PREVALENCE
AND
SEVERAL
POTENTIAL
SELF
PERCEPTION
ASSOCIATED FACTORS OF DEPRESSION AMONG SCHOOL GOING LATE
ADOLESCENTS IN THE GALLE MUNICIPALITY AREA OF SRI- LANKA
ABSTRACT
Depression remains a major public health problem. Recent reviews indicate that mental
health problems in the adolescents are increasing of which depression constitutes an
important part.
The objectives of the present study was to determine the prevalence, associated factors
and selected outcomes depression among late adolescents in Galle municipality area.
The study was a cross-sectional study of a two stage cluster sample of 412 late
adolescents utilizing a pre tested self administered questionnaire which included locally
adapted and validated CES-D scale to detect depression and two locally adapted scales
to determine the perceived stress (Perceived Stress Scale) and self esteem (Rosenberg
Self-Esteem Scale).
The data was processed and analyzed with the SPSS® statistical soft ware.
Data analysis was done using standard statistical methods such as Chi square, MannWhitney and Fisher’s Exact test. Uni-variate Odds ratios were calculated to quantify the
strengths of associations.
Only 98.3% of the sample completed the questionnaire sufficiently to allow computation
of a depression score. Forty two percent screened positive for depression (mild to
moderate - 21% of males and 21% of females, severe-21% of males and 21% of
females).
Respondents who were faced stressful events during the last year (OR= 3.0, 95%
CI=1.9-4.4), having high perceived stress (u=8573.5, z = -9.853, p < 0.001), having low
self esteem (u=10519.5, z= - 8.184, p < 0.001) were independently associated with a
high depression score. Stressors due to parents expectations (OR= 2.0, 95% CI=1.2-2.8)
were also associated with a high depression score.
Depression was independently associated with poor school performances (OR= 2.6,
95% CI=1.7-4.0).
The study reveals that the level of self reported depression was high among these
respondents and associated factors identified include high level of perceived stress and
low self esteem. Depression was independently associated with poor school
performances..
Key words: Adolescent, Depression, CES-D
1
INTRODUCTION
Adolescents
Adolescence is a transient period of physical and mental development that extends from
age 10 to19 years. Those who are in adolescence can be further categorized as early (1013 years), mid- (14-15 years), and late (16-19 years) adolescent (United Nations
Population Fund 2000, World Health Organization 2009).
Approximately 3.7 million adolescents live in Sri Lanka. They represent 19.7% of the
population (DCS 2002). Of those, 72.9% comprise of school-going adolescents
(MOHECA 2002).
Estimates show that the Sri Lankan adolescent population would remain around 3
million during the next few decades (Abeykoon 1998), highlighting the importance of
prioritizing and drawing increased attention to adolescents issues during the coming
years (Fig. 1)
.
Fig.
1
Projected adolescent population in Sri Lanka from 2005 – 2050
(Abeykoon 1998)
Depression and its impact on adolescence
Depressive disorders are identified as the fourth leading health problem in the world.
Recently published studies found that early onset depression often persists, recurs and
continues into adulthood (Strickland 2002). It also indicates that depression in youth may
predict more severe illness in adult life.
2
Justification
Today, Sri Lankan adolescents face many stressful situations which are rising both in
frequency and intensity, due to social, economic and political conflicts. In 1997, Sri
Lanka recorded the highest per capita suicide rate in the world. Deliberate self-harm and
suicide, and self-inflected injuries and poisoning was the main causes of mortality and
morbidity in the 15-24 year age group in Sri Lanka (Annual Health Bulletin 2001).
The aim of this study was to assess the prevalence and factors associated with
depression among school going late adolescents and to describe the association of
depression with school performance in terms of academic achievements and
extracurricular activities.
METHODOLOGY
Study Design
This study used a descriptive cross-sectional design.
Study Setting
This study was conducted among the grade 12 and 13 students from the schools of the
Galle Municipal Council area.
Target and Study Population
The target population was school going adolescents between 17 and 19 years in Sri
Lanka
Study Unit
A student of grade 12 or 13 studying in a Sinhala medium school was considered a study
unit from which an independent response was obtained.
