Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Feb 1.
Published in final edited form as: Int J Ment Health Nurs. 2020 Sep 24;30(1):334–345. doi: 10.1111/inm.12789

Cognitive reactivity among high-risk individuals at the first and recurrent episode of depression symptomology: A structural equation modelling analysis

Fei Fei Huang 1, Wei-ti Chen 2, Yu An Lin 1, Yu Ting Hong 1, Bin Chen 3
PMCID: PMC7855259  NIHMSID: NIHMS1641704  PMID: 32969568

Abstract

Understanding the particular mechanisms by which vulnerability and capability factors influence cognitive reactivity (CR) can contribute to an enhanced capacity to adequately react to depression. However, few studies have explored the CR model. The main aim of the present study was to develop a model that specifies the predictive effects of CR for depression among individuals at high risk for first-episode and recurrent depression. A national cross-sectional, online study using convenience sampling was conducted among 587 vulnerable healthy individuals and 224 depressed patients in China. A battery of indices, including measures of CR, social support, resilience, self-compassion, life events, neuroticism, sleep condition, and negative emotion, were collected. A structural equation model was applied to analyse the data. The final first-episode and recurrent depressive symptoms of the CR models showed good model fit. According to the models, 45%–52% of the variance in depressive symptom was predicted by CR. Social support, self-compassion, resilience, and positive life events directly influenced CR, with β values ranging from −0.18 to −0.24 (P < 0.01). Neuroticism, negative emotion, poor sleep conditions, and negative life events also directly and positively influenced CR (P < 0.01). The relationship between these negative or positive contributing factors and depression was also indirectly influenced by CR (P < 0.01). Our findings demonstrate the role of CR in the prevention and treatment of depression. The first-episode and recurrent depressive symptoms of the CR models considering both vulnerabilities and capabilities of CR in the psychopathology of depression provide a theoretical basis for interventions that reduce CR in high-risk populations.

Keywords: cognitive reactivity, depression, high risk, structural equation model, survey

INTRODUCTION

Depression is a common, costly, and disabling global mental health disorder (Smith 2014; World Health Organization 2017). Globally, more than 322 million people of all ages suffer from depression − 4.4% of the world’s population (Friedrich 2017). As reported by the institute for health metrics and evaluation, the prevalence of depression in China is higher than that of global but lower than some developed countries such as the United States and England (Ren et al. 2020). The population of China accounts for 18.4% of the global population, while nearly 37.9% of the Chinese population experience depressive symptoms, 3.5% experience lifetime major depression, and almost 10%−24% present subthreshold depressive symptoms (Gu et al. 2013; Qin et al. 2018), which significantly impairs individuals’ quality of life (Yang et al. 2018). Although the aetiology of depression is complex, over the past decades, there has been evidence that cognitive vulnerability to depression plays an important role in the aetiology of depression (Tehranchi et al. 2018). The ease with which such negative cognitions can be (re-)activated by sad mood states was labelled cognitive reactivity (CR) (Van der Does 2002). Understanding the particular mechanisms by which vulnerability and capability factors are related to CR can contribute to an enhanced capacity to adequately react to depression (Wang et al. 2016). However, previous studies focused on the interrelationships between CR and other psychological indicators that were limited to a single predictor or a single outcome variable (Barnhofer & Chittka 2010; Struijs et al. 2013; Tehranchi et al. 2018). The comprehensive direct or indirect relationships among vulnerability and capability factors, CR, and depression are still unknown. Therefore, we developed and tested first-episode and recurrent depressive symptoms of the CR models that consider depression, positive and negative contributing factors to CR, and individual sociodemographic characteristics.

BACKGROUND

Depression accounts for as much as 9.7% of the suicides in China (World Health Organization 2019) and significantly impairs individuals’ quality of life (Yang et al. 2018). Depressive episodes are chronic and recurrent (rates range from 60% to 90%) (Figueroa et al. 2018), and rates of depression are increasing, with 72.3% of subjects not even aware of their depression problems (Huang et al. 2019). Thus, understanding the aetiology of depression remains a clinical research priority.

As stated by Teasdale’s (1998) differential activation hypothesis (DAH), negative cognitions that remain latent in some individuals, even when depressive symptoms are in remission, maybe reactivated by life events, stress, or negative mood states (Teasdale 1988). There has increasing evidence on the role of CR in the development, maintenance, and relapse/recurrence of depressive symptoms or clinical depression (Figueroa et al. 2018; Struijs et al. 2013). For example, Kruijt et al. (2013) found that CR predicts the first onset of depression in never-depressed individuals (Kruijt et al. 2013), and Figueroa et al. (2015) also reported that CR predicted recurrence during a 3.5-year period in remitted patients (Figueroa et al. 2015). However, Tehranchi et al. (2018) found that CR has no direct influence on depression and displays indirect effects on depression through the mediation of negative affect (Tehranchi et al. 2018). Wojnarowski et al.’s (2019) meta-analysis indicated that CR was not a statistically significant predictor of relapse and/or recurrence after cognitive behavioural therapy for depression (Wojnarowski et al. 2019). It seems that the potential mechanisms accounting for the correlation between CR and depression remain mixed and need to be further explored.

