The impact of life events on the severity of depressive symptomatology in patients with a first major depressive episode

1. Introduction

The impact of stressful life events on the onset or worsening of depression has been extensively explored in the literature, but a considerable heterogeneity in the responsiveness to stress impedes the extraction of definitive conclusions on this topic⁽¹⁾. The role of stress in the development of depression is the result of multiple factors, including the chronic effects of environmental stimuli and other relevant stressors with onset during childhood, which are supposed to trigger hyperactivity of the hypothalamic-pituitary-adrenal axis⁽²⁾.

The cumulative effect of multiple life stressors was investigated, with the hypothesis that such events may lead to a higher risk for depression, beyond the effect of a single negative life event⁽²⁾. When all the severities of life events (assessed by interviews) were evaluated, the number of stressors significantly increased the risk of depression onset, with the risk following an exponential pattern (N=24,648 person-months)^(3,4). More exactly, the risk of a depressive episode onset related to negative life events during one month was 0.9% for no stressors, 3.4% for one stressor, 6.8% for two stressors, and 23.8% for three stressors⁽³⁾. However, the relationship between the severity and number of stressors and the onset of depression is far from being clear, and methodological aspects involved in different studies approaching this topic may significantly modify the results. For example, in the Camberwell studies, multiple severe life events were associated with an increased risk for depression only if the events were unrelated to each other, and, at any rate, the increase in the risk due to the second or third event was relatively reduced^(3,5,6).

The role of stressful events in the pathogenesis of depression is mediated by multiple individual variables, such as personality traits (e.g., openness to experience) and social support (i.e., absence of social support may be associated with the onset and relapse of depression)^(1,7). Multivariate models evaluated the effects of adverse life events in patients with major depressive disorder (MDD) and the mediation role of cognitive-personality style and type of events (interpersonal versus achievement), with 43 patients with MDD versus 43 healthy comparison subjects⁽⁸⁾. The authors of the respective study observed that adverse life events, sociotropy, and autonomy were significantly associated with depression onset and the event type impacted outcomes differently⁽⁸⁾. Also, specific life events interacted with cognitive-personality style when the prediction of response to treatment was evaluated⁽⁸⁾. Younger age, low social class, and negative and family-related stressful life events were associated with an increased risk of new-onset MDD, according to the results of a prospective study that included a structured evaluation of life events (by using the Munich Event Questionnaire) in a group of 2389 adolescents⁽⁹⁾. The absence of stressful events related to school and family was associated with the improvement in MDD symptoms, and the weighted overall number of life events predicted stable depression⁽⁹⁾. The severity of depression was positively correlated with the severity of stress, the last variable being defined by the life stress events 12 months prior to the current episode of depression⁽¹⁰⁾. When stress increases, the severity of symptoms also increases, and more physical and affective symptoms are observed in female patients, while more behavioral symptoms are reported in males⁽¹⁰⁾.

A study that evaluated the relationship between negative life events, depression, and academic performance (N=3629 Chinese college students) found that the effect of stressors on scholar performance is partly mediated by depressive symptoms⁽¹¹⁾. On the contrary, social support had a positive effect on academic performance and partially canceled the impact of negative life events⁽¹¹⁾. 

Regarding other variables that may impact the relationship between stressful life events and the risk of depression onset, a longitudinal evaluation of the two variables, which was conducted on 13,006 patients with a first episode of depression (FED) and a gender/age-matched control group of 260,108 subjects, concluded that the susceptibility to significant life stressors did not seem related to the life phases⁽¹²⁾. A recent divorce, recent unemployment and the suicide of a close relative were associated with an increased risk of FED, while the death of a relative due to other reasons than suicide had no significant impact on the risk for depression⁽¹²⁾. Stressful life events, such as imprisonment or recent divorce, were associated with a higher risk of suicide in a Danish national registers study (N=7115 suicides), with a five-time and 1.5-time higher risk than controls, respectively⁽¹³⁾.

Stressful life events have been demonstrated as causal factors for episodes of major depression, with an odds ratio (OR) for the onset of major depression in the four weeks after a stressful life event of 5.64 (overall, N=24,648 individuals)⁽¹⁴⁾. Monozygotic pairs had OR=3.58 in the same context, while dizygotic twins reached an OR of 4.52⁽¹⁴⁾.

