Journal of Affective Disorders 142 (2012) 143–149

Contents lists available at SciVerse ScienceDirect

Journal of Affective Disorders

journal homepage: www.elsevier.com/locate/jad

Research report

The development and validation of the Peradeniya Depression Scale

(PDS)—A culturally relevant tool for screening of depression in
Sri Lanka
D.R.R. Abeyasinghe a, S. Tennakoon b, T.N. Rajapakse a,n
a
Dept. of Psychiatry, Faculty of Medicine, University of Peradeniya, Peradeniya, Sri Lanka
b
Dept. of Community Medicine, Faculty of Medicine, University of Peradeniya, Peradeniya, Sri Lanka

a r t i c l e i n f o abstract

Article history: Background: Cultural factors may influence the manner in which a given population interprets and
Received 12 May 2011 conceptualizes their experience of depression. The aim of this study is to validate the Peradeniya
Received in revised form Depression Scale (PDS), a locally created, culturally relevant tool for detection of depression in
29 April 2012
Sri Lanka.
Accepted 29 April 2012
Method: Fifty currently depressed patients (diagnosed via the Structured Clinical Interview for DSM
Available online 9 August 2012
Disorders) and 50 (non-depressed) controls were administered the PDS.
Keywords: Results: At a score of 10/25 the PDS showed a sensitivity 88.5% and specificity of 85.0% with regards to
Depression the detection of depression. Culturally appropriate statements, which referred to international criteria
Sri Lanka
of depression and somatic symptoms, showed significantly higher odds of being positive in depressed
Screening
patients compared to controls.
Limitations: The PDS was validated among an outpatient population presenting to a psychiatry clinic in
a government hospital in Peradeniya, Sri Lanka. While this is fairly representative of patients presenting
to government hospital clinics in this country, further multi-centre studies in different areas of the
country maybe useful.
Conclusion: The PDS is the first screening tool for depression developed and validated in Sri Lanka,
written in Sinhalese, taking into account cultural expressions and idioms of the illness. It shows
satisfactory sensitivity and specificity as a screening tool for depression. The findings also suggest that
it maybe worthwhile for Asian countries to consider adopting scales which are based on internationally
accepted diagnostic criteria for depression, but which incorporate expressions that are more appro-
priate to their own culture and language.
& 2012 Elsevier B.V. All rights reserved.

1. Introduction useful in providing a ‘common language’ for the identification of

Depression is a major psychiatric disorder, with a lifetime pre-
valence of up to 16% (Kessler et al., 2003). Regional studies in Sri
Lanka have reported lifetime rates of depression of upto 6.6% (Ball
et al., 2010a). However the conceptualization of depression, particu-
larly cross-culturally, continues to be a source of discussion (Kinzie
et al., 1982; Manson, 1995). The term depression has been varyingly
used to describe a mood, a symptom, and a syndrome. Both the
DSMIV (American Psychiatric Association, 1994) and ICD10 (World
Health Organization, 1992) require the presence of either low mood,
or loss of interest/pleasure, in order to make a diagnosis of depres-
sion; other criteria include guilt, motor agitation/retardation et al.,
but do not include the presence of somatic manifestations of
distress or pain complaints. The DSMIV and ICD10 criteria are
n
Corresponding author. Tel.: þ94 777 386232.
E-mail address: gemba471@gmail.com (T.N. Rajapakse).
depression as an illness, and cross cultural studies suggest that
‘core’ depressive symptomology can be elicited in patients of
varying backgrounds, including Sri Lanka (Ball et al., 2010a;
Sartorius et al., 1980). However, international studies have also
reported that cultural factors may influence the manner in which a
given population may interpret and conceptualize their experi-
ence of depression (Chang et al., 2008; Kleinman, 2004; Parker
et al., 2001; Ryder et al., 2008). Similarly, Patel et al. (1995), Patel
and Mann (1997) have described the etic emic paradigm in cross-
cultural psychiatric research. Symptoms such as guilt, self-blame
and depressive affect have been reported to be less prominent in
depressed patients from non-western cultures (Kleinmann, 1977;
Marsella et al., 1974), which is similar to clinical experience
regarding depressed patients in Sri Lanka. Ball et al. (2010a), in
their study of depressed patients in Colombo, Sri Lanka, reported
that while psychosomatic symptoms were relatively more
commonly seen, ‘guilt’ items were less commonly reported. The
ICD10/DSMIV diagnostic criteria for depression often present

