Clinical Psychological Science

The potential hazards of endeavoring to recover ostensibly repressed memories of abuse in therapy have previously been documented. Yet no large survey of the general public about memory recovery in therapy has been conducted. In an age-representative sample of 2,326 adults in the United States, we found that 9% (8% weighted to be representative) of the total sample reported seeing therapists who discussed the possibility of repressed abuse, and 5% (4% weighted) reported recovering memories of abuse in therapy for which they had no previous memory. Participants who reported therapists discussing the possibility of repressed memories of abuse were 20 times more likely to report recovered abuse memories than those who did not. Recovered memories of abuse were associated with most therapy types, and most associated with those who reported starting therapy in the 1990s. We discuss possible problems with such purported memory recovery and make recommendations for clinical training.

In this comment on Patihis and Pendergrast (this issue, p. 3), we challenge an assumption that underpins recovered memory therapies: that there exists a close link of traumatic experiences with dissociation. We further suggest that (a) researchers examine how therapists who believe in repressed memories instill this belief in clients and establish expectations that current problems can be interpreted in light of past traumatic experiences, (b) recovered memories could be classified and studied as a function of how events come to light and are interpreted, and (c) therapists routinely provide informed consent regarding recovered memories and suggestive techniques.

The new national survey by Patihis and Pendergrast (this issue, p. 3) suggests that millions of people may have recovered traumatic memories that they spent large parts of their lives not thinking about. We wondered whether they are better off and suggest that more than a few may be worse off rather than better. Given this risk of therapy, should therapists be warning patients of the potential risks before conducting therapy? The answer is not clear as warning about risks can be risky itself. Overall, we propose that with so many people living with “recovered” memories, future research now needs to address whether they are indeed better off and which methods would help achieve that goal.

How often do clinical psychologists discuss with their adult clients the possibility that the clients might have been abused as children but had repressed the memory? If during the course of therapy clients remember being abused as children when the clients had no previous memories of such abuse, how likely is it that the memories are false? These questions underlie Patihis and Pendergrast’s Mechanical Turk survey study (this issue, p. 3). We discuss relevant scientific findings, including from longitudinal research on adults who as children experienced documented child maltreatment. We question inferences and generalizations resulting from the methodology Patihis and Pendergrast employed. We argue that clinicians are often justified in asking about past child abuse, remembered and forgotten, and that clinicians and researchers should strive to balance the risk of adults forming false memories with the need for adults to overcome childhood trauma.

We respond to various comments on our article (this issue, p. 3), which reported prevalence percentages of reports of recovered memories in therapy. We consider arguments against informed consent in therapy and conclude that we are in favor of informed consent that includes information about research on the malleability of memory. We note some useful suggestions from commentators, such as future research investigating iatrogenic outcomes of those who report recovered memories and investigating whether therapy-induced recovered memories are also an issue in various other countries. We understand that there are questions as to whether our sample was representative of the adult population of the United States, but we maintain that such questions can be investigated empirically and we could not find much evidence of systematic divergence. We investigated representativeness on gender, ethnicity, socioeconomic status, and age and made adjustments where possible. Future research should investigate reports of recovered memory in other general public samples.

Clinical science seems to have reached a tipping point. It appears that a new paradigm is beginning to emerge that is questioning the validity and utility of the medical illness model, which assumes that latent disease entities are targeted with specific therapy protocols. A new generation of evidence-based care has begun to move toward process-based therapies to target core mediators and moderators based on testable theories. This could represent a paradigm shift in clinical science with far-reaching implications. Clinical science might see a decline of named therapies defined by set technologies, a decline of broad schools, a rise of testable models, a rise of mediation and moderation studies, the emergence of new forms of diagnosis based on functional analysis, a move from nomothetic to idiographic approaches, and a move toward processes that specify modifiable elements. These changes could integrate or bridge different treatment orientations, settings, and even cultures.

In this invited commentary, I provide an historical and conceptual context for “The Future of Intervention Science: Process-Based Therapy” by Stefan Hofmann and Steven Hayes (this issue, p. 37). Special attention is paid to the importance of change mechanisms in psychological intervention, the use of functional analysis instead of categorical schemes such as the Diagnostic and Statistical Manual of Mental Disorders and treatment manuals in randomized clinical trials, and the challenge of applying nomothetic principles to individual cases. Historical context is provided for prior examination of these core issues in numerous articles and books extending back for over 60 years.

Hofmann and Hayes (this issue, p. 37) argue for a move away from a reliance on a medical illness model in which rigid therapy protocols are targeted at presumed disease entities in favor of a more process-based approach that targets core mediators and moderators based on testable theories. I could not agree more. I find their arguments persuasive and their implications exciting.

In this issue, Hofmann and Hayes (p. 37) shined a light on our field’s lack of progress in reducing the prevalence and burden of mental illness and suggested that a paradigm shift is needed to make more rapid progress. They argued that empirically supported processes of change should guide treatment planning, rather than treatment packages. This approach should allow for greater personalization of treatment and more iterative treatment planning and should increase the scalability of our intervention delivery models. In this commentary, I discuss some of the difficult questions that will arise when trying to make the shift from treatment packages to process-based therapy and highlight some steps that we can take now to facilitate the shift.

