Depression, sleep, body image, and lifestyle: Prevalence and associated factors of depression among school-going adolescents in Delhi National Capital Region

ABSTRACT

Background:

Adolescents are known to be vulnerable to depression, which worsened during the COVID-19 pandemic. They also exhibit poor sleep quality and body image issues (BIIs) which are considered risk factors for depression. There is a paucity of Indian studies regarding adolescent depression and its correlation with lifestyle.

Aims:

To detect the prevalence of depression in adolescents, and understand its association with sociodemographic variables, sleep quality, BIIs, and lifestyle factors.

Materials and Methods:

A cross-sectional study was conducted among 392 adolescents attending classes 8^th–12^th in two schools in Delhi National Capital Region. A pre-tested semi-structured questionnaire was used to collect sociodemographic and lifestyle data. Depression anxiety stress scale-21 was used to detect depression. Pittsburgh Sleep Quality Index and Stunkard Figure Rating Scale were used to evaluate sleep quality and BIIs. Data was analyzed using Statistical Package for Social Sciences Version-25.

Results:

Depression was present in 40.3% of students. It was more common in girls, older students, students in higher classes, who lacked siblings and whose mothers were engaged in professions as well as those with BIIs. However, after multivariate analysis, the association of depression with poor sleep quality, meal skipping, lack of exercise, family history of mental illness, and inverse association with open discussion of one’s thoughts and worries with family remained significant.

Conclusions:

Depression is common among adolescents and shows associations with poor sleep quality and lifestyle choices. Awareness of these associations may enable better screening and early intervention for vulnerable adolescents.

INTRODUCTION

Depression is one of the most common mental health disorders. Its prevalence has been increasing over the years, affecting an estimated 300 million people worldwide in 2017 and 4.5% of the Indian population in 2015.[1] Moreover, a study has predicted that mental illness is increasing at a faster pace in India than in other countries and that India will account for one-third of the global mental illness burden in 10 years.[2]

Although no age group is immune to depression, adolescents are particularly vulnerable.[3-5]

Merikangas et al. reported that the median age of onset for depression was 13 years.[6] A study among adolescents in South India found the prevalence of depression to be as high as 40.08%.[4] This is conceivably attributable to the biological and social turmoil faced by teenagers as they transition from childhood to adulthood. Changes in self-perception, social information processing mechanisms, increasing academic pressure, strained relationships with parents, and smoking or drinking are likely contributing factors.[3,5,7] This is consequential since adolescent depression affects academic performance, leads to social impairment, and is a major risk factor for suicide, substance abuse, and the development of major depressive disorders in adulthood.[3,4,8]

In Delhi, 12–13 years were identified as a period of transition in sleep pattern resulting in reduced sleep.[9] Sleep deprivation (<6 hours of sleep per night) has been associated with depressive symptoms as well as major depression. Similarly, body image issues (BIIs) have been associated with the development of major depression.[5,10] A study conducted in Tamil Nadu, India, reported that 77.6% of girls were dissatisfied with their body type and 42.7% resorted to skipping meals to lose weight.[11] Skipping meals, particularly breakfast, has also been reported as a risk factor for depression.[12] Meanwhile, higher levels of physical activity are associated with better mental health.[13]

The outbreak of COVID-19 pandemic coupled with necessary quarantine and isolation measures has also had adverse effects on mental health.[14] An Indian study conducted during the lockdown in 2020 reported the prevalence of depression as 25%.[15] Children and adolescents are at high risk of developing depression both during and after lockdown.[16] A global meta-analysis has reported a similar prevalence (25%) of depression among adolescents.[14]

India is home to 253 million adolescents, comprising one-fifth of the country’s population.[17] Yet, there is a paucity of data regarding their mental health and its correlation with lifestyle and sociodemographic factors. Better delineation of these associations could aid risk stratification and facilitate appropriate timely intervention. With this background, the index study was conducted aiming to estimate the magnitude of depression in Indian adolescents and study its association with sleep quality, BIIs, and lifestyle factors.

