Major depressive disorder in Parkinson’s disease: a cross-sectional study from Sri Lanka

Parkinson’s disease (PD) is the second commonest neurodegenerative disorder, its prevalence only being less than that of Alzheimer’s disease [1]. It affects approximately 1% of individuals older than 55 years [2]. Many patients with PD have clinically significant non-motor symptoms such as anxiety, depression, fatigue, sleep disturbance and sensory symptoms [3].

Up to 90% of patients with idiopathic PD experience psychiatric complications, including major depression [4]. Depression is the most common neuropsychiatric disturbance in PD [5] with its prevalence estimated to be around 35% [6]. It is also the most important predictor of poor quality of life in PD [1] and the most significant factor associated with experienced quality of life [7]. These findings emphasize the need to study the factors associated with depression in this population.

In comparison to the Western world there is a paucity of research on PD related psychiatric manifestations in Asia, especially in the South Asian region. Findings from Western populations cannot be directly extrapolated to the population of this region due to significant socio-cultural differences. This may also be supported by the observation that rates of depression have been shown to vary widely between different countries [1]. This underscores the importance of studying depression in PD in different countries and in a variety of socio-cultural contexts. To our knowledge, to date, only one study has been conducted in the South Asian region on this subject, which found a higher prevalence of depression, greater overall disability and inferior quality of life among PD patients compared to a control group of patients with other chronic medical illnesses [8]. The present study, the first of this nature in Sri Lanka, aims to estimate the prevalence of major depression in PD and identify potential risk factors in a local and regional context.

The study was conducted in the outpatient Movement Disorder and Neurology clinics of the premier hospital in Sri Lanka, a tertiary care centre, over a period of 3 months. One hundred and four consecutive patients with idiopathic PD, satisfying the required number derived from sample size calculation and diagnosed according to United Kingdom Parkinson’s Disease Society Brain Bank Diagnostic Criteria (UK-PDS-BB) [9], were recruited to the study after excluding patients with severe speech difficulty. Clinical records of the selected patients were perused to establish the initial presentation and to understand the initial assessor’s view about the diagnosis.

Informed consent was obtained from the study participants. Ethical clearance was obtained from the Ethics Committee of National Hospital of Sri Lanka.

Data related to socio-demographics, PD and other health associated factors of the study subjects were recorded using an interviewer-administered questionnaire. The severity of PD was graded according to Hoehn-Yahr Staging system [10]. Functional impairment was estimated using the Schwab and England Activities of Daily Living scale (SEADL) [11]. The rating was done by the first author based on the information obtained from the patient and caregiver. A semi-structured clinical interview was conducted by the first author and major depression was diagnosed according to DSM-IV-TR criteria [12]. The DSM-IV-TR diagnosis of depressive disorder was considered the gold standard. Irrespective of the presence or absence of depression all participants were rated using the Montgomery-Asberg Depression Rating Scale (MADRS) [13]. The MADRS scores were utilized primarily to rate the severity of depression in those diagnosed with the disorder. The MADRS cut-off scores used to rate severity were ≤20: mild depression, 21–34: moderate depression and above 34: severe depression [14]. The MADRS was not translated into the local languages as it was scored by the investigators. While the interview was conducted in local languages, the assessments were made in English by the investigators. The investigators are bilingual and conduct clinical assessment interviews in the local languages and record data and management plans in English. The validation of the MADRS score using a structured clinical interview are reported in the results section.

This study attempted to determine the prevalence of depression and identify potential risk factors in the PD population in Sri Lanka. The prevalence of major depressive disorder in the study population was 37.5%, similar to the rates reported in most Western populations [3],[5],[6]. The following factors had significantly increased the odds of developing depression: Hoehn-Yahr stage, functional impairment, civil status, level of education, caregiver dependence and concomitant diabetes mellitus. More than half of the depressed subjects were moderately depressed. Evidence suggests depression is usually milder in this population [17],[18].

In comparison to the majority of studies done in Asia which utilized assessment scales for the diagnosis of depression, the present study used a semi-structured DSM-IV-TR based clinical interview for this purpose. A study from Taiwan [19], which also used an interview based assessment, reported a frequency of major depressive disorder of 16.5%, and major and other depressive disorders, taken together, of 42.2%.

Studies using rating scales to diagnose depression report a higher prevalence than those which use a structured clinical interview and diagnostic criteria [1],[20]. A study from India using MADRS and the Beck Depression Inventory (BDI) found the rates as 54% and 49% respectively [8]. Other studies have reported rates of depression of 46% from the Philippines using the MADRS [21], 56% from Japan using the BDI-II [22] and BDI [23] and 65% using the Zung Self-Rating Depression Scale [24] and 21% in China using the Hamilton Rating Scale for Depression [25]. Interestingly in another study from Japan among the 38% identified as depressed by BDI-II, nearly 75% did not show depressed mood [26].

