Adverse childhood experiences are a key predictor of the development and relapse of adult depression. While current CBT-based treatments commonly target more proximal causes of depression such as maladaptive cognitions and biases, memory-focused interventions like EMDR and Imagery Rescripting directly target emotional memories of adverse childhood experiences, offering promising new avenues to treat depression. Few studies, however, have systematically investigated what aspects of emotional memory relate to psychopathology, and could thus be important targets for these interventions. We explored the relationship between phenomenological aspects of negative autobiographical memories and current depressive symptoms in a cross-sectional study.
Methods
In an online questionnaire, 119 first-year university students reported a formative negative emotional memory from childhood or adolescence and rated various aspects of this memory, such as the vividness, intrusiveness, and coherence. We also introduced the aspect of “emotional impact” that indexed effects of the memory on current mood, emotions, and identity.
Results
After correcting for multiple comparisons, only childhood maltreatment (CTQ-SF) and emotional impact predicted current depressive symptoms (PHQ-9). The effect of emotional impact remained significant when controlling for childhood maltreatment and the negativity of the memory.
Conclusion
These findings suggest that emotional impact represents a clinically relevant aspect of emotional memory, and highlight a potential pathway through which negative memories affect depressive symptoms.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Introduction
Depression is a major public health problem, with an estimated prevalence of 5% amongst adults (World Health Organization, 2023). Rates have substantially risen in the past decade, and depression is now the number one cause of disability worldwide (Friedrich, 2017). Current treatments often target depressive symptoms with pharmacological medication, through psychotherapeutic interventions such as cognitive behavioural therapy (CBT) and interpersonal psychotherapy (IPT), or a combination of both (Cuijpers et al., 2014; DeRubeis et al., 2008). The psychological interventions usually offer techniques to help a patient tackle factors that maintain the depression. While both psychological and pharmacological interventions can be effective in mitigating symptoms, 30–60% of patients experience relapse after initially responding to treatment (Bockting et al., 2015; DeRubeis et al., 2008; Vittengl et al., 2007). In fact, results from a longitudinal population-based study showed that only 32% of patients diagnosed with depression at baseline had fully recovered after nine years, highlighting the chronicity of depression (ten Have et al., 2022). We argue that the limited long-term effectiveness of CBT and IPT results from a focus on more proximal causes of depression including negative automatic thoughts, behaviours, cognitive biases, and interpersonal problems, respectively. Similarly, psychopharmacological treatments focus on manipulating neurotransmitters as proximal causes of depression, resulting only in a temporary reduction of depressive symptoms. In contrast, the field has recently seen a rising interest in the more distal causes of depression, including negative emotional memories resulting from adverse childhood experiences (Arntz, 2020; Gathier et al., 2023). We hypothesize that directly targeting negative memories of childhood or adolescence in therapy may address the underlying roots of depression, thereby diminishing the likelihood of relapse and treatment-resistant cases. However, to refine interventions targeting these memories, a deeper understanding of their connection to depression is essential. The present article therefore endeavours to delve deeper into this relationship.
One of the main consistent predictors of depression and relapse is the occurrence of adverse childhood experiences, such as parental abuse and neglect (Harkness et al., 2012; Kuzminskaite et al., 2022; Watters et al., 2023; Wojnarowski et al., 2019), but also experiences of loss or bullying by peers (Arseneault, 2017; Pham et al., 2018). Although it should be noted that depression is highly heterogeneous, and some individuals with depression may not have experienced adversities, these findings suggest that targeting memories of negative experiences could benefit those who have. According to cognitive theory, experiences from childhood or adolescence, such as bullying or rejection, can over time result in maladaptive core beliefs and schemas, which increase the vulnerability for depression (Bishop et al., 2022). Standard CBT, however, typically focuses on the more proximal or maintaining processes that stem from these experiences—such as automatic thoughts, dysfunctional beliefs, and biases—rather than directly addressing autobiographical memories themselves (Beck, 2021). In contrast, interventions such as eye-movement reprocessing and desensitization (EMDR) and imagery rescripting (ImRs) take a more direct approach to modifying the emotional memory. Recent evidence suggests that these interventions can indeed be effective to treat depression (Brewin et al., 2009; Dominguez et al., 2021; Kanczok et al., 2024; Ma & Lo, 2022; Moritz et al., 2018; Pile et al., 2021; Stavropoulos et al., 2023). Both EMDR and ImRs involve the reliving of a specific negative experience and changing the memory of this experience by making rhythmic eye movements or listening to sounds to reduce vividness (EMDR), or by having the therapist work together with the patient to target its content or meaning (ImRs). These interventions assume that the emotional processing of negative memories is a key working mechanism of the observed reductions in depression. However, there currently exist no independent clinically relevant measures of emotional memory, complicating attempts to assess whether symptom change is indeed achieved through emotional memory change. Moreover, it remains unclear what precise aspects of an emotional memory are related to depressive symptoms, and thus need to be specifically addressed. In order to optimise memory-based interventions for depression, further insights into emotional memory qualities that relate to depressive symptoms, and clinically relevant measures of these qualities, are required.
