Introduction
Emotion is a complex and multifaceted process that involves different evaluative components, including appraisal processes evaluating the meaning and relevance of actual or imagined events [Reference Peil1]. These appraisals may be consciously modulated to regulate emotions, such as during the reappraisal of negative emotions to neutral or positive terms. The ability to carry out this process is an important factor for determining well-being and the presence of psychopathology [Reference Aldao, Gee, De Los Reyes and Seager2,Reference Thompson-Schill, Bedny and Goldberg3]. One way to modulate such emotion appraisals is via cognitive reappraisal, an antecedent-focused cognitive control strategy that allows reframing emotion-inducing stimuli or scenarios in positive terms, which leads to decreased sympathetic activity and negative affect, better interpersonal functioning, and increased physical and psychological well-being [Reference Gross and John4].
In neurobiological terms, emotion regulation is characteristically implemented by the circuits linking different regions of the prefrontal cortex (PFC) with subcortical structures, such as the amygdala, related to emotional responding [Reference Goldin, McRae, Ramel and Gross5–Reference Picó-Pérez, Radua, Steward, Menchón and Soriano-Mas7]. More specifically, regulatory input to subcortical structures is assumed to be mediated by activity in the dorsolateral prefrontal cortex (dlPFC) and orbitofrontal cortices (OFCs), with recent evidence supporting that disturbances in these circuits are associated with emotion regulation behavior [Reference Scheuerecker8,Reference Steward, Davey, Jamieson, Stephanou, Soriano-Mas, Felmingham and Harrison9]. Moreover, other PFC regions, such as the ventrolateral prefrontal cortex (vlPFC), have been shown to be implicated in selecting goal-appropriate responses and retrieving information from semantic memory, which can then be used to develop new appraisals [Reference Thompson-Schill, Bedny and Goldberg3,Reference Badre and Wagner10].
Several studies have demonstrated that patients with psychiatric disorders have difficulties in using cognitive reappraisal [Reference Campbell-Sills, Ellard, Barlow and Gross11], although the mechanisms of alteration may differ across conditions. Alterations in emotion regulation are central to both major depressive disorder (MDD) and borderline personality disorder (BPD); however, they may be underpinned by different pathological mechanisms. First, individuals with BPD show fluctuations in subcortical system functioning, which results in failure to habituate and hypersensitivity to threat cues [Reference Gunderson, Herpertz, Skodol, Torgersen and Zanarini12,Reference Sicorello and Schmahl13]. Importantly, this has been suggested to underlie several of the pathological manifestations of this disorder, including affective instability, intense and tumultuous relationships, difficulty controlling anger, impulsivity, suicidal tendencies, and deliberate self-harm (thought to serve an emotion-regulating function) [Reference Harford, Chen, Kerridge and Grant14,Reference Schulze, Domes, Krüger, Berger, Fleischer and Prehn15]. Patients with MDD, in contrast, present a distinctive clinical profile, and portray cognitive impairments related to basic elements of emotional processing [Reference Heinzel, Northoff, Boeker, Boesiger and Grimm16]. These have been linked to decreased prefrontal recruitment during the explicit voluntary control of emotions [Reference Rive, van Rooijen, Veltman, Phillips, Schene and Ruhé17]. Nevertheless, such putatively distinct neurobiological mechanisms of altered emotion regulation have not been directly compared. This comparison may be, however, of great interest not only to further understand the different mechanisms of psychological maladjustment in BPD and MDD, but also to develop disorder-specific approaches to improve emotion regulation capacity.