We could not find an association between major depression in PD and age, gender, employment status, occupational class, reasons for unemployment, income, past psychiatric history, age of PD onset, duration of PD, levodopa dose, ropinirole treatment or the presence of concomitant other major medical illnesses. Among the anti-Parkinsonian agents ropinirole was given special consideration as it is commonly used to treat disabling dopa induced dyskinesia [27],[28], but also appears to ameliorate depressive symptoms in PD [27].

The prevalence of depression was higher in Hoehn-Yahr stages III, IV and above. Similar findings have been reported previously [2],[17]. It was highest in stage III possibly because this is the stage where the debility becomes marked. Though the rise in the rate of depression would be expected to be linear as the disability increases this was not observed. This is possibly due to the inability of the severely disabled to physically attend the clinic resulting in a lesser number of patients in severe stages participating in the study which may have influenced the results. Existing views about the association between the severity of PD and depression are contradictory: some studies reported it as positive [2],[17],[19], while others showed no association [29].

The putative association between functional impairment and depression has been a source of debate for long. Depression has been commonly found to be correlated with increased impairment in activities of daily living [2],[17],[19],[20] and self-rated disability has been identified as the greatest predictor of depressive symptoms [29]. The relationship between depression and functional impairment in PD may be reciprocal.

Our results indicate that depression in PD is significantly associated with a state of dependence in the affected population. There could be two explanations for this: while dependence due to physical debility can give rise to depression, depression with low motivation and reduced activity could also make people dependent on others. Among the large number of patients who were caregiver dependent all but one had informal caregivers, which highlights the customs of Sri Lankan society where most of its elderly and feeble members are cared for within a close family network, without being institutionalized.

The association between diabetes mellitus and depression in PD is an important finding from this study, which has not been widely reported. Depression and diabetes commonly co-occur in the general population with rates of depression twice as high in people with diabetes compared to those without [30]. The rate of depression among PD patients with diabetes in our study is 75%. Despite the small number of patients in this group this very high rate merits further exploration regarding the possible mechanisms underlying this comorbidity.

About 30% of our study population were categorised under early-onset PD (onset ≤ age 50) and about 10% developed the illness before 40 years. Most studies reported rates between 3%-5% for an onset ≤ 40 years, while in Japan higher rates, up to 10%, were reported [31]. It would appear that the rates observed in our study are elevated in comparison to Western studies. We speculate that the complexity of early onset PD leads to increased referrals to tertiary care centres, like our study centre; hence the over representation. However, a distinct type of PD prevalent among Asian populations is also a possibility, given the high rates noted in the Japanese study.

Only 20% of depressed subjects were on antidepressants at the time of assessment which draws attention to how depression is poorly identified and treated in neurology settings. The actual number treated might have been lower as it is possible that low dose antidepressants were prescribed for chronic pain instead of depression in some patients. Depression in PD is commonly found to be under-recognized and under-treated [3],[5].

Overall our findings are in keeping with what is reported in Western populations, in relation to prevalence and associated factors of depression in PD. Our study population showed some unique or less commonly reported characteristics such as a higher proportion of early onset PD and association of depression with diabetes. These characteristics merit further inquiry to ascertain whether they are replicated regionally and globally. Further, the differences in depression between cultures may be in the expression, and studying the phenomenology of depression in PD across cultures maybe useful.

In conclusion this first study in Sri Lanka on PD population found that a significant proportion of these patients suffer from major depression. Depression was significantly associated with a number of socio-demographic and clinical factors including PD stage and functional impairment; and was predominantly found in mild to moderate severity. In addition, the study population displayed some unique features such as depressive vulnerability in the presence of diabetes and a higher proportion of early onset PD. Whether there is a regional effect in these findings and possible underlying mechanisms for them need further exploration.

TK conceived the study and involved in design, organization and execution; design and execution of the statistical analysis; and writing of the first draft and finalizing the draft. RH was involved in the conception and supervision; and review and critique of the draft. RW was involved in the conception, design and supervision of the study; and review and critique of the draft. VAde S was involved in the conception and design of the study; design, review and critique of the statistical analysis; and review and critique of the draft. All authors read and approved the final manuscript.

TK is a consultant psychiatrist at Melbourne Health, Australia; RH is a professor of psychiatry at University of Colombo, Sri Lanka; RW is a consultant psychiatrist and senior lecturer at Sri Jayawardenepura University, Sri Lanka; VAdeS is a professor of psychiatry at University of Colombo, Sri Lanka.