Introduction

Internalizing problems, including anxiety and depression, are recognized as pervasive affective disorders (Costello et al. 2003). Given the reported highest prevalence of anxiety and depression among 970 million individuals with mental health challenges and their implications for psychosocial dysfunctions (Voltas et al. 2017), coupled with a substantial risk of handicaps and suicide (Malhi and Mann 2018; WHO 2022), it is crucial to identify associated risk factors.

Childhood trauma, encompassing both abuse and neglect dimensions (Bernstein et al. 1994), presents as a pivotal factor in increasing the likelihood of internalizing problems. Meta-analytic studies consistently show that adverse childhood experiences (ACEs) substantially increase the risk of depression (Chapman et al. 2004; Mandelli et al. 2015; Nanni et al. 2012) and the comorbidity of depression and anxiety (de Graaf et al. 2004). Aligned with the latent vulnerability theory, the damages to the cortico-limbic network induced by ACEs may not immediately manifest as symptoms but leave victims more vulnerable in aversive situations (De Bellis and Keshavan 2003; McCrory and Viding 2015). Therefore, understanding the underlying mechanisms of childhood traumatic experiences on the formation and development of internalizing problems is crucial for detecting and preventing psychopathological conditions.

Recently, a growing body of researches have investigated the mediating mechanisms behind the influence of ACEs on internalizing problems through the lens of cognitive inhibition, emotional regulation, and social cognition (Fuchshuber et al. 2018; Panagou and MacBeth 2022; Schierholz et al. 2016; Whiffen and Macintosh 2005). Nonetheless, these higher-order functions alone fail to encompass the full scope of developmental psychopathology intricately tied to environmental alterations.

Sensory perception, particularly olfaction, presents a noteworthy alternative pathway that displays remarkable sensitivity to stressful environmental changes. There is substantial neuroscience evidence supporting the relationship between ACEs and olfactory function. Specifically, the primary olfactory cortex, a highly flexible neural network, undergoes continuous reshaping in response to demanding conditions such as ACEs. This reshaping affects both the morphology of the olfactory bulb (e.g., volume) and subsequent olfactory function (e.g., odor recognition and discrimination) (Croy et al. 2013, 2014b; Negoias et al. 2010). Other research indicates that individuals with a history of childhood abuse exhibit significantly impaired olfactory ability, characterized by prolonged P2 latency of event-related potentials compared to controls without such a history (Croy et al. 2013). Additionally, individuals with an abuse history demonstrated heightened reactions to olfactory threatening cues, accompanied by increased amygdala activity (Maier et al. 2020). Therefore, ACEs may result in abnormal olfactory function, making it highly sensitive to external changes and potentially predictive of internalizing problems.

On the other hand, olfactory function may significantly impact human well-being, including stress responses and depressive symptoms (Bratman et al. 2024). Studies using animal models suggest that olfactory bulb removal leads to major dysfunctions in the cortical-hippocampal-amygdala circuit, as well as distortions in serotonin and dopamine concentrations, which further result in behaviors resembling depression, such as reduced food intake, memory deficits, and distorted adaptation to novel environments (Masini et al. 2004; Song and Leonard 2005). Beyond food detection and danger avoidance, olfactory function crucially influences social interactions (Croy et al. 2014a; Hummel and Nordin 2005; Rochet et al. 2018). Evidence from humans and rodents shows that olfaction plays a key role in social functions such as emotional contagion and reproduction (Croy et al. 2014a; Stevenson 2010), closely related to depressive symptoms and the emergence of social anxiety (Croy et al. 2014a; Kupferberg et al. 2016; Richey et al. 2019; Rochet et al. 2018).