Inclusion and Exclusion Criteria
Inclusion Criteria-Students of all grades 12 & 13, in the selected classes who were
present on the day of data collection were included in the study.
Exclusion Criteria- students of international schools; students absent in school on the
day of data collection; and students whose parents or the student himself/herself did not
provide consent to be included in the study were excluded.
Sampling techniques
The sample was a two staged systematic random cluster sample, taken according to the
probability proportional to the size of the school enrollments of the grade 12 and 13 in
the schools of Galle Municipality Area.
3
The primary sampling unit (PSU) was a school in the Galle Municipality Area. The
secondary sampling unit (SSU) was a grade 12 or 13 class of the selected school which
was also the cluster. All the students in a selected cluster were included in the sample.
(Average cluster size is 40).
Data collection methods and tools
The data of this study was collected using a self-administered questionnaire (SAQ). It
also included the following standard scales:
Scale 1:
Scale 2:
Scale 3:
Already translated into Sinhala and validated version of
CES-D (Center for Epidemiological Studies-Depression)
(Ferdinando 2006)
Rosenberg Self-Esteem Scale
Perceived Stress Scale
CES-D (Center for Epidemiological Studies-Depression) scale
Depression was assessed using a summated rating scale; the Center for Epidemiological
Studies for Depression (CES–D) scale, which had been already validated for the Sri
Lankan context by Ferdinando 2006. CES–D scale comprises of a 20-item self-reported
rating scale designed to measure symptoms of depression. CES-D can classify
adolescents as not depressed, mildly depressed or severely depressed.
Rosenberg Self-Esteem Scale
Self-esteem was measured using the Rosenberg Self–Esteem Scale (RSES). This scale
has ten items with possible responses on a four point Likert scale ranging from strongly
agree to strongly disagree. The possible scores range from 0 to 30. Scores between 15
and 25 are considered normal while those below 15 suggest low self–esteem.
Perceived stress scale
Perceived Stress Scale (PSS) was used to measure the perception of stress. It is a
measure of the degree to which situations in one’s life are appraised as stressful. This
scale consists of 10 items that are responded to on a 5 point Likert scale that asks the
respondents to answer a series of questions as they pertained to them in the previous
month.
The scale yields a single score and a higher score is indicative of greater levels of
perceived stress.
Pre-Testing
The self-administered questionnaire was pre-tested among 40 adolescents to check the
comprehension, readability and understanding of wording and meaning of the
questionnaire.
4
Data Collecting Personnel
Data collectors were the principal investigator (PI) and another medical officer.
Data Processing and Analysis
All the open ended questions were coded before data entry. The data was then entered
into the computer using Epi Data software. Data was processed and analyzed using a
personal computer with the SPSS® statistical soft ware.
Ethical issues and clearance
Informed Consent and Freedom of Choice-Informed consent was obtained from
principals of selected schools, parents and study subjects. Before obtaining consent, the
purpose of the study and the subject’s right to refuse to participate in the study, and the
fact that the decision would not adversely affect the subject were clearly explained.
Confidentiality and Privacy-Confidentiality and anonymity of the information obtained
from the subjects was assured and protected. The students who were identified as having
psychological problems were personally attended by the principal investigator and
referred to mental health professionals through the principals of the schools
.
Permission for the Study-Permission from the Provincial Director of Education,
Provincial Education Office and Galle was taken prior to commencement of study.
Ethical Clearance-The principal investigator (PI) obtained ethical clearance from
Ethical Review Committee, Faculty of Medicine, University of Ruhuna, Karapitiya
,Galle.
RESULTS
Profile of Study Sample
This study included both male and female adolescents that were recruited from the grade
12-13 in schools from Galle Municipality area. From the sample of 445 adolescents only
412 adolescents were included in the study due to none response and non participation.
Age and Sex
Seventeen-year olds formed the modal age group. Forty seven point three percent (195)
were 17 years of age while 43.7% (180) were 18 years. Only 9% (37) were in the 16 year
old age category.
