A study of psychiatric morbidity among school going adolescents.

AIM: This study aims to study the prevalence of psychiatric morbidity among adolescents and compare its distribution in the urban and rural areas. STUDY
DESIGN: This was a cross-sectional study.
MATERIALS AND METHODS: One thousand adolescents aged 11 to 16 years studying in various private and government schools in urban and rural areas in district Patiala, Punjab were studied. Stratified cluster sampling was used considering the type of school as strata and sections of each standard as clusters. The study was conducted in two steps; in the first step, self-designed sociodemographic questionnaire and socioeconomic status scale, Parekh's method of socioeconomic classification for rural area, and Kuppuswamy's revised method of social classification for urban areas. To study the psychiatric morbidity, the strength and difficulties questionnaire (SDQ) self-report version and parent version was used. Students who scored borderline or abnormal on SDQ, were further evaluated in second stage by clinical interview, detailed case history, and mental state examination; psychiatric disorders were diagnosed following International Classification of Diseases-10 (ICD-10) criteria. STATISTICAL ANALYSIS USED: Chi-square, Student's t-test.
RESULTS: The prevalence ranges from 17.94 in the private school in the urban area and 20.96% in government schools in the urban area to 20.61% in private schools in the rural area and 22.17 in government school of the rural area. The overall prevalence of psychiatric disorders is higher among adolescents in the rural area (21.38%) as compared to the urban area (19.43%). Rural adolescents had significantly higher rates of somatoform disorders (4.45%), conduct disorder (3.78%), dysthymia (1.11%), and other mood disorders (0.89%) whereas higher rates of depression (3.88%), anxiety (3.67%), and hyperkinetic disorders (3.02%) were found in urban counterparts.
CONCLUSION: An alarming number of adolescents suffer from different emotional and behavioral problems, but there is no excess of formal mental illness reaching the psychiatrist. This should help us formulate a rational basis for deploying our resources for the treatment and prevention of mental illness in tomorrow's adults.

INTRODUCTION

The term adolescence meaning “to emerge” or “achieve identity” is a relatively new concept, especially in developmental thinking. Its origin is from a Latin word, “adolescere” meaning “to grow, to mature” indicates the defining features of adolescence. In India, age limits of adolescents have been fixed differently under different programs keeping in view the objectives of that policy or program, like in the National Youth Policy, it is 13–19 years; in ICDS, it is 11–18 years; and in Reproductive and Child Health Programme, it is 10–19 years.[1]

Adolescence is often described as a phase of life that begins in biology and ends in society. The experience of adolescents during teen years would vary considerably according to the cultural and social values of the network of social identities they grow in.[2] There are around 230 million adolescents in India presently. Over the next two decades, the number of adolescents is likely to increase further, but their share to population will decrease marginally as per the projections.[3] Mental disorders and mental health problems seem to have increased considerably among adolescents in the past 20–30 years. The impact of changing youth subcultures on behavior and priorities can also make it difficult to define mental health and mental health problems in adolescents.[4]

Most children and adolescents have good mental health, but studies have shown that 1 in 10 children and adolescents suffer from mental health disorders severe enough to cause impairment. Mental health disorders in children and young people can damage self-esteem and relationships with their peers, undermine school performance, and reduce the quality of life, not only for the child or young person but also for their parents or caregivers and families. The majority of illness burden in childhood and more so in adolescence are caused by mental health disorders, and the majority of adult mental health disorders have their onset in adolescence. Mental health disorders in adolescence are the most powerful predictor of mental health disorders in adulthood.[5]

Adolescence has been viewed as a time of overwhelming turmoil and thus increase in psychiatric morbidity is expected. But adolescents are known to under consult their doctors, and also hesitate in taking their problems to an adult. Thus, unfortunately, there is a paucity of information. While there are a number of comprehensive studies on the prevalence of psychiatric illness in a community, there are few which have examined the teenage years themselves and fewer in India in which age-specific rates are available for a period in life when so many biological and emotional changes are taking place.[6]

Adequate data on the prevalence of psychiatric morbidity among adolescents in India are still lacking; thus, this study was conducted to determine the true prevalence of psychiatric symptoms among adolescents and the characteristics of high- and low-risk groups in Indian society.

MATERIALS AND METHODS

The population for the study comprised of children aged 11–16 years, studying in various government and public schools located in the urban and rural areas in the district of Patiala. The study was conducted with the help of Department of Psychiatry and Department of Community Medicine of Rajindra Hospital, Patiala.