Furthermore, understanding the particular mechanisms by which vulnerability and capability factors are related to CR can contribute to an enhanced capacity to adequately react to depression (Wang et al. 2016). Regarding vulnerability factors, Barnhofer and Chittka (2010) reported that neuroticism predisposes individuals to depression by generally increasing their CR levels (Barnhofer & Chittka 2010). Tehranchi et al. (2018) found that CR is positively related to negative affect, the relationship between CR and depression is also mediated by negative affect, and CR mediates the relationship between character strengths and depression (Tehranchi et al. 2018). Struijs et al. (2013) indicated that negative life events interact with CR to predict depressive symptoms (Struijs et al. 2013). For the factor associated with reduced CR, it was found that CR is negatively associated with resilience (Cladder-Micus et al. 2018; Huang et al. 2020), self-compassion, and social support (Huang et al. 2020). Besides, the levels of CR could also be affected by sociodemographic variables, such as body mass index (BMI) (Paans et al. 2016), depression status, and episode numbers (Elgersma et al. 2015; Jarrett et al. 2013).

Regarding the studies mentioned above, we found some interesting phenomena. First, the majority of studies were conducted in Western countries, and relatively few studies were conducted in Asian countries, especially in China, which has a high disease burden of depression. Second, researchers have mainly concentrated on negative psychological outcomes. A limited number of studies have been developed to examine the association between CR and positive psychological resources (e.g. social support and resilience), which are considered protective factors for preventing depression (Liang et al. 2019). Third, the interrelationships between CR and other psychological indicators were limited to a single predictor or a single outcome variable. The comprehensive direct or indirect relationships among vulnerability and capability factors, CR, and depression are still unknown. Therefore, it is imperative to identify pathways through which factors exert their effects on CR, which can then be applied to design integrated interventions to reduce CR among individuals vulnerable to depression.

To reduce the high disease burden of depression globally, especially in China, we need to concerns the individuals vulnerable to depression, including the first-episode and recurrent depression, and strictly perform the tertiary management policy of depression, that is, early detection, early diagnosis, and early treatment (Huang et al. 2020). Therefore, based on Teasdale’s (1998) DAH and literature reviews, we developed and tested first-episode and recurrent depressive symptoms of the CR models (Fig. 1) that consider depression, positive and negative contributing factors to CR, and individual sociodemographic characteristics. Both models suggest the following hypotheses:

  1. Social support, resilience, self-compassion, and positive life events have a direct negative relationship with CR and indirectly influence depression through CR.

  2. Neuroticism, poor sleep conditions, negative life events, and negative emotion have a direct positive relationship with CR and indirectly influence depression through CR.

  3. Social support, resilience, self-compassion, neuroticism, poor sleep conditions, both negative and positive life events, and CR have a direct positive relationship with depression.

FIG. 1.

FIG. 1

Open in a new tab

Hypothesized model of depression cognitive reactivity model.

METHOD

Study design

A cross-sectional study was conducted from January 2018 through January 2019. The study proposal adhered to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement (Vandenbroucke et al. 2014) and was approved by the ethical committee of the University. The participants read and signed the written informed consent form on the platform before they completed the questionnaire in approximately 20–30 min.

Participants

A total of 2330 healthy individuals and clinical patients participated in the study. Healthy individuals (n = 2000; aged 18–65 years) who did not meet criteria for any psychiatric disorder according to the Diagnostic and Statistical Manual of Mental Disorders, 5th edition (DSM-V) were recruited by nationwide convenience sampling through an online survey on Wenjuanxing (http://www.wjx.cn). Depressed patients in remission (n = 330) who had experienced more than two depressive episodes and had been in stable remission for at least eight weeks were recruited from the outpatient department of the Fourth Hospital (Fuzhou neuropsychiatric hospital), Fuzhou, China. Participants with current under treatment of depression were excluded from both groups. The detailed recruitment process and inclusion and exclusion criteria of participants were described in a previously published manuscript (Huang et al. 2019).

Furthermore, we included only depression-vulnerable individuals in this study; that is, healthy individuals who had a total score of the modified Chinese version of the Leiden Index of Depression Sensitivity (LEIDS-RR-CV) ≥60 were considered individuals vulnerable to first-episode depression. In addition, the depressed patients in remission who had a total score on the LEIDS-RR-CV ≥ 55 were considered individuals vulnerable to the recurrence of depression. Finally, a total of 587 vulnerable healthy individuals and 224 vulnerable depressed patients were included in the final analysis.

Typically, a median sample size of 200 is recommended for SEM analysis (Kline 2011; Wu 2009). An a priori sample size calculator for SEM was also applied (Soper 2015), which is popular and public especially designed for calculating sample size of SEM (https://www.danielsoper.com/statcalc/calculator. aspx?id=89). The minimum sample size with a moderate effect (0.3), at a power value of 0.95, including 5 latent and 24 observed variables (all observed indicators and sociodemographic variables), and with a value of 0.05 was calculated as 223. Hence, we chose a minimum sample size of 223.

Measures

A series of questionnaires that aimed to measure the constructs depicted in the hypothesized model, including five latent variables (CR, social support, resilience, self-compassion, and depression) and four observed variables (positive and negative life events, neuroticism, poor sleep conditions), were administered. Additionally, sociodemographic characteristics of the participants were also collected.