As a recent conceptual framework suggested, the transmission of olfactory information from the external environment, combined with an individual’s subjective perception, jointly influences human well-being such as depressive symptoms (Bratman et al. 2024). Moreover, olfactory function is thought to be affected by individuals’ past experiences of stressful events, such as ACEs (e.g., Croy et al. 2013; Maier et al. 2020). Thus, olfactory function may be a valuable factor for examining the relationship between childhood trauma and internalizing problems. However, research on its role in the association between ACEs and internalizing problems is limited. Previous studies using tools like Sniffin’ Sticks have yielded inconsistent findings (Naudin et al. 2012; Colle et al. 2022; Croy et al. 2010a, b). Although objective measures are crucial, they may not fully capture the nuanced aspects of olfactory dysfunction. Self-reported measures, which capture subjective olfactory experiences, can provide a more direct illustration of the interplay between personal perceptions of olfactory function and the psychological impact of ACEs (Liu et al. 2021). This approach may also enhance ecological validity, offering a more thorough understanding of olfactory dysfunction in mental health.

Moreover, based on the dimensional model of adversity and psychopathology (DMAP) (Sheridan and McLaughlin 2014), various adversities uniquely affect neurogenesis and neuropsychological development. Childhood neglect may impact higher-level cognitive functions, while childhood abuse predominantly disrupts lower-level emotion systems. Intrigued by distinct pathways linking childhood abuse and neglect to internalizing problems, we hypothesized that childhood abuse, rather than neglect, contributes more to internalizing problems through olfactory dysfunction.

Another intriguing factor to consider is sex, as it has been demonstrated to impact the prevalence of internalizing problems and relevant brain maturation (Achenbach et al. 2016; Gong et al. 2022; Lee et al. 2023; Xu et al. 2020), and olfactory function performance, manifested as a more sensitive sense of smell in females (Sorokowski et al. 2019). Females with a maltreatment history seemed to be more vulnerable to the accumulated effects of traumatic events (Chen et al. 2021; De Bellis and Keshavan 2003). Therefore, we also investigated how sex may modulate the mediating model of olfactory function.

Based on previous studies and theories, this study aimed to examine: 1) the role that subjective olfactory dysfunction played in the association between childhood trauma and internalizing problems, while also testing the stability of the model using a longitudinal design; 2) how the two dimensions of childhood trauma (i.e., abuse and neglect) differently affected depression through subjective olfactory dysfunction; 3) the extent to which sex modulated the associations between childhood trauma and internalizing problems.

We hypothesized that: 1) Childhood trauma would increase the likelihood of subjective olfactory dysfunction, subsequently intensifying internalizing problems with stable effects; 2) Childhood abuse would have a stronger impact on internalizing problems through olfactory dysfunction compared to childhood neglect; 3) Sex would moderate the proposed mediating model, with females who experienced childhood maltreatment exhibiting more pronounced internalizing problems directly and indirectly through impaired olfactory function compared to males.

Method

Participants

Through online advertisements, participants (N = 1661, 860 females; Meanage = 21.67 years, SD = 1.67) were recruited from mainland China. The inclusion criteria were young adults aged 18–25 who had no history of mental illness, respiratory diseases, or temporary loss of smell. The study at the first wave initially received feedback of 1835 participants. Due to the following exclusion criteria, 174 participants were excluded from the data analysis, including 1) scoring all items but with abnormally short completion time (less than 300 s), 2) incomplete survey responses, and 3) not being able to pass any of the three attention check items (e.g., not being able to select a specified option). To test the stability of the findings, we conducted a second-wave survey three months later in which participants were required to report their internalizing symptoms again. Despite inviting all participants from the first wave to participate in the second round, we only received responses from 327 individuals (152 females; Meanage = 21.81, SD = 1.67). The majority of the respondents declined to participate in the second-wave survey for several potential reasons, including scheduling conflicts (e.g., heavy academic assignments), personal health issues (e.g., COVID-19 infections), or a lack of motivation. Despite this, there were no significant differences in sex distribution or age between the two waves (χ2 = 0.322, p = 0.565; t = −1.203, p = 0.229).