The study was conducted in eight schools in district Patiala. To get a representative sample of all socioeconomic classes of the society, two government schools and two public schools were chosen by simple randomization, in urban and rural areas of district Patiala, respectively. Permission to conduct the study was taken from the principals of the concerned school.

From the above population, children aged 11–16 years studying in VII–X classes who satisfied the selection criteria and whose parents/guardian gave informed consent were included in the sample for the study. Stratified cluster sampling was used considering the type of school as strata and sections of each standard as clusters. One section from each class from each school was selected randomly covering at least 30 students of each class in a school and covering 120 students in all the classes in a school.

The study was conducted in two steps; in the first step, self-designed questionnaire consisting of questions pertaining to sociodemographic data of the children which was prepared separately and pretested before final administration was used along with socioeconomic status scale, Parekh's method of socioeconomic classification for rural area, and Kuppuswamy's revised method of social classification of an individual for urban areas. To study the psychiatric morbidity, the strength and difficulties questionnaire (SDQ) self-report version and parent version was used. The students of each section were asked to fill the questionnaires at a time in the presence of the researcher. Supervision by the teacher was avoided to enable the students to answer the questions. SDQ parent version was given to the students to be filled by their parents and was collected on the next working day. Students who scored borderline or abnormal on SDQ either version formed the sample for the second stage, and further, 5% cases were randomly selected out of the students with normal score, which were followed by clinical interview, detailed case history, and mental state examination; psychiatric disorders were diagnosed following International Classification of Diseases-10 (ICD-10) criteria. The diagnoses were crosschecked by a senior psychiatrist.

Multiinformant strengths and difficulties questionnaires (SDQs)[78910] can identify individuals with a psychiatric diagnosis with a specificity of 94.6% (95% confidence interval 94.1%–95.1%) and a sensitivity of 63.3% (59.7%–66.9%)

To find the association between sociodemographic factors and psychiatric morbidity, Chi-square test was applied. Student's t-test was used for analyzing scores of questionnaire. P < 0.05 was considered as statistically significant.

RESULTS

Numbers of students enrolled in the study were almost equal in urban area and rural area, as depicted in Table 1. This table shows that in both the groups, the number of total students is almost the same and the number of male and female students is also comparable in number with some male dominance, but the difference is not significant (P < 0.54). The numbers were equal as children were taken from school sections and all the schools had almost equal number of children in each section.

Table 1

Number of students enrolled in the study

	Urban schools(Group A)	Rural schools(Group B)	Total	P, χ2
n	463	449	912
Sex, n (%)
Male	260 (56.16)	254 (56.57)	514 (56.35)	P>0.05 (NS)0.37
Female	203 (43.84)	195 (43.43)	398 (43.64)
Total	463 (50.76)	449 (49.23)	912 (100)

NS – Not significant

Screening of students was done using strengths and difficulties questionnaire, and it was identified that 14.7% of the individuals were abnormal, 25.5% as borderline, and 59.8% as normal [Table 2]. Individuals who scored abnormal or borderline were further evaluated for the diagnosis of psychiatric morbidity. The mean scores on SDQ in the normal group were 10.61 in urban and 10.52 in rural; in the borderline group, 14.33 in urban and 13.52 in rural whereas it was 17.68 in urban and 17.42 in rural in the abnormal group. The difference between the scores of the three groups was significant on t-test (P < 0.01) as shown in Table 3.

Table 2

Screening of students using strengths and difficulties questionnaire

SDQ	Urban private schools, n (%)	Urban government schools, n (%)	Rural private schools, n (%)	Rural government schools, n (%)	Total, n (%)
Abnormal	34 (14.5)	28 (12.2)	38 (16.7)	34 (15.4)	134 (14.7)
Borderline	50 (21.4)	59 (25.8)	63 (27.6)	61 (27.6)	233 (25.5)
Normal	150 (64.1)	142 (62)	127 (55.7)	126 (57)	545 (59.8)

SDQ – Strengths and difficulties questionnaire

Table 3

Mean scores of strengths and difficulties questionnaire: Student’s t-test

SDQ scores	Mean	SD	Mean	SD	SE	t-test
Urban schools
Abnormal versus borderline	17.68	2.42	14.33	1.49	0.33	10.15**
Abnormal versus normal	17.68	2.42	10.61	2.06	0.32	22.09**
Borderline versus normal	14.33	1.49	10.61	2.06	0.17	5.24**
Rural schools
Abnormal versus borderline	17.42	2.32	13.52	1.14	0.30	13.00**
Abnormal versus normal	17.42	2.32	10.52	1.97	0.31	22.26**
Borderline versus normal	13.52	1.14	10.52	1.97	0.17	17.64**

