Functional Impairment and Painful Physical Symptoms in Patients with Major Depressive Disorder Treated with Antidepressants: Real-World Evidence from the Middle East

2.1. Study Design

This study used data from a 6-month, multi-national, prospective, non-interventional, observational study. It was primarily designed to examine treatment-emergent sexual dysfunction (TESD) and other treatment outcomes among patients with MDD who were treated with either an SNRI or a selective serotonin reuptake inhibitor (SSRI) in actual clinical practice. A total of 1,549 patients, out of 1,647 patients who were enrolled between Nov 2007 and Nov 2008, were classified as “sexually active patients without sexual dysfunction at study entry”. This post-hoc study focused only on those patients from the Middle East (n=314) (Saudi Arabia [n=179] and United Arab Emirates (UAE) [n=135]).

This study followed the ethical standards of responsible local committees and the regulations of the participating countries. Ethical review board (ERB) approval was obtained as required by local laws. All patients provided informed consent for the provision and collection of the data. This study was also conducted in accordance with the ethical principles that have their origin in the Declaration of Helsinki and that are consistent with Good Pharmacoepidemology Practices (GPPs), as well as applicable laws and regulations of the countries where the study was conducted, as appropriate.

Further details of the study design have been published elsewhere [22-26].

2.2. Study Sample

Patients were eligible to participate in the study if they met the following inclusion criteria: (1) at least 18 years of age; (2) presenting with an episode of MDD in an outpatient setting, with MDD diagnosed according to the International Statistical Classification of Diseases 10^th revision (ICD-10) [27]or Diagnostic and Statistical Manual of Mental Disorders-4th edition text revision (DSM-IV-TR) [28] criteria; (3) the Clinical Global Impression-Severity (CGI-S) score of ≥4 (i.e., at least moderately ill) [29]; (4) initiating or switching to any available SSRI or SNRI antidepressant in accordance with a treating psychiatrist’s discretion; (5) being sexually active without sexual dysfunction, as defined by Arizona Sexual Experience Scale (ASEX) [30]; and (6) providing consent to release data. The study excluded patients who (1) were participating in another study; (2) had a history of treatment-resistant depression; (3) had a past or current diagnosis of schizophrenia, schizophreniform or schizoaffective disorder, bipolar disorder, dysthymia, mental retardation or dementia; or (4) received any antidepressant within 1 week (1 month for fluoxetine) prior to study entry, with the exception of patients receiving an ineffective treatment for whom the immediate switch to an SSRI or an SNRI antidepressant was considered to be the best treatment option.

Patients were, however, not required to continue their initial medications. Changes in medication and dosing were possible at any time during follow-up, as determined by their treating psychiatrist.

2.3. Data Collection and Outcome Assessment

Data were collected during visits within the normal course of care. Subsequent data collection was targeted at week 8, week 16, and week 24 from the baseline visit (i.e., study entry). Patient socio-demographics and clinical history were recorded at the baseline assessment. Clinical severity of depression was assessed at each visit using the CGI-S [29] and the 16-item Quick Inventory of Depressive Symptomatology Self-Report (QIDS-SR₁₆) [31].

Functional impairment was also assessed at each visit using the Sheehan Disability Scale (SDS) [32]. It is a brief self-report inventory that assesses functional impairment in work/school, social life, and family life. The level of functional impairment in each of the three domains is rated from 0 to 10. The level of global functional impairment is determined with the sum of the three subscores. A higher score indicates a higher level of functional impairment.

Depression-related pain severity was measured using the pain-related items of the Somatic Symptom Inventory (SSI), which includes headaches, heart/chest pain, neck pain, abdominal pain, lower back pain, joint pain, and muscular soreness [33]. PPS status was also assessed as painful physical symptom negative (PPS-) or positive (PPS+); PPS+ was defined as a mean score of at least 2 for the seven pain-related items of the SSI.

2.4. Statistical Analysis

This post-hoc study included 275 patients, out of the 314 patients from Saudi Arabia and UAE, who (1) initiated either duloxetine or an SSRI as monotherapy at baseline, and (2) who had non-missing data on the QIDS-SR₁₆ score at baseline with at least one non-missing QIDS-SR₁₆ score during follow-up (n=152 in the duloxetine cohort and n=123 in the SSRI cohort). This study analysed the patient observations up to the point where their initial medications were discontinued (n=127 [83.6%] in the duloxetine cohort and n=97 [78.9%] in the SSRI cohort available at 24 weeks).

