Intermediary roles of prospective memory and retrospective memory in the comorbidity of depression and pain

Participants

This cross-sectional cohort study was conducted at Anhui Mental Health Center (AMHC) from January 2019 to January 2022. The medical ethics committee of AMHC approved the study. In accordance with the Helsinki declaration, all participants provided written consent before participating in this study. Two professional psychiatrists used the Chinese version of the Mini International Neuropsychiatric Interview (C-MINI) 5.0.0 to evaluate participants. A total of 218 participants were initially screened: 156 depressed inpatients and outpatients were screened from AMHC and 62 healthy individuals were screened from the hospital’s physical examination centre. Among the depressed inpatients/outpatients, 16 failed to complete the assessment, 15 declined to provide written informed consent, 25 did not meet the study criteria and 18 withdrew consent. Meanwhile, among healthy controls, 3 failed to complete the assessment, 13 declined to provide written informed consent, 2 did not meet the study criteria and 2 withdrew consent. Ultimately, 124 participants were included and divided into the following groups: a comorbidity group (patients with MDD and CP, n=40), a depression group (patients with depression without CP, n=42) and a control group (healthy controls, n=42) (figure 1).

Request permission

Inclusion criteria for patients in the comorbidity group were as follows: (1) met the Diagnostic and Statistical Manual of Mental Disorders, fifth edition (DSM-5) criteria for MDD and the CP standards defined by the International Association for the Study of Pain, as assessed by two independent senior psychiatrists; (2) aged between 18 and 65 years; (3) chronic pain was not due to traumatic physical injury, serious somatic disease or inflammatory disease; (4) pain intensity numerical rating scale (PI-NRS) ≥3. The inclusion criteria for the depression group were as follows: (1) met the DSM-5 criteria for MDD as assessed by two independent senior psychiatrists; (2) aged between 18 and 65 years; (3) PI-NRS <3. The inclusion criteria for the control group were as follows: healthy people screened from the physical examination centre of AMHC. Exclusion criteria for all participants were as follows: (1) history of severe craniocerebral trauma or somatic trauma; (2) history of severe tumour-related, neurological or inflammatory diseases; (3) history of substance use or other mental disorders; (4) history of serious physical disease; (5) pregnant or lactating women; (6) administration of modified electroconvulsive therapy within three months before enrolment.

Assessment instruments

Statistical analysis

SPSS V.22.0 was used for statistical analyses. Values conforming to a normal distribution were presented as mean and standard deviation (SD), otherwise, as median and interquartile spacing (IQR: 25%, 75% percentiles). Student’s t-tests, χ² test and ANOVA were performed to compare differences in continuous or categorical parameters among the three groups. We used one-way ANOVA to evaluate the symptoms among the three groups and the Least-Significant Difference (LSD) (homogeneous variances) method for intergroup comparison. As for non-homogeneous variables, we used the Games-Howell test to estimate the intergroup comparison. We adopted Spearman correlation to assess the correlations among cognitive, memory, clinical symptoms, symptom intensity and pain perception. Stepwise regression analysis demonstrated that the independent determinants of pain-related factors were the main influential factors for patients with comorbidity. Finally, mediation analysis was performed to test further whether the relationship between pain and depressive symptoms was mediated by PM and RM. This analysis was performed by quantifying the direct and indirect relationships among the independent variable (pain), mediators (PM and RM) and dependent variables (symptoms of depression). In the mediation models, all paths were reported as unstandardised ordinary least-squares regression coefficients. Significance analysis was based on 5 000 bootstrap realisations, and a significant indirect effect was indicated when the bootstrap 95% confidence interval (CI) was not zero. All tests were two-tailed, and differences were considered statistically significant at a p value of <0.05.

Demographic, depression psychometric properties and pain intensity of the comorbidity, depression and control groups

The demographic data, psychometric properties of depression and pain intensity of the three groups are shown in table 1. A total of 124 individuals were included in the final analyses. We recruited 40 patients with comorbid MDD and CP (27 males and 13 females), 42 patients with MDD but without CP (24 males and 18 females) and 42 healthy controls (18 males and 24 females). No remarkable differences in age, years of education, employment status, family patterns, marital status, income level and the number of outpatients/inpatients were found among the groups (p>0.05). The clinical properties of depression were comprehensively assessed by a self-rated BDI-II scale and a psychiatrist-evaluated HAMD scale. Scores from the two scales in the comorbidity group were significantly higher than those in the depression group, regardless of whether the depression was self-evaluated by patients or evaluated by a psychiatrist (t=2.519, p=0.014; t=2.378, p=0.020). Pain intensity was also found to be more severe in the comorbidity group than in the depression group (t=9.343; p<0.001).