Patients with MDD who had severe life event(s) preliminary to the onset of depression had higher levels of depressive symptoms severity, more cognitive and somatic manifestations of depression, and lower levels of functioning, compared to patients with MDD without severe life events (N=100 participants)⁽¹⁵⁾. Single negative life events had a more than five-fold higher risk and multiple events had more than eight-fold higher risk for the onset of depressive disorders independently of previous depressive or anxiety symptoms in a cohort study of 1947 adolescents⁽¹⁶⁾.

An analysis of the National Survey of American Life cross-sectional study (N=5899 adults) reported that stressful life events were associated with the presence of a major depressive episode (MDE) independent of covariates, including race (African-Americans, Caribbean Blacks, and Non-Hispanic Whites were participating in this survey)⁽¹⁷⁾. In men, however, an interaction between race and stressful life events was detected, with a stronger association with MDE being reported in White men⁽¹⁷⁾.

The results of a systematic literature review, including papers published over more than two decades, exploring the relationship between life stressors and depression, showed that empirical findings mostly supported the previously mentioned relationship⁽¹⁸⁾. Therefore, the investigation of the association between life stressors and FDE and MDD is granted by the existing data.

This study aims to investigate the relationship between the number of life events and the severity of depressive symptoms in individuals experiencing their first depressive episode. Understanding this relationship is crucial for developing targeted interventions and for improving treatment outcomes for individuals with depression^(19-22). On a theoretical level, understanding the complex interplay between early and late stressors and the onset of major depression has important consequences for the creation of new pathophysiological models. Such new interpretative models are needed in the context of high rates of treatment resistance and partial responsiveness reported in patients with MDD^{(19,20,23-25)}.

2. Materials and method

This case-control study was conducted at the Psychiatric Department of the “Carol Davila” Central University Military Emergency Hospital, Bucharest, Romania. The study focused on exploring the relationship between the first episode of major depression (FEMD) and contributing factors such as negative life events.

Between February 2022 and January 2024, a total of 93 cases and 93 controls were enrolled in the study. The “cases” group consisted of individuals seeking treatment for severe FEMD and treated with selective serotonin reuptake inhibitors (SSRIs). The “controls” were selected from the general population through rigorous screening to ensure no history of mental illness. Both groups were matched by sex and age, adhering to an age difference criterion of ±2 years.

The inclusion criteria were represented, for cases, by the presence of a first severe depressive episode, which implies the absence of prior depressive episodes during their lifetime. The diagnosis was confirmed through the administration of the Structured Clinical Interview for DSM IV Disorders (SCID-I) and a Hamilton Depression Rating Scale (HAM-D17) score of 24 or higher. For controls, a HAM-D17 score of ≤7 was required, along with the absence of psychiatric disorders in personal history and no Axis I diagnosis at the time of the interview.

Regarding the assessment tools and their administration, the HAM-D17 was employed to measure the severity of depressive symptoms. Life events were elicited through unstructured interviews, where participants were asked to detail significant events in their lives occurring in the past six months. These interviews were designed to capture the type and number of negative life events experienced.

For data analysis, R Software (version 4.3.2) was used. All statistical analyses were performed using R software, version 4.3.2, ensuring robust and replicable statistical procedures. Statistical analyses focused on the impact of life events on HAM-D17 scores were performed using Analysis of Variance (ANOVA) and post hoc tests.

The impact of life events on baseline HAMD-17 scores in the cases group was analyzed using ANOVA and linear regression approaches. An extended linear regression analysis was conducted to examine the relationship between baseline HAMD-17 scores and the number of life events among cases, while controlling for additional demographic and socioeconomic predictors.

3. Results

The age distribution of participants in both cases and control groups was analyzed to ensure proper matching. For the case group, the mean age was 37.75 years old (SD=11.91), with a median age of 36 years old (range: 18-62). For the control group, the mean age was 37.49 years old (SD=11.80), with a median age of 35 years old (range: 18-62). The Mann-Whitney U test did not show significant differences regarding this variable (p=0.885), which confirmed that the groups were well matched (Figure 1).
 

The gender distribution between cases and controls did not show a statistically significant difference (Chi-square=0.000, p=1.000). Both groups had similar proportions of males and females, indicating no gender-related bias in the sample selection (Figure 2).
 

