The Association Between Childhood Trauma and Selective Serotonin Reuptake Inhibitor Treatment Outcomes in Adult Patients With Major Depressive Disorder

Article information

Psychiatry Investig. 2025;22(8):886-896

Publication date (electronic) : 2025 July 31

doi : https://doi.org/10.30773/pi.2024.0364

Yutong Deng¹, Wen Bian¹, Jingwen Sun¹, Lijun Liu¹, Shuzhe Zhou¹, Qi Liu¹, Tianmei Si¹, Jing Wang¹, Hongjun Tian², Kerang Zhang³, Jing Wei⁴, Gang Wang⁵, Qiaoling Chen⁶, Gang Zhu⁷, Xueyi Wang⁸, Nan Zhang⁹, Xiaozhen Lv^,1, Xin Yu^,1

¹Peking University Sixth Hospital, Peking University Institute of Mental Health, NHC Key Laboratory of Mental Health (Peking University), National Clinical Research Center for Mental Disorders (Peking University Sixth Hospital), Beijing, China

²Tianjin Fourth Central Hospital, The Affiliated Hospital of Tianjin Medical University, Tianjin, China

³Department of Psychiatry, First Hospital of Shanxi Medical University, Taiyuan, China

⁴Department of Psychological Medicine, Peking Union Medical College Hospital, Beijing, China

⁵Beijing Anding Hospital, Capital Medical University, Beijing, China

⁶Department of Psychiatry, Dalian Seventh People’s Hospital, Dalian, China

⁷Department of Psychiatry, The First Affiliated Hospital of China Medical University, Shenyang, China

⁸Department of Psychiatry, The First Hospital of Hebei Medical University, Mental Health Institute of Hebei Medical University, Shijiazhuang, China

⁹Department of Neurology, Tianjin Medical University General Hospital, Tianjin, China

Correspondence: Xin Yu, MD Institute of Mental Health, Peking University, Huayuanbei 51, Haidian District, Beijing 100191, China Tel: +86-10-82801999, E-mail: yuxin@bjmu.edu.cn

Correspondence: Xiaozhen Lv, PhD Institute of Mental Health, Peking University, Huayuanbei 51, Haidian District, Beijing 100191, China Tel: +86-10-62723705, E-mail: lvxiaozhen@bjmu.edu.cn

Received 2024 December 2; Revised 2025 May 8; Accepted 2025 May 16.

Abstract

INTRODUCTION

Major depressive disorder (MDD) is a chronic mental disorder characterized by high prevalence and recurrence, and it imposes a significant burden on individuals and society [1]. MDD is the leading cause of disability globally, affecting more than 350 million individuals [2]. Selective serotonin reuptake inhibitors (SSRIs), such as escitalopram, are the primary pharmacological treatments for MDD. However, the Sequenced Treatment Alternatives to Relieve Depression (STAR*D) trial indicated that only 36.8% of patients achieved remission following acute-phase treatment with escitalopram [3]. An urgent need exists to understand the factors associated with treatment outcomes, and childhood trauma (CT) may be a crucial factor [4].

CT encompasses abuse, neglect, and adverse living conditions during childhood. It is highly prevalent, with approximately 60% of individuals experiencing at least one type of CT [5]. In a Chinese study, 80.9% of 11,972 individuals reported exposure to at least one type of CT [6]. Research has shown that CT may alter the development of the nervous, endocrine, and immune systems [7,8], leading to impairments in cognitive [9], social [10], emotional, and reward functions [11], and increasing the risk of MDD [4]. Several studies have examined the association between CT and MDD treatment, but the results remain controversial. Nanni et al.’s [12] meta-analysis of 10 clinical trials (3,098 participants) revealed that CT was associated with nonresponse or relapse, particularly in patients receiving pharmacotherapy and combined treatments. Similarly, Nelson et al.’s [13] meta-analysis of 5 clinical studies (1,229 participants) revealed that CT was related to poorer treatment outcomes. Conversely, a trial by Medeiros et al. [14] involving 665 MDD patients did not show significant association between CT and symptom improvement after 12 weeks of treatment, regardless of whether CT occurred before age 7. Moreover, Childhood Trauma Meta-Analysis Study Group’s meta-analysis of 29 studies (6,830 participants) concluded that CT did not diminish the benefit of treatment for MDD [15].