Previous studies on the role of autobiographical memories in the aetiology of depression focused predominantly on the role of coherence and specificity of personal narratives. In these studies, personal narratives are recalled either verbally or in writing, and coded for various aspects of coherence, including contextual (where and when), chronological (clear order of events) and thematic (creating a meaningful narrative around a central theme) coherence (Reese et al., 2011). While some studies show that low coherence of a personal narrative was related to depression (Vanderveren et al., 2019), others found this effect only for thematic coherence (Vanaken et al., 2021). In contrast, Vanderveren et al. (2020) found no relationship between memory coherence and depressive symptoms at all. Two other phenomena that are often observed in patients with depression are the overgeneral memory bias, where patients provide fewer details when recalling a memory, and the negative memory bias, where patients tend to recall fewer positive memories compared to healthy controls (Duken et al., 2024; Raes et al., 2006; Williams et al., 2007). These effects are frequently studied in the context of depression and have resulted in novel potential interventions such as memory specificity training (MEST; Neshat-Doost et al., 2013). Yet experimenter rated memory coherence and detail are rather objective indices of autobiographical memories, while current memory-based interventions such as EMDR and ImRs address the more subjective experience of the memories (e.g., the emotionality or the vividness). We therefore take a broader perspective and aim to investigate the effect of the phenomenology (i.e., the subjective experience) of negative memories on depression.
As a validated measure of subjective memory phenomenology, Sutin and Robins (2007) developed the memory experiences questionnaire, which asks participants to describe a key self-defining autobiographical memory, to then rate various aspects of this memory, including the subjective vividness, coherence, and visual perspective. A cross-sectional study in a healthy sample confirmed that most of these phenomenological aspects were related to depression, with lower self-reported vividness, coherence and accessibility of memories predicting more depressive symptoms (Luchetti & Sutin, 2016). It is unclear, however, whether participants in this sample reported negative or positive memories, limiting the interpretation of the results. For example, while the vividness of positive memories is indeed found to be reduced in formerly-depressed patients, vividness of negative memories tends to be stronger, although this effect is only observed after a negative mood induction (Werner-Seidler & Moulds, 2011, 2012). Interestingly, treatments such as EMDR specifically aim to decrease the vividness of negative memories, and therefore further elucidating the effect of experienced memory vividness on depression is needed (Houben et al., 2020; Leer et al., 2014). In addition to the existing subscales of the Memory Experiences Questionnaire we included two additional indices in the current study: “Emotional impact” and “Intrusiveness”. Emotional impact refers to the effect of the memory on current mood, emotions, and identity. We deem this quality of memories to be strongly related to mental health and an important mechanism of change. Memory intrusions were traditionally strongly linked to post-traumatic stress disorder (PTSD), but have recently been found to play an important role in depression (Patel et al., 2007; Payne et al., 2019). The extent to which negative memories are intrusive could therefore be an important determinant in the development of depression. In sum, the current literature concerning the subjective experience of negative autobiographical memories in depression is limited and often focused on a single memory quality. This narrow focus may result in the neglect of other important aspects of emotional memory that may, at least in part, account for the relationship between childhood maltreatment and depression.
