Urbanisation negatively impacts sleep health and mood in adolescents: a comparative study of female students from city and rural schools of North India

Abstract

This study investigated the impact of social settings on sleep, physical and mental health in female adolescents of North India (latitude 29.5 ^oN; longitude 77.5 ^oE). Using a battery of questionnaires, we compared the chronotype, sleep–wake pattern, sleep health (e.g. sleep quality, daytime sleepiness and fatigue) and mood (via depression, anxiety and stress symptoms) in female students (age 14–18 years) from rural (N = 719) and urban (N = 1033) schools separated by about 35 km, but families had almost similar socio-demographic details. The morning type was prevalent amongst rural, whilst the evening type was prevalent amongst urban students who also had access to smart phones, suggesting a possible greater use of the internet. There were greater negative sleep effects, daytime sleepiness, overall poor sleep quality, higher fatigue and anxiety levels in urban than the rural cohort. Interestingly, these measures also differed between school days and free days, suggesting an impact of the conflict between internal biological and social timings (= social jet lag). We also found a significant relationship between chronotype, internet addiction, mood-related parameters and measures of sleep health. Overall, these results suggest a possible impact of social settings on sleep health and mood-related behaviours in female adolescents.

Introduction

The circadian rhythm dependence of behavioural manifestations in humans leads to their categorization broadly into timed phenotypes, called chronotypes [1]. A chronotype refers to an individual’s propensity for being a better performer during the 24-h day. Thus, an evening chronotype individual tends to be relatively more active and alert in the evening (hence a delayed night sleeper), and a morning chronotype individual tends to be more active and alert in the morning (hence an early night sleeper). These two extremes, however, exclude a larger proportion of human population, as shown by their flexibility in both performance and timing of the night sleep [2].

Synchrony of daily sleep–wake pattern, which directly affects both activity and feeding behaviours, with an individual’s chronotype being vital for both health and optimal performances as indicated by overall physical activity and fitness levels [3, 4]. A social setting can impact this, as can be discerned by differences in daily sleep and activity schedules of individuals from urban and rural environments. The urban environment offers a better livelihood opportunity but at the same time promotes reduction in physical activity and night sleep, probably because of longer indoor stays, and spending time on television viewing and in engaging with electronic gadgets, such as computers and smartphones. The screen time or the time spent with electronic gadgets has significantly increased amongst youth, in particular [5–7]. For example, Rideout et al. [8] found > threefold (~ 7.5 h/day instead of recommended 2 h/day) increase in the average screen time in the United States teens [9]. As a consequence, there is the possibility of delayed sleep timing, which is characteristic of an evening chronotype [10]. An association of screen time with the duration and quality of sleep (i.e. irregular bedtimes and night time awakenings) has also been reported in both school age children and adolescents [11, 12]. A study on Brazilian Quilombos community has reported a shorter sleep duration in individuals living with an access to electricity, compared to those living without electricity [13].

The consequence of the disruption in daily activity and sleep schedules due to an urbanised environment is also exhibited in the levels of fatigue and daytime sleepiness, and mood-related (anxiety, depression and stress-like) behaviours [14, 15]. In a recent study, Braçe et al. [16] have reported poor mental health with a shorter night sleep in a Spanish urban population. A higher risk of internet addiction amongst urban, compared to rural, has also been amongst school students reported in Poland [17].

It is also suggested that as compared to men, women are more likely (about 40% higher) to report insomnia symptoms and have worse sleep [18, 19], owing in large part due to differences in sex hormones [20]. There is also a gender difference in prevalence of common mental disorders with their increased frequencies amongst women [21]. Therefore, the present study studied the impact of social settings in school-going female adolescents of North India since such studies are warranted in a developing country like India with consistently increasing rates of urbanisation. In particular, using a battery of questionnaires, we compared the chronotype, sleep–wake pattern, and the overall sleep, physical and mental health of students from rural and neighbouring urbanised locations. The schools varied in terms of their catchment area environments (e.g. access to better electricity, electronic gadgets like television, computers and smartphones), but not in the language and cultural background of students.

Materials and methods

Data collection

This study on school-going female students from two different environmental settings (i.e. a rural and an urbanised area) was conducted in accordance with the guidelines of the Human Ethics Committee, University of Delhi, India (Ref. No. SAT/DU/IECHR/2018-1/12). In total, the data from 5 schools (3 urban and 2 rural areas) were collected during 4 h around middle of the day (1000–1400 h) in August 2018 and December 2018, where there are large numbers of schools for female students in cities, and good-quality schools are fewer in villages. So, for data on rural students, we approached nearby schools in a small town area, which provided an identical village like social environment and where girl students commuted daily for class studies. However, regardless of the locations, both urban and rural schools started at 7:30 a.m.