The present study is, to the best of our knowledge, the first to investigate the neurobiological underpinnings of disrupted emotion regulation in MDD and BPD using a cognitive reappraisal paradigm and the concurrent evaluation of regional brain activity with functional magnetic resonance imaging (fMRI). Moreover, given the central role of interregional connectivity alterations in neurobiological models of emotion regulation, we decided to not only assess task-related activations, but also task-modulations of interregional connectivity. We hypothesized that both MDD and BPD groups would differ from health controls (HCs) in measures of brain activity and connectivity during cognitive reappraisal. More specifically, we hypothesized that patients with BPD would show increased subcortical activations related to inefficient regulatory input from prefrontal areas, while patients with MDD would present reduced recruitment of prefrontal areas during cognitive reappraisal. Finally, we also anticipated that these neurobiological alterations would be correlated to core clinical measurements in both patient groups.
Methods
Sample
The study included three groups of participants: patients with BPD (n = 19), patients with MDD (n = 20), and HCs (n = 19), which were recruited at Fundación Lucha contra las Enfermedades Neurológicas en la Infancia (FLENI Foundation) in Buenos Aires, Argentina. The groups consisted of 19 males and 39 females, ranging from 21 to 63 years of age (mean = 41.26; SD = 13.11). Patients were consecutively recruited when attending the Department of Psychiatry at FLENI Foundation if they met DSM-5 diagnostic criteria for BPD or MDD. All participants were evaluated via a clinical interview in order to confirm their DSM-5 diagnosis (patients) or the absence of any present or past diagnosis of a psychiatric disorder (HCs). Table 1 summarizes the sociodemographic and clinical features of the study participants. As can be seen in this table, mean Hamilton Depression Rating Scale (HDRS) scores indicated that most patients with MDD were in remission at the time of study. Specifically, 65% of patients with MDD, and also 68% of patients with BPD, scored below the 7-point cutoff value for depression, and, therefore, results of this study should be considered as related to stable depression features more than to transient alterations related to fluctuations in mood state. Further information regarding the sample and exclusion criteria can be found in the Supplementary Material.
Demographic and clinical characteristics of the study group.
Notes: Symbol references for significant pair-wise between-group differences: *HC–BPD; °HC–MDD; ^BPD–MDD.
Abbreviations: BPD, borderline personality disorder; CERQ, Cognitive Emotion Regulation Questionnaire; DERS, Difficulties in Emotion Regulation; HARS, Hamilton Anxiety Rating Scale; HCs, healthy controls; HDRS, Hamilton Depression Rating Scale; MDD, major depressive disorder.
a Although no across-group differences were observed in these variables, in clinical groups, these values significantly differed (one-sample t-tests) from those obtained in a population of nondepressed older adults [Reference Carvajal, Molina-Martínez, Fernández-Fernández, Paniagua-Granados, Lasa-Aristu and Luque-Reca18]: CERQ reappraisal (lower values in clinical groups), HCs: p = 0.882 (n.s.), BPD: p = 0.098 (trend-level), MDD: p = 0.017; CERQ rumination (higher values in all groups), HCs: p = 0.019, BPD: p < 0.0005, MDD: p = 0.001.
The present study was carried out in accordance with the latest version of the Declaration of Helsinki. The Ethics Committee in clinical research of FLENI Foundation approved the study. Signed informed consent was obtained from all participants.
Psychometric assessment
All participants completed the validated Spanish versions of the Cognitive Emotion Regulation Questionnaire (CERQ) and the Difficulties in Emotion Regulation Scale (DERS) to evaluate emotion dysregulation. Likewise, all subjects also completed the HDRS and the Hamilton Anxiety Rating Scale (HARS) to assess severity of depression and anxiety symptoms, respectively [Reference Carvajal, Molina-Martínez, Fernández-Fernández, Paniagua-Granados, Lasa-Aristu and Luque-Reca18–Reference Hervas and Jodar24].