Cognitive reactivity (CR)- was assessed by the modified Chinese version of the Leiden Index of Depression Sensitivity (LEIDS-RR-CV), which is a 26-item self-report measure of cognitive reactivity to sad moods (Huang et al. 2019). The participants were asked to imagine the last time they felt a mild state of dysphoria and then to indicate the degree to which a list of statements described their typical cognitions and behaviours in response to a sad mood. The LEIDS-RR-CV contains 26 items with five subscales. All of the items were rated using a 5-point Likert-scale (0 = not at all to 4 = very strongly). A higher total score indicated stronger CR. In this study, Cronbach’s α coefficient was 0.95 for the overall scale. Huang et al. (2019) identified a cut-off score of 60 for the LEIDS-RR-CV to screen for healthy individuals at risk for depression in China (Huang et al. 2019).

Depression-

The Chinese version of the 20-item Center for Epidemiological Studies Depression Scale (CES-D) (Zhang et al. 2010) has been used to screen for depressive symptoms in general and clinical populations (Radloff 1977), and it has been commonly used in clinical assessments (Yang et al. 2018). The participants were asked to indicate how frequently they experienced depressive feelings and behaviours during the past week on a four-point scale, ranging from 0 (rarely or none of the time) to 3 (almost or all of the time). In this study, Cronbach’s α coefficient was 0.96.

Social support-

The Multidimensional Scale of Perceived Social Support (MSPSS) is a 12-item self-report scale used to measure perceived social support from family, friends, and significant others (Jiang 2001; Pedersen et al. 2009). The scale employed a 7-point rating scale ranging from 1 (very strongly disagree) to 7 (very strongly agree). In this study, Cronbach’s α coefficient was 0.97 for the overall scale.

Neuroticism-

The Neuroticism Subscale of the Chinese Big Five Personality Inventory (NEO-CBF-PI), which is the most comprehensive self-report questionnaire measuring the five dimensions of personality, including neuroticism. The CBF-PI consists of 40 items and has been extensively validated (Wang et al. 2011). In this study, Cronbach’s α coefficient was 0.87 for the NEO-CBF-PI.

Resilience-

The Chinese version of the 14-item Resilience Scale (RS-14) developed by Wagnild and Young is one of the most reliable tools in measuring resilience in various age groups and different conditions (Tian & Hong 2013; Wagnild & Young 1993). It is composed of 14 items representing the ‘Personal Competence Factor’ and ‘Acceptance of Self and Life Factor’. Each item was graded from 1 (strongly disagree) to 7 (strongly agree). Cronbach’s α for the RS-14 was 0.96 in this study.

Self-compassion-

The 26-item Self-Compassion Scale (SCS) has been the most commonly used scale to measure self-compassion at times of perceived difficulty (Neff 2003). In this study, Cronbach’s α of the scale was 0.77.

Life events-

The 48-item Life Events Scale (LES) has been used to evaluate negative and positive life events that had occurred during the previous year or longer, including family, work or study, and social-related events (Mo et al. 2019). Each of the 48 life event items was anchored to four questions: (i) when it happened, measured by `never’, ‘in the past 1 month’, and ‘in the past 1 year or longer’; (ii) whether it was positive or negative for the target person; (iii) the impact on the target person’s mental health, measured by a 5-point scale ranging from ‘no impact’ to ‘very severe impact’; and (iv) the duration of the event, measured by a 4-point scale ranging from ‘3 months’, ‘6 months’, ‘≤1 year’ to ‘longer’. The intensity of each life event was calculated by the impact multiplied by the duration and then by the timeframe in which it occurred. The total intensities of positive and negative life events were summed by the score of each positive or negative life event. In this study, Cronbach’s α of the scale was 0.94.

Sociodemographic characteristics-

The participants were asked about their residential area and location, age, sex, marital status, educational level, religion, monthly household income (yuan, RMB), employment status, smoking status, living status, BMI, family history of mental illness, previously experienced depression, frequency of sad mood in the past month, and sleep quality.

Statistical analysis

Data analyses were conducted using SPSS 24.0 and AMOS 23.0 (IBM, Chicago, IL, USA). In this study, the data met the assumptions of normality (one-sample Kolmogorov–Smirnov test was not statistically significant). The continuous variables were expressed as the mean and standard deviation (SD). Categorical variables were expressed as proportions or percentages.

We took three steps to test the proposed models. First, Pearson correlation analysis was conducted to examine the relationships between all variables. Second, SEM was used to test the fit of the hypothesized models (Fig. 1) with the maximum likelihood (ML) method, using the following model fit indices (Kline 2011): normed chi-square (χ2/df, 1.0 ± 3.0, P > 0.05), root mean squared error of approximation (RMSEA < 0.08), comparative fit index (CFI > 0.9), normed fit index (NFI > 0.9), and goodness of fit index (GFI > 0.9). The bootstrap method (repeated 1000 times) was applied in the SEM to obtain more stable and valid standard errors of the estimates. The standardized regression coefficient (β) and P value for β of the direct, indirect, and total effects were identified and reported by path analysis. Finally, post hoc model modifications were performed to obtain better fitting models by using the critical ratios (if value <±1.96) and modification indices (MI) (Tehranchi et al. 2018). Furthermore, the Akaike information criterion (AIC) was applied to compare the modified models with the initial model. Missing data were replaced using full information ML, and P < 0.05 was considered significant.