The baseline socio-demographic and clinical characteristics of the two treatment cohorts were summarised and compared using the Chi-square test (for categorical variables) and the Mann-Whitney test (for continuous variables). The level of functional impairment (both SDS total scores and SDS subscores) at each visit was also described and compared using the Mann-Whitney test. The treatment difference in functional impairment at each visit was also expressed in terms of the effect size, using Cohen’s d [34].

The mixed effects models for repeated measures (MMRM) analysis was used to estimate and compare the levels of functional impairment (SDS total, work, social life, and family life) during follow-up between the two treatment groups. The models were adjusted for age, gender, SSI-pain score at baseline, QIDS-SR₁₆ score at baseline, the baseline value of the outcome modelled, and visit number. The models were further adjusted for the following variables if they appeared to be significant (p<0.1) when each one of them was added to the above model: independent living (living in his/her own house), living with a spouse/partner, employment status, having MDD episodes in the 24 months prior to study entry, MDD hospitalisations in the 24 months prior to study entry, and the number of any significant pre-existing comorbidities. Finally, the models also included the interaction term between time (visit number) and treatment if it appeared to be significant (p<0.05) in the full model.

All analyses were repeated for each subgroup of patients treated with duloxetine and SSRIs to examine whether the association between depression-related pain (SSI-pain score) and functional impairment (SDS total score) was mediated or moderated by the type of treatment. In addition, the MMRM analysis was further conducted to compare the impacts of duloxetine and SSRIs on depression-related pain during follow-up. The same list of independent variables aforementioned were considered in the model.

All statistical analyses were performed using SAS version 9.3 for Windows (SAS Institute, Cary, NC, USA).

3.1. The Characteristics of Patients at Baseline

A total of 275 patients were included in the final analysis. Overall, the mean (standard deviation [SD]) age of these patients was 37.3 (SD=8.4) years, and only about one-third of them (34.6%) were female (27.1% in Saudi Arabia and 44.2% in the UAE, p=0.003). Slightly more than half of the patients (56.4%) were from Saudi Arabia, and the rest were from the UAE (43.6%). The majority of the patients (76.7%) were living with a spouse. About two-thirds of the patients (67.3%) had full-time employment, and more than half of the patients (57.8%) completed at least a tertiary or university degree. In addition, nearly two-thirds of these MDD patients (61.5%) also suffered from PPS at baseline. Only 20% of MDD patients had taken antidepressants, and 48.4% (21.1% for benzodiazepines and the 27.3% for other medications) had taken other psychiatric medications in the previous 24 months.

Of the 275 patients, 152 initiated duloxetine at baseline and the rest (n=123) initiated an SSRI antidepressant at baseline. The most common SSRIs prescribed at baseline were escitalopram (38.2%), paroxetine (27.6%), and sertraline (20.3%). The median daily doses (interquartile ranges [IQRs]) of these antidepressants at baseline were 10.0mg/d (IQR: 10.0mg/d – 10.0mg/d) for escitalopram, 22.5mg/d (IQR: 20.0mg/d – 25.0mg/d) for paroxetine, 50.0mg/d (IQR: 50.0mg/d – 100.0mg/d) for sertraline, and 60.0mg/d (60.0mg/d – 60.0mg/d) for duloxetine.

Table (1) summarises the baseline patient characteristics by treatment cohorts. A higher proportion of the SSRI cohort were female (p=0.015) and from the UAE (p=0.003), whereas a higher proportion of the duloxetine cohort had completed a tertiary or university degree (p=0.013). In addition, a higher proportion of the SSRI cohort had taken other psychiatric medications in the previous 24 months (37.4% vs. 19.1%), whereas a higher proportion of the duloxetine cohort had taken benzodiazepines in the previous 24 months (26.3% vs. 14.6%) (p=0.001). Nevertheless, the levels of depression severity and depression-related pain at baseline were similar for the two cohorts, and the proportion of patients exhibiting PPS at baseline was also similar for the two (63.8% in the duloxetine cohort and 58.5% in the SSRI cohort, p=0.371).