MRI acquisition
This information can be found in the Supplementary Material.
fMRI task, cognitive reappraisal paradigm
We used a well-validated paradigm to evaluate brain activations during emotion regulation with fMRI using negative images and in-scanner behavioral ratings [Reference Phan, Fitzgerald, Nathan, Moore, Uhde and Tancer6,Reference Albein-Urios, Verdejo-Román, Soriano-Mas, Asensio, Martínez-González and Verdejo-García25]. Picture stimuli were obtained from the International Affective Picture System [Reference Lang, Bradley and Cuthbert26]. The task consisted of three conditions (“observe,” “maintain,” and “regulate”) presented in an ABC design with four blocks per condition (i.e., a total of 12 blocks). An additional description of the task is detailed in the Supplementary Material. Like in most previous research, participants were instructed to use distancing or reinterpretation as reappraisal strategies. These are two antecedent-focused strategies acting before emotional responses have been completely generated. The former refers to rationalizing the content of a situation by adopting the perspective of an uninvolved observer, while the latter refers to changing the meaning of stimuli in order to view the outcome of a situation in a more positive light [Reference Dörfel, Lamke, Hummel, Wagner, Erk and Walter27]. All blocks consisted of two consecutive images (each image was presented on screen for 10 s, with no interstimulus interval), and each block was followed by 10s of baseline during which a cross fixation was presented on the screen to minimize carryover effects [Reference Steward, Picó-Pérez, Mata, Martínez-Zalacaín, Cano and Contreras-Rodríguez28].
fMRI preprocessing and analysis
A thorough description of this section can be found in the Supplementary Material.
Results
The demographic and clinical characteristics of the sample are displayed in Table 1. Because patients were consecutively recruited, groups significantly differed in age (patients with MDD were older). For this reason, age was introduced as a nuisance covariate in all analyses.
Intrascanner ratings
Overall, in-scanner emotion ratings were the highest during the maintain blocks (mean = 3.08; SD = 0.89), followed by regulate (mean = 2.56; SD = 0.90) and observe (mean = 1.860; SD = 0.99) blocks (F = 4.30; p = 0.02). We did not observe, however, significant across-group differences or a group × condition interaction in these ratings.
fMRI task-related activations
In the regulate versus maintain contrast, a direct between-group comparison showed that when compared to HCs, individuals within the BPD and MDD groups presented overlapping decreased activations in the right vlPFC during cognitive reappraisal. In addition, patients with MDD, also in comparison to HCs, showed decreased activations in the left dlPFC and in the bilateral OFC (Figure 1A and Table 2). These results remained significant when controlling for sex in addition to age (Supplementary Table 1). Moreover, we observe no significant across-group differences during the Maintain > Observe contrast, although we observed a significant (at an uncorrected level) across-group activation in the right amygdala in this contrast, indicating successful emotion induction (Supplementary Table 2 and Supplementary Figure 1). Interestingly, in the Maintain > Observe contrast, we also observed that both patient groups tended to activate the vlPFC and the bilateral OFC more than HCs during the maintain condition, although at an uncorrected significance level (Supplementary Table 3).
Between-group differences in task-related activations. (A) Patients with borderline personality disorder (BPD; cyan) and patients with major depressive disorder (MDD; blue) showed overlapping decreased activations in comparison to healthy controls (HCs) during emotion regulation in the right ventrolateral prefrontal cortex (vlPFC; red circle). Patients with MDD showed additional hypoactivation in the left dorsolateral prefrontal cortex and the orbitofrontal cortex (bilaterally). Top color bar: MDD versus HC TFCE (Threshold Free Cluster Enhancement) values; bottom color bar: BPD versus HC TFCE (Threshold Free Cluster Enhancement) values. (B) Correlation between Cognitive Emotion Regulation Questionnaire rumination scores and right vlPFC activation in patients with MDD.
Regions showing significant activation differences during Regulate > Maintain.