MATERIALS AND METHODS

A cross-sectional analytical study was conducted over a period of two months from July to September 2021. From a previous Indian study considering the prevalence of depression as 40.8% with 5% absolute precision and 5% alpha level, the sample size was calculated to be 386 using the formula for cross-sectional studies n = 4 pq/d2.[4] Taking into account a 10% non-response rate, 425 students were proposed to be enrolled.

It was proposed to include co-education senior secondary schools so that both boys and girls aged 11–18 could be included in the study and multistage sampling was done.

Stage-1: Selection of schools

Out of 148 senior co-education, secondary schools in Delhi NCR, contact details were found for 60. The schools were contacted over phone and email due to restricted movement during COVID-19 pandemic. Positive responses were received from ten schools out of which two were selected by simple random sampling. One was a government school while the other was private. The schools were visited, and the study was explained to the principals, who permitted to conduct the same.

Stage-2: Selection of classes

Students aged 11–18 study in classes 5^th–12^th. However, students studying in classes 6^th–8^th May have been too young to answer the study questionnaires accurately, and hence, classes 8^th–12^th were selected by purposive sampling.

Stage-3: Selection of students

All the students in the selected classes were briefed about the study and written informed assent was taken from those who agreed to participate. They were also given informed consent forms and subject information sheets to take home to their parents. On the following days, students who gave assent and whose parents gave informed consent were enrolled in the study. The study questionnaires were given to them, and responses were collected. It was proposed to include students evenly distributed between the two schools, and hence, 426 students were enrolled. After excluding students who did not give assent or whose parents refused informed consent (n = 23) and incomplete questionnaires (n = 11), 392 responses were obtained (Response rate = 92%).

Data was collected using a pre-tested semi-structured questionnaire containing the following parts

1. Data about substance abuse, physical activity, meal skipping, and sociodemographic characteristics including age, sex, parents’ education, family type, existence of siblings, history of physical or mental illness in parents, substance abuse, substance abuse in parents, etc. Modified Kuppuswamy scale was used to calculate socioeconomic status.[18] BMI was classified as per Asian criteria.[19] Substance abuse was defined as using drugs (including alcohol, glue sniffing, chewable tobacco, cannabis, cigarettes, and e-cigarettes) more than once during the past month.[20] Daily physical activity of 60 minutes was considered sufficient as per WHO criteria.[21] Meal skipping was defined as skipping two or more meals in a week.[22] Students were asked if they had faced stressful events in the past one year, such as death, disease, or separation in the family.

2. DASS-21 was used to screen for depression, anxiety, and stress (DAS). Based on self-reporting, this scale consists of 21 items, seven each for DAS, which are scored from 0 to 3. The scores are multiplied by 2 and graded into mild, moderate, severe, and extremely severe.[23] It has been validated among adolescents in India in both English and Hindi.[8] The values of Cronbach’s alpha for depression (English = 0.76; Hindi = 0.70) have been found acceptable.[8]

3. Pittsburgh Sleep Quality Index (PSQI) is based on self-reporting and assesses sleep quality over the past month. It generates 7 component scores and a global PSQI score. A global PSQI score greater than 5 indicates poor-quality sleep.[24] It has been validated in adolescents in India with high internal consistency (Cronbach’s alpha = 0.83).[25]

4. Stunkard Figure Rating consists of nine silhouettes of different body types for both males and females.[26] Students were asked to choose their perceived figure as well as the figure they considered ideal. A discrepancy between the two indicated body image dissatisfaction.[11]

The questionnaire was pilot tested among 20 students studying in the tenth class from one of the schools which gave permission but was not included in the study. All the questionnaires were translated from English to Hindi and back to English to ensure accuracy. Students were given a choice to fill out the questionnaires in Hindi or English.