3.2. The Level of Functional Impairment over Time by Treatment Cohort

Table (2) demonstrates the mean levels of functional impairment at baseline and at 24 weeks, respectively, by treatment cohort. At baseline, the level of functional impairment was slightly higher in the duloxetine cohort than in the SSRI cohort: the mean SDS total score at baseline was 17.36 (SD=6.00) in the former and 15.66 (SD=6.28) in the latter (p=0.012). Nevertheless, both treatment groups, on average, had a moderate level of functional impairment in all three subdomains (SDS work, social life, and family life), with a mean score of 5.35 (SD=2.66) (for work) to 6.24 (SD=2.16) (for social life) in the duloxetine cohort and 5.11 (SD=2.41) (for work) to 5.35 (SD=2.17) (for social life) in the SSRI cohort.

Despite a higher level of baseline functional impairment, the duloxetine cohort achieved a greater level of functioning at 24 weeks, compared to the SSRI cohort. At 24 weeks, the mean SDS total score was 3.41 (SD=3.06) in the former and 4.60 (SD=3.21) in the latter (p=0.008). Similar patterns were also observed in all three SDS subdomains. This treatment difference was equivalent to the effect size of -0.38 for SDS total (subdomains: -0.25 for SDS social life to -0.47 for SDS work).

The level of functional impairment at 24 weeks was still lower in the duloxetine cohort than in the SSRI cohort even when the adjustment of baseline differences was made through the MMRM, as shown in Table (3) and Fig. (1). At 24 weeks, the estimated mean SDS total score (i.e., least squares [LS] means) was 3.13 (standard error [SE]=0.26) in the former, which was lower than that of 4.96 (SE=0.29) in the latter (p<0.001) (i.e. better functioning in the duloxetine cohort) (Fig. 1).

Notably, the level of depression-related pain at baseline, as measured by SSI-pain scores, was also statistically significantly associated with the level of functional impairment during follow-up (coefficient=0.10, p=0.030). This positive association between the two was even more apparent in the subgroup of SSRI-treated patients (coefficient=0.23, p=0.003), but not in the subgroup of duloxetine-treated patients (coefficient=0.04, p=0.500) (Fig. 2). In relation to this finding, an additional analysis was conducted to confirm the differential impacts of duloxetine and SSRIs on depression-related pain. The MMRM results showed that the level of depression-related pain at 24 weeks was statistically significantly lower in duloxetine-treated patients (LS means [SE]: 8.31 [0.24]) than in SSRI-treated patients (LS means [SE]: 11.14 [0.28]) (p<0.001) (data not shown).

4.1. Effects and Roles of Antidepressants in the Treatment of MDD in The Middle East

Treatment with antidepressants, both duloxetine and SSRIs, was found to be effective in improving functioning of MDD patients in the Middle East. These findings are highly consistent with those of our previous studies with East Asian patients, which examined the effects of duloxetine with a daily dose of ≤ 60mg and SSRIs, respectively, on various clinical and humanistic outcomes [24, 26]. There are, however, no studies in the Middle East to have evaluated the effects of antidepressants in the treatment of MDD, as confirmed in the recent systematic review by Travers et al. [21]. This is rather surprising given the role of antidepressants in modern psychiatry and the high level of MDD burden (in terms of DALYs) reported in this region [3, 4].

No regional treatment guidelines for MDD are available either, except for the two published in a neighbouring country, Turkey [35, 37]. While both Turkish and other international guidelines [38, 39] recommend the use of antidepressants as first-line treatment of MDD, available evidence suggests that the rate of antidepressant use is low, but that of benzodiazepine is relatively high in the treatment of MDD in the Middle East [21]. In our study, 21.1% and 20.0% of the MDD patients were reported to have taken benzodiazepines and antidepressants, respectively, in the 24 months prior to the study entry. The lack of regional clinical evidence might have contributed in part to the absence of treatment guidelines and the lack of appropriate treatments for MDD in this region.

While our study found both duloxetine and SSRIs to be effective in improving functional impairment, duloxetine-treated patients achieved a higher level of functioning than SSRI-treated patients during follow-up. However, the superiority of SNRIs including duloxetine over SSRIs has not been well established in the literature [40], although limited evidence supports the advantage of SNRIs in subgroups of MDD patients with PPS and/or more severely ill [18, 41]. Notably, the majority of MDD Middle Eastern patients in this study exhibited PPS at baseline (63.8% in the duloxetine cohort and 58.5% in the SSRI cohort, p=0.371).