Abbreviations: BPD, borderline personality disorder; dlPFC, dorsolateral prefrontal cortex; FWE, family-wise error; HCs, healthy controls; kE , cluster extent; MDD, major depressive disorder; MNI, Montreal Neurological Institute; OFC, orbitofrontal cortex; vlPFC, ventrolateral prefrontal cortex.
gPPI analyses
Generalized psychophysiological interactions (gPPIs) revealed between-group differences when assessing right vlPFC connectivity. Specifically, in comparison to HCs, individuals from both the BPD and MDD groups showed a similar pattern of reduced connectivity with right posterior temporal areas (although peak coordinates of the different group comparisons were located in different gyri, all these clusters overlapped in the posterior temporal cortex). Nevertheless, the BPD group showed an additional cluster of decreased connectivity with the right vlPFC involving the left inferior temporal cortex. In addition, when directly comparing both clinical groups, patients with BPD showed decreased connectivity values in comparison to patients with MDD within these same clusters (Figure 2A and Table 3).
Between-group differences in generalized psychophysiological interaction (gPPI) analyses from the right vlPFC seed. (A) In comparison to healthy controls (HCs), patients with major depressive disorder (MDD; blue) showed a decreased connectivity with right posterior temporal areas, involving the medial temporal gyrus and the parahippocampal gyrus. Patients with borderline personality disorder (BPD; cyan) showed also decreased connectivity with posterior temporal areas (in this case, with peak differences in the fusiform gyrus) and, specific to these subjects, with the left inferior temporal gyrus (ITG). Moreover, patients with BPD showed, in comparison to the MDD group (yellow), decreased connectivity between the right vlPFC and the left ITG and the right fusiform gyrus. Top color bar: MDD versus HC TFCE (Threshold Free Cluster Enhancement) values; middle color bar; BPD versus HC TFCE (Threshold Free Cluster Enhancement) values; bottom color bar; BPD versus MDD TFCE (Threshold Free Cluster Enhancement) values. (B) The Cognitive Emotion Regulation Questionnaire reappraisal scores correlated positively with right vlPFC–left ITG connectivity (red circle) in the BPD group.
Regions showing significant connectivity differences during Regulate > Maintain.
Abbreviations: BPD, borderline personality disorder; FG, fusiform gyrus; FWE, family-wise error; gPPI, generalized psychophysiological interaction; HCs, healthy controls; ITG, inferior temporal gyrus; kE , cluster extent; MDD, major depressive disorder; MNI, Montreal Neurological Institute; MTG, medial temporal gyrus; PG, parahippocampal gyrus; vlPFC, ventrolateral prefrontal cortex.
Correlations between clinical and imaging data
We observed a significant negative correlation between right vlPFC activation and CERQ rumination scores in the MDD group (Pearson’s r = −0.505; p = 0.023; Figure 1B), which significantly differed (z = −1.881; p = 0.03) from the same correlation in the BPD group (Pearson’s r = 0.099; p = 0.686). Nevertheless, the difference with the correlation in HCs (Pearson’s r = −0.22; p = 0.365) did not reach statistical significance (z = −0.954; p = 0.17). Furthermore, we observed that patients with BPD showed a significant positive correlation between CERQ reappraisal scores and right vlPFC–left inferior temporal gyrus (ITG) connectivity (Pearson’s r = 0.644; p = 0.003; Figure 2B). This correlation differed from what was observed in MDD (Pearson’s r = 0.004; p = 0.986) and HC (Pearson’s r = −0.112; p = 0.648) groups (z = −2.208; p = 0.014; and z = −1.846; p = 0.032, respectively). We found no further correlations between clinical and imaging data, including depression (HDRS) and anxiety (HARS) scores.
Discussion
Our results showed that both individuals with MDD and individuals with BPD display decreased activation in the vlPFC during cognitive reappraisal. Nevertheless, such hypoactivation was more extensive in the MDD group, who also showed a negative correlation between reappraisal-related vlPFC activity and rumination. Likewise, patients with MDD displayed other clusters of significant hypoactivation during cognitive reappraisal, including the left dlPFC and the bilateral OFC. Conversely, patients with BPD showed greater connectivity decreases between the vlPFC and left inferior and right posterior temporal regions during reappraisal. Furthermore, these connectivity alterations were significantly associated with psychometric measures of cognitive reappraisal. Overall, this pattern of results confirms our a priori hypotheses, since patients with MDD seem to recruit prefrontal regions to a lesser extent during regulation of emotions, while the BPD group displayed inefficient regulatory input from prefrontal areas.