Characteristics of cognition, prospective memory, retrospective memory and pain-related features in the comorbidity, depression and control groups

The MoCA-BC and PRMQ were used to evaluate the multidimensional characteristics of cognition and memory. Results for the seven factors of the MoCA-BC (visuospatial/executive functions, naming, attention, language, abstraction, delayed recall and orientation) can be seen in table 1. The delayed recall score varied among the groups (F=5.764, p=0.004), and the depression and comorbidity groups were more seriously impaired than the control group (p=0.016 and p=0.002, respectively). Meanwhile, the total MoCA-BC score differed among the three groups (F=3.486, p=0.034), and the impairment of the comorbidity group was more severe than the control group (p=0.011).

Furthermore, we used the PRMQ to evaluate memory functioning. Two components of the PRMQ (PM and RM) were evaluated, and distinguishable divergence was observed in the three groups (F=7.221, p<0.001; F=7.408, p<0.001). We also found that impairment of PM and RM in the comorbidity group was more severe than in the depression group (p=0.002 and p=0.014, respectively) and the control group (p<0.001 and p<0.001, respectively). Finally, we used the SF-MPQ-2-CN to assess pain-related features in the three groups, including continuous, intermittent, and neuropathic pain, and then examined affective descriptors. All the components of the SF-MPQ-2-CN differed among the groups (F=32.186, p<0.001; F=20.268, p<0.001; F=27.394, p<0.001; F=55.836, p<0.001). The comorbidity group had significantly higher scores in all SF-MPQ-2-CN factors than the depression and control groups (all p<0.05).

Factors relevant to impaired cognition and memory with depressive symptoms and pain-related features

To address potential risk factors for impairments of cognition and memory with depressive symptoms and CP, we performed correlation analysis to determine whether the factors with impairments of PM and RM are associated with MDD and CP. The total score of the patient self-rated BDI-II scale was positively correlated with neuropathic pain, affective descriptors and the total SF-MPQ-2-CN score (r=0.317, p=0.004; r=0.565, p<0.001; r=0.466, p<0.001). The total HAMD score was positively correlated with neuropathic pain, affective descriptors and the total SF-MPQ-2-CN score (r=0.262, p=0.018; r=0.249, p=0.024; r=0.269, p=0.015). We discovered that prospective memory was positively associated with continuous pain, neuropathic pain and the total SF-MPQ-2-CN score (r=0.431, p<0.001; r=0.253, p=0.022; r=0.320, p=0.003). Retrospective memory was positively correlated with continuous pain, neuropathic pain and the total SF-MPQ-2-CN score (r=0.415, p<0.001; r=0.247, p=0.025; r=0.314, p=0.004). In addition, we found that impairment in PM was positively correlated with the total BDI-II score and the total HAMD score (r=0.293, p=0.007; r=0.234, p=0.034). Impairment of RM was positively correlated with the total BDI-II score (r=0.355, p<0.001). We also found that marital status was negatively correlated with continuous pain and neuropathic pain (r=−0.233, p=0.035; r=−0.275, p=0.012). The results are shown in table 2 and online supplemental table 1.

Independent predictors of chronic pain and depression symptoms

We performed stepwise regression analyses of sociodemographic characteristics, depression symptoms and memory impairment to predict pain-related factors. As shown in table 3, the final model from forward regression indicated that remarkable explanatory variables accounted for 28.8% (R²=0.288) of the variance in the total SF-MPQ-2-CN scores. A significant positive effect was found between the total SF-MPQ-2-CN score and total BDI-II score (β=0.397, t=4.009, p<0.001), and a negative effect on marital status was found as well (β=−0.297, t=−2.976, p=0.004). Furthermore, the regression model accounted for 21.8% (R²=0.218) of the variation in neuropathic pain. A significant positive effect was found between total neuropathic pain and the total BDI-II score (β=0.300, t=2.889, p=0.005), and a negative effect on marital status was also found (β=−0.312, t=−2.977, p=0.004). The regression model accounted for 35.3% of the variation in affective descriptors. A significant positive relationship was found between affective descriptors and the total BDI-II score (β=0.594, t=6.600, p<0.001). Finally, an examination of the mediator pathways revealed the indirect role of PM and RM in CP. Bootstrapping results indicated that the indirect effect was significant (indirect effect: 0.144 and 0.186, respectively) and the 95% CI had no zero values, confirming the significant mediating effect of PM and RM for CP prediction in patients with depression (figure 2).