**P<0.01, P<0.001. SDQ – Strengths and difficulties questionnaire; SD – Standard deviation

The overall rates of psychiatric disorders were higher among rural adolescents (21.38%) as compared to urban (19.43%). Children from rural areas had higher odds for the overall rates of dysthymia, any other mood disorder, conduct, somatoform, adjustment, and other behavioral disorders whereas the reverse was true for anxiety, hyperkinetic disorders, and depression among urban students. However, the difference between rural and urban was not found to be statistically significant, i.e., P > 0.05 [Table 4].

Table 4

Psychiatric disorders in rural versus urban areas

DiagnosisICD-10	Urban area, n (%)			Rural area, n (%)			OR (rural vs. urban)	CI
	Male (n=260)	Female (n=203)	Total (n=463)	Male (n=254)	Female (n=195)	Total (n=449)
Depression	5 (1.92)	13 (6.40)	18 (3.88)	4 (1.57)	8 (4.10)	12 (2.67)	0.62	0.29-1.33
Dysthymia	1 (0.38)	2 (0.98)	3 (0.64)	0	5 (2.56)	5 (1.11)	1.72	0.41-7.26
BAD	2 (0.76)	0	2 (0.43)	2 (0.78)	1 (0.51)	3 (0.67)	0.51	0.04-5.69
Any other mood disorder	0	4 (1.97)	4 (0.86)	0	4 (2.05)	4 (0.89)	1.03	0.25-4.15
Anxiety	6 (2.30)	11 (5.41)	17 (3.67)	4 (1.57)	11 (5.64)	15 (3.34)	0.91	0.45-1.84
Adjustment	1 (0.38)	2 (0.98)	3 (0.64)	2 (0.78)	2 (1.02)	4 (0.89)	1.03	0.21-5.13
Somatoform	3 (1.15)	8 (3.94)	11 (2.37)	7 (2.75)	13 (6.66)	20 (4.45)	1.81	0.85-3.86
Hyperkinetic	11 (4.23)	3 (1.47)	14 (3.02)	9 (3.54)	3 (1.53)	12 (2.67)	0.88	0.40-1.92
Conduct	11 (4.23)	2 (0.98)	13 (2.80)	16 (6.29)	1 (0.51)	17 (3.78)	1.36	0.65-2.83
PDD	1 (0.38)	1 (0.49)	2 (0.43)	0	0	0	0	0 to NaN
Behavioral	2 (0.76)	1 (0.49)	3 (0.64)	3 (0.78)	1 (0.51)	4 (0.89)	1.03	0.21-5.14
Total	43 (16.53)	47 (23.15)	90 (19.43)	47 (18.50)	49 (25.12)	96 (21.38)	1.13	0.82-1.56

Rural versus urban χ2=0.35, df=1, P<0.55. OR – Odds ratio; CI – Confidence intervals; BAD – Bipolar affective disorder; PDD – Pervasive developmental disorder; NAN – Not a number; ICD – The international classification of diseases

Distribution of psychiatric illness according to socioeconomic status

Table 5 shows that maximum number of diagnosed children, i.e., 27 belonged to the upper lower class (5.83%) and second in the rank were children from upper class, i.e., 25 (5.39%) in the urban area. Whereas in the rural area, the highest number belonged to the middle class, i.e., 24 (5.35%). It was seen that adolescents in lower socioeconomic classes (6 out of 23 students in urban area and 14 out of 48 students in rural area) had higher psychiatric morbidity as compared to upper class in both rural (18 out of 72 students) and urban areas (25 out of 122 students) and was significant statistically (P < 0.05).