Our findings reporting decreased vlPFC activation during emotional processing in both patient groups are in agreement with previous research [Reference Chechko, Kellermann, Augustin, Zvyagintsev, Schneider and Habel29]. The vlPFC plays a crucial role in response selection and inhibition [Reference Aron, Robbins and Poldrack30], and, in particular, in the inhibition of emotional appraisals [Reference Wager, Davidson, Hughes, Lindquist and Ochsner31]. Our present results are, therefore, indicative of cognitive reappraisal impairments in MDD and BPD that may be partly a consequence of ineffective management of inhibitory resources. Notably, the vlPFC has been related to the use of reinterpretation strategies during reappraisal, as opposed to the use of distancing strategies engaging parietal regions [Reference Picó-Pérez, Radua, Steward, Menchón and Soriano-Mas7,Reference Ochsner, Silvers and Buhle32]. Consequently, MDD and BPD seem to share a diminished capacity to reinterpret negative emotions. Moreover, among MDDs, decreased vlPFC activation was inversely related to rumination scores, a core feature of the depression phenotype. This concurs with reports in HC samples [Reference Hooker, Gyurak, Verosky, Miyakawa and Ayduk33] and other findings indicating that the lateral PFC plays a general inhibitory role limiting the impact, or carryover effects, of an emotional state onto emotional states evoked by subsequent events [Reference Waugh, Shing, Avery, Jung, Whitlow and Maldjian34]. On the other hand, the lateralization of this finding to the right hemisphere is in agreement with previous reports in MDD [Reference Bruder, Alvarenga, Abraham, Skipper, Warner and Voyer35] and BPD [Reference Visintin, De Panfilis, Amore, Balestrieri, Wolf and Sambataro36] samples, although not with recent meta-analytic evidence in anxiety and depression groups [Reference Picó-Pérez, Radua, Steward, Menchón and Soriano-Mas7]. The recruitment of the right vlPFC has been associated with particular features of cognitive reappraisal, such as reiteratively implementing the same reappraisal strategies to counteract negative affect [Reference Perchtold, Weiss, Rominger, Fink, Weber and Papousek37], or the implementation of operations needed to maintain strategies in working memory and to monitor its success during the late phases of emotional situations [Reference Kalisch38]. Therefore, the right hemisphere lateralization of findings observed here may stem as a consequence of the specific instructions given to participants in this study, whom may have probably implemented a limited choice of strategies, in interaction with a decreased capacity in the maintenance and monitoring of ongoing reappraisal efforts in patients.
Decreased activation of the dlPFC has been also previously reported in clinical samples, including not only depression and anxiety patients [Reference Campbell-Sills, Ellard, Barlow and Gross11,Reference Johnstone, Walter and Gross39]. This region contributes to different executive functions [Reference Wager and Smith40], and, in the context of emotion regulation, its role seems to be related to the active manipulation of information to reappraise emotional stimuli [Reference Ochsner, Silvers and Buhle32]. This alteration concurs with the executive function alterations commonly described in depression samples [Reference Rock, Roiser, Riedel and Blackwell41]. In this case, however, findings were lateralized to the left hemisphere. This may be partially accounted for by the role of right dlPFC activation in negative emotion appraisal, which is related to depression severity and may, therefore, compensate for the executive function-related hypoactivations allegedly occurring during reappraisal [Reference Phan, Fitzgerald, Nathan, Moore, Uhde and Tancer6,Reference Grimm, Beck, Schuepbach, Hell, Boesiger and Bermpohl42].