Table 1: Distribution of respondents according to age and sex
Sex
Age
Female
Male
5
Total
(Years)
16
17
18
N (%)
16 (7.8)
90 (43.7)
100 (48.5)
N (%)
21 (10.2)
105 (51.0)
80 (38.8)
N (%)
37 (9.0)
195 (47.3)
180 (43.7)
Total
206 (100.0)
206 (100.0)
412 (100.0)
The sample of 412 adolescents was comprised of equal number females 50.0% (206) and
males 50.0% (206). The majority of males 51.0% (105) were in the 17-year old age
group while a large proportion 48.5% (100) of females was 18 years old.
The majority of the sample of adolescents, 95.6% (197) of the females and 98.1% (202)
of the males were Buddhists. Only 2.4% (5) of the females and 1.0% (2) of the males
were Christians. None of the males and 0.5 %(1) of the females were Hindus and 1.5%
(3) of the females and 1% (2) of the male were Islam.
The majority 98.1% (403) of the sample was comprised of Sinhalese adolescents, while
only small percentages were Tamils 0.5% (2), Moors 0.7%(3), Burghers 0.2%(1) and
Malay 0.5%(2). Therefore, the findings of the present study would mostly reflect the
perspective of Sinhalese Buddhist adolescents.
Presence of Depression
Of the respondents, only 98.3% completed the depression questionnaire sufficiently to
allow computation of a depression score.
CES-D score was used as a proxy indicator of depression. The possible range of scores is
0 to 60. The threshold value used for operationally defining either mild or moderate
depression was a score of 16. A score of 21 was the threshold for differentiating to
moderate and severe depression.
A total of 42% (172) of the respondents had a depression score above the threshold score
of 16, indicating that a significant percentage of students were having at least mild
depression based on the CES-D (figure 4).
6
Substantial proportion of adolescents also had severe depression, and there were no
difference between male and female adolescents in terms of prevalence of severe
depression (21% of males and 21% of females) (Table 14).
Associated Factors for Depression
Several potential self perception risk factors were evaluated to determine whether they
were in fact associated with depression. These included: stressors at home and school,
body image, self-esteem and self-confidence.
Stressful events
Feeling unsafe while at home, feeling that schooling is not enjoyable, experiencing
stressful events (death of mother, death of father, separation from the family/friends)
during the previous year were considered as stressful events that are potentially related to
depression. Table 19 presents the details of these associations.
Table 2: Risk of depression by stressful events factors
Risk Factor
Feel unsafe at home
Yes
No
Depression Level
Depressed
NonDepressed
(N=171)
(N=234)
N (%)
N (%)
Odds Ratio
(95%
Confidence
Interval)
19 (11.1)
9 (3.8)
3.1 (1.4-7.1)
152 (88.9)
225 (96.2)
1.0
Degrees
of
Freedom
p-Value
χ2
8.103
1
0.004
18.681
1
<0.01
(reference)
Schooling is enjoyable
Rarely/never
Always/most of the time
28 (16.4)
9 (3.8)
143 (83.6)
225 (96.2)
7
5.0 (2.2-10.6)
1.0
(reference)
Underwent stressful
event during the
previous year
Yes
No
27.624
105 (61.4)
82 (35.0)
66 (38.6)
152 (65.0)
1
<0.01
3.0 (1.9-4.4)
1.0
(reference)
Adolescents who felt unsafe at home showed an over three times greater risk of having
depression than those who did not feel unsafe (OR=3.1; 95% CI=1.4-7.1; p=0.004).
Feeling that schooling was enjoyable was also significantly associated with having
depression among adolescents. Those adolescents who rarely/never felt that schooling
was enjoyable had a 5 times greater risk of developing depression than those, who
always/most of the time felt schooling was enjoyable (OR=5.0; 95% CI=2.2-10.6;
p<0.01). The adolescents who underwent stressful events during the previous year were
at 3 times greater risk of developing depression than the adolescents who did not
underwent stressful events during the previous year (OR=3.0; 95% CI=1.9-4.4; p<0.01).