These discrepancies may arise from the heterogeneous natures of CT and MDD. CT includes various adversities, such as abuse, including sexual abuse (SA), emotional abuse (EA), and physical abuse (PA), as well as neglect, including emotional neglect (EN) and physical neglect (PN) [16]. Different types of CT may be associated with distinct treatment outcomes. Some studies suggest that specific types of CT, rather than a general history of trauma, may be more strongly associated with the treatment response. For example, Williams et al. [17] assessed 1,002 MDD patients treated with escitalopram, sertraline, or venlafaxine and reported that SA, PA, and EA, rather than a general history of CT or other types of CT, were associated with poorer outcomes. A study of 334 adolescents with MDD receiving either pharmacotherapy or combined pharmacotherapy and cognitive behavioral therapy (CBT) indicated that patients with a history of SA responded similarly to both treatments, whereas those with a history of PA showed a significantly better response to pharmacotherapy alone than to the combined treatment [18]. Additionally, some research has suggested that EN and EA may have stronger links to MDD than PA or SA does [13,16], whereas EN and EA, particularly EN, have often been overlooked in their relationships with treatment outcomes, highlighting the need for further analysis. These findings suggest that different types of CT may exert varying influences on treatment outcomes. MDD, similarly, encompasses various dimensions, including anhedonia and insomnia, which may respond differently to treatment [19]. Anhedonia, which affects 37%–72% of MDD patients, is a core symptom of the disorder [20] and is linked to reward-related neural circuit dysfunction [21]. Similarly, insomnia, which affects more than 80% of MDD patients, is a key component of the disorder and significantly contributes to the overall severity of depression [22]. It has a profound impact on patients’ quality of life and is strongly associated with an increased risk of relapse [23]. Research has suggested that CT, especially EN, worsens anhedonia by altering reward-related neural circuits [11,24]. Similarly, CT may exacerbate insomnia [25], which partly mediates more severe depressive symptoms [26]. Several studies have examined the relationships between different CT types and treatment outcomes for anhedonia or insomnia in MDD patients. Harkness et al. [27] found that severe EA and PA correlated with smaller improvements in anhedonia after 8 weeks of escitalopram treatment. Pigeon et al. [28] reported that while depressive symptoms improved in female MDD patients with SA, insomnia symptoms remained unchanged. However, these studies focused only on abuse and did not explore the role of neglect.

Given the controversies in prior research, our study aims to explore the association between CT and treatment outcomes in MDD patients, with a particular focus on examining the associations between different types of CT and treatment outcomes in the symptom dimensions of anhedonia and insomnia. By conducting a detailed investigation, we seek to deepen our understanding of the associations between CT and SSRI treatment outcomes, providing insights to guide personalized treatment strategies for MDD.

METHODS

RESULTS

Baseline characteristics and treatment outcomes of MDD patients receiving 8 weeks of SSRI therapy

Among the 567 MDD patients who participated in the 8-week SSRI antidepressant treatment, 4 were excluded for not completing the CTQ-SF at baseline, and 5 were excluded due to hypomanic episodes. Finally, 285 MDD patients who completed the CTQ-SF and the 8-week SSRI treatment follow-up were included in the study (Figure 1).

Figure 1.

Patient selection process. MDD, major depressive disorder; CTQ-SF, Childhood Trauma Questionnaire-Short Form; SSRI, selective serotonin reuptake inhibitor.

No significant differences in sociodemographic characteristics or CTQ-SF scores were observed between the MDD patients who completed and dropped out of the study. However, MDD patients who were included were more likely to have recurrent MDD and had relatively lower baseline HRSD-17 and HAMA scores (Supplementary Table 1).

Table 1 presents the sociodemographic characteristics, clinical characteristics, and HRSD-17 scores at follow-up of the participants. Among the 285 MDD patients analyzed, 74 (25.96%) were male, with a mean age of 39.30±10.60 years. The average total score for the CTQ-SF was 39.11±10.12. Among the subscales, EN had an average score of 12.15±4.89, and PN had an average score of 8.95±3.31. The median scores for EA, PA, and SA were 6.00 (IQR, 5.00–8.00), 5.00 (IQR, 5.00–6.00), and 5.00 (IQR, 5.00–5.00), respectively. The average baseline scores were 20.94±4.90 for the HRSD-17 and 19.14±7.25 for the HAMA. At weeks 2, 4, and 8 of follow-up, the HRSD-17 scores were 14.11±6.14, 10.25±5.32, and 7.01±4.56, respectively.

