessed using the Toronto Alexithymia Scale, the Linguistic Inquiry Word Count, the Derogatis Affect Balance Scale, and the Problem Solving Inventory. Participants who showed greater improvements in affect balance, problem solving, and the ability to identify and describe emotions showed greater improvements on treatment outcome, with affect balance remaining statistically significant under the most conservative conditions. The results provide preliminary evidence to support the theory that specific improvements in emotion and cognitive processes are associated with positive treatment outcomes (symptom distress, interpersonal functioning) in BPD. The implications for treatment are discussed.

本探索性研究檢視邊緣性人格疾患(BPD)的特定情緒與認知問題解決歷程，並評 估這些歷程的改變與治療結果的關係。情緒和認知問題解決歷程係以多倫多情 緒表達障礙量表(Toronto Alexithymia Scale)，語言詢問字句計算(Linguistic Inquiry Word Count)，德若伽提斯情感平衡量表(Derogatis Affect Balance Scale)， 以及問題解決量表(Problem Solving Inventory)等工具加以評估。研究發現情緒平 衡、問題解決、以及辨識和描述情緒的能力較佳的參與者在治療結果上呈現較 大幅度的進步，即使在最保守的狀況，這些參與者的情緒平衡仍具有統計的顯 著性。這些結果初步支持邊緣性人格疾患的特定情緒和認知歷程的進步與正向 治療結果有關(憂傷症狀、人際功能)。進一步討論研究對於治療的意義。

Questo studio ha esaminato i processi emotivi specifici e i processi cognitivi di problem solving in soggetti con disturbo borderline di personalità e ha valutato la relazione tra questi cambiamenti e gli esiti di trattamento.

L'emozione e i processi di problem solving sono stati valutati con l'utilizzo della Toronto Alexithymia Scale, il Linguistic Inquiry Word Count, la Derogatis Affect Balance Scale, e il Problem Solving Inventory. I soggetti che hanno fatto registrare un ottimo miglioramento nella modulazione dell'emozione, nella gestione delle situazioni problematiche e nell'abilità di identificare e descrivere le emozioni presentavano anche uno degli ottimi esiti di trattamento.

I risultati mostrano una prima conferma dell'ipotesi secondo la quale specifici miglioramenti nei processi emotivi e cognitivi sono associati a risultati positivi nel trattamento in pazienti con disturbo borderline di personalità.

Diese explorative Studie untersucht spezifische emotionale Prozesse und kognitive Problemlöseprozesse bei Individuen mit Borderline Persönlichkeitsstörung (BPKS) und misst den Zusammenhang diesbezüglicher Veränderungen und dem Therapieergebnis. Emotionen und kognitive Problemlöseprozesse wurden mittels der Toronto Alexithymia Skala, dem Linguistic Inquiry Word Count, der Derogatis Affect Balance Skala und dem Problem Solving Inventar erhoben. Studienteilnehmer, die stärkere Verbesserungen bezüglich Affektregulierung, Problemlösung und der Fähigkeit zur Identifikation und Beschreibung von Emotionen aufwiesen, zeigten stärkere Verbesserungen im Therapieergebnis, wobei die Affektregulation auch unter den konservativsten Maßstäben signifikant blieb. Die Ergebnisse liefern vorläufige Evidenz zur Unterstützung der Theorie, dass spezifische Verbesserungen in emotionalen und kognitiven Prozessen assoziiert sind mit einem positiven Therapieergebnis (Symptombelastung, interpersonelles Funktionsniveau) bei BPKS. Die Implikationen für die Behandlung werden diskutiert.

Este estudo exploratório examinou processos emocionais específicos e processos cognitivos de resolução de problemas em indivíduos com perturbação de personalidade borderline (PPB) e avaliou a relação entre estas mudanças e o resultado terapêutico. Os processos emocionais e cognitivos de resolução de problemas foram avaliados através da Escala de Alexitimia de Toronto (Toronto Alexithymia Scale), do Inquérito Linguístico de Contagem de Palavras (Linguistic Inquiry Word Count), da Escala de Equilíbrio Afetivo de Derogatis (Derogatis Affect Balance Scale) e do Inventário de Resolução de Problemas (Problem Solving Inventory). Os participantes que demonstraram maior progresso no equilíbrio afetivo, na resolução de problemas e na capacidade de identificar e descrever emoções revelaram maior evolução em termos de resultado terapêutico, sendo que o equilíbrio afetivo se mostrou estatisticamente significativo mesmo sob as condições mais conservadoras. Os resultados fornecem evidência preliminar de suporte à teoria de que progressos específicos ao nível dos processos emocionais e cognitivos se encontram associados a resultados terapêuticos positivos (distress sintomático, funcionamento interpessoal) em PPB. As implicações para o tratamento são discutidas.

Keywords:

• borderline personality disorder
• emotion processes
• cognitive problem-solving processes

Acknowledgments

The authors extend sincere gratitude to all the patients and therapists who participated in this trial, without whom this research would not have been possible. We also extend thanks to our colleagues at the Centre for Addiction and Mental Health, St. Michael's Hospital, and the University of Toronto, Department of Psychiatry, for their support of this work. Finally, we thank Ryan Barnhart, who provided us with statistical consultation.

This study was supported by the Ontario Mental Health Foundation and the Canadian Institutes for Health Research, Grant #200204MCT-101123.

Notes

¹ The Holm sequential Bonferroni adjustment reduces alpha (.05) by the number of hypotheses remaining to be tested. The results are ordered by the observed p-values from smallest to largest and each observed p-value is compared to the adjusted alpha in sequence. The smallest p-value is compared to alpha (.05) divided by the total number of hypotheses. The next smallest p-value is then compared against alpha (.05) divided by the total number of hypotheses less 1 (the total number remaining). This continues until non-significance against the adjusted alpha is observed. All others are then considered non-significant.

² The imputation models contained more and different predictor variables than the substantive regression models of interest (Kenward & Carpenter, 2007).

³ REML is a maximum likelihood method and so it does not estimate the coefficients by minimizing the variance, rather it maximizes the likelihood function. The R2 statistic is therefore not calculable.