Ethical clearance certificate was obtained from the Institutional Ethics Committee (IEC) prior to the start of the study. Students who screened positive for depression were referred to a tertiary care center and counseled to seek treatment. Data was entered into MS Excel and analyzed using Statistical Package for Social Sciences (SPSS) Version-25. Categorical variables were analyzed as proportions. Univariate analysis was used to find the association between different socio-demographic/lifestyle-related variables and DAS. P value less than 0.05 was considered significant. Variables thus found to be significant were entered into binary logistic regression model. Results of multivariate analysis have been reported as odds ratio with 95% confidence interval.

RESULTS

The study was conducted among 392 school-going adolescents in Delhi NCR. The participants’ age ranged from 12 to 18 years with a mean (SD) age of 15.25 (1.43) years. Table 1 depicts the proportion of sociodemographic variables such as gender, age, family type, and history of physical or mental illness in the subject or families.

Table 1.

Descriptive and univariate analysis of sociodemographic variables between patients with and without depression

Variable	Categories	Total (n=392)	Depression present (n=158)	Chi-squared (df=1)	P
Gender	Female	194 (49.5%)	88 [45.4%]	4.08	0.043
	Male	198 (50.5%)	70 [35.4%]
Age group	16 to 18	183 (46.7%)	85 [46.4%]	5.38	0.020
	12 to 15	209 (53.3%)	73 [35%]
Class	11th, 12th	151 (38.5%)	78 [51.7%]	13.15	<0.001
	8th to 10th	241 (61.5%)	80 [33.2%]
Family	Nuclear	283 (72.2%)	115 [40.6%]	0.46	0.83
	Joint	109 (27.8%)	43 [39.4%]
Siblings	Absent	26 (6.7)	16 [61.5%]	5.22	0.022
	Present	366 (93.4%)	142 [38.8%]
Mother’s occupation	Working	76 (19.4%)	39 [51.3%]	4.75	0.029
	Housewife	316 (80.6%)	119 [37.7%]
Socioeconomic status	Lower middle or less	138 (35.2%)	59 [42.8%]	0.53	0.467
	Upper middle and above	254 (64.8%)	99 [39%]
History of somatic illness in subject	Present	15 (3.8%)	6 [40%]	0.001	0.980
	Absent	377 (96.2%)	152 [40.3%]
History of mental illness in subject	Present	44 (11.2%)	34 [77.3%]	28.15	<0.001
	Absent	348 (88.8%)	124 [35.6%]
History of somatic illness in family	Present	64 (16.3%)	31 [48.4%]	2.10	0.147
	Absent	328 (83.7%)	127 [38.7%]
History of mental illness in family	Present	62 (15.8%)	49 [79%]	45.90	<0.001
	Absent	330 (84.2%)	109 [33%]

df=Degree of freedom. Values are presented as number (%). P values are calculated from Chi-squared tests. Simple brackets () represent column-wise percentage. Square brackets [] represent row-wise percentage

The participants’ mean (SD) BMI was 21.57 (2.3). Three-fifth of the subjects engaged in adequate exercise, defined as at least one hour of physical activity everyday over the past two weeks. The average time spent exercising was 6.5 hours a week. Almost two-thirds of them reported poor sleep quality and the mean (SD) PSQI score was 6.635 (3.28). Of the 271 (69.1%) participants who had BIIs, almost half of them wanted to have a larger body type while half wanted a smaller body type. Boys more commonly wanted to have a larger body, whereas females expressed a desire to be thinner more frequently. Cigarettes were the most commonly abused substance, followed by alcohol. However, in participants’ parents, alcohol was the most common substance abuse. Poor sleep quality, meal skipping, and substance abuse were more common in girls, whereas BIIs were more common in boys. Adequate physical activity was also more common in boys. However, these differences were not statistically significant [Table 2].

Table 2.