RESULTS

A total of 811 individuals participated in the study, including 587 healthy individuals who were at high risk for first-episode depression and 224 depressed patients in remission who had a high risk for recurrent depression. Table 1 shows the sociodemographic characteristics of all participants.

TABLE 1.

Sociodemographic characteristics of participants

Variables RFD (n = 587) RRD (n = 224)
n% n%
Age (years, mean ± SD) 31.13 ± 13.96 33.41 ± 11.43
BMI (kg/m2, mean ± SD) 23.19 ± 6.15 22.23 ± 6.72
Resident area
 Northeast area 75 (12.80) 28 (12.40)
 Eastern area 232 (39.60) 80 (35.70)
 North area 79 (13.40) 33 (14.80)
 South central area 77 (13.20) 31 (13.80)
 Southwest area 63 (10.70) 24 (10.68)
 Northwest area 60 (10.20) 28 (12.62)
Sex
 Male 181 (30.90) 82 (36.50)
 Female 406 (69.10) 142 (63.50)
Resident location
 Urban 355 (60.40) 142 (63.50)
 Suburban 62 (10.60) 21 (9.40)
 Rural 170 (29.00) 61 (27.10)
Religion
 None 471 (80.20) 177 (78.80)
 Had 116 (19.80) 47 (21.20)
Education level
 Less than high school degree 29 (5.10) 18 (8.10)
 High school degree (including technical training) 77 (13.10) 65 (29.10)
 Bachelor’s degree or higher 481 (81.90) 141 (62.80)
Monthly household income (yuan, RMB)
 <1000 22 (3.80) 9 (4.10)
 1000–2999 146 (24.80) 28 (12.30)
 3000–4999 200 (34.10) 74 (32.90)
 5000– 219 (37.30) 114 (50.70)
Employment status
 Students 136 (23.10) 70 (31.40)
 Full-time employment 379 (64.50) 96 (43.00)
 Unemployment 7 (1.20) 11 (4.70)
 Farmer 5 (0.80) 11 (4.70)
 Others (e.g. retired and homemaker) 60 (10.30) 37 (16.30)
Marital status
 Married 327 (55.70) 109 (48.80)
 Unmarried 247 (42.00) 104 (46.50)
 Others (e.g. divorced and widowed) 14 (2.30) 11 (4.70)
Family history of mental illness
 No 534 (91.00) 155 (69.00)
 Unclear 44 (7.50) 40 (17.90)
 Yes 9 (1.50) 29 (13.10)
Living status
 Living by oneself 76 (12.90) 55 (24.70)
 Living with spouse 65 (11.00) 22 (9.60)
 Living with family 323 (55.00) 120 (53.40)
 Others 124 (21.10) 28 (12.30)
Smoking status
 Yes 65 (11.10) 53 (23.80)
 No 522 (88.90) 171 (76.20)
Sleep condition
 Very good 118 (20.10) 55 (24.70)
 Good 191 (32.50) 47 (21.20)
 General 211 (35.90) 84 (37.60)
 Bad 59 (10.10) 26 (11.80)
 Very bad 8 (1.40) 11 (4.70)
Open in a new tab

BMI, body mass index; VFD, vulnerable individuals for first episode of depression; VRD, vulnerable individuals for recurrent depression.

Bivariate analyses

Table 2 shows the relationships between CR and related variables among 587 healthy individuals vulnerable to depression and 224 patients vulnerable to recurrent depression.

TABLE 2.

Correlation coefficients among variables

Health vulnerable individuals (n = 587) Depression-vulnerable recurrent patients (n = 224)
Cognitive reactivity Depression Cognitive reactivity Depression
Cognitive reactivity 0.50** 0.61**
Social support −0.15** −0.39** −0.08** −0.30**
Resilience −0.06** −0.33** −0.05** −0.26**
Self-compassion −0.34** −0.47** −0.38** −0.46**
Positive life events −0.22** −0.11** −0.18** −0.12**
Neuroticism 0.47** 0.66** 0.58** 0.69**
Negative emotion 0.42** 0.43** 0.42** 0.45**
Negative life events 0.19** 0.10** 0.22** 0.19**
Poor sleep conditions 0.23** 0.26** 0.26** 0.33**
Residential area 0.07** / 0.06* /
Residential location 0.14** / 0.10** /
Age 0.12** / 0.17** /
Marital status 0.06** / 0.15** /
Educational level 0.06** / 0.17** /
Religion 0.05* / / /
Employment status 0.13** / 0.16** /
Body mass index 0.05* / 0.06** /
Family history of mental illness 0.10** / 0.16** /
Open in a new tab
*

P < 0.05;

**

P < 0.01.

Structural equation models

Controlling for the statistical significance of the sociodemographic variables related to CR, including participants’ residential area and location, age, marital status, educational level, religion, employment status, BMI, and family history of mental illness, the initial first-episode depression CR model (χ2/df = 4.81, P = 0.12; CFI = 0.89; TLI = 0.88; RMSEA = 0.08) and recurrent depression CR model (χ2/df = 5.81, P = 0.34; CFI = 0.88; TLI = 0.87; RMSEA = 0.08) were both indicated as having an acceptable model fit. To obtain better fitting models, model modification was performed.