Procedure

In the first wave, all respondents were asked to provide answers regarding demographic information and completed the questionnaires. The aims of the current study were briefly introduced to all participants, and they were fully informed that their anonymity was assured, their data would be utilized only for research purposes, and the risk of participating was merely a “very small risk” (e.g., slight discomfort to some questions). Then, all participants provided their written informed consent before the start of this study. Each participant who completed the questionnaires received a payment of CNY 10.00 (approximately USD 1.38). In the second wave, conducted three months later, we contacted participants to track their internalizing problems (i.e., depressive and anxious symptoms) to ensure the stability of the model. All processes and standards in the second wave were consistent with those in the first wave. The procedures in this study comply with the ethical standards for human experimentation established by national and institutional committees, as well as the revised 1975 Helsinki Declaration. This study was approved by the Ethics Committee of East China Normal University [Approval number: HR2-0249-2021; Approval date: 21st, Dec., 2021].

Measures

The Chinese version of the Childhood Trauma Questionnaire (CTQ) was used to assess participants’ adverse childhood experiences (ACEs) (Bernstein et al. 1994; Zhao et al. 2004). This 25-item scale, covering physical abuse, emotional abuse, sexual abuse, physical neglect, and emotional neglect, gauges adverse experiences with family members before age 16. Respondents rated each item on a 5-point Likert scale (from “1 = none to minimal” to “5 = severe to the extreme”). Composite scores, indicating the frequency of adverse experiences, were obtained by summing item scores. The CTQ demonstrated high internal consistency (Cronbach’s alpha = 0.935), with subscale coefficients ranging from 0.735 to 0.935, indicating acceptable to excellent reliability.

The Olfactory Disorder Questionnaire (SODQ) by Liu et al. (2021) assessed olfactory disorder in this study with 10 items exploring daily situations related to olfactory functioning (e.g., ‘I can’t smell the aroma of the food when eating.’). Each item used a 4-point Likert-type scale, and the total score, indicating olfactory disorder severity, was the sum of individual item scores. The SODQ exhibited good internal consistency (Cronbach’s alpha = 0.953), confirming its reliability.

The Chinese version of the Beck Depression Inventory (BDI) (Beck et al. 1961; Yeung et al. 2002), which consists of 21 items, was used to assess the level of depressive symptoms among participants, with higher scores indicating higher levels of depressive symptoms. The BDI demonstrated excellent internal consistency in both waves of data collection (Wave 1: Cronbach’s α = 0.959; Wave 2: Cronbach’s α = 0.953).

Anxiety was assessed using the Generalized Anxiety Disorder-7 (GAD-7) scale (Spitzer et al. 2006; Sun et al. 2021). Participants rated each item on a 4-point scale ranging from 0 to 3, corresponding to the frequency of the described situation (e.g., ‘Not being able to stop or control worrying.’). Higher scores indicate more frequent experiences and higher severity of general anxiety symptoms. The scale revealed strong internal consistency for both Wave 1 (Cronbach’s α = 0.923) and Wave 2 data collection (Cronbach’s α = 0.918).

Statistical analysis

Descriptive statistics and Pearson correlation analyses of the study variables were conducted using SPSS 26 (IBM 2019). To confirm the stability of the measurements, the longitudinal factorial invariance of the scales (i.e., the BDI and GAD-7) was examined using Mplus 8.1 (Muthén & Muthén). Four nested models were analyzed to assess factorial invariance, incorporating increasingly strict constraints: (1) configural invariance, examining the overall patterns of the scales; (2) weak invariance, with factor loadings constrained across the two waves; (3) strong invariance, imposing equality constraints on intercept terms; and (4) strict invariance, additionally constraining residual variances. Model fit was evaluated using Chi-square analysis to degrees of freedom (χ²/df), with p values above 0.05 considered acceptable. Multiple comparisons were adjusted using the Bonferroni correction approach.