Regarding the hypoactivations also observed in the medial OFC in patients with MDD, it should be noted that this region, like other medial prefrontal structures, has been shown to downregulate activity in subcortical structures [Reference Steward, Davey, Jamieson, Stephanou, Soriano-Mas, Felmingham and Harrison9], and, indeed, its functional connectivity with the amygdala is increased during threat-induced anxiety in HCs [Reference Gold, Morey and McCarthy43]. According to our findings, the medial OFC of patients with MDD is probably not properly exerting this downregulatory input into emotion-processing structures, including, but not limited to, the amygdala [Reference Picó-Pérez, Radua, Steward, Menchón and Soriano-Mas7,Reference Steward, Davey, Jamieson, Stephanou, Soriano-Mas, Felmingham and Harrison9]. In addition, it is important to note that both the OFC and the vlPFC showed a trend to increase activation in patient groups during the Maintain condition (as compared to the control Observe condition). This result suggests that patients engage in a regulation effort during the Maintain condition, which may also contribute to the decreased activation levels in these same regions (especially in the MDD group) observed when contrasting Regulate and Maintain conditions.
Patients with BPD did, however, not show such extended prefrontal hypoactivation, but rather decreased functional connectivity between the vlPFC and visual association cortices of the ventral stream, implicated in complex visual feature detection and recognition of facial expression [Reference Im, Albohn, Steiner, Cushing, Adams and Kveraga44]. Different studies have consistently described hyperresponsiveness of the visual system in BPD patients when processing emotional information, especially, emotional faces, extending from primary cortices to association cortices of the temporal lobe [Reference Chechko, Kellermann, Augustin, Zvyagintsev, Schneider and Habel29,Reference Guitart-Masip, Pascual, Carmona, Hoekzema, Bergé and Pérez45–Reference Scherpiet, Brühl, Opialla, Roth, Jäncke and Herwig48]. Although we here have not observed such increased activation in the visual system, the regulatory input from the vlPFC cortex was diminished in patients with BPD, which seems to be a plausible mechanism to account for the visual hyperresponse described with other emotional tasks in the above studies. Likewise, although patients with MDD also showed some degree of decreased connectivity from the vlPFC to early visual perception areas, their clusters were less extended, and, at least for some of these clusters (i.e., right posterior fusiform gyrus), we also observed a significant difference between the clinical groups, with MDD patients showing significant connectivity increases in comparison to the BPD group.
According to these results, it can be concluded that alterations in cognitive reappraisal in patients with BPD (and to a lesser extent, in patients with MDD) could start at perceptive stages, before information reaches emotion-processing structures. Nevertheless, since information is conveyed from these visual association to limbic cortices [Reference Cardoner, Harrison, Pujol, Soriano-Mas, Hernández-Ribas and López-Solá49,Reference Dolan50], it is expected that such (lack of) modulation of perceptive input will indirectly weaken the regulatory input from prefrontal structures to emotion-processing regions. In this sense, it is worth mentioning that only patients with BPD, but not patients with MDD, showed decreased connectivity between the vlPFC and more rostral parts of the ITG. It can, therefore, be suggested that as information progresses through the ventral stream, alterations in prefrontal modulation of visuo-emotional processing are exclusively observed in patients with BPD. Interestingly, alterations in the white matter tracts linking anterior brain areas with visual association cortices (i.e., the inferior fronto-occipital and the inferior longitudinal fasciculi) have been described in patients with BPD [Reference Ninomiya, Oshita, Kawano, Goto, Matsuhashi and Masuda51].