0165-0327/$ - see front matter & 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.jad.2012.04.019
144 D.R.R. Abeyasinghe et al. / Journal of Affective Disorders 142 (2012) 143–149
indirectly in Sri Lanka—for example depressed mood may often be
represented by cultural idioms of distress and low mood, rather
than by a direct expression of sadness (Obeysekere, 1985).
2. Aims of the study
The aim of this study is to develop and validate a locally
created, culturally relevant screening tool for detection of depres-
sion in Sri Lanka as it presents in this country. We have also
attempted to make the scale easy to use for clinicians not
specialized in psychiatry.
3. Methods
3.1. Development of the Peradeniya Depression Scale (PDS)
The PDS was created by a senior consultant psychiatrist, based
on verbatim statements recorded in clinical notes of Sri Lankan
patients who presented with depression. Over the period of 10
years, he became familiar with cultural idioms and expressions
used by outpatients with depression presenting to Teaching
Hospital Peradeniya. Subsequently, over a prospective 2 year
period, the presentations of outpatients clinically diagnosed to
have depression were recorded verbatim in their case notes.
These case notes (n 192) were then perused and all consistently
occurring statements were extracted and used in the construction
of the initial version of the PDS. The symptoms thus represented
in the PDS fell into one of the following categories: somatic
symptoms, biological symptoms of depression, statements related
to mood (including culturally constructed expressions of affect),
depressive cognitions and depressive behaviours. Cultural idioms
of distress were used to denote some of these symptoms, where
relevant. The PDS was then reviewed by a second psychiatrist to
ensure that the tool would identify the features of depression.
The PDS is written in Sinhalese (the language used by a
majority in Sri Lanka), using easy to understand wording, and
where appropriate, colloquial terms. Attempts were made to
make each statement in the scale as brief and unambiguous as
possible, and each statement requires a yes/no response only. The
wording and nature of the statements in the Peradeniya Depres-
sion Scale (PDS) was considered carefully, keeping in mind that
patients in Sri Lanka, similar to those from other Asian popula-
tions (Kinzie et al., 1982) tend to express their symptoms
indirectly, and to put more emphasis on the distress caused by
oneself to the group as a whole, rather than focusing on the
individual. This is in contrast to the more self-declarative indivi-
dualistic statements seen in tools such as the BDI (Beck et al.,
1961), which have been developed for a Western population. The
pilot version of the PDS thus created consisted of 30 items.
3.2. Validation of the PDS
Outpatients aged over 18 years, presenting to the psychiatry
clinic, Teaching Hospital Peradeniya, Sri Lanka, who were pre-
senting from the community to the clinic for the first time, who
were fluent in spoken Sinhalese and who were suffering from a
current episode of major depression without psychotic symp-
toms, were considered eligible as cases for the study (N ¼50)
(Mumford et al., 2005). Depressed patients with co-morbid
psychotic disorders, those with cognitive impairment, mental
retardation and those who were not fluent in spoken Sinhalese,
were excluded.
A current major depressive episode was diagnosed by face to
face assessment by two consultant psychiatrists, working
independently using the Structured Clinical Interview for DSM
Disorders (SCID). The patients (cases) were also separately invited
to complete the PDS by a medical officer. The examining psy-
chiatrist was blinded to the outcome of the PDS, and likewise the
medical officer was blinded to the outcome of the SCID based
interview.
The controls consisted of a group from the community,
matched for age, sex, social class and level of education, who
did not give a current or past history of depression. Since the
cases were patients presenting from the community to the clinic
for the first time, it was considered appropriate to use matched
controls from the community, and this is similar to methods used
in previous studies (Mumford et al., 2005). The controls (N 50)
were examined for depression via the SCID based interview and
invited to complete the PDS, in the same manner as described
above (Mumford et al., 2005). If a control subject was found to be
depressed on the SCID assessment, he/she was excluded from
the study.
The first 20 consecutive cases were interviewed independently
by both consultant psychiatrists in order to identify agreement
between the two with regards to diagnosis. Each consultant was
blind to the diagnosis made by the other. Ethical approval for the
study was obtained from the Ethical Review Committee of the
Medical Faculty, Peradeniya, Sri Lanka.
3.3. Analysis
The odds of each item being positive for cases and controls (as
diagnosed according to the SCID interview) and the odds ratio for
cases vs. controls were calculated. All items that did not show
significantly higher odds of being positive in depression com-
pared to controls were excluded from the final version of the PDS.
Further to this, all items that showed statistically significant odds
of being positive in depression with a lower confidence limit of
below 1.5, were also excluded.
A ROC curve was fitted using the rocfit option of STATA 9.0
statistical package to estimate the general validity of the new
scale. An arbitrary cut off of 40.75 is considered as a highly
accurate tool (Swets, 1988).
The sensitivity and specificity at the point perpendicularly
furthest from the diagonal line on the curve was calculated
manually. Since all items in the PDS were considered to con-
tribute equally towards the detection of depression, all items in
the PDS were given equal weightage. The final cut off point was
decided based on consideration of the fitted ROC as well as
consideration of the cost of a false positive and a false negative
result (Greiner et al., 2000; Park et al., 2004).
4. Results
The mean age of the patient group was 38.8 years, and 60%
were female. The controls were matched for age and sex. There
was 100% agreement between the two clinicians in the diagnosis
of depression for the subsample examined by both.
4.1. Item selection for the PDS
The item wise breakdown of the PDS, showing the odds ratio
of each item for depression as diagnosed by the SCID based
clinician assessment, is shown in Table 1. All items referring to
somatic symptoms had significant odds of being positive in the
depressed group compared to the control group (Table 1). These
symptoms were burning sensation, headache, chest pain or
palpitations, limb pains, vatha amaru (an ayurvedic concept of
migratory pain), fullness of the abdomen, and feeling faintish.
 D.R.R. Abeyasinghe et al. / Journal of Affective Disorders 142 (2012) 143–149 145