Marital status showed a statistically significant difference between cases and controls (Chi-square=10.696, p=0.013). Detailed post hoc analysis revealed that this difference was primarily driven by the widowed category (Chi-square=8, p=0.0047). The distribution of married/in a relationship, divorced/separated, and never married categories did not differ significantly between the groups (Figure 3).
 

Education levels differed significantly between cases and controls (Chi-square=18.904, p=0.0008). Significant differences were observed in the following categories: 12 years or vocational training (Chi-square=4.571, p=0.033), university (Chi-square=5.538, p=0.019), and postgraduate studies (Chi-square=8.333, p=0.0039). These findings suggest that educational background may be an important factor associated with the cases in this study (Figure 4).
 

Occupation distribution also demonstrated a statistically significant difference between cases and controls (Chi-square=8.844, p=0.031). However, post hoc analysis did not reveal significant differences within individual occupation categories (employed, unemployed, student, retired). This indicates that, while the overall occupation distribution differed, no single category was responsible for this difference (Figure 5).
 

The analysis of depression severity as reflected by HAMD-17 scores revealed that, for the case group, the mean HAMD-17 score at baseline was 29.5 (Mdn=29, SD=3.91) – Figure 6. At the six-week follow-up, the mean was 13.9 (Mdn=14, SD=6.34). The Wilcoxon signed-rank test comparing baseline and six-week HAMD-17 scores within the case group was significant (V=4371, p<0.001).
 

For the control group, the mean HAMD-17 score at baseline was 3.58 (Mdn=4, SD=2.33). The Mann-Whitney U test comparing HAMD-17 scores between cases and controls was significant (W=8649, p<0.001).

Regarding the descriptive statistics for life events, the control group had a mean number of life events of 0.30, a median (Mdn) of 0.0, and a standard deviation (SD) of 0.59. The case group had a mean number of life events of 0.68, a median of 1, and a standard deviation of 0.74.

Due to the non-normal distribution of the data, the Mann-Whitney U test was used. The Mann-Whitney U statistic was 3044, with a p-value of 0.000054. This p-value indicates that there is a significant difference in the number of life events between the control and case groups (Figure 7).
 

The impact of life events on baseline HAMD-17 scores in the cases group was analyzed using ANOVA and linear regression approaches (Figure 8). An ANOVA analysis was conducted to evaluate the impact of the number of life events on baseline HAMD-17 scores. The results were marginally non-significant, F(3, 89)=2.49, p=0.0654, suggesting that the number of life events had a trend-level effect on the initial HAMD-17 scores.
 

The breakdown of observations in each category revealed that there were 43 participants with zero life events, 39 participants with one life event, nine participants with two life events, and two participants with three life events. Although the differences in HAMD-17 scores across life event categories did not reach the conventional threshold for statistical significance (p<0.05), the observed trend suggests a potential relationship. The small number of observations in the categories with two and three life events may have contributed to the marginally non-significant result. This trend indicates that participants with different numbers of life events might exhibit varying levels of depressive symptoms at baseline. Further investigation with larger sample sizes or more refined categorization of life events is warranted to determine whether the number of life events significantly influences baseline depressive symptoms. Future studies should aim to include more participants with higher numbers of life events to provide a more robust analysis and potentially uncover statistically significant differences.

A linear regression analysis was conducted to examine the relationship between the number of life events and baseline HAMD-17 scores among cases only (Figure 9). The results indicated a significant positive relationship, with each additional life event associated with an increase of approximately 1.33 points in HAMD-17 scores (p=0.0148). The regression model suggests that life events have a significant impact on the severity of depressive symptoms at baseline among cases, with a higher number of life events contributing to increased HAMD-17 scores.
 

Diagnostic plots were generated to assess the assumptions of the linear regression model examining the relationship between the number of life events and baseline HAMD-17 scores among cases only (Figure 10).
 

The Residuals versus Fitted plot indicated that the residuals are approximately randomly distributed around zero, suggesting that the model’s predictions are unbiased. The Normal Q-Q plot revealed that the residuals are mostly normally distributed, with some deviations at the extremes. The Scale-Location plot suggested a reasonably constant variance of the residuals across fitted values, indicating homoscedasticity.