Descriptive and univariate analysis of lifestyle variables between patients with and without depression

Variable	Categories	Total (n=392)	Depression present (n=158)	Chi-squared value (df=1)	P
Substance abuse by subject	Present	13 (3.3%)	10 [76.9%]	7.49	0.006
	Absent	379 (96.7%)	148 [39.1%]
Substance abuse in parents	Present	95 (24.2%)	52 [54.7%]	10.85	0.001
	Absent	297 (75.8%)	106 [35.7%]
Exercise	Inadequate	154 (39.3%)	90 [58.4%]	34.67	<0.001
	Adequate	238 (60.7%)	68 [28.6%]
Sleep quality (PSQI)	Poor	287 (73.2%)	145 [50.5%]	46.48	<0.001
	Good	105 (26.8%)	13 [12.4%]
BMI	Overweight and above	161 (41.1%)	75 [46.6%]	4.47	0.034
	Underweight and normal	231 (58.9%)	83 [35.9%]
Body image issues (Stunkard Figure Rating)	Present	271 (69.1%)	125 [46.1%]	12.36	<0.001
	Absent	121 (30.9%)	33 [27.3%]
Meal skipping	Present	99 (25.3%)	59 [59.6%]	20.48	<0.001
	Absent	293 (74.7%)	99 [33.8%]
Can discuss thoughts and worries with family	No	123 (31.4%)	78 [63.4%]	39.78	<0.001
	Yes	269 (68.6%)	80 [29.7%]
Recent stressful events	Present	167 (42.6%)	88 [52.7%]	18.56	<0.001
	Absent	225 (57.4%)	70 [31.1%]

df=Degree of freedom, PSQI=Pittsburgh Sleep Quality Index. Values are presented as number (%). P values are calculated from Chi-squared tests. Simple brackets () represent column-wise percentage. Square brackets [] represent row-wise percentage

Of the total participants, 158 (40.3%) screened positive for depression. Meanwhile, 186 (47.5%) screened positive for anxiety and 87 (22.2%) for stress. Fifty (12.8%) students were suffering from both depression and anxiety, and 23 (5.6%) had depression as well as stress. Twelve (3%) students were diagnosed with all three. Table 3 shows the distribution of various categories of DAS in the population.

Table 3.

Distribution of depression, anxiety, and stress in males and females as per DASS-21

Disorder	Mild (DASS score 10-13)	Moderate (DASS score 14-20)	Severe (DASS score 21-27)	Extremely severe (DASS score >27)	Total
Depression
Males (%)	41 (20.7%)	13 (6.6%)	9 (4.5%)	7 (3.5%)	70 (35.4%)
Females (%)	32 (16.5%)	29 (14.9%)	15 (7.7%)	12 (6.2%)	88 (45.4%)
Total (%)	73 (18.6%)	42 (10.7%)	24 (6.1%)	19 (4.8%)	158 (40.3%)
Anxiety
Males (%)	25 (12.6%)	37 (18.7%)	11 (5.6%)	9 (4.5%)	82 (41.4%)
Females (%)	30 (15.5%)	37 (19.1%)	17 (8.8%)	20 (10.3%)	104 (53.6%)
Total (%)	55 (14%)	74 (18.9%)	28 (7.1%)	29 (7.4%)	186 (47.5%)
Stress
Males (%)	15 (7.6%)	8 (4%)	8 (4%)	0 (0%)	31 (15.7%)
Females (%)	16 (8.2%)	22 (11.3%)	13 (6.7%)	5 (2.6%)	56 (28.9%)
Total (%)	31 (7.9%)	30 (7.7%)	21 (5.4%)	5 (1.3%)	87 (22.2%)

DASS-21=Depression anxiety stress scale-21. Values are presented as number (%)

Depression was significantly higher in girls, older students, and those studying in higher classes. Participants whose mothers were engaged in jobs and those who lacked siblings or had a history of diagnosed mental illness in themselves or their family showed higher levels of depression [Table 1].

Depression was also found to have positive correlations with all the lifestyle variables assessed, namely substance abuse in subject or parents, inadequate exercise, poor sleep quality, BIIs, and skipping meals. Those who discussed their thoughts and worries openly with their family showed lower levels of depression [Table 2].