Perceived Stress
Adolescents’ perceived stress was measured using the Perceived Stress Scale (PSS).
Figure 3 presents the box plots of PSS scores among depressed and non depressed
adolescents.
Figure 3: Distribution of Perceived Stress Scale (PSS) score among depressed and
non-depressed
As PSS scores were not approximately normally distributed based on the KolmogorovSmirnov test (p<0.001), the Mann-Whitney U-test was used to compare them between
those depressed and not depressed. The Mann-Whitney U-test showed that overall, nondepressed adolescents had relatively less perceived stress scores compared to those with
depression (U=8573.5; Z=-9.853; p<0.001).
8
Self-Esteem
Self-esteem of the adolescents was also evaluated as a potential associated factor for
depression. Self-esteem was measured using the Rosenberg Self-Esteem Scale (RSES).
Figure 4 displays the distribution of RSES scores of depressed and non depressed
adolescents.
Figure 4: Distribution of Rosenberg Self-Esteem Scale (RSES) Score among the
depressed and non-depressed
Similar to the PSS, the RSES scores were also found to be a non-normally distributed
based on the Kolmogorov-Smirnov test (p<0.001). The Mann-Whitney U-test showed
that overall, the self-esteem score distribution of non-depressed adolescents was
significantly higher than that of the depressed, and that the differences were statistically
significant (U=10519.5; Z=- 8.184; p<0.001).
Body Image
Perceived self-body image was also evaluated as a potential risk factor. The body image
was assessed using Body Image Scale (BIS). There was no evidence to suggest body
image was significantly associated with the depression in adolescents.
9
School-related factors were considered as consequences of depression. School
performance, parents’ expectations regarding performance and participation in extracurricular activities were assessed for potential associations with depression among
adolescents.
Adolescents with poor/average school performance were significantly more likely to be
depressed than the adolescents with above average school performance, with over twice
the risk (OR = 2.6; 95% CI=1.7-4.0; p<0.01). adolescents who felt stressed due to
parents’ expectations were nearly twice as likely to be depressed as those who did not
feel stressed (OR=2.0; 95% CI=1.3-2.9; p=0.002). In contrast there was no association
between engagement in extracurricular activities and depression (OR =1.0; 95% CI=0.51.0; p<0.065).
DISCUSSION
Prevalence of depression
Though CES-D is considered as an internationally accepted instrument for screening for
depression, strict application of these types of western classifications with their
underlying concepts to screen for depression in our culture might lead to
misinterpretations. According to Wickramasingha (2000), this is evidenced in past
community surveys conducted in Sri Lanka which have yielded very low community
prevalence rates such as 0.13% (Jayasundara cited in Ferdinando 2006) and 0.19 %
(Wijesingha cited in Ferdinando 1978).
Therefore the present study used a locally adapted and validated Sinhala version of CESD questionnaire to determine self reported depression among this cohort of adolescents.
This instrument demonstrated good psychometric properties during the original
validation study by Ferdinando (2006):with a sensitivity of 91.3%, specificity of 90%,
positive predictive value of 90.14%, and negative predictive value of 91.30%. Reliability
assessed using level of agreement between psychiatrists’ diagnosis and CES-D diagnosis
(kappa) was 0.814. Internal consistency measured using Cronbach alpha reliability
coefficient was found 0.7936. Guttman split half reliability was 0.75992.Inter-rater
reliability varied from 0.831 to 0.894. Therefore it would have captured the prevalence of
depression in this sample in a fairly valid and reliable manner.
The cut off values used in this study according to the ROC curve was 15/16,thus identify
those who score 15 or less as non depressed & those who score 16 or above as depressed.
Using the same cut off values the findings of the present study suggest that about 42. %(
n = 172) of the participants could be diagnosed with depression, a similar percentage as
reported in previous study of the prevalence of depression among adolescents. This
finding is consistent with the findings by Lawlor and James (2000) which showed, using
the Youth Self – Report questionnaire, that 39.2%, of an Irish 17-year-olds have
psychological problems.