Overall, the above notions concur with recent reports suggesting that, in comparison to patients with MDD, patients with BPD show an exaggerated response during emotional induction paradigms [Reference Schulze, Schulze, Renneberg, Schmahl and Niedtfeld52]. It is also noteworthy that correlations between interregional connectivity alterations and emotion regulation scores were only observed in patients with BPD. Specifically, we observed a positive association between reappraisal scores and vlPFC–rostral ITG connectivity, indicating that prefrontal input at this particular stage of visuo-emotional processing within the ventral stream may critically determine emotion regulation success in this clinical group. In sum, these prefrontovisual association connectivity alterations observed in BPD are likely to account for the increased sensitivity to emotional aspects of the environment [Reference Herpertz, Gretzer, Steinmeyer, Muehlbauer, Schuerkens and Sass53] and the general higher sensitivity to emotional stimuli and slow return of emotional arousal to baseline that characterize patients with BPD [Reference Johnson, Shea, Yen, Battle, Zlotnick and Sanislow54].
Importantly, the correlations between imaging and psychometric data observed in the clinical groups should be interpreted with caution, because CERQ scores did not differ across the study groups (although they significantly differed from those of a reference population; see Table 1). Nevertheless, our findings suggest that the extent of alterations in vlPFC activation (patients with MDD) and vlPFC–rostral ITG connectivity (patients with BPD) seems to be preferentially accounting for, among all the emotion regulation facets, interpatient variability in rumination and reappraisal, respectively. Anyhow, these do not exclude that these same neurobiological changes may be also related, to a lesser extent, to other emotion regulation facets, and this is probably the reason why these imaging alterations are significant at the group level.
The results of this study have to be interpreted in the context of the following limitations. First, our overall sample was small (n = 58, with 19/20 subjects per group), which may have limited the power of our analyses to detect additional significant findings. Nonetheless, we would like to stress that our subjects were carefully recruited according to strict inclusion criteria, and we have obtained several significant differences between the study groups at a strict family-wise error threshold. Likewise, the sample size of our study is similar to sample sizes in most neuroimaging emotion regulation studies in MDD or BPD samples (see, e.g., the studies included in the reviews and meta-analyses by Sicorello and Schmahl [Reference Sicorello and Schmahl13], Rive et al. [Reference Rive, van Rooijen, Veltman, Phillips, Schene and Ruhé17], or Picó-Pérez et al. [Reference Picó-Pérez, Radua, Steward, Menchón and Soriano-Mas7]). We acknowledge, however, that there are some notable exceptions to this trend, such as the studies by van Zutphen et al. (n = 55 BPD patients) [Reference van Zutphen, Siep, Jacob, Domes, Sprenger and Willenborg55] and Silvers et al. (n = 60 BPD patients) [Reference Silvers, Hubbard, Biggs, Shu, Fertuck and Chaudhury56]. Second, patients with MDD were older than the other two groups, although this is reflective of the clinical setting in which this study took place and adds external validity to our study. Moreover, we controlled for age in all our analyses. Anyhow, future studies should try to provide a more accurate matching across study groups. This is especially important when comparing clinical groups overlapping in clinically relevant variables such as depression or anxiety. Although our clinical groups did not differ in these variables, an accurate matching of nuisance factors is expected to provide clearer results and, consequently, allow drawing straightforward conclusions. Third, we observed no significant across-group differences in intrascanner ratings, although this is commonly observed in emotion regulation studies and should not be interpreted as evidence of similar cognitive reappraisal implementation [Reference Picó-Pérez, Radua, Steward, Menchón and Soriano-Mas7]. Indeed, our groups differed from HCs in psychometric measurements of emotion regulation (i.e., DERS scores). Moreover, intrascanner ratings showed that negative emotion reactivity was successfully induced in all groups. There are, nevertheless, different reasons for this lack of between-group differences in intrascanner ratings, such as the inherent limitations of subjective behavioral assessments, social desirability effects, or impaired self-awareness of emotional experience, as suggested by Zilverstand, Parvaz, and Goldstein [Reference Zilverstand, Parvaz and Goldstein57]. Finally, we did not collect any measure of BPD severity [Reference Pfohl, Blum, St John, McCormick, Allen and Black58], and we assessed dispositional use of emotion regulation strategies with a retrospective self-report measure. Although previous research has shown that such measurements may significantly predict real-life outcomes, such as well-being and depressive symptomatology [Reference Gross and John4], future research can benefit from real-time and real-life approaches, such as ecological momentary assessments.