For the Wave 1 data, mediation analyses were conducted using Mplus 8.1 to examine the mediating effect of subjective olfactory dysfunction. This involved constructing a second-order latent variable mediation model and a moderated mediation model. In the mediation model, childhood trauma (i.e., the CTQ scores) was the independent variable, and internalizing problems (i.e., the BDI and GAD-7 scores) were the outcome variables, with olfactory dysfunction (i.e., the SODQ score) serving as the mediator. Childhood trauma was treated as a second-order latent variable, with abuse and neglect as first-order latent variables. Physical abuse, emotional abuse, and sexual abuse were the observed variables for abuse, while physical neglect and emotional neglect were the observed variables for neglect. A residual correlation between physical abuse and sexual abuse was established to improve model fit. Depression and anxiety symptoms were used as observed variables for the latent variable of internalizing problems. A bootstrap procedure with 5000 samples was used to generate a 95% confidence interval (CI) for effect estimation. If the 95% CI included zero, the effect was deemed nonsignificant; otherwise, it was considered significant.

Regarding the moderated mediation model, sex was included as a moderator to examine its influence on the associations. To clarify the differences between abuse and neglect, mediating models for the two dimensions of child trauma were separately constructed to examine whether their mediating effects were statistically significant. Further, Wald tests were carried out to compare the differing effects between childhood abuse and neglect.

To assess the stability of the mediating effect, we replicated the mediation models using scores of the CTQ from the first wave as the independent variable and internalizing problems (i.e., BDI and GAD-7 scores) measured in the second wave as the outcome variable. The Wald test was repeated to check the difference between child abuse and neglect.

Results

Factorial invariance

To compare the measurement constructs of the instrument measures between the two waves, models involving depressive symptoms and general anxiety symptoms were examined. The results, as shown in Table 1, indicated that the goodness of fit for the various levels of invariance was acceptable (BDI: Δχ2: 25.32–57.74, Δdf: 20–21; ΔCFI: 0.0–0.001; ΔRMSEA: 0.001–0.002; GAD: Δχ2: 5.31–11.81, Δdf: 6–7; ΔCFI: 0.0–0.001; ΔRMSEA: 0.003–0.004).

Table 1 Measurement invariance testing of BDI and GAD.
Full size table

Correlation results

Table 2 shows the descriptive statistics of the study variables in wave 1. The results indicated positive correlations between all observed variables included in the analysis. Specifically, elevated scores on the BDI and GAD were found to be positively associated with CTQ (rs = 0.609–0.650, ps < 0.001) and its subscale scores (rs = 0.323–0.629, ps < 0.001). The scores of SODQ were not only positively correlated with CTQ and its subscales but also with BDI and GAD (rs = 0.246–0.663, ps < 0.001).

Table 2 Descriptive statistics and correlations of scales scores at Wave 1.
Full size table

Mediation effect of olfactory function

As depicted in Fig. 1, the results presented a significant mediating effect of olfactory dysfunction in wave 1 between childhood trauma and internalizing problems (CFI = 0.961, RMSEA = 0.098). Specifically, childhood trauma was found to increase the risk of olfactory dysfunction (standardized parameter B = 0.721, 95% CI [0.671, 0.724]), and subsequently, olfactory dysfunction further exacerbated internalizing problems (B = 0.311, 95% CI [0.226, 0.408]). The direct effect of childhood trauma on internalizing problems was found to be significant (B = 0.544, 95% CI [0.444, 0.630]), with a higher level of childhood trauma experiences associated with increased internalizing problems in adulthood. The direct effect remained significant after incorporating the mediator into the model, with the indirect effect accounting for 22.4% of the total effect.

Fig. 1: The moderated second-order latent variable mediation model.
figure 1

The direct and indirect effect are both significant, but the moderating effect was not significant. Estimations are standardized betas and residuals are hidden from this figure for clarity. ***p < 0.001. CTQ-1 childhood trauma questionnaire at wave 1, ea emotional abuse, pa physical abuse, sa sexual abuse, en emotional neglect, pn physical neglect, SODQ-1 self-reported olfactory dysfunction questionnaire at wave 1, IP-1 internalizing problems at wave 1, DD-1 depression disorder at wave 1, GAD-1 general anxiety disorder at wave 1.