The first-episode depressive symptoms of the CR model

Seven structural paths were found to have a critical ratio <±1.96, including the paths of (i) negative life events to CR, (ii) age to CR, (iii) BMI to CR, (iv) religion to CR, (v) resilience to depression, (vi) self-compassion to depression, and (vii) positive life events to depression, suggesting that these paths could be eliminated from the model. In addition, six covariances among the residual errors of the CR and the self-compassion, neuroticism, and poor sleep conditions were added as suggested by the MI. Compared to the initial model, the model fit indices of the final model were improved (χ2/df = 2.51, P = 0.22; CFI = 0.93; TLI = 0.92; RMSEA = 0.06). The final model was further supported by smaller AIC values (160.78 & 145.32), indicating better fit. The coefficients for all paths are shown in Figure 2.

FIG. 2.

FIG. 2

Open in a new tab

The first-episode depression cognitive reactivity model.

Table 3 summarizes the standardized direct, indirect, and total estimates of the paths in the first-episode depressive symptoms of the CR model. According to the model, social support (β = −0.24), self-compassion (β = −0.21), resilience (β = −0.18), and positive life events had negative direct effects on CR (P < 0.01 for model). However, neuroticism (β = 0.52), poor sleep conditions (β = 0.20), and negative emotion (β = 0.22) had positive direct effects on CR (P < 0.01 for model). Furthermore, social support (β = −0.13), CR (β = 0.34), neuroticism (β = 0.48), negative emotion (β = 0.21), and poor sleep conditions (β = 0.06) had direct impacts on depression. Social support, resilience, self-compassion, positive life events, negative emotion, and poor sleep conditions had small indirect influences on depression (β < 0.1, P < 0.01) and were medicated by CR. The relationship between neuroticism and depression was also mediated by CR (indirect β = 0.18).

TABLE 3.

Effect coefficients of the high risk of first and recurrent depression model

Endogenous variables Predicting variables The high risk of first depression model The high risk of recurrent depression model
Standardized direct effect β Standardized indirect effect β Standardized total effect β Standardized direct effect β Standardized indirect effect β Standardized total effect β
Cognitive reactivity Resilience −0.18* / −0.18* −0.13* / −0.13*
Self-compassion −0.21* / −0.21* −0.14* / −0.14*
Social support −0.24* / − 0.24* −0.15* −0.15*
Positive life events −0.18* / −0.18* −0.16* / 0.16*
Neuroticism 0.52* / 0.52* 0.61* / 0.61*
Poor sleep conditions 0.20* / 0.20* 0.28* / 0.28*
Negative emotion 0.22* / 0.22* 0.12* / 0.12*
Depression Cognitive reactivity 0.34* / 0.34* 0.41* / 0.41*
Resilience / −0.06* −0.06* / −0.05* −0.05*
Self-compassion / −0.07* −0.07* −0.13* −0.06* −0.18*
Social support −0.13* −0.08* −0.21* −0.06* −0.06*
Positive life events / −0.06* −0.06* / −0.02* −0.02*
Poor sleep conditions 0.06* 0.07* 0.13* 0.43* 0.25* 0.68*
Neuroticism 0.48* 0.18* 0.66* 0.15* 0.15* 0.30*
Negative emotion 0.21* 0.05* 0.26* 0.13* 0.05* 0.18*
Open in a new tab
*

P < 0.01.

Consistent with the hypothesis, 45% of the variance in depressive symptom was predicted by CR. CR was predicted by negative contributing factors (social support, resilience, self-compassion, and positive life events), with 43% of the variance in CR being accounted for by this predictor, while 32% of the variance in CR was predicted by positive contributing factors (neuroticism, negative emotion and poor sleep conditions).

The recurrent depressive symptoms of the CR model

Six structural paths were found to have a critical ratio <±1.96, including the paths of (i) negative life events to CR, (ii) area to CR, (iii) age to CR, (iv) resilience to depression, (v) self-compassion to depression, and (vi) positive life events to depression, suggesting that these paths could be eliminated from the model. As suggested by MI, six covariances among the residual errors of the CR and the self-compassion, neuroticism, and poor sleep conditions were added in the model. Compared to the initial model, the model fit indices of the final model were improved (χ2/df = 2.31, P = 0.12; CFI = 0.91; TLI = 0.90; RMSEA = 0.07), which was further supported by smaller AIC values (170.43 & 150.33). The coefficients for all paths are shown in Figure 3.

FIG. 3.

FIG. 3

Open in a new tab

The recurrent depression cognitive reactivity model.

Table 3 summarizes the standardized direct, indirect, and total estimates of the paths in the recurrent depressive symptoms of the CR model. According to the model, social support, self-compassion, resilience, and positive life events were negatively related to CR, with β ranging from −0.13 to −0.16 (P < 0.01). Neuroticism (β = 0.61), negative emotion (β = 0.12), and poor sleep conditions (β = 0.28) were positively associated with CR (P < 0.01). The relationship between these negative or positive contributing factors and depression was also found to be mediated by CR (P < 0.01).

In this model, 52% of the variance in depressive symptom was predicted by CR. CR was predicted by negative contributing factors (social support, resilience, self-compassion, and positive life events), with 25% of the variance in CR being accounted for by this predictor, while 43% of the variance in CR was predicted by positive contributing factors (neuroticism, negative emotion and poor sleep conditions).