In the current study we aimed to explore the relationship between childhood maltreatment, subjective phenomenological aspects of a negative emotional memory, and depressive symptoms. We used an adapted version of the Memory Experiences Questionnaire, focussing on negative memories from an event that occurred during childhood or adolescence. We exploratively tested the effects of all subscales of the Memory Experiences Questionnaire on depressive symptoms, and further explored those relationships that were significant after correcting for multiple comparisons. In doing so, we aim to take a first step in the development of a clinically relevant measure of emotional memory, that represents memory qualities that should be targeted during memory-based interventions for depression.
Methods
Participants
Participants were 123 students from the University of Amsterdam, recruited through an internal website to participate in an online study on emotional memory and mental health. Next to being a student at the psychology or communication faculty, the only two inclusion criteria were the ability to understand English and age ≥ 18. Recruitment took place between 11-12-2023 and 07-01-2024. Anticipating a medium effect size for simple correlations, we aimed to recruit at least 80 participants, but we left recruitment open as long as possible and analysed all completed datasets. Participants received course credits for their participation, and were offered to also join a raffle in which they could win a 50-euro voucher. Qualtrics was used to collect the questionnaire data. Ethical approval was granted by the ethical review board of the clinical psychology department at the University of Amsterdam (project nr. FMG-5667). Data collection was fully anonymous and data from unfinished questionnaires were deleted after one week. All participants were informed about the procedure of the study and signed informed consent.
Measures
Depressive Symptoms
Depressive symptoms were measured with the Patient Health Questionnaire-9 (PHQ-9; Kroenke et al., 2001) consisting of 9 items asking participants how frequently in the past two weeks they have experienced certain problems related to depression (e.g., “Feeling tired or having little energy”), including 1 item on suicidal ideation. The items are each scored on 4-point Likert scale ranging from 0 “not at all” to 4 “nearly every day”. Total scores range between 0 and 36 and a cut-off of 10 has been established to indicate the potential presence of clinical depression. Previous studies (Kroenke et al., 2001) reported good psychometric properties and internal consistency of the PHQ-9 (Cronbach α = 0.89). The internal consistency in our sample was comparable (Cronbach α = 0.87).
Childhood Maltreatment
Childhood maltreatment was measured with the Childhood Trauma Questionnaire Short Form (CTQ-SF; Bernstein et al., 2003), a 28-item questionnaire indexing childhood maltreatment in five domains: emotional abuse, physical abuse, sexual abuse, emotional neglect, and physical neglect. Each subscale consists of 5 items that are scored from 1 to 5 and are summed to calculate a score per subscale ranging from 5 to 25. Subscale scores were added to calculate an overall childhood maltreatment score. The other three items measure minimization/denial and are used to detect underreporting of maltreatment. These items were not used in the current analyses. Internal consistency of the CTQ-SF subscales ranges from α = 0.61 to α = 0.95 depending on subscale and sample (lowest reliability for the physical neglect subscale; Bernstein et al., 2003). In our sample, internal consistency was low for physical neglect (α = 0.5) but good for other subscales (α = 0.75–0.96). The internal consistency of the total scale was 0.91.
Emotional Memory
To measure memory phenomenology, we asked participants to describe a negative emotional memory from childhood or adolescence that is still important to them, and that they still think about often. We did not restrict these memories to certain experiences (e.g., parental abuse) as a wide variety of negative experiences could be related to the development of depression. Various aspects of emotional memory were measured with the short form of the memory experiences questionnaire (MEQ; Luchetti & Sutin, 2016). The MEQ indexes 9 phenomenological aspects of a self-defining autobiographical memory: vividness, coherence, accessibility, time perspective, sensory details, visual perspective, emotional intensity, sharing, and distancing. Items in the subscales are rated on a scale from 1 (strongly disagree) to 5 (strongly agree). We have added two subscales that we felt were important in depression, namely an intrusiveness and an emotional impact scale, thus creating a total of 11 subscales. The intrusiveness scale consists of 3 items measuring how intrusive the memory is (e.g., “This memory often pops up in my mind involuntary”). The emotional impact scale includes 5 items measuring effects of the memory on current mood, emotions, and identity (e.g., “I can feel my body get tense when I think about this memory”). See Supplementary Materials for full description of the added items. In previous research, the internal consistency of the subscales ranges from α = 0.41 to α = 0.88 (median 0.79), with lower alphas for the subscales sensory details and coherence (Luchetti & Sutin, 2016). In our sample, Cronbach’s alpha for the subscales ranged from 0.66 to 0.89. Importantly, the internal consistencies of the added subscales Emotional Impact (α = 0.77) and Intrusiveness (α = 0.88) were acceptable to good, respectively. Because the number of items in the subscales ranged from 3 to 5, we averaged the scores per subscale.