Other items that showed significant odds of being positive in the
depressed group compared to the controls included biological
symptoms, and items related to depressive mood, depressive
cognitions and depressed behaviours (Table 1). Of the original
30 item scale, all but four items showed a significantly higher
Table 1
Item wise breakdown of the PDS, showing odds ratio for depression as diagnosed
by the SCID based clinician assessment.
odds of being positive in depression as diagnosed by the Struc-
tured Clinical Interview for DSM Disorders (Table 1). Furthermore
the item related to sexual difficulties in depression was also
excluded from the final version of the scale since the lower
confidence limit for this item was 1.2. Therefore the 30 item pilot
version was reduced to 25 items in the final version (Appendix),
based on odds ratios between patients and controls (Mumford
et al., 2005).
4.2. Validation of the PDS

Statement in the PDS OR (95% confidence The area under the ROC curve was 0.95 (95% CI 0.91–0.99)
interval) (Fig. 1). The highest sensitivity and specificity as calculated by
using the fitted ROC curve were 87.5% and 88.0%, respectively,
Related to biological symptoms:
which corresponded to a PDS score of between 9 and 10.
Insomnia 4.9 (2.0–11.9)
Loss of weight 4.7 (1.7–12.4)
Loss of appetite 12.0 (4.3–33.6)
Related to somatic symptoms: 5. Discussion
Burning sensation of the body 5.3 (2.1–13.3)
Headache 8.6 (3.5–21.7) 5.1. Validation
Chest discomfort 4.3 (1.8–9.9)
Aching pain/discomfort of limbs or joints 4.0 (1.8–9.2)
Waatha amaaru-generalized discomfort 4.1 (1.7–10.3) The area under the ROC for the PDS was 0.95, which is
Abdominal fullness 6.1 (2.4–15.7) considered highly accurate according to arbitrary guidelines
Feeling faintish 10.4 (4.0–26.9) developed based on the work by Swets (1988). Therefore the
Related to depressed effect PDS can be considered a highly accurate diagnostic tool for the
‘‘Saankaawa’’ (term implying sadness & grief) 9.4 (3.8–23.4)
‘‘Bade gindara’’—‘‘fire in the abdomen’’ 5.6 (2.3–13.6)
detection of depression in Sri Lanka.
Anger/irritability 5.9 (2.4–14.2) According to the fitted ROC curve, the highest sensitivity and
Sadness 9.5 (3.5–26.1) specificity were 87.5% and 88.0%, respectively. The corresponding
Related to depressed cognitions: score of the PDS for this sensitivity and specificity lies between 9
Ideas that life is not worth living 10.4 (4.0–27.0)
and 10. Having considered the fitted ROC curve, as well as the
Fearfulness of the future 4.2 (1.8–9.9)
Suspiciousness 4.3 (1.8–9.9) importance of having a relatively high sensitivity for the detection
Loss of self confidence 11.4 (4.3–29.7) of depression, and since the PDS is scored in discrete numbers, we
Preoccupation that this situation (illness) is due to: chose a cut off of 10/25 (Greiner et al., 2000). The sensitivity and
Past bad karma. 3.4 (1.5–7.7) specificity corresponding to a score of 10 on the fitted ROC curve
An astrologically ‘bad time’. 2.2 (1.8–9.9)
Evil charms. 3.8 (1.2–11.7)
was 88.5% and 85%.
Related to depressive behaviours:
Neglect of duties at home/towards family 11.6 (4.6–29.6) 5.2. Item analysis
Neglect of duties at work 5.2 (2.1–13.0)
Social withdrawal 6.4 (2.6–15.6)
5.2.1. Somatic symptoms
Seeking treatment for a medical illness where the 5.6 (1.9–16.9)
cause could not be found
All items referring to somatic symptoms had significant odds
of being positive in the depressed group compared to the control