Finally, the Residuals versus Leverage plot identified a few points with higher leverage, but none appeared to unduly influence the model. Overall, these diagnostics support the linear regression model’s appropriateness for the cases dataset, with minor concerns regarding normality at the extremes.

The significant trend observed in the ANOVA analysis is supported by the linear regression results, highlighting the importance of considering life events in the assessment of depressive symptoms. Further investigation with larger sample sizes or more refined categorization of life events is warranted to determine whether the number of life events significantly influences baseline depressive symptoms.

An extended linear regression analysis was conducted to examine the relationship between baseline HAMD-17 scores and the number of life events among cases, while controlling for additional demographic and socioeconomic predictors (Table 1). The results indicated a significant positive relationship between the number of life events and HAMD-17 scores, with each additional life event associated with an increase of approximately 1.43 points in HAMD-17 scores (p=0.0144).
 

The analysis also revealed that having a high school/post-secondary education was associated with significantly lower HAMD-17 scores compared to having less than 12 years of education (Estimate = -2.69, p=0.0219). Being retired was also associated with significantly lower HAMD-17 scores compared to being a student (Estimate = -7.80, p=0.0264). Other predictors, including age, gender, marital status, and other education and occupation categories, did not have a statistically significant impact on HAMD-17 scores. These findings suggest that the number of life events and certain sociodemographic factors, such as education level and occupation, play an important role in the severity of depressive symptoms at baseline among cases.

To ensure the validity of the extended linear regression model, several diagnostic tests were conducted. The Shapiro-Wilk test indicated that the residuals are normally distributed (W=0.98505, p=0.3698), meeting the assumption of normality.

The Breusch-Pagan test for heteroscedasticity showed no significant heteroscedasticity (BP=13.829, df=13, p=0.386), indicating that the variance of residuals is constant. The Durbin-Watson test suggested no significant autocorrelation in the residuals (DW=1.7684, p=0.1348). Variance Inflation Factor (VIF) values for all predictors were below the threshold of 10, indicating no significant multicollinearity (all GVIF^(1/(2*Df)) values close to 1). These diagnostic tests support that the assumptions of linear regression are reasonably met, validating the extended model for further analysis and interpretation.

4. Conclusions

In summary, the analysis of categorical variables indicates significant differences in marital status, education levels, and occupation distributions between cases and controls. Gender distribution did not show any significant differences. These findings highlight the potential influence of marital status, education, and occupation on the studied outcomes and suggest areas for further research. The differences in the HAMD-17 scores between the groups and between the baseline and follow-up scores in the study group supported the distinct characterization of the two groups and the efficacy of the treatment in patients with FED.

A significant difference in the number of life events between the control and cases groups was reported, also confirming the difference in the potential significance of stressors for the onset of FED. The analysis of the impact of life events on baseline HAMD-17 scores in the cases group suggested that the number of negative events had a trend-level effect on the initial HAMD-17 scores. This trend indicates that participants with different numbers of life events might exhibit varying levels of depressive symptoms at baseline. The linear regression analysis conducted to examine the relationship between the number of life events and baseline HAMD-17 scores among cases revealed a significant positive relationship, with each additional life event being associated with an increase of approximately 1.33 points in HAMD-17 scores. Also, the extended linear regression that controlled additional demographic and socioeconomic predictors supported the existence of a significant positive relationship between the number of life events and HAMD-17 scores, with each additional life event associated with an increase of approximately 1.43 points in HAMD-17 scores.

The current analysis also revealed that having a high school/post-secondary education was associated with significantly lower HAMD-17 scores compared to having less than 12 years of education. Being retired was also associated with significantly lower HAMD-17 scores compared to being a student. These findings suggest that the number of life events and certain sociodemographic factors, such as education level and occupation, play an important role in the severity of depressive symptoms at baseline among cases.  

Institutional Review Board Statement: The study was conducted in accordance with the Declaration of Helsinki. The “Carol Davila” Central University Military Emergency Hospital Ethics Committee granted ethical approval (number 549/10.02.2022).

Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.

 

Corresponding author: Andrei-Gabriel Mangalagiu E-mail: andrei.mangalagiu@drd.umfcd.ro

Conflict of interests: none declared.

FINANCIAL SUPPORT: none declared.

This work is permanently accessible online free of charge and published under the CC-BY.