The Relationship Between Baseline Clinical Symptom Characteristics and Working Ability in Japanese Patients Treated for Major Depressive Disorder and Painful Physical Symptoms

Introduction

Major depressive disorder (MDD) is a chronic, recurrent disease that adversely affects quality of life (QoL).^1,2 The median age of onset of mood disorders, including MDD, ranges between the late 20s and the early 40s,³ which is the primary working-age population. MDD can lead to loss of working ability, which has effects not only on individuals but also on social and economic activity,^4,5 because MDD is a major reason for long-term sick leave and permanent work disability.⁶ Vietri et al reported that approximately 20% of MDD patients complained of painful physical symptoms (PPS) without clear physical causes.⁷ Further, MDD patients with PPS had a higher burden from the perspective of healthcare resources and costs, health-related QoL, and work productivity.^7,8 Any characteristics that help identify those patients with MDD and PPS who can return to work after treatment, especially if assessed at an early stage of treatment, would be useful. However, there have been limited reports regarding the relationship between patients’ characteristics and working ability in patients with MDD,^9,10 and no reports in patients with MDD and PPS.

A 12-week, prospective, observational study was previously conducted in Japanese patients with MDD and PPS to investigate the effectiveness of duloxetine or selective serotonin reuptake inhibitor (SSRI) monotherapy for PPS, depressive symptoms, and QoL.^11,12 In a post hoc analysis of the study, a 17-item Hamilton Rating Scale for Depression (HAM-D17) score ≤6 after 12 weeks of treatment corresponded best with psychiatrists’ judgment of working ability.¹³ The objective of the current post hoc analysis was to explore the relationship, including changes over time, between baseline clinical symptom characteristics and working ability, as judged by investigators, after 12 weeks of antidepressant monotherapy in Japanese patients with MDD and PPS in a real-world clinical setting. An additional aim was to determine whether there were any differences in baseline clinical symptom characteristics between patient subgroups based on working ability in the presence or absence of previous depressive episodes before study entry.

Patients and Methods

Study Design

This was a prospective, observational study in Japanese patients with MDD and PPS treated with duloxetine or SSRIs for 12 weeks. The study was conducted at 39 psychiatric and psychosomatic outpatient/inpatient clinics/hospital sites from February 2014 to February 2016. Full study methods and results of the primary efficacy outcome analyses, preplanned subgroup analyses, and post hoc analyses have been published previously.^11–13

Patients

In brief, patients were ≥20 years old, resided in Japan, and presented with a depressive episode of MDD without psychotic traits.¹⁴ Patients diagnosed with at least moderate levels of depression (Quick Inventory of Depressive Symptomatology-Self Report ≥16) and PPS (Brief Pain Inventory-Short Form [BPI-SF], average pain ≥3) were included. The duloxetine and SSRI (escitalopram, sertraline, paroxetine, or fluvoxamine) treatment groups were pooled before being divided into 2 groups, “working ability recovered” or “working ability not recovered,” based on working ability judged by investigators after the 12-week treatment period. Patients were also divided into 4 subgroups by the presence or absence of previous depressive episodes before starting the study (ie, the patient had experienced episodes before the presenting episode or the presenting episode was the patient’s first episode) and by working ability. Patients for whom the presence or absence of previous depressive episodes was unknown were excluded from this analysis.

Outcome Measures

Outcome measures for this post hoc analysis included baseline demographics and clinical characteristics, HAM-D17¹⁵ (total and subscale [core, insomnia, anxiety, somatic]¹⁶ scores) for depressive symptoms, BPI-SF (average pain score) for PPS, Social Adaptation Self-evaluation Scale (SASS) and the Global Assessment of Functioning (GAF) for social functioning, and EuroQol-5 Dimensions questionnaire (EQ-5D) for QoL at baseline and Weeks 2, 4, 8, and 12.

Statistical Analysis

All patients were included in the analysis except those who were able to work at baseline, those with working ability data missing at baseline or after 12 weeks of treatment, or those who were excluded from the effectiveness analysis population (eg, did not meet entry criteria, case report form not retrievable, administration of study drug could not be confirmed, no post-baseline visit, no post-baseline effectiveness data).

Data were summarized descriptively as mean ± standard deviation for continuous variables and percentages for categorical variables. Statistical tests were performed using Welch’s t-test for differences in means between groups for continuous variables, a chi-square test for differences in percentages in each category between groups for categorical variables, and the Wilcoxon rank sum test for the number of MDD episodes experienced (1, 2, 3 or more; excluding “unknown”) as the ordinal scale between groups, with a two-sided significance level of 0.05. No adjustments for multiplicity were made. Statistical analyses were performed using SAS^® version 9.4 (SAS Institute Inc., Cary, NC, USA).