Table 5

Distribution of psychiatric illness according to socioeconomic status

SES	Urban area		Rural area
	Normal (n=373)	Abnormal (n=90)	Normal (n=353)	Abnormal (n=96)
Upper class, n (%)	97 (20.95)	25 (5.39)	54 (12.03)	18 (4.01)
Upper middle class, n (%)	95 (20.52)	21 (4.54)	81 (18.04)	21 (4.68)
Lower middle class, n (%)	60 (12.96)	11 (2.38)	116 (25.84)	24 (5.35)
Upper lower class, n (%)	104 (22.46)	27 (5.83)	69 (15.37)	19 (4.23)
Lower class, n (%)	17 (3.67)	6 (1.29)	34 (7.57)	14 (2.67)
χ2	11.54** (P<0.01)		5.63* (P<0.05)

**χ2 = 11.54 (P<0.01), P<0.05. SES – Socioeconomic status

Sex-wise distribution of psychiatric diagnosis

Table 6 shows that females have higher odds of the prevalence of depression (odds ratio [OR] 3.52), dysthymia (OR 9.18), somatoform disorders (OR = 3.13), and anxiety (OR 3.44) as compared to males. Whereas the reverse was observed for hyperkinetic disorders, conduct disorder, and other behavioral disorders. Odds of having psychiatric disorders were more among female children (24.12%) as compared to male children (17.50%), OR = 1.49. Psychiatric morbidity was higher among females as compared to males and this was found to be statistically significant (P < 0.05).

Table 6

Sex-wise distribution of psychiatric diagnosis

DiagnosisICD-10	Males (n=514), n (%)	Females (n=398), n (%)	Total (n=912), n (%)	OR (females vs. males)	CI
Depression	9 (1.75)	21 (5.27)	30 (3.28)	3.52	1.54-8.04
Dysthymia	1 (0.19)	7 (1.75)	8 (0.87)	9.18	1.12-7.95
BAD	4 (0.77)	2 (0.50)	6 (0.66)	0.64	0.12-3.54
Any other mood disorder	0	8 (2.01)	8 (0.87)	-	-
Anxiety	10 (1.94)	22 (5.52)	32 (3.50)	3.44	1.57-7.52
Adjustment	3 (0.58)	4 (1.005)	7 (0.76)	1.72	0.47-14.26
Somatoform disorders	10 (1.94)	21 (5.27)	31 (3.39)	3.13	1.41-6.90
Hyperkinetic disorders	20 (3.8)	6 (1.50)	26 (2.85)	0.37	0.15-0.95
Conduct	27 (5.25)	2 (0.50)	29 (3.17)	0.13	0.04-0.46
PDD	1 (0.19)	1 (0.25)	2 (0.21)	1.29	0.08-20.72
Behavioral	5 (0.97)	2 (0.50)	7 (0.76)	0.64	0.12-3.53

Males versus females χ2=3.96, df=1, *P=0.04. OR – Odds ratio; CI – Confidence intervals; BAD – Bipolar affective disorder; PDD – Pervasive developmental disorder; NAN – Not a number; ICD – The international classification of diseases

DISCUSSION

In the total sample of 960 adolescents who gave consent for the participation in the study and filled the pro forma in the first step, 11 children from urban schools and 18 students from rural schools did not return their parent version of SDQ and 19 students with their parents could not be contacted for the interview and diagnosis in the second step.

There are not many studies which have studied the prevalence of psychiatric disorders among adolescents in the community. In this study, both private and government schools were included to represent both the upper and lower socioeconomic classes in the society. There was no significant difference (P > 0.05) in the distribution of population in different socioeconomic classes of rural and urban areas, and thus, our sample was representative of the whole rural and urban society in Punjab, India.

The results of screening instrument show that on applying SDQ, 14.7% of the individuals were identified as abnormal, 25.5% as borderline, and 59.8% of the students had normal scores. The mean scores on SDQ in the normal group were 10.61 in urban and 10.52 in rural; in borderline group, 14.33 in urban and 13.52 in rural whereas it was 17.68 in urban and 17.42 in rural in the abnormal group. The difference between the scores of the three groups was significant (P < 0.001) according to Student's t-test, and thus, our screening instrument had good sensitivity.

The prevalence in our study ranges from 17.94 in the private school in the urban area to 22.17 in government school of the rural area whereas the prevalence among government schools in urban was 20.96% and private schools in the rural area was found to be 20.61%. The prevalence in rural adolescents was higher as compared to the urban ones.

This was found to be similar to a study conducted by Gau et al.[11] in 1995 in Taiwan in which the overall prevalence of psychiatric disorders was found to be 20.3% which over 3 years decreased to 14.8% in the 3rd year among adolescents. The overall rates of mental disorders were generally higher in rural than in urban youths.

Our results also coincided with the study conducted by Anita et al.[12] in 2001 in Rohtak in which overall prevalence for psychiatric disorders among 6–14 year olds, according to ICD-10, was 17.5% in urban and 16.5% in rural areas of Rohtak.