Other Questionnaires
As part of a larger project, we also collected data on generalised anxiety (GAD-7), early maladaptive schemas (Young Schema Questionnaire 3; YSQ-3, only defectiveness/shame and social isolation schemas), attachment (Experiences in Close Relationships Scale; ERC-12), neuroticism (Big Five Inventory Short Form; BFI-SF), and self-compassion (Self-Compassion Scale Short Form; SCS-SF). These measures were not used in the current analyses.
Procedure
After showing interest in participating in the study, participants were redirected to an online Qualtrics questionnaire. They were first asked to read the online information about the study and to provide digital informed consent. Upon consenting to their participation, participants were asked to give some demographic information, and filled out a few questions about their history of mental health and treatment, as well as current treatment and psychotropic medication use. Then the GAD-7 and the PHQ-9 were filled out. Next, participants were asked to describe a negative emotional memory from childhood or adolescence that was still important to them. After doing so, participants were asked to rate how negative and how easy to recall this memory was, and then they filled out the subscales of the MEQ including the added scales of emotional impact and intrusiveness. Then participants completed the final questionnaires (CTQ-SF, ERC-12, BFI-SF, SCS-SF, YSQ-3) and were given the opportunity to leave their email address in a separate questionnaire if they wanted to participate in the raffle. Finally, participants were debriefed about the study and shown information on where to get mental health support if they felt they needed this. The entire questionnaire took on average about 20 min to complete.
Statistical Analyses
All analyses were performed in RStudio version 4.4.1 (R Core Team, 2021). We qualitatively inspected the written memories and excluded those participants (n = 4) who did not write down a memory. No other participants were excluded from the analyses. Given that the study was largely explorative, we first computed a correlation matrix between the phenomenological subscales of the MEQ and the PHQ-9 scores. All correlations were corrected for multiple comparisons by dividing the alpha cut-off value with the number of comparisons made. As scores for the PHQ-9 and CTQ-SF are often highly skewed in the general population, we used bootstrapping methods with 5000 samples and report the bias-corrected bootstrapped confidence intervals (BCa CIs) in addition to p-values. We also performed robustness tests for the correlations that survived the correction. We identified outliers using the robust Mahalanobis distance with a cutoff of 5.99 (probability = 0.95 and df = 2) and reran the correlations without those outliers. To investigate the unique effect of memory characteristics on depressive symptoms we also performed partial correlations where we controlled for childhood trauma scores. An exploratory network analysis was performed including childhood maltreatment and all subscales of the MEQ that had a significant relation with PHQ-9 scores before correction. For the network analysis, the MGM package was used, allowing us to inspect which subscales have a direct relation with PHQ-9 scores, and which subscales have an indirect effect. The MGM package (Haslbeck & Waldorp, 2020) also allows to estimate the predictability, a metric that quantifies how much variance of each variable is explained by all other variables in the network. Finally, we used the relaimp package (Groemping, 2007) to test differences in unique variance each of the variables share with the PHQ-9 score.
Data Availability
All data (except the written memories, as full anonymisation of the memories cannot always be guaranteed) and analysis code are available at the Open Science Framework and can be accessed at https://osf.io/ncmx2/. The design and analysis of the current study were not pre-registered.
Results
Sample Characteristics
Upon visual inspection of the data, we identified 4 participants who did not (want to) fill in a negative emotional memory. Given that this was our main measure of interest these participants were excluded from further analyses, leaving a final sample of 119. Collected sample characteristics can be found in Table 1. Of all participants, 10% reported receiving psychological treatment at the moment of study, and 3% was currently taking antidepressant medication. Importantly, 30% (n = 36) scored above the cut-off for being currently depressed (PHQ > 10, see Supplementary Fig. 1), indicating that although participants were sampled from a student population, the level of depressive symptoms is relatively high compared to the general population. Total scores on the CTQ-SF were mostly determined by the subscales of emotional abuse, emotional neglect, and physical neglect, with very few participants reporting sexual or physical abuse (see Table 2).