.75
Sensitivity

.5

.25

0
0 .25 .5 .75 1
1-specificity
Area under curve = 0.9507 se(area) = 0.0215

Fig. 1. Receiver operator curve (Roc) for the PDS against the SCID based clinician assessment (gold standard).
146 D.R.R. Abeyasinghe et al. / Journal of Affective Disorders 142 (2012) 143–149
group. Particular care was taken to preserve the phrasing of the
statements in the manner that described by the patients, as
recorded in the clinical notes. For example, the statement with
regard to headache in the PDS is closer in meaning to ‘head
ailment’; likewise the statement with regard to chest pain is
closer in meaning to ‘chest ache’. The phrase related to limb pain
has cultural connotations, and relates to the concept of having
‘tired limbs’. The concept of ‘tired limbs’ and ‘faintishness’ may
also be cultural expressions of increased fatigability. This in
keeping with previous work by Ball et al. (2010b), Leone (2010),
which reported a significant overlap between abnormal fatig-
ability and depression in Sri Lanka.
Fullness of the abdomen is again a term with cultural con-
notations, and refers to a state of nervous bloating of the abdo-
men. The statement referring to vatha amaru refers to an
ayurvedic concept of migratory pain. Burning sensation is a
common complaint of distress in depressed patients in this
country and appears similar to hwa-byung, a Korean term mean-
ing ‘‘fire illness’’ (Pang, 1990).
Interestingly, the somatic symptoms described above are not
included in the DSMIV or ICD10 diagnostic criterion for depres-
sion. As suggested by recent studies (Simon et al., 1999) somatic
symptoms reported by patients in our study could be part of their
depressive illness itself, albeit expressed in cultural terms.
Furthermore it maybe a reflection of the traditional doctor-
patient relationship, where patients are expected to report
‘physical symptoms’ to doctors, rather than psychological symp-
toms (Simon et al., 1999). Sumathipala et al. (2004), in their study
of outpatients presenting to a Sri Lankan hospital with multiple
complaints, found that the patients are unlikely to volunteer
psychological symptoms such as suicidal ideation or life weari-
ness to doctors. Ball et al. (2010a) reported that psychomotor
symptoms were relatively more commonly reported by depressed
patients from Colombo. Similarly the study by Sumathipala et al.
(2008) of patients presenting with medically unexplained symp-
toms in Colombo found that only 7.4% gave a psychological
explanation for their symptoms. Ball et al. (2011), based on their
work in Colombo, suggest a complex aetiological relationship
between psychological symptoms, fatigue and depression. The
tendency towards externally orientated thinking, and the per-
ceived stigma related to psychological symptoms and illness too
may also be influencing this presentation (Ryder et al., 2008).
5.2.2. Biological symptoms
As expected, biological symptoms such as poor sleep, loss of
weight and poor appetite showed significantly higher odds of
being positive in depression. Again these items were worded in
the way that they are commonly expressed culturally. For
example, poor appetite was expressed as kama aruchiya in
Sinhalese, which is equivalent to ‘disgust with food’ rather than
loss of desire for food. Ball et al. (2010a) in their study of
depressed patients in Colombo found that patients endorsed loss
of weight and appetite (rather than gain), which they surmise
maybe influenced by the Sri Lankan cultural perception that loss
of weight as being unhealthy. This too may have influenced our
findings.
5.2.3. Items related to mood and culturally constructed expressions
of affect
Even though depression is readily used and understood as a
metaphor for situational changes in one’s emotions or as a
diagnostic term in Europe and North America, this term does
not always have the same significance in other cultures (Shweder
and Haidt, 2000). In Sri Lankan culture there is no equivalent term
for depression in the Sinhalese language and this has been
reported from other non-western countries as well (Kinzie et al.,
1982). As a result clinicians have devised terms such as visadaya
or manasika avapidanaya to refer to depression. These terms
sound artificial and contrived. In Sri Lanka, the closest commonly
used term dukka is a term derived from Buddhist religion that
transcends feeling low or sad, and is described in Buddhism as a
universal state of suffering experienced by mankind. Hence it is
both a personal emotion as well as a universal human condition.
Obeysekere (1985), arguing against the universality of depression,
made out that affective states such as depression relate to larger
cultural and religious frames. His arguments were based on his
ethnographic work in Sri Lanka, where he showed that a Buddhist
society may actually value certain cultural concepts such as
denigration of the body and expression of dukka. Despite this
assertion however, the results of our study show that the direct
expression of sadness (or dukka) is an important statement that
differentiates between depressed patients and healthy controls
(Table 1). Clearly the depressed patients in our study appear to
recognize an excess personal sadness that goes beyond that of the