Table 1.

Baseline characteristics and HRSD-17 scores at follow-up of 285 MDD patients

After 8 weeks of treatment, the average percentage reduction in the HRSD-17 score was 66.02%±20.85%. For anhedonia and insomnia scores, the median percentage reductions were 60% (IQR, 33.33–83.33) and 75% (IQR, 50.00–100.00), respectively. Remission was achieved in 181 patients (63.51%), 224 patients (78.60%) achieved a response, and 112 patients (39.30%) achieved a substantial improvement.

Associations between CT and SSRI treatment outcomes in MDD patients

Our univariable linear regression analysis revealed a significant negative association between the EN subscale score and the percentage reduction in the HRSD-17 score after 8 weeks of treatment (β=-2.779, p=0.025), which remained significant after adjusting for baseline confounders (β=-3.035, p=0.019). The total CTQ-SF score and scores for other subscales (EA, PA, SA, and PN) were not significantly associated with the HRSD- 17 score in either the univariable or multivariable analyses.

Both SA and EN were negatively associated with the percentage reduction in the anhedonia symptom score in the univariable analysis, but after adjusting for confounders, only EN remained significantly associated (β=-4.227, p=0.044). The total CTQ-SF score and scores for the other subscales (EA, PA, and PN) showed no significant associations with the percentage reduction in anhedonia symptoms.

The EN subscale score was significantly negatively associated with the percentage reduction in the insomnia symptom score in both the univariable (β=-6.635, p=0.048) and multivariable (β=-7.054, p=0.045) analyses. The CTQ-SF total score was significantly associated in the univariable analysis but lost significance after adjustment. No other subscales showed significant associations (Table 2).

Table 2.

Linear regression results showing the associations of the CTQ-SF and its subscale scores on the percentage reduction in the HRSD-17 score and in anhedonia and insomnia symptoms after 8 weeks of treatment

Using the “very much improved” status as the outcome, the univariable logistic regression analysis revealed that a higher EN score significantly reduced the likelihood of achieving this status after 8 weeks of treatment (OR=0.719, 95% CI [0.556, 0.921], p=0.010), and this association remained significant in the multivariable analysis (OR=0.733, 95% CI [0.553, 0.960], p=0.027). The CTQ-SF total score or scores for other subscales (EA, PA, SA, or PN) showed no significant associations with the “very much improved” outcome. Neither the CTQ-SF total score nor the scores for any of its subscales had a significant association with remission or a response (Table 3).

Table 3.

Univariable and multivariable logistic regression results for remission, response, and very much improved

DISCUSSION

Our findings indicate that EN is significantly associated with less improvement in overall symptoms and anhedonia and insomnia in particular following 8 weeks of SSRI treatment. EN was also significantly linked to “very much improved” outcomes after treatment. Neither the total CTQ-SF score nor the scores for subtypes of CT (EA, PA, SA, and PN) showed significant associations with treatment outcomes for overall symptoms or the dimensions of anhedonia and insomnia.

Our findings align with those of previous studies, which revealed no statistically significant correlation between a general history of CT and improvements in depressive symptoms in MDD patients after short-term SSRI treatment [15,44]. However, some studies have shown that CT significantly impacts MDD treatment outcomes [12,13]. This discrepancy may be due to differences in treatment modalities and the methods used to assess both CT and MDD symptoms. As an early-life experience, CT is temporally distant from the onset of adult MDD, and its complex relationship with treatment outcomes may be influenced by various intervening factors, which may explain the inconsistent findings across studies.