After binary logistic regression, significant higher odds ratios were seen between depression and family history of mental illness, open discussion of thoughts and worries with family, exercise, skipping meals, and poor sleep quality. However, the confidence intervals are wide, indicating that margin of error may be high [Table 4].

Table 4.

Binary logistic regression depicting factors significantly associated with depression

Category		Significance (P)	OR	95% CI lower limit	95% CI upper limit
Family history of mental illness	Present	0.002	3.439	1.593	7.425
Discuss thoughts and worries with family	No	<0.001	3.362	1.911	5.915
Recent stressful events	Present	0.051	1.735	0.997	3.016
Exercise	Inadequate	<0.001	4.053	2.313	7.103
Meal skipping	Present	0.014	2.128	1.163	3.895
Sleep quality	Poor	<0.001	3.962	1.932	8.125

Nagelkerke R²=0.471

DISCUSSION

In the index study, 40.3% participants screened positive for depression, 47.5% for anxiety, and 22.2% for stress. Depression was more common in females and older adolescents. Parental education, family type, and socioeconomic status did not show significant correlations. However, participants who lacked siblings and had working mothers showed higher levels of depression, possibly reflecting a less robust familial support system. Meanwhile, depression was less common in participants who felt confident about discussing their thoughts and worries with their family. A personal or family history of mental illness as well as facing adverse events also emerged as risk factors. Almost three-fourths of the participants had poor sleep quality. It was worse in older adolescents and correlated with depression. BIIs, meal skipping, and low levels of exercise too showed positive associations with depression. Although depression was higher in subjects who consumed drugs, this was not identified as an independent risk factor.

The prevalence of depression in the present study is similar to a study conducted in Bangalore.[4] Arguably, the most consistent finding in studies evaluating depression is its greater prevalence in females than males.[3,6,7,16,22,27,28] A meta-analysis by Salk et al. concluded that the female preponderance of depression emerges by puberty and subsequently declines.[29] Thus, it has been attributed tentatively to menarche.[8] Thapar et al. suggested that estrogen increases stress sensitivity in the prefrontal cortex, placing girls at greater risk of depressive disorders. This is supported by the finding that depression is associated with hormonal changes rather than chronological age in adolescent girls.[3] Additionally, gender inequality could also lead to increased stress exposure and lower self-esteem in females, both of which contribute to depression.[14]

A global-level meta-analysis carried out by Racine et al. during COVID-19 pandemic reported that depression became more common with increasing age.[14] Likely, older adolescents rely more on social interaction with peers for support, thereby experiencing greater distress during isolation.[14] Studies conducted before the pandemic had similar findings, possibly attributable to academic stress.[5,7] Concurrently, the index study found higher levels of depression in students of 11^th and 12^th, who face more academic stress. In India, higher levels of depression have been reported in students of board classes (10^th and 12^th) compared to non-board classes.[7,28] Contrarily, these differences were not significant in the present study, possibly because the cancellation of board exams due to the pandemic abated this difference.

Contrary to the present study, Bhasin et al. did not find an association between depression and mother’s occupation. However, they concurred that students who could express their thoughts and worries openly with their families had lower levels of DAS.[28] This indicates a pressing need for families to nurture an environment in which children can comfortably express their thoughts without fear of judgment or retribution. Previous studies have found higher levels of depression in adolescents with a past history and family history of mental illness.[3] Bhasin et al. also agreed that higher levels of depression were significantly associated with a history of facing adverse events in the past year.[28] Events such as the death of a loved one, or a relationship breakup have been noted as risk factors for depression.[1] Such events are associated with first onset of depression more than recurrence.[3] Girls and adolescents who face multiple such events are at greater risk.[3] It is imperative that schools recognize such events as risk factors for poor mental health and give special care to students who face this risk.