existential sadness learned within religious and socio-moral
frames. We used another similar term in the PDS—saankaawa, a
word that is close in meaning to sadness, but which also implies
elements of grief. Saankaawa and dukka (Table 1) had equivalence
in the response pattern providing further confirmation that
depressed patients were able to identify personal sadness as a
distressing emotion, rather than a universal and existential entity
as defined by the Buddhist culture.
In the PDS, we also attempted to explore depressed affect more
broadly, through inclusion of indirect cultural idioms which are
reflective of low mood. The statement querying the presence of
‘bade gindara’, if literally translated into English asks the person if
he has ‘fire in the stomach.’ However culturally, in Sinhalese, this
idiom is an expression of sad mood and distress, and this item too
showed significantly higher odds of being positive in depression
(Table 1), though to a lesser degree than the direct statements
referring to sad mood.
Increased anger (Table 1) was significantly associated with
depression. It is interesting that similar findings have been
reported in depressed patients in Vietnam as well (Kinzie et al.,
1982). While anger is not a DSMIV or ICD10 criterion for the
diagnosis of depression, this finding reflects local clinical
experience—where patients often voluntarily describe increased
anger and irritability in the context of depression. This could
partly be a reflection of stigma issues—reporting a sense of anger
is likely to be less stigmatizing than direct description of low
mood and suicidal ideas. Another explanation is that in an
environment where the concept of depression as a medical illness
is unfamiliar, anger maybe more noticed and reported, in contrast
to a depressed and subdued mood alone. A further possibility is
that anger is less acceptable in Asian cultures, where more
emphasis is laid on cultural harmony. However, it is worth stating
that anger maybe a universal feature of depression that requires
consideration as a criterion in diagnosis, as suggested for male
depression (Walinder and Rutz, 2001). Irritability maybe the
emotion that is at the heart of anger issues.
5.2.4. Depressive cognitions and cultural idioms of distress
Several statements included in the PDS reflect depressive
cognitions, phrased in the way they are often expressed in the
Sri Lankan cultural context. In the initial scale there were five
such items, namely statements which attributed the current
distress (i.e., depressive symptoms) to past karma, bad astrological
period, evil charms, evil spirits, and vatha pitha (an aryurvedic
concept). Of these, only the items related to past karma, bad
 D.R.R. Abeyasinghe et al. / Journal of Affective Disorders 142 (2012) 143–149
astrological period and evil charms reached a level of significance
with regards to the detection of depression (Table 1).
The belief that one’s previous karma is now causing distress is
a culturally acceptable form of guilt, which is believed to span
across several previous births. This could be described as an
indirect exploration of guilt, through cultural norms. Interestingly
Ball et al. (2010a) found that Sri Lankan depressed patients are
less likely to endorse direct expression of guilt, and they too
hypothesis that this maybe a reflection of the Buddhist traditions
in this country.
Evil charms in contrast is a belief projected onto evil doers,
thus exonerating oneself from any guilt. The bad astrological
period or belief in astrological planetary alignment is a form of
passive acceptance of one’s fate. These cognitions suggest that
depressed patients in this study may use a form of personal guilt,
passivity and projection to others as coping mechanisms.
In contrast, the terms related to evil spirits and vatha pitha (an
aryurvedic concept) did not significantly differentiate between
depressed subjects and non-depressed controls, and hence were
removed from the final version of the PDS. A possible explanation
is that many of the controls too (who were negative for DSMIV
depression) responded positively to these items, probably
because the controls too experience cognitions related to evil
charms and vatha pitha in the absence of clinical depression. Such
items could be described as an ‘over culturalization’—these items
appear to reflect cultural beliefs that are widely prevalent among
the community, and therefore do not discriminate between the
depressed and non-depressed in the community.
Although cultural idioms of distress did significantly differ-
entiate between depressed and non-depressed subjects, the
results of our study show that the more universally accepted
depressive cognitions have a much higher rate of success in
detecting depression. We included three such depressive cogni-
tions in the PDS—namely suicidal ideas, pessimism and loss of
self confidence (Table 1). These items were clearly superior to the
culturally expressed idiom of distress in the detection of depres-
sion. Therefore it appears that culturally acceptable idioms of
distress need to be used prudently in screening tools of this
nature. Our results show that only certain idioms of distress are