EN seems to have a significant association with less of improvement in symptoms. As previously discussed, a single type of CT may have a more noticeable association with poorer treatment outcomes than a general history of CT [17]. EN, which is characterized by a lack of adequate love, belonging, nurturing, and support from parents or guardians, leads to feelings of abandonment and helplessness in adolescents, heightening their risk of developing mental health issues, including MDD [45]. EN is the most prevalent form of CT [46], yet its subjective nature and difficulty in detection often result in it being overlooked [47]. Research has shown that EN is significantly associated with detrimental outcomes, including increased childhood aggression, social withdrawal, and associations with mental health issues such as anxiety, MDD, dissociation, borderline personality disorder, posttraumatic stress disorder, and somatoform dissociation [48-50]. From a neurobiological perspective, EN is linked to heightened amygdala reactivity to threats, which is mediated by the hypothalamic-pituitary-adrenal (HPA) axis [51]. Few studies have focused on the association between EN and outcomes of treatment with SSRIs in MDD patients, and our study addresses this gap. Notably, the association between EN and SSRI treatment outcomes may not be fully captured by traditional endpoints such as remission or response but rather is evident when percentage reduction or very much improved is used as the outcome variables PN was not associated with MDD treatment outcomes. Some studies suggest that PN has a weaker connection with mood disorders [52,53]. To date, no studies have thoroughly examined the differential associations of EN and PN with antidepressant treatment outcomes. Our study further highlighted the distinct role of EN.

EA, PA, and SA were not significantly associated with treatment outcomes in our study. Previous research has suggested that abuse and neglect are differently associated with psychopathology [54]. Abuse is typically seen as a threat to an individual’s physical integrity, potentially affecting the HPA axis and leading to emotional dysregulation, hypervigilance, and fear responses [54]. Moreover, neglect is characterized by a lack of environmental input, which undermines reward learning processes and impairs reward function, contributing to core depressive symptoms such as anhedonia [55]. Research has shown that HPA axis hyperactivity can be reversed with SSRI treatment [56]; however, SSRI treatment has shown limited effectiveness in addressing reward function and anhedonia symptoms [57]. Notably, the scores for EA, PA, and SA in our data were low and exhibited limited variability, and thus, the findings should be interpreted with caution.

The analysis of anhedonia and insomnia symptoms yielded similar results to the overall depressive symptoms: EN was correlated with lower improvements in anhedonia and insomnia after 8 weeks of SSRI treatment, whereas the CTQ-SF total score, abuse score, and PN score were not significantly associated with improvements in anhedonia or insomnia. Few studies have explored the relationship between CT and improvements in anhedonia symptoms in MDD patients following treatment. One study examined the association between abuse and changes in anhedonia following antidepressant treatment in MDD patients but focused solely on abuse [27]. In the study, 164 patients underwent 8 weeks of escitalopram treatment, and the results revealed that while no type of abuse (EA, PA, or SA) predicted the overall severity of depression after treatment, both EA and PA were associated with more severe posttreatment anhedonia symptoms. The discrepancy in findings may be due to differences in the methods used to assess CT and anhedonia. Notably, the study examined only abuse among the types of CT. Currently, no studies have specifically investigated the association between childhood neglect and changes in anhedonia symptoms. Our study fills this gap by demonstrating that EN is negatively associated with improvements in the anhedonia symptoms of MDD patients. Anhedonia, a central symptom in most patients with MDD [20], is closely associated with dysfunction in reward-related neural circuits [21,24]. This dysfunction, which affects key brain areas involved in reward processing, such as the medial prefrontal cortex, anterior cingulate cortex, striatum, and insula, leads to reduced neural activation and altered connectivity, ultimately impairing reward anticipation, perception, and reception [11,58]. Based on prior research, we identified work and interest, somatic symptoms (general), and genital symptoms from the HRSD-17 as indicators of anhedonia [38,39]. Among all CT types, EN appears to have the strongest negative association with anhedonia symptoms in patients with MDD [59]. From a neurobiological perspective, EN has been linked to blunted development of reward-related activity in the ventral striatum during adolescence, which plays a crucial role in the experience of pleasure and motivation [60]. Structural MRI studies have also revealed reductions in gray matter volume in the orbitofrontal cortex, subgenual prefrontal cortex, amygdala, striatum, and hippocampus in adolescents exposed to EN [61], brain areas that are intricately involved in processing reward, emotional salience, and contextual memory. Moreover, functional imaging has demonstrated disrupted connectivity in the prefrontal-limbic-thalamic-cerebellar circuits among individuals with MDD and a history of childhood neglect, further implicating widespread network dysfunction in EN-related anhedonia [62]. These structural and functional abnormalities are thought to impair dopaminergic signaling within the reward circuitry, thereby contributing to more persistent anhedonic symptoms [20].