Table 1
Overview of demographic, clinical, and negative memory characteristics
Mean or % (n)
Sd
Min
Max
Questionnaire range
Gender
Male
28% (33)
Female
71% (85)
Other
1% (1)
Age
20.0
2.8
18
32
Relationship
Single
60% (71)
Long-term
29% (35)
Other
11% (13)
Nationality
Dutch
47% (56)
German
16% (19)
Other
37% (44)
Clinical indices
Treatment history
50% (59)
Treatment current
10% (12)
Mediation current
3% (4)
PHQ > 10
30% (36)
Questionnaires
PHQ-9
7.9
5.5
0
24
0–27
GAD-7
6.9
4.9
0
19
0–21
CTQ-SF
37.8
12.5
25
95
25–125
Emotional memory
Negativity
72.7
23.5
3
100
0–100
Recall ease
74.1
28.6
0
100
0–100
Vividness
3.5
0.9
1
5
1–5
Coherence
3.5
1.0
1.25
5
1–5
Accessibility
3.9
0.9
1
5
1–5
Time perspective
2.6
1.1
1
5
1–5
Visual perspective
3.2
1.2
1
5
1–5
Sensory details
2.9
0.9
1
4.75
1–5
Emotional intensity
3.5
1.1
1
5
1–5
Sharing
2.4
1.1
1
4.7
1–5
Distancing
3.6
1.1
1
5
1–5
Emotional impact
3.1
0.9
1
5
1–5
Intrusiveness
2.6
1.1
1
5
1–5
Table 2
Frequencies of different types of reported childhood maltreatment in the current sample
None % (n)
Low % (n)
Moderate % (n)
Severe % (n)
Emotional abuse
52% (62)
29% (34)
12% (14)
7% (9)
Emotional neglect
54% (64)
29% (34)
7% (9)
10% (12)
Physical abuse
90% (107)
6% (7)
2% (3)
2% (2)
Physical neglect
70% (83)
23% (27)
4% (5)
3% (4)
Sexual abuse
87% (103)
4% (5)
3% (4)
6% (7)
Cut-off scores are based on Bernstein and Fink (1998)
Memory Characteristics
The memories that participants recalled were on average rated as moderately to highly negative and easy to recall (Table 1, see Supplementary Fig. 3 for visualised distribution of memory characteristics). In general participants provided detailed descriptions of the memory, including episodic characteristics about the time and place. About half of the participants (n = 58, 49%) described negative experiences with parents or teachers, whereas the other half (n = 52, 44%) described negative experiences with friends or classmates. The remaining participants (n = 9, 7%) recalled memories in which either the participant was alone (e.g., a panic attack), or that were not detailed enough to categorise.
Relationships Between Emotional Memory Characteristics and Depressive Symptoms
We first computed correlations between depression scores, childhood maltreatment scores, and all measured aspects of the reported negative memory (see Fig. 1). The uncorrected correlations showed that scores on the CTQ-SF (r(117) = 0.32, p < 0.001, 95% BCa CI [0.11, 0.53]) and the intrusiveness (r(117) = 0.29, p = 0.001, 95% BCa CI [0.09, 0.45]), emotional intensity (r(117) = 0.25, p = 0.005, 95% BCa CI [0.06, 0.41]), and emotional impact (r(117) = 0.45, p < 0.001, 95% BCa CI [0.26, 0.58]) of a negative memory were positively related to PHQ-9 scores. Distancing was negatively related to depressive symptoms (r(117) = − 0.22, p = 0.015, 95% BCa CI [− 0.38, − 0.04]). After Bonferroni correcting for multiple comparisons by dividing the alpha cut-off with 91 (14 variables that are all correlated with each other gives 91 comparisons when excluding duplicates and self-correlations), only childhood maltreatment (CTQ-SF) and the emotional impact of a negative memory remained significantly associated with PHQ-9 scores (see Fig. 2). Based on the Malahanobis distance we identified 15 (CTQ and PHQ) and 12 (Emotional impact and PHQ) bivariate outliers for these relationships, but removing these outliers from the analyses did not change the results (see Supplementary Materials).