Taken together, our findings indicate that MDD and BPD share an altered neural response during cognitive reappraisal involving the right vlPFC, indicating that this region is implicated in the emotion regulation shortcomings that characterize both disorders. Nevertheless, MDD patients showed a more widespread pattern of reduced prefrontal activation, which may be interpreted in the context of a pervasive alteration in executive functioning probably stemming from a primary deficit in the strength of prefrontal activations. On the other hand, BPD patients showed a more extended pattern of dysfunctional connectivity between prefrontal areas and visual association cortices that may lead to the higher sensitivity to emotional stimuli typically observed in these patients. These findings substantiate in neurobiological terms the existence of dissimilar profiles of emotion regulation alteration between these disorders, and may ultimately be of relevance for the development or optimization of clinical interventions aimed at restoring emotion regulation capacities.
Acknowledgments
We would like to thank CERCA Program/Generalitat de Catalunya for institutional support. We would also like to thank all the participants who contributed to this study.
Supplementary Materials
To view supplementary material for this article, please visit http://dx.doi.org/10.1192/j.eurpsy.2021.2231.
Data Availability Statement
The data that support the findings of this study and the brain maps of all analyses are available from the corresponding author upon reasonable request.
Author Contributions
Conceptualization: X.G., N.C., S.M.G., and C.S.-M.; Data curation: X.G., S.M.G., and C.S.-M.; Formal analysis: V.D.P.-A., M.B.-S., and T.S.; Funding acquisition: J.M.M., S.M.G., and C.S.-M.; Investigation: V.D.P.-A., M.B.-S., X.G., A.W., and C.A.; Methodology: V.D.P.-A., M.B.-S., T.S., I.M.-Z., and A.W.; Project administration: S.M.G. and C.S.-M.; Resources: A.W., C.A., M.N.C., and M.V.; Software: I.M.-Z.; Supervision: N.C., M.N.C., M.V., S.M.G., and C.S.-M.; Validation: A.W., C.A., S.M.G., and C.S.-M.; Visualization: V.D.P.-A. and M.B.-S.; Writing & original draft: V.D.P.-A., S.M.G., and C.S.-M.; Writing, review, & editing: V.D.P.-A., M.B.-S., T.S., I.M.-Z., X.G., A.W., C.A., N.C., M.N.C., M.V., and J.M.M.
Financial Support
This project has received funding from the European Union Horizon 2020 research and innovation program under the Marie Sklowdowska Curie grant agreement No. 714673 and Fundación Bancaria “la Caixa.” This study was supported in part by grants PI16/00889 and PI19/01171 (Carlos III Health Institute, Spain), co-funded by European Regional Development Fund-ERDF, a way to build Europe, the Departament de Salut, Generalitat de Catalunya (PERIS SLT006/17/249), the Agència de Gestió d’Ajuts Universitaris i de Recerca (2017 SGR 1247), and the Agencia Nacional de Promoción de la Investigación, el Desarrollo Tecnológico y la Innovación PICT-2019-02328 (Argentina). V.D.P.-A. was supported by “la Caixa” Foundation (ID 100010434, fellowship code LCF/BQ/IN17/11620071). T.S. was supported by the University of Melbourne McKenzie Fellowship and an NHMRC/Medical Research Future Fund (MRFF) Investigator Grant (MRF1193736). I.M.-Z. was supported by a PFIS grant (FI17/00294) from the Carlos III health Institute. X.G. was supported by the Health Department of the Generalitat de Catalunya (SLT002/16/00237).
Conflict of Interest
The authors declare none.
Ethical Standards
The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national and institutional committees on human experimentation and with the Helsinki Declaration of 1975, as revised in 2008.
Open access
Comments
No Comments have been published for this article.