DISCUSSION

To our knowledge, this is the first study identify and elucidate the pathways linking the positive and negative contributing factors of CR, CR, and depression among individuals at risk for first episode and relapse of depressive symptoms. The structural models further support a direct effect of CR on depression for both first-episode and recurrent depression. In addition, CR significantly mediated the associations between depression and vulnerability and capability factors. The findings suggest that health professionals should emphasize the important role of CR and its relationship to vulnerabilities and capabilities in the prevention and treatment of depression.

In our study, the hypotheses regarding the relationships among study variables were supported, except for the direct effect of positive life events, resilience, and self-compassion on depression. First, we found that higher levels of CR are associated with lower levels of social support, resilience, self-compassion, and positive life events both in the first-episode and recurrent depression CR models. As positive psychology theory states, each individual has inherent capacities for growth, well-being, and warmth. The aetiology of depression can be viewed as a lack of capabilities and strengths. The findings of the present study demonstrate the predictive effects of capacities on CR. Although there is limited evidence, these findings are also consistent with previous research demonstrating that CR is related to individuals’ self-compassion skills (Kuyken et al. 2010). It has been suggested that an individual’s internal factors (e.g. resilience and self-compassion), interpersonal factors (e.g. social support), and external factors (e.g. positive life events) may help an individual recognize and reduce any CR that they may exhibit.

On the other hand, our study also indicated that CR plays an important mediating role between these capacities and depression. Individuals with high levels of positive psychological resources or capabilities, such as resilience and self-compassion, tend to have low levels of CR and may have a decreased risk of depression. A possible explanation may be that positive psychological resources or capabilities have been considered positive adaptations to help a person hold their feelings of suffering with a sense of warmth, connection, and concern, negotiating, managing, and adapting to significant sources of stress and trauma (Shi et al. 2015). The results further suggest that for individuals with low levels of positive psychological capacities, interventions targeted at reducing CR may benefit them in preventing episodes of depression.

Second, consistent with previous research (Barnhofer & Chittka 2010; Struijs et al. 2013; Tehranchi et al. 2018), vulnerabilities such as high neuroticism, poor sleep conditions, negative life events, and negative emotion are directly related to CR, and the indirect effect of these vulnerabilities on depression via CR are also significant in individuals at high risk for first-episode and recurrent depression. As supported by the hopeless theory (Abramson et al. 1978), individuals with these depression vulnerabilities, such as higher neuroticism and negative emotion, poorer sleep quality, or more negative life events, were more likely to respond to mildly negative moods with the reactivation of thoughts relating to hopelessness (or other negative states), which are in turn related to depression. The findings provide evidence that CR constitutes a useful intervention target for individuals with these vulnerabilities.

Finally, we found that CR, social support, neuroticism, sleep condition, and negative life events have direct positive relationships with depression. The findings are in line with those of previous research (Li & Fu, 2017; Struijs et al. 2013), further highlighting the important role of CR, social support, neuroticism, sleep condition, and negative life events in the development and maintenance of depression. However, an unexpected finding was that some positive psychological capabilities, including positive life events, resilience, and self-compassion, have no direct effect on depression. This finding is consistent with that of previous research that did not find any significant relationships between character strengths and depression (Huta & Hawle 2010; Tehranchi et al. 2018). Thus, the protective factors of depression were further explored.

Limitations

Although the current study provides evidence in support of the proposed models of depression, it does have several limitations. First, the proposed models did not explain the majority of the variance in depression; that is, other variables, potentially a third class of variables, are likely influential in depression but were not considered in our models. Second, although both models showed appropriate fit, still needed to be examined by larger sample sizes. For example, less than 0.05 of the RMSEA will suggest a good fit. Third, the convenience sampling method and recruitment of nonclinical young adults may impact the generalization of the findings. Fourth, the subjectivity associated with the use of self-reported questionnaires may also pose limitations, which should be confirmed by the use of objective measurements in the future, such as the use of clinician-administered assessments. Finally, the data were cross-sectional, and causality cannot be inferred. Research with time-series and experimental methods is needed to further support the hypotheses that follow from the analyses in this paper.

CONCLUSIONS

In conclusion, the structural equation modelling analyses demonstrate an adequate fit between the proposed model and the data. Our study further presented the role of CR in the prevention and treatment of depression. The CR model provides a theoretical basis for interventions designed to reduce CR in high-risk populations. Overall, effectively establishing public health interventions should focus on improving positive psychological capabilities (e.g. social support, positive life events, resilience, and self-compassion) and reducing vulnerabilities (e.g. neuroticism, negative life events, and poor sleep conditions), which will directly or indirectly affect depression.

RELEVANCE FOR CLINICAL PRACTICE.

Considering both vulnerabilities and capabilities and their relationships to CR in the psychopathology of depression is one of the major strengths of the present study. The findings of the current study may have clinical implications for healthcare providers involved with depression prevention or care. The first-episode and recurrent depressive symptoms of the CR model could provide theoretical evidence for the development of a multifaceted intervention to reduce CR to depression for individuals at high risk for first-episode depression. For example, to prevent the relapse of individuals who have experienced depression but are now in remission, we can take targeted measures to reduce individuals’ levels of CR by enhancing their self-compassion, resilience, and social support, improving the quality of their sleep, decreasing their neuroticism, and decreasing the impact of experienced negative life events and emotions.