In this rejoinder, we discuss the commonalities and differences of the commentaries to our target article. Each commentary agreed with our basic message that intervention science needs to move from the DSM-governed protocols-for-syndrome approach to process-based treatments. Functional analysis has been a guiding principle since the early days of behavior therapy but lost its dominance with the rise of the latent disease model of psychiatry. This model gave rise to disorder-specific treatments with limited benefit to patients and science. We now have the tools and expertise to study human complexity grounded in an understanding of processes of change drawn from and fully applicable to the psychological level of analysis.

One of the primary goals of cognitive therapy for depression is to teach an individual a new, more adaptive way of thinking about stressful life experiences. We argue that this process of supplanting a deeply engrained way of negative thinking with a more adaptive style of thinking is similar to learning a second language. The purpose of this article was to use a second-language acquisition framework to evaluate the strategies typically used in cognitive therapy for changing depressogenic cognitive patterns and to propose new strategies that may make the therapy more effective.

The categorical model of classification in the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders is sorely problematic. A proposed solution is emerging in the form of a quantitative nosology, an empirically based dimensional organization of psychopathology. More specifically, a team of investigators has proposed the Hierarchical Taxonomy of Psychopathology (HiTOP). The purpose of this article is to discuss the potential role, importance, and implications of personality within the HiTOP dimensional model of psychopathology. Suggested herein is that personality provides a foundational base for the HiTOP dimensional model of psychopathology. Implications concern the potential value of the early assessment of and screening for personality as well as the development of protocols for the treatment of personality trait domains, which may in turn contribute to substantial improvements in quality of life as well as mental and physical health.

It is unclear whether impaired cognition is a risk factor for depression or a consequence of depression, or whether both depression and impaired cognition are caused by a third underlying process (e.g., stress). These three hypotheses were tested in 523 adolescents assessed annually for depression, attentional functioning, and childhood/recent life stress. Baseline switching, sustained, and selective attention did not predict first onset of depression or depressive symptoms. Divided attention predicted depressive symptoms only. Piecewise growth modeling indicated that the trajectory of switching attention declined prior to first onset of depression; there was evidence of significant recovery in switching attention following first onset of depression. Structural equation modeling indicated that impaired switching attention prospectively predicted higher depressive symptoms and that higher depressive symptoms predicted worse selective and switching attention. Further, childhood stress prospectively predicted higher depressive symptoms via switching attention and worse switching attention via depressive symptoms.

An influential evolutionary model proposed that social anxiety biases people to treat social interactions as competitive struggles with the primary goal of avoiding status loss. Among subordinate nonhuman primates in highly hierarchical social groups, this goal leads to adaptive submissive behavior; for humans, however, affiliative responses may be more effective. We tested three predictions about social anxiety and social cognitions, emotions, and behavior that Trower and Gilbert advanced. College students (N = 122) whose self-reported social anxiety ranged from minimal to extremely high played the Prisoner’s Dilemma game three times. Consistent with two model-based predictions, social anxiety was positively associated with self-reported competitive goals and with nervousness during game play. Unexpectedly, however, social anxiety was associated with a tendency to engage with coplayers in an ostensibly hostile, rather than appeasing, manner. We discuss implications of these findings for updated models of socially anxious behavior.

Multiple executive functions (EFs) are associated with trait anxiety and depression symptoms, but it is unclear how genetic and/or environmental factors account for these associations and whether they are explained by general variance underlying multiple EFs (i.e., common EF). In this study, 1,207 male twins completed seven EF tasks and measures of trait anxiety and depression symptoms at an average age of 62 years. The common EF factor was associated with both anxiety (r = −.24) and depression (r = −.32) symptoms. Anxiety and depression had a near-complete genetic overlap (r_g = .96). Genetic influences shared with depression accounted for 83% of the phenotypic correlation between anxiety and common EF. For depression, the phenotypic correlation with common EF was explained by genetic influences shared with anxiety (69%) but also nonshared environmental influences distinct from anxiety (20%). These results suggest that genetic influences on common EF play a role in the observed genetic overlap between anxiety and depression.

Difficulties with emotion regulation in depression may be linked not only to emotion regulation strategies but also to the motivation to experience certain emotions. We assessed the degree of motivation to experience happiness or sadness in major depressive disorders outside the laboratory and prospective links to clinical outcomes over time. Depressed individuals were consistently less motivated to experience happiness and more motivated to experience sadness than nondepressed individuals. The less motivated participants were to experience happiness, the less they tried to upregulate happiness in an emotion regulation task and downregulate negative emotions during real-life stress. Importantly, the less motivated depressed participants were to experience happiness, the more clinical symptoms they exhibited months later during a stressful period, even after controlling for initial levels of symptoms. These findings demonstrate that individual differences in the degree of motivation to experience happiness in depression may carry clinical implications.

Obsessions are commonly described as intrusive, ego-dystonic, and coming out of nowhere. This might reflect an experience of a low sense of agency (SoA; i.e., the experience of being the source of one’s own thoughts). In this study, we investigated the relationship between obsessive-compulsive (OC) symptoms and the SoA over thoughts. Participants were told that subliminal auditory primes (actually sham) can insert thoughts into their minds, and their experiences of inserted thoughts were collected online. The results of three experiments showed reduced SoA over thoughts in individuals with subclinical OC symptoms, regardless of the valence of the thoughts. Several potential confounding factors (e.g., suppression, vigilance, general anxiety) were ruled out. The experience of inserted thoughts was related to experiencing thoughts as “out of context,” which partially explained reduced SoA in participants with high levels of OC symptoms. These experiments highlight the importance of focusing on the low-level, contextual, and phenomenological characteristics of intrusive thoughts in addition to their content and appraisals.