useful in detecting pathological states such as depression. In
comparison, when standard or ‘core’ depressive criteria were
translated into culturally acceptable terms, these criteria appear
more effective at differentiating between the depressed and non-
depressed. This is in keeping with previous studies which have
reported that the ‘core’ symptomology of depression is shared by
patients from different social and cultural backgrounds (Sartorius
et al., 1980).
5.2.5. Depressed behaviours
Another area of culturally relevant distress, which was
screened in the PDS relates to behavioural changes in depression,
particularly the neglect of one’s duty towards the family and the
workplace. The statements relating to behaviour in the PDS
included neglect of duty at home and towards family, neglect of
duty at work, and reduction of social contacts (Table 1). These
items had a high validity in detecting depression. One possible
explanation lies in the Asian tendency to discern one’s value and
well being in terms of the overall well being of the group that one
belongs to (Kitayama and Markus, 2000; Rothbaum et al., 2000).
Therefore a depressed patient may express distress in terms of
his/her inability to contribute to the successful functioning of the
group, i.e., either the family or the workplace.
Interestingly, the statement relating to sexual problems
showed only a low level significance of being positive in depres-
sion, and was thus excluded from the final version of the PDS.
147
Cultural reticence and taboos towards discussing sexual problems
(more so in the females) may have influenced this finding. This is
a significant departure from the findings of semen loss syndrome
called dhat as a significant finding in depression in South Asia
(Mumford, 1996). Sumathipala (2004), in their critical review of
dhat as a culture bound syndrome, suggest that as South Asia
becomes more industrialized and urbanized, symptoms due to
semen-loss anxiety may also diminish in this region. This may
have been a factor influencing our findings.
The statements referring to increased alcohol use and under-
going of rituals did not show significantly higher odds of being
present in depression, and were also excluded from the final
version of the PDS. The alcohol related item may have been non-
significant since our patient population was 60% female, most of
whom would avoid drinking alcohol as is customary in Sri Lanka.
The other non-significant item, which related to undergoing of
shamanic rituals, is an unexpected finding. It may be a reflection
of changing cultural patterns in Sri Lanka. Or it could be that such
rituals are commonly done for physical illnesses and hence do not
distinguish between depression and controls.
6. Summary of main findings
As shown by the area under the fitted ROC curve, the PDS is a
highly accurate tool for the detection of depression, with a
sensitivity of 88.5% and a specificity of 85.0% at a cut off of 10/
25. Item analysis of the PDS scale shows that biological symp-
toms, and statements related to mood and depressive cognitions,
which are stated in the DSM IV and ICD10 diagnostic criteria for
depression are valid and useful in the detection of depression in
the Sri Lankan culture. Cultural idioms of distress in depression is
of limited use for screening purposes in this country, as non-
depressed controls too use them in the absence of depression. In
comparison, statements which were phrased in a culturally
appropriate manner, and which referred to somatic symptoms,
biological symptoms, depressed mood, depressive cognitions and
the depressive behaviour were more significant in the detection
of depression. Thus the internationally accepted criteria of
depression, together with somatic symptoms, are applicable and
useful in the diagnosis of depression when translated into
culturally appropriate statements.
Etic and emic approaches to cross-cultural psychiatric research
have been described Patel et al. (1995), Patel and Mann (1997).
Etic refers to the use of diagnostic criteria and instruments
developed in Euro American cultures whereas the emic approach
uses diagnostic criteria developed within the culture. Patel dis-
cusses the limitations of both approaches: mainly the lack of
cultural validity in etic approach verses the lack of comparability
in the emic approach (Patel et al., 1995). In the design of PDS we
have used a dual approach in that culturally expressed content
was used to assess DSM IV and ICD10 criteria. This dual approach
has been justified as discussed above.
Although this study was conducted in a Sinhalese population
living in Sri Lanka, the scale may have wider application in Tamil
populations living in both Sri Lanka and the Indian subcontinent.
However, it is applicability in Muslim populations maybe limited,
since clinical experience suggests that their belief systems and
cultural expressions tend to be quite different from the Sinhala
and Tamil populations.
6.1. Strengths
The PDS is the first screening tool for depression developed
and validated for a Sri Lankan population, taking into account
cultural expressions and idioms of the illness. As shown by the
148 D.R.R. Abeyasinghe et al. / Journal of Affective Disorders 142 (2012) 143–149