Results

Patient Recruitment and Baseline Characteristics

Of 596 patients enrolled, 213 patients (122 working ability recovered, 91 working ability not recovered) were included in the analysis (Figure 1). Numbers of patients with working ability recovered without or with previous depressive episodes were 66 and 36, respectively. Numbers of patients with working ability not recovered without or with previous depressive episodes were 40 and 35, respectively.

Figure 1 Patient recruitment. aMultiple reasons for exclusion from the analysis were possible.

The baseline characteristics of patients with working ability recovered were similar to those of patients with working ability not recovered, except for the presence of complications and the use of psychotherapy (Table 1). The percentage of patients with complications was significantly lower in the working ability recovered group than the working ability not recovered group (8.2% vs 24.2%, p=0.0012). All complications were reported in ≤3 patients in each group, except for hypertension (2 patients [1.6%] with working ability recovered, 5 patients [5.5%] with working ability not recovered) (Supplementary Table). The percentage of patients with working ability recovered who had used psychotherapy was significantly greater than in patients with working ability not recovered (50.8% vs 28.6%, p=0.0011).

Table 1 Demographics and Baseline Clinical Characteristics

Working Ability Recovered Group Compared with Working Ability Not Recovered Group

Depressive Symptoms

Baseline HAM-D17 total, insomnia, and anxiety scores were significantly higher (p<0.05), but the somatic score was significantly lower (p<0.05), in the working ability recovered group than in the working ability not recovered group (Supplementary Figure 1a–1e).

HAM-D17 total and all subscale (core, insomnia, anxiety, and somatic) scores at Weeks 8 and 12 were significantly lower in the working ability recovered group than in the working ability not recovered group (p<0.01 for all scores). HAM-D17 total and insomnia scores at Week 4, and core and somatic scores at Weeks 2 and 4, in the working ability recovered group were also significantly lower than in the working ability not recovered group (p<0.05 for core score at Week 2 and insomnia score at Week 4, p<0.01 for total and core scores at Week 4, and p<0.01 for somatic score at Weeks 2 and 4).

PPS

BPI-SF scores were lower in the working ability recovered group than in the working ability not recovered group throughout the treatment, with no significant difference at baseline but with significant differences at Weeks 4, 8, and 12 (p<0.05 for Week 4, p<0.01 for Weeks 8 and 12) (Supplementary Figure 1f).

Social Functioning

SASS scores were higher in the working ability recovered group than in the working ability not recovered group throughout the treatment, with no significant difference at baseline but with significant differences at Weeks 4, 8, and 12 (p<0.01 for all) (Supplementary Figure 1g).

GAF scores were higher in the working ability recovered group than in the working ability not recovered group throughout the treatment, with no significant difference at baseline but with significant differences at Weeks 2, 4, 8, and 12 (p<0.01 for all) (Supplementary Figure 1h).

QoL

EQ-5D scores were significantly higher in the working ability recovered group than in the working ability not recovered group at Weeks 4, 8, and 12 (p<0.01 for all), with no significant difference at baseline (Supplementary Figure 1i).

Working Ability Recovered Compared with Working Ability Not Recovered in Subgroups with and without Previous Depressive Episodes

Depressive Symptoms

Significant differences between working ability groups were observed in several baseline HAM-D17 subscales (Supplementary Figure 2a–2e). In patients without previous depressive episodes, insomnia score was higher in patients with working ability recovered than in patients with working ability not recovered (p<0.05). In patients with previous depressive episodes, somatic score was lower (p<0.05), but anxiety score was higher (p<0.01), in patients with working ability recovered than in patients with working ability not recovered.

At Weeks 8 and 12, HAM-D17 total and all subscale scores were lower in the working ability recovered group than in the working ability not recovered group, regardless of the presence or absence of previous depressive episodes.

PPS

BPI-SF scores did not differ at baseline or during the early part of treatment between patients with working ability recovered and patients with working ability not recovered, regardless of the presence or absence of previous depressive episodes (Supplementary Figure 2f).

Social Functioning

SASS scores were higher at Weeks 4, 8, and 12 in patients with working ability recovered than in patients with working ability not recovered, regardless of the presence or absence of previous depressive episodes (Supplementary Figure 2g).

GAF scores were consistently higher in the working ability recovered group than in the working ability not recovered group throughout the treatment, regardless of the presence or absence of previous depressive episodes, with no significant differences at baseline (Supplementary Figure 2h).

QoL

EQ-5D scores were higher at Weeks 8 and 12 in patients with working ability recovered than in patients with working ability not recovered, regardless of the presence or absence of previous depressive episodes, with no significant differences at baseline (Supplementary Figure 2i).