Singh et al. reported a higher prevalence of sleep deprivation, but a similar association with age.[9] Moitra et al. reported a lower prevalence of the same using a novel questionnaire.[5] Sleep deprivation and depression proposedly have a bidirectional relationship, with poor sleep being a symptom of as well as antecedent to depression.[30] Both Indian and international studies have documented a similar correlation in adolescents and young adults.[5,9,10,13] Pubertal phase delays combined with lifestyle changes such as earlier school timings in teenage years lead to changes in sleep habits.[10,31] This can be detrimental since sleep plays an essential role in cognition and emotional processing.[31] Sleep-deprived adolescents also face more difficulty with social interaction, which can lead to isolation from peers and subsequent loneliness. Moreover, cognitive bias causes these adolescents to selectively remember negative experiences, culminating in depressed mood.[32]

Body image refers to a person’s perception of their body.[11] A previous Indian study found a higher prevalence of BIIs than the present study. In past studies, boys were more commonly dissatisfied with thinness, whereas girls majorly expressed a desire to be thinner.[11,33] Indian and international studies have also correlated BIIs with depression reproduced in the past.[11,27,34] It has been suggested that depression and body image dissatisfaction often act as coexisting mental disorders.[33] Adolescents who are more satisfied with their body type also exhibit more prosocial behavior, which is linked with lower levels of depression.[8] Additionally, depressed individuals often have a relatively sedentary lifestyle and poorer food choices, which predispose them to BIIs.[33]

Prior studies have also correlated skipping breakfast with depression in adolescents and adults.[12,34] Food insecurity has also been identified as a significant predictor of depression. A possible explanation for this protective effect of meals is that breakfast increases blood glucose which reduces adrenaline and cortisol.[12] A study also showed that adolescents who consume a high-quality breakfast report lower levels of depression than their counterparts who have a poor-quality breakfast.[35] This suggests that healthy eating, not just eating is essential for mental wellbeing. A previous European study found a lower prevalence of adequate physical activity than the index study, but a similar positive association between lack of exercise and depression.[13,21] Studies have suggested a bidirectional relationship between depression and obesity, which may also reflect unhealthy lifestyle.[36]

CONCLUSION

Depression affects a high number of adolescents and shows a significant correlation with biological factors like gender as well as facing adverse events. Lifestyle factors such as poor sleep quality, BIIs, and unhealthy habits such as meal skipping and lack of exercise were also associated with depression.

Limitations

The present study carries certain limitations. Since it is a cross-sectional study, causal relations cannot be established. Factors such as sleep quality and drug use are suggested to have a bidirectional relationship with disorders like depression. Cause-consequence discrepancies could not be clarified by the present study. As the study was carried out in the aftermath of COVID-19, several schools were not open and attendance was limited, restricting the number of students enrolled. Moreover, since the study was originally a Short-Term Studentship project under the Indian Council of Medical Research, it was conducted in 2 months as per their guidelines. The time constraint also limited the number of schools that could be contacted, and consequently, only two schools were included in the study. Like most interview-based studies, the participants had to answer several questions about their recent habits and lifestyle based on recall, potentially increasing recall bias. Further, screening tools based on self-reporting were used for assessment and could not be followed up by a clinical interview. Screening tools are known to overestimate depression, and thus, the actual prevalence may be lower than reported.[37]

Implications and recommendations

This study sheds light upon certain factors that are likely associated with depression in adolescents. Awareness of these associations among families and teachers may enable timely identification of teenagers who are at high risk of depression. Such students could subsequently be advised to carry out lifestyle changes as well as seek appropriate counseling or clinical help, thereby reducing the morbidity of depression in this age group.

Directions for future research

Future research could further explore the correlates of depression in adolescents as well as their healthcare-seeking behavior. Cohort studies may be carried out to determine causal associations between depression and the lifestyle factors suggested by the index study. Experimental studies can also be carried out to design and test innovative intervention strategies to combat this issue.

Key messages

The establishment of poor sleep quality, meal skipping, etc., as risk factors for depression should enable timely identification of high-risk adolescents. This could help predict and prevent depression in a vulnerable population.

Financial support and sponsorship

This study was approved and financially supported by the Indian Council of Medical Research under Short-Term Studentship program.

Conflicts of interest

There are no conflicts of interest.