10
The prevalence results, however, were somewhat higher than findings of another Sri
Lankan study , performed in Rathnapura district on 445 adolescents of ages 14-18yrs
using CES-D questionnaire, which reported that 36% of adolescents had symptoms of
depression (Rodrigo et al. 2010).
This present study showed mild to moderate depression in 21.2% adolescents and severe
depression in 21% of adolescents. The above mentioned study by Rodrigo et al. (2010)
has shown a prevalence of mild to moderate depression of 17% and severe depression of
19%.
Rodrigo’s study used a Sinhala version of CES-D questionnaire without validation where
as the current study used the locally adapted and validated CES-D questionnaire. This
difference in scale characteristics may be partially attributable to the differences in
proportions observed in the two studies. The locally adapted and validated version used
in this study may have been more effective in capturing the changes in mental health
among this group of adolescents.
These findings were different from an Northern Iran study, which was performed among
high school and pre-university students, using Beck’s self administered depression
questionnaire that reported mild to moderate depression in 35.7% and severe in 0.3%
(Mohammad et al 2007). Considering the similarity of the age and gender in the above
mentioned studies, it seems that cultural differences regarding areas such as
psychological stress and different using of concepts such as self evaluation, self
confidence and adaptive behavioral styles and the two different type of scales may
explain these differences.
The high prevalence of depressive symptoms among school going adolescents revealed
in this study indicate the importance of focusing attention on the mental health problems
of adolescents in the school setting.
As the majority of the sample consisted of Sinhalese (98.1%) and Buddhist (95.6%)
adolescents, the genralizabilty of findings may be somewhat restricted to adolescents
from these ethnic religious sectors. The prevalence of depression and patterns in
associations of other ethnic and religious types might be little different from these trends.
Associated factors of depression
According to the present study depressive respondents had significantly higher levels of
perceived stress than the non depressive respondents as indicated by the perceived stress
scale scores. The Mann-Whitney U-test showed that overall, non-depressed adolescents
had relatively less perceived stress scores compared to those with depression(U=8573.5;
Z=-9.853; p<0.001) , which is similar to Windle s’ (1992) findings who used the same
instrument to measure stress. that showed stress as the strongest predictor of adolescent
depression.
Self esteem measured by the Rosenberg Self-esteem Scale found to have a significant
association with depression. Those adolescents with a high depression score had a lower
11
self – esteem score than those with a low depression score(U=10519.5; Z=- 8.184;
p<0.001).
These results were broadly in keeping with a study carried out in Ireland (Farrell et al
1994) using the same scale to measure self esteem. Similar results were found by
Mehmet et al., ( 2008).,Burwell and Shirk, (2006).
Though previous studies found strong association between depression and negative body
image (McCabe and Marwit 1993, Van et al. 2007 this study failed to identify significant
association between negative body image and depression which is another aspect of self
perception.
Outcomes of depression
This study showed more depressive symptoms among adolescents with poor or average
school performance (OR = 2.6; 95% CI=1.7-4.0; p<0.01) and those who felt stressors
due to parents’ expectations regarding studies (OR = 2.0; 95% CI=1.3-2.9; p<0.01).
So this study indicates that depressive symptoms are found to have adolescents and it
significant associations with low self esteem and stressors at home and school.
Limitations
There are limitations in this study that must be acknowledged. Because of the cross
sectional design the present study only indicates point affective situation rather than its
trends in past and future. On the other hand the direction of the associated factors and
depression may sometimes be difficult to ascertain the data on both dependant and
independent variables were collected simultaneously.
The students absent in school on the day of data collection and students whose parents or
the student himself/herself did not provide consent to be included in the study were
excluded from the study. Therefore non participation rate was somewhat higher in this
study (7%).
Conclusion
This study has shown a high prevalence (42%) of self-reported depression as measured
CES-D scale. Moreover, it was also shown that depression was associated with stressors
at home and feeling low self esteem. Poor school performances were identified as
outcomes of the depression.