Discussion

In this post hoc analysis, we compared baseline clinical symptom characteristics of Japanese patients with MDD and PPS between those with or without working ability, as judged by investigators, after 12 weeks of antidepressant treatment in a real-world clinical setting. Patients who recovered working ability had fewer complications, used psychotherapy more frequently, and, unexpectedly, had higher baseline HAM-D17 total, insomnia, and anxiety scores, but lower somatic scores, than patients who did not recover working ability. Among patients without previous depressive episodes, the baseline HAM-D17 insomnia score was higher in those with working ability recovered, whereas among patients with previous depressive episodes, the anxiety score was higher, but the somatic score was lower, in those with working ability recovered.

Our comparison between the working ability recovered group and the working ability not recovered group showed that HAM-D17 total, insomnia, anxiety, and somatic scores at baseline were significantly different. However, these statistical differences may not be applicable clinically. Nevertheless, these outcome measures improved over time, and greater differences between the working ability groups were observed with longer treatment up to 12 weeks, with significant differences seen at Weeks 8 and 12. This suggests that continuous treatment and observation by healthcare professionals are important to improve depressive symptoms and to accurately judge working ability in patients with MDD and PPS.

The baseline characteristics of patients with working ability recovered were similar to those of patients with working ability not recovered, except for the presence of complications and the use of psychotherapy. No specific complications that may affect working ability between the groups were identified, although the impact of complications on working ability could not be excluded. Psychotherapy reduces sick leave in patients with common mental disorders, including depression, or musculoskeletal disorders, including low back pain and muscle pain,¹⁷ and early access to psychotherapy reduces the duration of sick leave in patients with mental disorders.^18,19 On the other hand, one study reported that psychological care combined with antidepressants did not reduce medium- or long-term sick leave.²⁰ Although the use of psychotherapy may be positively associated with working ability recovery, further research is needed. Other previous research reported that patients whose depressive symptoms clinically improved during initial treatment were more likely than nonresponders to have increased work productivity.²¹ In addition, some previous studies in patients with MDD reported that severity of depression at baseline predicted functional recovery,^9,22,23 although a systematic review including 35 controlled studies reported that baseline disease severity did not predict functional improvement.²⁴ Although these reports did not specifically target patients with MDD and PPS, an analysis of treatment history and the recovery of working ability by severity of MDD may also be worth considering in this patient population.

Baseline differences between working ability groups were seen in HAM-D17 insomnia score in patients without previous depressive episodes, and in HAM-D17 anxiety and somatic scores in patients with previous depressive episodes. The results of these subgroup analyses, including significant differences, were mostly the same as the results by working ability alone. Incomplete resolution of previous depressive episodes is a predictor of poor prognosis, including the severity, relapsing, and chronic course of depression.²⁵ However, as with the main analysis, despite the observed statistical differences, these results may not be clinically useful for predicting recovery of working ability between patient subgroups in the presence or absence of previous depressive episodes.

A previous post hoc analysis of this study reported that a HAM-D17 total score ≤6 after 12 weeks of treatment corresponded with patients’ working ability.¹³ The results of the current analysis also showed that mean HAM-D17 total score in the group with working ability recovered was less than 6 only at Week 12, although improvements of all outcome measures were observed over time in all groups. In contrast, SASS and GAF scores for social functioning were significantly higher in the working ability recovered group than in the working ability not recovered group throughout the treatment up to Week 12. It is important to judge full recovery of social functioning including working ability, as well as remission of depressive symptoms, by continuous treatment and observation.

Several limitations need to be considered. This was a post hoc analysis of a nonrandomized observational study, and no adjustments for multiplicity were made. Working ability after 12 weeks of treatment was judged by investigators, and their judgment may have been influenced by the results of the HAM-D17 total score. The study enrolled Japanese patients with MDD and PPS; therefore, the results may not be generalizable to MDD patients without PPS, and there may be potential cultural or ethnic differences. In particular, it may be difficult to extrapolate the somatic score results seen in patients with MDD and PPS to patients without PPS.

Conclusions

This post hoc analysis suggests that most baseline clinical characteristics with significant differences were not clinically applicable to the prediction of working ability recovery, nor was the presence or absence of previous depressive episodes. However, the use of psychotherapy in parallel with antidepressant monotherapy may be positively associated with of working ability recovery, as judged by investigators after 12 weeks of antidepressant treatment. All outcome measures improved over time regardless of the baseline characteristics, reinforcing the importance of continuous treatment and observation to improve and accurately judge working ability in patients with MDD and PPS.