Full size image

Further analysis of specific dimensions of childhood trauma

Both childhood abuse and neglect demonstrated significant mediating effects on internalizing problems through olfactory dysfunction when examined separately. Specifically, the pathways from childhood abuse and neglect to olfactory dysfunction subsequently resulted in internalizing problems (see Table 3). The model fit parameters for both the abuse model (CFI = 0.974, RMSEA = 0.097) and the neglect model (CFI = 0.973, RMSEA = 0.094) were considered acceptable.

Table 3 Mediating models testing for two dimensions of child trauma separately.
Full size table

As expected, the Wald test results revealed that childhood abuse and childhood neglect had significantly different effect on olfactory function (χ2(1) = 56.012, p < 0.001) (see Fig. 2). When comparing the two dimensions in the first wave (see Fig. 2a), although the direct effects of abuse and neglect on internalizing problems were both significant (ps < 0.001), only the indirect effect of childhood abuse was strong and significant (p < 0.001), by contrast the influence of childhood neglect on olfactory dysfunction was statistically non-significant (p = 0.340). The model fit parameters were all acceptable (CFI = 0.964, RMSEA = 0.097).

Fig. 2: The comparison model of childhood abuse and childhood neglect in both two waves.
figure 2

The upper panel is constructed by data in wave 1 (a), and the bottom panel is constructed by data in wave2 (b). The influence of childhood abuse is strong while that of childhood neglect is faint. Estimations are standardized betas and residuals are hidden from this figure for clarity. ***p < 0.001. CTQ-1 childhood trauma questionnaire at wave 1, ea emotional abuse, pa physical abuse, sa sexual abuse, en emotional neglect, pn physical neglect, SODQ-1 self-reported olfactory dysfunction questionnaire at wave 1, IP-1 internalizing problems at wave 1, DD-1 depression disorder at wave 1, GAD-1 general anxiety disorder at wave 1, IP-2 internalizing problems at wave 2, DD-1 depression disorder at wave 2, GAD-1 general anxiety disorder at wave 2.

Full size image

The stability of mediating effect model

To assess the stability of the mediating effect, the same procedures of data analysis were replicated with the internalizing problems data in wave 2. The predicting effect of childhood trauma on internalizing problems remained significant (B = 0.651, 95%CI [0.484, 0.874]). The mediating effect of olfactory dysfunction remained similar between childhood trauma and internalizing problems in wave 2. The model fit parameters also met the acceptable criteria (CFI = 0.969, RMSEA = 0.085). Furthermore, the Wald test results continued to show significant differences between childhood abuse and neglect (χ2 = 5.156, p = 0.023). Notably, childhood abuse maintained a significant and strong influence in this model (p < 0.001), whereas the impact of childhood neglect on olfactory dysfunction remained non-significant (p = 0.355) (see Fig. 2b).

Moderating effect of sex

The moderating effect of sex on the mediating model was found to be non-significant in wave 1. Specifically, there was no statistically significant interaction between sex and childhood trauma in predicting olfactory dysfunction (B = 0.011, 95% CI [−0.072, 0.089]). Additionally, the interaction between sex and olfactory dysfunction did not yield significant results in predicting internalizing problems (B = 0.012, 95% CI [−0.040, 0.061]). Lastly, the moderation of the direct effect was also non-significant (B = −0.012, 95% CI [−0.082, 0.058]) (see Fig. 1). Sex differences in the study variables are presented in Supplementary Materials.