Fig. 1
The A raw and B corrected correlation matrices for the associations between various aspects of a negative emotional memory and depressive symptoms as measured with the PHQ-9
Scatterplots of correlations between PHQ-9 scores and A CTQ-SF scores, and B Emotional Impact of a negative emotional memory. Dashed horizontal line indicates a PHQ-9 value > 10
To understand the relative effect of emotional impact of a negative memory on depressive symptoms when controlling for childhood maltreatment we performed partial correlations. These showed that the emotional impact is still significantly related to depression when controlling for CTQ-SF scores (r(116) = 0.40, p < 0.001, 95% BCa CI [0.22, 0.55]). Lastly, we controlled the effect of emotional impact for the negativity of the memory, to ensure that this effect is not driven by the fact that participants with higher depression scores simply recall more negative memories (r(116) = 0.44, p < 0.001, 95% BCa CI [0.24, 0.60]).
Lastly, to explore whether these memory characteristics are specific to depressive symptoms, we also evaluated the relationship between memory characteristics and GAD-7 (generalised anxiety disorder) scores. Uncorrected results showed that CTQ-SF scores (r(117) = 0.25, p = 0.005, 95% BCa CI [0.06, 0.44]), and the intrusiveness (r(117) = 0.25, p = 0.005, 95% BCa CI [0.08, 0.41]), emotional intensity (r(117) = 0.25, p = 0.006, 95% BCa CI [0.08, 0.39]) and emotional impact (r(117) = 0.40, p < 0.001, 95% BCa CI [0.21, 0.55]) of a negative memory were positively related to generalised anxiety symptoms. In line with results for depressive symptoms, only emotional impact survived when controlling for multiple corrections.
Individual Contributions of the Emotional Impact Items to Depressive Symptoms
Given that the emotional impact items have not been previously validated, we have additionally checked the individual relationships between these items and depressive symptoms. We found that, after correcting for multiple comparisons, all items except “Thinking about this memory does not impact my current feeling” were positively related to PHQ-9 scores (corrected ps < 0.013). However, when testing all items as predictors in a multiple regression, only the item “This memory has a strong impact on how I see myself” remained a significant predictor of depressive symptoms (b = 1.32, t = 3.07, p = 0.003).
Network Analyses
When conducting a network analysis of the variables that were significantly related to current depressive symptoms, we confirmed that both emotional impact and childhood maltreatment positively related to current depressive symptoms (Fig. 3), while distancing was negatively related to current depressive symptoms. Intrusiveness and emotional intensity appear to have an indirect effect on current depressive symptoms. All nodes together explained 27% of the variance of current depressive symptoms. Relative importance analysis revealed that childhood maltreatment now explained 6% of the variance, and emotional impact 13%, but this difference was not significant (Fig. 4).
Fig. 3
Network of depressive symptoms, childhood maltreatment and selected characteristics of emotional memory. Green connections represent positive associations, red connections negative associations. Thicker edges represent stronger associations. The blue colouring of the white circle around each node represents the amount of variance explained in that node by its neighbours
Relative importance (% of variance explained) of childhood maltreatment and selected characteristics of emotional memory on depressive symptoms. Error bars represent bootstrapped 95% CIs. Total variance explained by all variables is 27.09%
The current study aimed to explore the relationship between subjective phenomenological aspects of a formative negative childhood or adolescent memory and current depressive symptoms. Results showed that intrusiveness, emotional intensity, and emotional impact of a negative memory were positively related to current depressive symptoms, while distancing was negatively related to symptoms. When controlling for multiple comparisons only emotional impact remained significant, also after controlling for childhood trauma scores. Lastly, an explorative network analysis indicated a direct relationship between emotional impact, childhood maltreatment and depressive symptoms, and indirect effects of intrusiveness and emotional intensity.