Request permission

Main findings

The objective of this study was to investigate the predictive role of self-reported PM and RM in patients with comorbid MDD and CP. The results verified our hypotheses that patients with comorbid MDD and CP suffer more severe clinical symptoms than patients with MDD without CP and also have more obvious PM and RM impairments. Moreover, we found that deficiencies in PM and RM were positively associated with the severity of MDD, continuous pain, neuropathic pain and the total SF-MPQ-2-CN score. Furthermore, we found support for our hypothesis that PM and RM have mediating effects on the association between depression and pain outcomes. The results showed that impairments in PM and RM could be potential prognostic factors for comorbid MDD and CP, which should be considered in treatment and intervention.

Evidence of comorbid MDD and CP indicates that high levels of chronic pain are associated with high levels of severity in depressive symptoms and lack of remission from depressive disorders.24 Some studies have shown that this comorbid state can be stressful, lead to catastrophising, rumination and helplessness, and have deleterious effects on daily activities and workplace productivity.25 Patients with comorbid MDD and CP had more severe pain-related and MDD symptoms than patients in the depression and control groups. Currently, studies have focused on identifying predictive factors for the risk of recurrence, course of the disease, residual symptoms or therapeutic resistance. Several predisposing factors may be associated, and one or more may be present in patients with comorbidity. Unmarried patients or patients with low education had lower tolerance to CP. Moreover, developmental deprivation and early loss of a parent are vulnerabilities for MDD or CP.26 In these psychologically or physically distressed circumstances, persistent pain interacting with MDD can be one of the mechanisms that may cause comorbidity symptoms.

Patients with CP display altered brain structures and functions,27 and many of the altered brain regions and networks are involved in pain perception, attitude and beliefs. Cognitive and emotional factors have a surprisingly significant effect on the disease course.28 Consistently, we found that the total MoCA-BC scores and delayed memory in the comorbid group were significantly impaired compared with those in the healthy control group, especially the ability of delayed recall. Meanwhile, deficient cognitive function seemed to have a positive relationship with neuropathic pain. Moreover, the patients in the comorbid group suffered a higher degree of pain intensity than those in the other two groups and also suffered more severe pain-related clinical symptoms, such as persistent pain, intermittent pain, neuropathic pain and difficulties with emotional description. The severity of depression exhibited a positive correlation with pain intensity, neuropathic pain and difficulties with affective descriptors. These outcomes further confirmed the complex mechanism underlying psychological pain modulation, which incorporates emotional and cognitive factors and may alter how an individual feels pain. Thus, insights into the interactive mechanism of pain perception modulation and emotional states with comorbid depression and chronic pain are relevant.

The information-processing mechanisms indicated that patients with chronic pain recalled more pain words than the controls at immediate and delayed recall, suggesting a memory bias in chronic pain. The nodes of interaction between depression and chronic pain accumulate memories of pain experiences, even their sensory attributes, over time, as well as emotions associated with them and behavioural responses.29 PM refers to the ability to remember to carry out actions at some future point in time and is an important aspect of day-to-day memory function. By contrast, RM reflects how much information is retained from a prior episode.30 The successful implementation of a PM task entails taking action when correct circumstances are encountered (prospective component) and remembering the correct action to take (retrospective component). Multiple cognitive abilities associated with PM and RM, including attention, maintenance, working memory, executive functions, information processing speed and visuospatial ability, have high integrity.31 In this present study, the results showed that the impairments of PM and RM in the comorbid group with depression and chronic pain were more severe than those in the other two groups, and impaired PM and RM were positively correlated with persistent pain, pain intensity and neuropathic pain. These results indicate that psychological cognitive processes, such as memory impairment, are involved in the clinical symptoms of depression and chronic pain, which is consistent with previous studies.32 Moreover, deficits in PM and RM potentially exerted an indirect mediating effect on patients with the comorbidity. These findings provide new insights into memory dysfunction in people with chronic pain.

Limitations

The results of the current study should be interpreted with caution because of several of its limitations. First, the intensity and features of CP were evaluated through self-assessment; thus, in future studies, we should use laboratory-based pain-measuring instruments to evaluate the characteristics of CP. Furthermore, the study had a moderate sample size, which may have limited statistical power for detecting small effects. Further studies with larger sample sizes are needed to show the impact of PM and RM impairments in comorbid MDD and CP. In addition, this study had a cross-section design, and thus we were unable to determine the causal relationship between potential risk predictors and comorbid MDD and CP. Thus, longitudinal research on this topic is needed.

Implications

In summary, our findings indicated that the severe impairments of PM and RM may cause complex symptoms in patients with comorbid MDD and CP. However, more studies are needed to explore the long-term effects of the impairments of memory processes and other cognitive functions in this vulnerable population. Determining whether alleviating deficiencies in PM and RM can reduce the burden of pain and depression for pain management is crucial.