Discussion

The study investigated the link between two dimensions of childhood trauma and internalizing problems, focusing on the mediating role of subjective olfactory dysfunction. Results indicated that childhood trauma had enduring effects on internalizing problems through subjective olfactory dysfunction, with the mediation remaining significant over three months. Notably, childhood abuse had a more pronounced impact via the path of subjective olfactory impairment compared to neglect, while the moderating effect of sex was not significant. These findings enhance our understanding of how childhood trauma influences internalizing problems.

The mediating role of subjective olfactory dysfunction

Supporting our hypothesis, subjective olfactory dysfunction consistently mediated the relationship between childhood trauma and internalizing problems. This aligns with previous findings (Karaca Dinç et al. 2021; Mandelli et al. 2015) and further contributes to the latent vulnerability theory, expanding it to a sensory dimension which is often overlooked in prior research (McCrory and Viding 2015). While existing studies supported the theory at higher-order cognitive and emotional levels (Armbruster-Genç et al. 2022; Jeong et al. 2021), our research uniquely complements it by exploring the sensory perspective. We propose that ACEs may predispose individuals to a vulnerable olfactory system, thereby increasing the risk of future internalizing problems before clinical symptoms arise. This aligns with the most recent conceptual framework proposed by Bratman et al. (2024), which highlights the critical role that olfaction plays in the impact of the environment on human well-being.

At the neuroscience level, the mechanisms underlying olfactory dysfunction are likely to lie in the atypical function of primary olfactory cortex and the limbic system. Childhood trauma, a potent stressor, disrupts the normal processing of environmental information, and affects the primary olfactory cortex known for its high plasticity. The complex olfactory circuits are susceptible to alterations through both bottom-up inputs and top-down regulation (Negoias et al. 2010), allowing external adverse environment to modify the morphology of the olfactory bulb, leading to consequential olfactory function impairments (Croy et al. 2013, 2014b; Maier et al. 2020).

Olfactory information is primarily processed in the limbic system and its immediate vicinity, responsible for emotion generation and regulation (Croy et al. 2014b; Taalman et al. 2017). Therefore, deviations in olfactory function may indicate disruptions in the prefrontal-limbic-mesolimbic circuitry, which is involved in emotional regulation (e.g., amygdala and medial prefrontal cortex pathways), reward learning and decision making (e.g., hippocampus, ventral striatum, orbitofrontal cortex) (Masini et al. 2004; Song and Leonard 2005). These disruptions could potentially lead to future internalizing problems (Ahmed et al. 2015). Notably, our study reveals the reliable mediating effect of subjective olfactory function over three months, highlighting the enduring impact of chronic stress, particularly during childhood, on mental health in adulthood. This underscores the importance of minimizing the occurrence and severity of stressful events in childhood as a crucial aspect of early prevention.

The differential effect of abuse and neglect dimensions

By dissecting the impact of each type of childhood trauma separately, we gain a deeper understanding of how childhood trauma contributes to these difficulties. Specifically, the findings indicated a stronger and more significant association between childhood abuse and internalizing problems mediated by olfactory dysfunction, as compared to childhood neglect.

This observation aligns with the DMAP theory, supported by empirical evidence (Betz et al. 2022; Mattheiss et al. 2022; Sheridan and McLaughlin 2014). Childhood abuse disrupts emotion learning circuits, such as the amygdala and hippocampus, leading to fear generalization across stimuli (McLaughlin et al. 2016). Indeed, individuals with a history of childhood abuse may adopt avoidant strategies to cope with fear, associated with automatic negative thinking and mood distress (Compas et al. 2017; Milojevich et al. 2019). Abnormalities in emotional learning circuits, implicated in depression and anxiety disorders (France et al. 2022; Goltermann et al. 2022; Zhang et al. 2020), are also relevant to olfactory information processing (Fjaeldstad et al. 2017). Traumatized individuals, with hyperactive amygdala and its abnormal connections to the ventromedial prefrontal cortex (Herringa et al. 2013; Puetz et al. 2016), have been found to experience heightened negative emotions when encountering diverse odors in daily life (Kim et al. 2018). Notably, such negative emotions triggered by olfactory stimuli, are an important characteristic in depressive disorders (Twivy et al. 2023), which offers an explanation for the stronger mediating effect observed between childhood abuse and internalizing problems.