In line with previous literature, we found a strong relationship between childhood maltreatment and current depressive symptoms, even in a non-clinical student sample (Harkness et al., 2012; Kuzminskaite et al., 2022; Wiersma et al., 2009; Wojnarowski et al., 2019). Childhood maltreatment in the current sample consisted predominantly of emotional neglect and abuse, indicating that it is not merely physical and sexual traumas that result in adult psychopathology. In addition, several aspects of a negative autobiographical memory, including the intrusivity, emotional intensity and emotional impact, were also related to depressive symptoms. These findings thus support the idea that targeting these characteristics of negative memories could benefit patients suffering from depression, a key assumption of memory-focused treatments (Dominguez et al., 2021; Gathier et al., 2023). Imagery Rescripting and EMDR reduce the emotional intensity, distress, and intrusivity of negative autobiographical memories in both clinical (Rameckers et al., 2024) and non-clinical samples (Çili et al., 2017; Houben et al., 2020; Strohm et al., 2019). Although intervention studies are needed to establish causal effects, the current results do suggest that targeting these specific aspects of emotional memories through EMDR or Imagery Rescripting could benefit patients suffering from depression. Notably, further analyses showed that memory characteristics such as intrusivity, emotional intensity, and emotional impact were not exclusively related to depression, but also to generalised anxiety symptoms. This suggests that negative emotional memories may function more as transdiagnostic factors associated with symptom severity (Kredlow et al., 2024, 2025).
The largest proportion of variance in depressive (and anxious) symptoms was explained by the emotional impact of negative memories. This subscale consisted of five newly developed items indexing how negative memories influence someone’s current mood, emotions, and identity. Importantly, these items have not been previously validated, emphasizing the need for follow-up research on this memory quality. It could be argued that emotional impact is similar to an index of autobiographical memories called the centrality of events, measuring the extent to which a traumatic event is integrated in someone’s life story (Berntsen & Rubin, 2006). Stronger event centrality correlates with symptoms of PTSD, and to some extent depression and anxiety (Fitzgerald et al., 2016; Gehrt et al., 2018), and it is related to subjective characteristics of the negative memory such as the vividness and emotional intensity (Gehrt et al., 2018; Newby & Moulds, 2011). An important difference is, however, that the items on emotional impact ask specifically about the memory of the event rather than the event itself, providing more detailed insights into the subjective experience of memory. Interestingly, when including all five separate items in a regression, only the item “This memory has a strong impact on how I see myself” remained predictive of depressive symptoms, suggesting that the effect of negative memories on the representation of oneself plays an important role in the development and maintenance of depression. This is in line with ideas from cognitive theory that negative experiences and memories thereof over time contribute to the development of generalised negative beliefs about the self or the world, which subsequently affect depression. The construct of emotional impact is in that sense closely aligned with core beliefs, but it is specific to the memory. Whether changing emotional impact affects more general negative core beliefs (or potentially vice versa) remains to be investigated. Notably, the item we included only measured the strength of the impact, but not the direction (i.e., whether the recalled memory had a positive or negative impact on how participants see themselves).
Interestingly, our results do not completely follow findings by Luchetti and Sutin (2016), who found positive correlations between almost all phenomenological aspects of autobiographical memories (e.g., vividness, accessibility, coherence, time perspective) and current depressive symptoms. It is not clear, however, whether participants in their study reported positive or negative memories, and how old these memories were. Moreover, depressive symptoms were measured with the anhedonic depression subscale of the mini mood and anxiety symptom questionnaire (mini MASQ), which focuses specifically on signs of anhedonia. Given its association with low positive affect, anhedonia may relate to particular aspects of autobiographical memory, like reduced coherence and vividness, especially when participants report positive memories (Werner-Seidler et al., 2013). Especially the lack of a relationship between vividness of a negative memory and depressive symptoms is somewhat surprising (Houben et al., 2020; Leer et al., 2014). Previous studies show that formerly-depressed patients report higher vividness of negative memories, but only when a negative mood is induced (Werner-Seidler & Moulds, 2011, 2012), which could explain the absence of a relationship in the current study. Importantly, even though vividness was not directly related to depressive symptoms, it did correlate with both the emotional impact and the emotional intensity of the reported memory. It could thus be that vividness of negative memories is indirectly related to depressive symptoms, but may not necessarily be a key target for treatment. In further contrast to previous literature, self-reported memory coherence was not related to depressive symptoms either. Studies on overgeneral memory bias and coherence suggest that patients with depression tend to retrieve less coherent and detailed memories (Raes et al., 2006; Vanaken et al., 2021; Vanderveren et al., 2019; Williams et al., 2007), although recent research suggests that negative memories are in fact retrieved with more detail (Duken et al., 2024). In the current study, we measured coherence using three self-report items from the Memory Experiences Questionnaire, and these items reflect subjective contextual and chronological coherence (clarity of time and place, and order of events). In contrast, previous studies used experimenter-rated measures of coherence, and in one study only a lack of thematic coherence (creating a meaningful narrative) was found related to depression, which could explain the discrepant results (Vanaken et al., 2021). In sum, our results suggest that in understanding the effect of emotional memory on depressive symptoms, qualities such as vividness and coherence may be less critical than the emotional impact of memories.