area under the fitted ROC curve, the PDS is a highly accurate tool (SACTRC), particularly for providing access to the Structured Clinical Interview for
DSM Disorders. The authors also thank the staff of the Psychiatry Clinic, Teaching
for the detection of depression, with a sensitivity of 88.5% and a
Hospital Peradeniya, and particularly Dr B.M.B Kumara, and Dr M.G.M.A. Bandara, of
specificity of 85.0% at a cut off of 10/25. Importantly, The PDS is the Dept of Psychiatry, Medical Faculty, Peradeniya, for their support and help.
written in direct, easy to understand terms, with use of colloquial
terms where relevant. It is relatively easy and quick to use, and
responders were able to complete the scale without difficulty, Appendix. The PDS translated into English
often within about 10 minutes. The scale can be self-administered
or administered by a rater—for e.g., for patients who speak but The following questions are about you have been feeling over
cannot read Sinhalese. Therefore illiteracy was not a limitation for the past 2 weeks. Please consider how you have felt over the past
this study. No special training is required to use or administer the two weeks, and circle the most appropriate response for each
scale. We believe that the PDS can be used as a screening tool for statement.
the detection of depression in Sri Lanka.
Have you been experiencing burning Yes/No
6.2. Limitations pains of your body?
Have you been having difficulty Yes/No
The PDS was validated among an outpatient population pre- sleeping at night?
senting to a psychiatry clinic in a government hospital in Have you lost weight, without an obvious illness? Yes/No
Peradeniya, Sri Lanka. This consists mostly of patients from the Have you been experiencing loss of appetite? Yes/No
lower socio economic groups, of both urban and rural back- Have you often been experiencing any of the
grounds. While we believe that these patients are fairly repre- following?
sentative of patients presenting to government hospital clinics in Headache Yes/No
this country, further multi-centre studies of patients of in differ- Chest ache or palpitations Yes/No
ent areas of the country maybe useful. Aches of the arms or legs Yes/No
Only 20 cases were interviewed independently by both con- Aches of the body (waatha amaaruwa) Yes/No
sultant psychiatrists to identify agreement between the two Fullness of the abdomen Yes/No
regards to diagnosis, and this is a limitation. Furthermore, the Faintishness Yes/No
same patient sample was used for both item analysis/selection Have you been suffering from any of the following?
and scale validation, which may have led over-optimistic estima- Grief-sadness (saankaawa) Yes/No
tion of scale metrics. However, this technique has been used Distress-low mood (bade gindera) Yes/No
previously in similar situations (Kinzie et al., 1982). Increased anger or irritability Yes/No
The final version of the PDS does not include a statement Sadness (dukka) Yes/No
related to possible increased alcohol use in depression. Clinical Have you been feeling sad that there is no point in Yes/No
experience and current research suggests that alcohol misuse is living anymore?
likely to be associated with depression (Pinquart and Gamble, Have you been fearful about the future of yourself or Yes/No
2009). As discussed earlier, the finding in this study related to your family?
alcohol may have been due to the fact that 60% of our sample was Have you felt uncertain, or felt less confident in Yes/No
female, since alcohol use in this country is mostly seen in males. yourself?
The PDS is written in Sinhalese, and the study was conducted Have you felt suspicious at times? Yes/No
in subjects fluent in Sinhalese. As Sinhalese is the most commonly Have you been worrying that the above difficulties
spoken language in the country, this would have included a are due to any of the following?
majority of the patients presenting to the outpatient clinics. Due to the affect of a previous karma Yes/No
However, this did exclude the minority of patients who are only Because of an astrologically bad period Yes/No
familiar in Tamil (or Tamil and English), and this is a limitation. Because of evil charms cast against you Yes/No
Have you been unable to attend to your usual Yes/No
6.3. Implications for future research everyday work at home?
Have you been unable to do your job as usual? Yes/No
The construction of the PDS and its validation shows that Have you found yourself avoiding the company of Yes/No
depression can be detected in a clinical setting in a non-western other people?
country, by using questions that refer to usual criteria in the For some time now, have you been seeking treatment Yes/No
international classificatory systems. However, it is worth explor- for an illness that could not be identified?
ing the possibility that instead of using direct translations of
questionnaires that are used in the Western world, countries
should adopt scales based on internationally accepted diagnostic
criteria but which incorporate expressions that are more appro- References
priate to their own culture and language.
American Psychiatric Association, 1994. Diagnostic and Statistical Manual of
Mental Disorders, fourth ed. (DSM-IV). American Psychiatric Association.
Role of funding source Beck, A.T., Ward, C.H., Mendelson, M., Mock, J., Erbaugh, J., 1961. An inventory for
Nothing declared. measuring depression. Archive General Psychiatry 4, 561–571.
Ball, H.A., Siribaddana, S.H., Kovas, Y., Glozier, N., McGuffin, P., Sumathipala, A.,
Hotopf, M., 2010a. Epidemiology and symptomatology of depression in Sri
Lanka: a cross-sectional population-based survey in Colombo District. Journal
Conflict of interest
of Affective Disorders 123, 188–196.
No conflict declared.
Ball, H.A., Siribaddana, S.H., Sumathipala, A., Kovas, Y., Glozier, N., Rijsdijk, F.,
McGuffin, P., Hotopf, M., 2011. Genetic and environmental contributions to the
overlap between psychological, fatigue and somatic symptoms: a twin study