Insomnia is common among MDD patients and often shows limited improvement after acute-phase antidepressant treatment [22,23]. Addressing factors that affect the treatment response for insomnia is essential. The association between CT and improvements in insomnia symptoms following antidepressant treatment in MDD patients has been understudied. One study revealed that most female MDD patients with SA continue to experience insomnia symptoms after treatment, but a control group was lacking and the study focused only on female patients with SA [28]. Our findings that EN is linked to poorer treatment outcomes for insomnia underscore the importance of addressing EN in MDD treatment. CT can trigger a strong stress response, increasing the activity of the HPA axis and the sympathetic nervous system, which leads to sleep disturbances [63]. Research suggests that EN, rather than other types of CT, is more significantly associated with insomnia symptoms [64]. From a neurobiological perspective, EN has been associated with hyperactivation of the HPA axis, resulting in heightened physiological arousal that can interfere with sleep initiation and maintenance [51,65]. In addition, EN may impair the brain’s emotional regulation circuits, including the prefrontal cortex, hippocampus, and amygdala, which are also implicated in the development of insomnia [61,66]. Recent findings suggest that individuals with insomnia exhibit fragmented and restless rapid eye movement sleep, which disrupts overnight emotional processing. This may be particularly relevant for individuals with a history of EN, who are already vulnerable to emotional dysregulation and distress retention [67]. The failure to adequately resolve emotional distress during sleep may contribute to chronic hyperarousal and persistent insomnia symptoms despite antidepressant treatment.

Our findings show that EN is significantly associated with poorer treatment response to SSRIs in Chinese adults with MDD, particularly in anhedonia and insomnia. This highlights the clinical value of EN as a predictor of suboptimal antidepressant response, underscoring the importance of its routine assessment in psychiatric practice. Despite being often overlooked due to its subjective nature, EN has a measurable impact on treatment outcomes [47]. Systematic screening for CT could help identify at-risk individuals, facilitating more proactive individualized treatment planning. The limited improvement in anhedonia and insomnia symptoms among patients with EN suggests the need for more tailored treatments, including adjunctive therapies such as trauma-focused CBT [68]. More broadly, these findings support a shift toward trauma-informed and symptom-dimension-based approaches in MDD management, recognizing the lasting neurobiological effects of early adversity and the heterogeneous etiologies of depression. Our study has several strengths that contribute to a deeper understanding of the association between CT and MDD treatment outcomes. First, our study fills this gap by providing a detailed analysis of the associations between different types of CT and MDD treatment outcomes, particularly in terms of anhedonia and insomnia. Second, we utilized sensitive measures, including HRSD-17 scores, CTQ-SF scores and the percentage reduction in HRSD-17 scores, which allowed us to adequately illustrate the associations between CT and SSRI treatment outcomes.

Our study has several limitations. First, given the post hoc nature of the analysis, we acknowledge that the findings are limited by the lack of detailed information on the timing, intensity, and duration of CT. Additionally, CT was assessed using a retrospective self-report questionnaire, which might lead to recall bias. Second, the CTQ-SF alone is insufficient to capture the full scope of CT, as it only includes neglect and abuse, and does not cover other common childhood adversities in the Chinese population, such as family conflict and overprotection. Future research should incorporate these factors for a more comprehensive assessment of CT. Third, the use of HRSD-17 items to evaluate anhedonia is relatively crude. Future studies should consider the use of more comprehensive and systematic scales, such as the Snaith–Hamilton Pleasure Scale or the Dimensional Anhedonia Rating Scale, to assess anhedonia symptoms. Fourth, the low levels of EA, PA, and SA experienced by participants in our sample, along with the concentration of scores at the lower end, suggest that conclusions regarding the lack of associations between these types of CT and changes in MDD symptoms after treatment should be interpreted with caution. Future research should optimize the sample by including MDD patients with more severe levels of EA, PA, and SA to further investigate the associations between childhood abuse and MDD treatment outcomes. Finally, we did not account for certain confounding factors, including variations in genes such as brain-derived neurotrophic factor, the promoter region of the serotonin transporter gene, and the glucocorticoid receptor, as well as epigenetic modifications, which have been found to interact with CT and MDD, potentially affecting our findings [69].

In summary, EN is significantly associated with poorer SSRI treatment outcomes in Chinese adult patients with MDD, with notable associations observed between anhedonia and insomnia. These findings underscore the need for tailored treatment strategies that address EN, highlighting its importance in improving patient outcomes and enhancing MDD management.

Supplementary Materials

Article information Continued