ACKNOWLEDGEMENTS

We would like to thank the healthy individuals and remission depression patients who participated in this study and the mental health providers who completed the assessments.

FUNDING INFORMATION

The study was funded by a grant from the Natural Science Foundation of Fujian Province, China (Grant No. 2017J05133), Startup Fund for scientific research, Fujian Medical University (Grant No. 2016QH017), and Startup Fund for High-level talents of Fujian Medical University (Grant No. XRCZX2017011).

Footnotes

Declaration of conflict of interest: We indicate that all of us had knowledge of and adherence with the Journal’s Conflict of Interest policy, and we do not have any conflict of interest in this study.

Ethics approval statement: This study was approved by the ethical committee of Fujian Medical University (NO: FMU2017024).

Fei Fei Huang,.

Wei-ti Chen,.

Yu An Lin,.

Yu Ting Hong,.

Bin Chen,.

REFERENCES

  1. Abramson LY, Seligman ME & Teasdale JD (1978). Learned helplessness in humans: Critique and reformulation. Journal of Abnormal Psychology, 87, 49–74. [PubMed] [Google Scholar]
  2. Barnhofer T & Chittka T (2010). Cognitive reactivity mediates the relationship between neuroticism and depression. Behaviour Research and Therapy, 48, 275–281. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Cladder-Micus MB, van Aalderen J, Donders ART, Spijker J, Vrijsen JN & Speckens AEM (2018). Cognitive reactivity as outcome and working mechanism of mindfulness-based cognitive therapy for recurrently depressed patients in remission. Cognition & Emotion, 32, 371–378. [DOI] [PubMed] [Google Scholar]
  4. Elgersma HJ, de Jong PJ, van Rijsbergen GD et al. (2015). Cognitive reactivity, self-depressed associations, and the recurrence of depression. Journal of Affective Disorders, 183, 300–309. [DOI] [PubMed] [Google Scholar]
  5. Figueroa CA, Mocking RJT, Mahmoud GA et al. (2018). The measurement of cognitive reactivity to sad mood in patients remitted from major depressive disorder. British Journal of Clinical Psychology, 57, 313–327. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Figueroa CA, Ruhe HG, Koeter MW et al. (2015). Cognitive reactivity versus dysfunctional cognitions and the prediction of relapse in recurrent major depressive disorder. Journal of Clinical Psychiatry, 76, e1306–e1312. [DOI] [PubMed] [Google Scholar]
  7. Friedrich MJ (2017). Depression is the leading cause of disability around the world. JAMA, 317:1517. [DOI] [PubMed] [Google Scholar]
  8. Gu L, Xie J, Long J et al. (2013). Epidemiology of major depressive disorder in mainland china: a systematic review. PLoS One, 8, e65356. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Huang FF, Li H, Lei Y, Li Q & Chen B (2019). Adaptation and validation of the Chinese version of the modified Leiden index of depression sensitivity. Journal of Affective Disorders, 256, 458–67. [DOI] [PubMed] [Google Scholar]
  10. Huang FF, Wen ZP, Li Q, Chen B & Weng WJ (2020). Factors influencing cognitive reactivity among young adults at high risk for depression in China: a cross-sectional study. BMC Public Health, 20, 703. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Huta V & Hawle L (2010). Psychological strengths and cognitive vulnerabilities: Are they two ends of the same continuum or do they have independent relationships with well-being and ill-being? Journal of Happiness Studies, 11, 71–93. [Google Scholar]
  12. Jarrett Robin B., Minhajuddin Abu, Gershenfeld Howard, Friedman Edward S. & Thase Michael E. (2013). Preventing depressive relapse and recurrence in higher-risk cognitive therapy responders: a randomized trial of continuation phase cognitive therapy, fluoxetine, or matched pill placebo. JAMA Psychiatry, 70, 1152–1160. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Jiang QJ (2001). Perceived Social Support Scale. Chinese Journal of Behavioral Medical Science, 10, 41–43. [Google Scholar]
  14. Kline RB (2011). Principles and Practice of Structural Equation Modeling. New York: Guilford Press. [Google Scholar]
  15. Kruijt AW, Antypa N, Booij L et al. (2013). Cognitive reactivity, implicit associations, and the incidence of depression: a two-year prospective study. PLoS One, 8, e70245. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Kuyken W, Watkins E, Holden E et al. (2010). How does mindfulness-based cognitive therapy work? Behaviour Research and Therapy, 48, 1105–1112. [DOI] [PubMed] [Google Scholar]
  17. Li L & Fu H (2017). China’s health care system reform: Progress and prospects. The International Journal of Health Planning and Management, 32, 240–253. [DOI] [PubMed] [Google Scholar]
  18. Liang D, Teng M & Xu D (2019). Impact of perceived social support on depression in Chinese rural-to-urban migrants: The mediating effects of loneliness and resilience. Journal of Community Psychology, 47, 1603–13. [DOI] [PubMed] [Google Scholar]