Recommendation
Considering the high prevalence, it is warranted that appropriate screening programs for
depression and school mental health development programmes concentrating more on
stress factors, low self esteem.
12
Acknowledgement
Completing this research has been a life changing experience for me. Therefore number
of people who generously provided assistance and actively took part in the wonderful
success of this research should be appreciated with heart felt gratitude.
I would like to acknowledge my supervisor, Dr. Neil Thalagala, Consultant Community
Physician, Family Health Bureau for his professional advice, encouragement, kindness
and patience which made this task fruitful. And also my special thank is rendered to Dr.
Champa Wijesinghe, Senior Lecturer, Department of community medicine, University of
Ruhuna for encouraging me and directing me during the final stages of the study.
LIST OF REFERENCES
1. Abeykoon ATPL.,Demographic projects for Sri Lanka.. Research paper series No.9, population
information Center, Population Division Ministry.
2. Abeykoon AT PL. & Wilson P. Emerging population issues among adolescent and youth. Sri Lankan
Journal of population Statistics. 1998, Vol.1, PP.25-34.
3. Bernard HR. Direct reactive observation. Research methods in Anthropology, qualitative and quantitative
approach, 2nd ed. Sage publications, 1995. United States.
4. Cohen S, Kamarck T & Mermelstein, R., ‘A Global Measure of Perceived Stress’, Journal of Health and
Social Behavior, 24 (1983), 385-396.
5. Department of Health services, Annual Health Bulletin 2001. Department of Health Services, Colombo.
6. Department of Health Services, Annual Health Bulletin 2004, Department of Health Services, Colombo,
2005.
7. Farell Ao, Flangan, E, Bedford D, James D. and Howell F. Factors associated with self reported
depression and self – esteem among school going adolescents from a geographically defined region in
Ireland.. Irish Journal of Medical Science.1994. vol. 174, no 4, pp 17 – 22.
8. Ferdinando KDAR.. Prevalence and correlates of depression, in adults in Kaluthara DDHS area. Thesis
(MD in Community Medicine. Postgraduate Institute of Medicine, University of Colombo.2006.
9.Jayasinghe JARC.Parenting and Mental Wellbeing of Adolescents’ between 15-16 years,MD.
Dissertation(Psychiatry),Postgraduate Institute of Medicine,University of Colombo.2009.
10.. MOHECA (Ministry of Human Resource Development, Education & Cultural Affairs, 2002, School
census 2002, Preliminary Report, MOHECA, Sri Lanka.
11..Mehmet NE, Kamil E, Hacer H,Dereboy,C.,2008,Prevalence of and factors related to depressionin high
school students.Turkish Journal of Psychiatry,2008;19((4).
13
12 . Mohomed JM, Tehrani HS, Moodari, S, Asli NJ and Fallahi M. The prevalence of depression among
high school and pre-university Adolescents. Archives of Iranian Medline.2007. vol.110, no.2, pp.141-146.
13. Rushton JL, Forcier M, Schectman RM.. Epidemiology of depressive symptoms in the National
Longitudinal study of Adolescent Health.
J. Am A, viewed 20 th April 2010, cad Child Adolescent
Psychiatry, vol.41, pp.199-205.
14.Strickland PL.Bio-social origins of depression in the community. The British Journal of Psychiatry.
vol.180, pp, 168-173.
15. Talagala NI. Attempted suicides antecedent and consequences. MD (Community Medicine) Thesis.
2000. Postgraduate Institute of Medicine. University of Colombo.
16. UNFPA (United Nations Population Fund). State of the World Population 2003. United Nations
Population Fund. New York.
17. UNICEF.National Survey on emerging issues among adolescents in Sri Lanka .2004. UNICEF. Sri
Lanka.
18. WHO (World Health Organization), 2009. 1998
Adolescent Health, viewed 20 th April 2010,
http://www,int/mentalhealth/meia/en/62pdf.
19. WHO 2010, ‘Adolescent health’, viewed 20 th April 2010
http://www.who.int/features/facilities/adolescents health/en/index/.html.
14