Acknowledgements in Sri Lanka. Twin Research and Human Genetics 14, 53–63.
The authors wish to acknowledge the support of Prof Andrew Dawson, and Dr Ball, H.A., Sumathipala, A., Siribaddana, S.H., Kovas, Y., Glozier, N., McGuffin, P.,
Indika Gawarammana, of the South Asian Clinical Toxicology Research Collaboration Hotopf, M., 2010b. Aetiology of fatigue in Sri Lanka and its overlap with
 D.R.R. Abeyasinghe et al. / Journal of Affective Disorders 142 (2012) 143–149
depression. The British Journal of Psychiatry: The Journal of Mental Science
197, 106–113.
Chang, S.M., Hahm, B.-J., Lee, J.-Y., Shin, M.S., Jeon, H.J., Hong, J.-P., Lee, H.B., Lee, D.-
W., Cho, M.J., 2008. Cross-national difference in the prevalence of depression
caused by the diagnostic threshold. Journal of Affective Disorders 106,
159–167.
Greiner, M., Pfeiffer, D., Smith, R., 2000. Principles and practical application of the
receiver-operating characteristic analysis for diagnostic tests. Preventive
Veterinary Medicine 45, 23–41.
Kessler, R.C., Berglund, P., Demler, O., Jin, R., Koretz, D., Merikangas, K.R., Rush, A.J.,
Walters, E.E., Wang, P.S., Replication, N.C.S., 2003. The epidemiology of major
depressive disorder: results from the National Comorbidity Survey Replication
(NCS-R). Journal of the American Medical Association 289, 3095–3105.
Kinzie, J.D., Manson, S.M., Vinh, D., Tolan, N., Anh, B., Pho, T., 1982. Development
and validation of a Vietnamese-language depression rating scale. American
Journal of Psychiatry 139, 1276.
Kitayama, S., Markus, H.R., 2000. The Pursuit of Happiness and the Realization of
Sympathy; Cultural Patterns of Self, Social Relations and Well being. In
Subjective Well-being Across Cultures. MIT Press.
Kleinmann, A.M., 1977. Depression, somatization and the new cross-cultural
psychiatry. Social Science & Medicine 11, 3–10.
Kleinman, A., 2004. Culture and depression. The New England Journal of Medicine
351, 951–953.
Leone, S.S., 2010. A disabling combination: fatigue and depression. The British
Journal of Psychiatry: The Journal of Mental Science 197, 86–87.
Manson, S.M., 1995. Culture and major depression: current challenges in the
diagnosis of mood disorders. Psychiatric Clinics of North America 18, 487–501.
Marsella, A.J., Kinzie, D., Gorden, P., 1974. Ethnic variations in the expression of
depression. Journal of Cross-Cultural Psychology 4 (4), 433–458.
Mumford, D.B., 1996. The Dhat syndrome: a culturally determined symptom of
depression? Acta Psychiatrica Scandinavica 94 (3), 163–167.
Mumford, D.B., Ayub, M., Karim, R., Izhar, N., Asif, A., Bavington, J.T., 2005.
Development and validation of a questionnaire for anxiety and depression in
Pakistan. Journal of Affective Disorders 88, 175–182.
Obeysekere, G., 1985. Depression, Buddhism and the Work of Culture in Sri Lanka.
In Culture and Depression: Studies in the Anthropology and Cross-cultural
Psychiatry of Affect and Disorder. University of California Press, Berkely.
Pang, K.Y., 1990. Hwabyung: the constructioin of Korean popular illness among
Korean elderly immigrant women in the United States. Culture, Medicine and
Psychiatry 14, 495–512.
Park, S.H., Goo, J.M., Jo, C.H., 2004. Receiver operating characteristic (ROC) curve:
practical review for radiologists. Korean Journal of Radiology 5, 11.
149
Parker, G., Cheah, Y.C., Roy, K., 2001. Do the Chinese somatize depression? A cross-
cultural study. Social Psychiatry and Psychiatric Epidemiology 36, 287–293.
Patel, V., Gwanzura, F., Simunyu, E., Lloyd, K., Mann, A., 1995. The phenomenology
and explanatory models of common mental disorder: a study in primary care
in Harare, Zimbabwe. Psychological Medicine 25, 1191–1200.
Patel, V., Mann, A., 1997. Etic and emic criteria for non-psychotic mental disorder:
a study of the CISR and care provider assessment in Harare. Social Psychiatry
and Psychiatric Epidemiology 32, 84–89.
Pinquart, M., Gamble, S.A., 2009. Meta-analysis of depression and substance use
among individuals with alcohol use disorders. Journal of Substance Abuse
Treatment 37, 127–137.
Rothbaum, F., Pott, M., Azuma, H., Miyake, K., Wisz, J., 2000. The development of
close relationships in Japan and United States: paths of symbiotic harmony
and generative tension. Child Development 71, 1121–1142.
Ryder, A.G., Yang, J., Zhu, X., Yao, S., Yi, J., Heine, S.J., Bagby, R.M., 2008. The cultural
shaping of depression: somatic symptoms in China, psychological symptoms
in North America? Journal of Abnormal Psychology 117, 300–313.
Sartorius, N., Jablensky, A., Gulbinat, W., Ernberg, G., 1980. Application of WHO
scales for the assessment of depressive states in different cultures. Acta
Psychiatrica Scandinavica 62, 204–211.
Shweder, R.A., Haidt, J., 2000. The Cultural Psycholog of Emotions: Ancient and
New. In the Handbook of Emotions. Guildford.
Simon, G.E., VonKorff, M., Piccinelli, M., Fullerton, C., Ormel, J., 1999. An interna-
tional study of the relation between somatic symptoms and depression. The
New England Journal of Medicine 341, 1329–1335.
Sumathipala, A., 2004. Culture-bound syndromes: the story of dhat syndrome. The
British Journal of Psychiatry: The Journal of Mental Science 184, 200–209.
Sumathipala, A., Siribaddana, S., Hewege, S., Sumathipala, K., Prince, M., Mann, A.,
2008. Understanding the explanatory model of the patient on their medically
unexplained symptoms and its implication on treatment development
research: a Sri Lanka study. BMC Psychiatry 8, 54.
Sumathipala, A., Siribaddana, S., Samaraweera, S.D., 2004. Do patients volunteer
their life weariness and suicidal ideations? A Sri Lankan study. Crisis 25,
103–107.
Swets, J.A., 1988. Measuring the accuracy of diagnostic systems. Science 240, 1285.
Walinder, J., Rutz, W., 2001. Male depression and suicide. International Clinical
Psychopharmacology 16, S21–S24.
World Health Organization, 1992. ICD-10 Classification of Mental and Behavioural
Disorders: Clinical Descriptions and Diagnostic Guidelines. World Health
Organization.