  19. Mo Q, Zhou L, He Q, Jia C & Ma Z (2019). Validating the Life Events Scale for the Elderly with proxy-based data: A case-control psychological autopsy study in rural China. Geriatrics & Gerontology International, 19, 547–51. [DOI] [PubMed] [Google Scholar]
  20. Neff KD (2003). The development and validation of a scale to measure self-compassion. Self and Identity, 2, 223–250. [Google Scholar]
  21. Paans NP, Bot M, Gibson-Smith D et al. (2016). The association between personality traits, cognitive reactivity and body mass index is dependent on depressive and/or anxiety status. Journal of Psychosomatic Research, 89, 26–31. [DOI] [PubMed] [Google Scholar]
  22. Pedersen SS, Spinder H, Erdman RA & Denollet J (2009). Poor perceived social support in implantable cardioverter defibrillator (ICD) patients and their partners: cross-validation of the multidimensional scale of perceived social support. Psychosomatics, 50, 461–467. [DOI] [PubMed] [Google Scholar]
  23. Qin X, Wang S & Hsieh CR (2018). The prevalence of depression and depressive symptoms among adults in China: Estimation based on a National Household Survey. China Economic Review, 51, 271–282. [Google Scholar]
  24. Radloff LS (1977). The CES-D scale: A self-report depression scale for research in the general population. Applied Psychologicak Measurement, 1, 385–401. [Google Scholar]
  25. Ren X, Yu S, Dong W, Yin P, Xu X & Zhou M (2020). Burden of depression in China, 1990–2017: Findings from the global burden of disease study 2017. Journal of Affective Disorder, 268, 95–101. [DOI] [PubMed] [Google Scholar]
  26. Shi M, Liu L, Wang ZY & Wang L (2015). The mediating role of resilience in the relationship between big five personality and anxiety among Chinese medical students: a cross-sectional study. PLoS One, 10, e0119916. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Smith K (2014). Mental health: a world of depression. Nature, 515, 180–181. [DOI] [PubMed] [Google Scholar]
  28. Soper DS (2015). A-priori sample size calculator for structural equation models. [Cited 20 January 2017]. Available from URL: http://www.danielsoper.com/statcalc3/calc.aspx
  29. Struijs SY, Groenewold NA, Oude Voshaar RC & de Jonge P (2013). Cognitive vulnerability differentially predicts symptom dimensions of depression. Journal of Affective Disorders, 151, 92–99. [DOI] [PubMed] [Google Scholar]
  30. Teasdale JD (1988). Cognitive vulnerability to persistent depression. Cognition and Emotion, 2, 247–274. [Google Scholar]
  31. Tehranchi A, Neshat Doost HT, Amiri S & Power MJ (2018). The role of character strengths in depression: A structural equation model. Frontiers in Psychology, 9, 1609. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Tian J & Hong JS (2013). Validation of the Chinese version of the resilience scale and its cutoff score for detecting low resilience in Chinese cancer patients. Supportive Care in Cancer, 21, 1497–502. [DOI] [PubMed] [Google Scholar]
  33. Van der Does W (2002). Cognitive reactivity to sad mood: structure and validity of a new measure. Behaviour Research and Therapy, 40, 105–120. [DOI] [PubMed] [Google Scholar]
  34. Vandenbroucke JP, von Elm E, Altman DG et al. (2014). Strengthening the Reporting of Observational Studies in Epidemiology (STROBE): explanation and elaboration. International Journal of Surgery, 12, 1500–24. [DOI] [PubMed] [Google Scholar]
  35. Wagnild GM & Young HM (1993). Development and psychometric evaluation of the Resilience Scale. Journal of Nursing Measurement, 1, 165–78. [PubMed] [Google Scholar]
  36. Wang MC, Dai XY & Yao SQ (2011). Development of the Chinese Big Five Personality Inventory (CBF-PI) III: Psychometric Properties of CBF-PI Brief Version. Chinese Journal of Clinical Psychology, 19, 454–57. [Google Scholar]
  37. Wang KC, Wang T, Meng J, Peng X & Qiu J (2016). Neural mechanisms underlying susceptibility factors in depression. Chinese Science Bulletin, 61, 654–667. [Google Scholar]
  38. Wojnarowski C, Firth N, Finegan M & Delgadillo J (2019). Predictors of depression relapse and recurrence after cognitive behavioural therapy: a systematic review and meta-analysis. Behavioural and Cognitive Psychotherapy, 47, 514–29. [DOI] [PubMed] [Google Scholar]
  39. World Health Organization. (2017). Depression. [Cited 20 January 2017]. Available from URL: https://www.who.int/news-room/fact-sheets/detail/depression
  40. World Health Organization. (2019). Suicide rate estimates, crude - Estimates by country. [Cited 09 September 2019]. Available from URL: http://apps.who.int/gho/data/node.main.MHSUICIDE
  41. Wu M (2009). Structural equation modeling—the operation and application of AMOS Chongqing. China: Chong qing University Press.. [Google Scholar]
  42. Yang X, Lau JTF & Lau MCM (2018). Predictors of remission from probable depression among Hong Kong adolescents - A large-scale longitudinal study. Journal of Affective Disorders, 29, 491–97. [DOI] [PubMed] [Google Scholar]
  43. Zhang J, Wu ZY, Fang G, Li J, Han BX & Chen ZY (2010). Development of the Chinese age norms of CES-D in urban area. Chinese Journal of Mental Health, 24, 139–43. [Google Scholar]

RESOURCES