4.2. Implications of Depression-Related Pain in the Treatment of MDD in the Middle East

Our study also found differential impacts of depression-related pain on functioning by treatment cohorts. Duloxetine-treated patients achieved functioning independently of pain severity, but SSRI-related patients did not. That is, SSRI-treated patients who had a higher level of depression-related pain at baseline did not achieve the same level of functioning as those who had a lower level of depression-related pain at baseline, confirming the negative impact of pain on improvement in functioning in this cohort. This finding adds to the existing evidence that supports the advantages of SNRIs in the subgroup of patients with PPS, as discussed above, and also reaffirms the importance of pain control in the treatment of MDD.

Painful physical symptoms seemingly play dual roles in the treatment of MDD as both a trigger for help-seeking and a risk factor for delayed diagnosis. Its dual role is particularly relevant in the Middle East, where mental illness carries an intense and enduring stigma [42]. Patients in such cultures often tend to somatise their emotional problems and express them in terms of physical symptoms. This may partly explain the high percentage of painful physical symptoms (61.5% at baseline) found in our study, although it is not clear whether this is a manifestation of depressive symptoms or a coexisting pain condition with an unknown aetiology.

Okasha reported that up to 80% of psychiatric patients have a tendency to somatise their emotional and psychiatric problems [43]. Many of these patients therefore do not seek help in a psychiatric setting, but instead present to primary care with physical complaints. This may explain the over-representation of depressed patients in primary care in the Middle East. Using the Beck Depression Inventory, Hamdan et al. [12] found that nearly half of women attending primary care clinics (n=224) in the UAE were mildly (14.3%), moderately (14.7%), or severely (17.9%) depressed (12). Similarly, Becker et al. reported that about one in five patients attending primary care clinics (n=431) in Saudi Arabia had depression and 67.4% of them were female, using the Patient Health Questionnaire (PHQ) [44]. The authors also suggested that the restrictive Arab culture may have placed women at higher risk for somatic presentations of psychological distress in primary care settings. This postulation is in line with the low proportion of women (34.6%) included in our study, which was conducted in an outpatient psychiatric setting in the region. Becker further examined whether there is a discrepancy in the detection rates of depression and somatisation between primary care physicians and the use of the PHQ screening instrument [45]. The study reported that primary care physicians were aware of psychiatric disorders, but their diagnostic skills for these conditions were poor.

Taken together, a substantial proportion of MDD patients in the Middle East also suffer from painful physical symptoms, which are likely to negatively impact on clinical outcomes, including functioning. Many of these patients tend to seek help first from primary care physicians. It is therefore important to make sure that primary care physicians, who are at the front-line of the healthcare system, correctly diagnose and treat MDD even in the presence of painful physical symptoms.

4.3. Study Limitations

While this study provides meaningful clinical insights into MDD in the Middle East, its findings should be interpreted in the context of the following study limitations. Firstly, data were taken from an observational study. Although the MMRM analysis adjusted for the baseline imbalance between duloxetine- and SSRI-treated patients, there still exists the unobserved imbalance between the two cohorts. Secondly, the primary objective of the observational study was to assess TESD in the treatment of MDD. The study, therefore, included only those patients who were sexually active and did not have sexual dysfunction at baseline. It has been reported that sexual dysfunction is two to three times more prevalent in patients with depression than in the general population [46, 47]. Our findings may therefore not be immediately generalizable to MDD patients as a whole. In addition, sexual dysfunction was measured using the Arizona Sexual Experience Scale [30]. It is not clear whether patients, especially women, were willing to accurately report their sexual dysfunction, particularly in this restrictive culture. Further research is needed to examine whether our findings can be replicated in MDD patients without such inclusion/exclusion criteria. Thirdly, even within the SSRI class, there exist differences in efficacy and/or tolerability between the individual drugs [48]. Therefore, our findings cannot represent the differences in effectiveness, in terms of functional improvement, between duloxetine and any specific SSRIs. Finally, this post-hoc analysis included only 275 patients from the Middle East, and it may therefore not be representative of all patients with MDD in the region. Our results should be interpreted with caution until further replication is available.