Although our findings point towards an important role for emotional impact of negative autobiographical memories in the development and maintenance of depression, there are some important limitations to the current design. First of all, the study merely provides cross-sectional insights into the investigated relationships. Although this is, at least in explorative stages, quite common in the field, we cannot interpret the causality of the observed relationships. Negative mood and cognitive biases are core characteristics of depression that can affect the subjective re-experiencing of autobiographical memories (Mitchell, 2016), meaning that qualities of emotional memories may have been affected by depressive symptoms instead of the other way around. To fully appreciate the clinical relevance of these findings it is necessary to understand the causal effects of emotional memory on depressive symptoms. Moreover, the current sample existed of first-year psychology students. Even though levels of depression were substantial, the results may not be generalizable to a clinical population. Given that the average age of the sample was 20, the memories the participants reported were in some cases more recent than one would observe in the general population (e.g., merely a few years ago). Lastly, for practical reasons we asked participants to report a single memory, yet in reality it is likely that multiple memories play a role in the development of depression. Furthermore, the reported memories encompassed negative experiences from childhood or adolescence that participants perceived as still significant in their present lives. However, it was not a prerequisite for these memories to reflect instances of maltreatment within family environments, which is the most frequently examined form of childhood adversity in depression research. This approach was adopted to acknowledge that other adverse experiences, such as bullying or loss, may also contribute to the onset of depression. Notably, the memories recalled by participants were often highly negative and described in considerable detail, in some cases resembling those targeted in trauma-focused interventions. Therefore, while preliminary, these findings offer valuable insights into the potential impact of such negative memories on depressive symptoms.
In conclusion, to gain further insights into how memories of adverse childhood experiences affect current depressive symptoms, and to understand what qualities of these memories should be targeted to reduce symptoms, we conducted a cross-sectional questionnaire study in a student sample. The emotional impact of a negative memory was most strongly related to depressive symptoms, even when correcting for childhood maltreatment and the negativity of the memory. On the one hand, these results provide support for the view that negative memories of adverse experiences affect depressive symptoms during (early) adulthood, suggesting that these memories could be potentially valuable targets for memory-focused treatments. On the other hand, these findings do not align with the more traditional view of autobiographical memories and depression, where vividness and coherence play a key role. Moreover, we found that these effects are not specific to depression, indicating a potential transdiagnostic role for emotional memories. Although the exact effect of emotional impact and the further specification of this construct in relation to one’s current emotional state and core beliefs requires additional investigation, these findings provide novel insights into the potential mechanisms underlying the relationship between adverse experiences, autobiographical memories, and current depressive symptoms.
Declarations
Ethics Approval
Ethical approval for the study was granted by the Ethical Review Board of the clinical psychology department at the University of Amsterdam (project nr. FMG-5667).
Competing interests
The authors declare no competing interests.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
share
DELEN
Deel dit onderdeel of sectie (kopieer de link)
Optie A:
Klik op de rechtermuisknop op de link en selecteer de optie “linkadres kopiëren”
Met BSL Psychologie Totaal blijf je als professional steeds op de hoogte van de nieuwste ontwikkelingen binnen jouw vak. Met het online abonnement heb je toegang tot een groot aantal boeken, protocollen, vaktijdschriften en e-learnings op het gebied van psychologie en psychiatrie. Zo kun je op je gemak en wanneer het jou het beste uitkomt verdiepen in jouw vakgebied.