Childhood neglect primarily contributes to anhedonia symptoms of depression by impacting reward circuits, such as the ventral striatum, and disrupted associative learning (Sheridan et al. 2018; Wang et al. 2022). Distorted hedonic experiences may affect olfactory perceptions, increasing emotional disturbance (Atanasova et al. 2010). The weaker mediating effect between neglect and internalizing problems may be explained by the involvement of nucleus accumbens (NAcc) deficits in anhedonia, as the distance between NAcc and the olfactory system is not as close as with the limbic system (Forbes and Dahl 2012). The olfactory dysfunction tool used in this study did not measure olfactory hedonic experience, suggesting the need for future research to explore its relationship with childhood abuse and neglect separately. In summary, it’s essential to recognize that both childhood abuse and neglect contribute to the exacerbation of internalizing problems. Childhood abuse harms the emotion-learning system, particularly the limbic system, leading to olfactory function deficits and increased susceptibility to internalizing problems. These findings may shed light on the distinct effects of different aspects of childhood trauma, enhance our understanding of the mechanisms underlying internalizing problems, and offer new indicators for the early detection and intervention of depression and anxiety from a sensory perspective.

The moderating role of sex

Unexpectedly, we did not find evidence supporting the moderating effect of sex. Sex did not moderate either the indirect effect through subjective olfactory dysfunction or the direct effect. This finding differs from previous studies suggesting a higher susceptibility in females to stress-related disorders like depression and anxiety following cumulative traumatic events (Kessler et al. 2012; Purtle et al. 2016). However, some studies found no differences between the sexes (Arnow et al. 2011; Cutajar et al. 2010). One potential explanation for this inconsistency is that resilience may aid in better recovery for females from childhood trauma over time, leading them to report fewer trauma-related experiences (Wei et al. 2021). Another explanation could be related to the different coping strategies employed by the two sexes. In contrast to girls, boys typically show a lower inclination to openly express distress or seek social support, often concealing their emotions (Chaplin and Aldao 2013), potentially canceling out other protective factors.

Limitations and future directions

Several limitations need to be noted. Firstly, while our study demonstrates consistency between the two waves of data, it is limited by the relatively short three-month interval. The inability to track data from childhood into adulthood hinders a comprehensive understanding of long-term mediating effects and patterns of psychopathological development. Future studies are needed to examine the longitudinal trajectory and mediating effect of subjective olfactory function on the relationship between ACEs and internalizing problems from childhood through adolescence. Secondly, relying on retrospective reports in the CTQ introduces potential memory distortions and biases. Future studies should consider multiple evaluation sources, including assessments from participants and others, for a more accurate understanding of childhood adversity. Thirdly, our study exclusively enrolled healthy individuals, limiting the generalizability of findings. Future research should include a broader participant range, incorporating clinical and sub-clinical populations, such as individuals with major depressive disorder or depression proneness. Lastly, this study did not control for age, as the age range in our sample was relatively narrow (young adults: 18 to 27 years) and did not encompass childhood and adolescence. Future researches may expand to include a broader range of age groups and incorporate long-term longitudinal designs to observe changes across different age ranges and understand how age-related factors influence outcomes. This approach would provide deeper insights into the developmental aspects and potential age-related variations that might affect the findings, ultimately leading to more comprehensive and generalizable results.

Conclusions

This study provides insights into the relationship between childhood trauma and internalizing problems from the perspective of sensation and perception. It concluded that childhood trauma might contribute to the severity of olfactory dysfunction (a latent vulnerability in the limbic system), which concurrently and prospectively predicts internalizing problems. Furthermore, the findings that abuse dimensions of childhood trauma can result in more severe internalizing problems through the olfactory functioning path provide experiential evidence that supports the DMAP theory.