Robert et al.[13] in 2000 found the prevalence of psychiatric morbidity to be 17.1% in USA, and also, a comprehensive review of studies from other countries by Robert et al.[14] concluded that prevalence rates from studies using earlier versions of the DISC and DSM criteria were in the 18%–20% range. However, Srinath et al.[15] in 2000, Bengaluru reported the prevalence rates among the 4–16 years group to be 12% overall, which was lower compared with our findings and from other community-based studies in Western countries. The urban slum areas had the lowest total prevalence rates whereas urban middle class reported the highest prevalence rates.

Furthermore, the overall prevalence of psychiatric disorders is higher among adolescents in the rural area (21.38%) as compared to the urban area (19.43%). This was similar to findings in other studies where the rural area has been reported to have comparatively higher rates of psychiatric morbidity as compared to urban areas.[1617181920]

Compared to their urban counterparts, rural adolescents had significantly higher rates of somatoform disorders (4.45%), conduct disorder (3.78%), dysthymia (1.11%), and other mood disorders (0.89%) whereas higher rates of depression (3.88%), anxiety (3.67%), and hyperkinetic disorders (3.02%) were found in urban counterparts.

Our findings of higher rates of conduct disorder in rural areas and anxiety disorders in urban areas, although contrary to that in the study by Robert et al.,[20] and Srinath et al.[14] where both anxiety and conduct disorders were more in the urban area (0.5%) each and were in accordance with those from other studies.[1112] Anita et al.[8] found similar results of increased anxiety (4%) in urban and increased conduct disorder (4.75%) among rural students. Gau et al.[11] also reported higher prevalence of disruptive behavior disorders among adolescents from rural area as compared to urban (OR 1.5) and lower odds of anxiety among rural area.

The overall prevalence among boys was 17.50%, and among girls, it was 24.12%. It was higher among girls which was found to be similar to the study conducted by Jaju et al.[21] in Oman which reported that female gender was a strong predictor of lifetime risk of major depressive disorder (OR 3.3, 95% CI 1.7–6.3, P = 0.000), any mood disorder (OR 2.5, 95% CI 1.4–4.3, P = 0.002), and specific phobia (OR 1.5, 95% CI 1.0–2.4, P = 0.047).

However, it was in contrast to the other studies[111220] where males had higher psychiatric morbidity as compared to girls. Anita et al.[12] reported prevalence among males as 18.37% and females as 14.44%. In an another study by Roberts et al.,[13] male adolescents had higher odds of having psychiatric morbidity as compared to females (OR 1.15, CI 0.96–1.37). However, in contrast to all other studies, there was no gender difference seen in the study by Srinath et al.[14]

This difference was most probably due to inclusion of substance use disorders in all other studies which is mostly seen in male gender, and in our study, substance use disorder was not included, thus giving higher prevalence rate among females.

The prevalence rate of psychiatric morbidity ranging from 17.94% to 22.17% for the entire sample (11–16 years) validates the conclusion that prevalence rates among adolescents are increasing over time. Furthermore, as the prevalence is higher in the rural area, one might speculate that low awareness of the importance of psychiatric disorders, poor living conditions, and the presence of multiple stressors could have combined to increase the magnitude of adolescent's problems.

Limitations

There are a few limitations pertaining to this study. The results should be interpreted in the context of these limitations.

The limited sample size of the study was due to time-limited nature of the study. Thus, there is a need for a larger sample size to accurately assess prevalence

The study had cross-sectional research design and thus sample was not followed up

Comorbid diagnosis was not made at present and as there is evidence to suggest that single disorders often progress to complex comorbid disorders that are impervious to treatment and more likely to recur than less complex conditions. Therefore, our individuals need to be reassessed at a later period for a meaningful understanding of the impact of the present labeling

All the variables were assessed cross-sectionally; hence, answers to cause-effect relationship between variables cannot be given. Longitudinal studies should be carried out to look for correlations between changes in impact (variables) with changes in severity of illness

Furthermore, it is possible that the present survey may have omitted those who had dropped out from school as a result of mental ailments and also those who were nonschool going for other reasons.

CONCLUSION

The conclusion to be made from these data is that turmoil there may be, but there is no excess of formal mental illness reaching the psychiatrist. Conceivably, there are many teenagers who experience symptoms but who do not report these. This should help us formulate a rational basis for deploying our resources for the treatment and prevention of mental illness in tomorrow's adults.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.