The three urban area schools situated in Muzaffarnagar, Uttar Pradesh (27.12° N; 79.78° E) represented fully private, private/partially government-aided and fully government-aided as well as co-education and exclusively girl schools. Established in 1963, the first Holy Angels’ Convent School is a private, co-education school providing education from grade lower kindergarten (LKG) to XII. The data from students of this school were collected over 2 days, on the last day of August 2018 and first day of September 2018. However, established in 1955, the second Shiksha Sadan Kanya Inter College is a private/partial government-aided exclusively girl’s school and provides education from grade I to grade XII. The data from students of this school were collected in the first week of December 2018. Likewise, established in 1956, the third Sanatan Dharm Kanya Pathshala Inter College is a fully government-aided girl’s school providing education from grade VI to XII standard. The data from students of this school were collected at the end of the first week of December 2018.

Likewise, the two rural area schools situated in Budhana, Uttar Pradesh (29.17° N; 77.28° E) represented one each government-aided and private school. Established in 1990, the first Swami Kalyan Dev School is a government-aided Senior Secondary girl’s school providing education from grade VI to XII. The data from students of this school were collected in the first week of December 2018. The second school established in 1965 is a private Senior Secondary girl school and provides education from grade VI to XII. The data from students of this school were collected in mid-December 2018.

Subjects

The study sample comprised a total of 1753 school-going female adolescents of age 14–18 years (mean age = 16 ± 2 years) studying in IX, X, XI and XII standards. We were confident that our 14–18-year volunteers easily understood and comprehended all questions of the questionnaire that we used in this study. The participants comprising different subject streams (arts, commerce and science) were regular day scholars with a healthy menstrual cycle. Before the study began, the participants were briefed about the study aim, and their anonymity and confidentiality of data from them were assured. They were also informed that their participation was voluntary (no compensation allocated to subjects for participation in the study), and that they could withdraw from study any time if they so wished.

Protocols and measures

The study protocol was non-invasive and consisted of a battery of already validated and published questionnaires. Using these questionnaires, we assessed the chronotype, sleep-related variables (sleep–wake pattern, sleep health, daytime sleepiness, sleep quality, physical fatigue and social jetlag) and mood-related behaviours (depression, anxiety and stress) (see Fig. 1).

Fig. 1.

Experimental details and measured variables: Upper panel shows urbanisation of our environment and our study participants (school-going female adolescents) chosen from two distinct environments viz. urban and rural (A), and their social schedules (B). Lower panel indicates study variables viz. chronotype (i), mood-related parameters and internet dependence (ii), sleep parameters (iii) and sleep health (iv)

Socio-demographic and general information

A descriptive questionnaire recorded the socio-demographic information. The participants were asked to identify themselves (name and school), and to fill out the information about their age, gender, height (metre), weight (kg), place and time of birth, present inhabiting place (rural or urban), grade (class or standard), and class and commuting times. They were then also asked to rate their academic performance (poor, average, good or excellent), although this was not verified lest it could possibly demotivate them from being part of this study. Finally, they were asked about regularity of their menstrual cycle and if they were on any form of medication.

Munich chronotype questionnaire

Munich chronotype questionnaire was used to assess the chronotype and sleep-related variables. This self-reported questionnaire developed by Till Roenneberg and colleagues [1] consists of 12 questions, and extracts information from participants regarding their sleep–wake schedules both on work and free days. From the reported information, we assessed various primary sleep variables, such as the sleep latency, sleep onset and offset, sleep inertia, duration of sleep, and total time spent in bed. These primary sleep variables led to the determination of secondary variables like the chronotype, MSF_SC (mid-sleep phase of free days), social jet lag (difference in mid-sleep phase between school and free days) and weekly sleep loss (average sleep loss during the week). Based on mid-sleep-phase score of 0–6, the volunteer subjects were categorised into morning type (score 0–2), neither type (score 3) and evening type (score 4–6).

Epworth sleepiness scale

Epworth Sleepiness Scale (ESS) was used to calculate the accumulation of daytime sleepiness. This scale has been used successfully for adolescents in several studies [22, 23], although generally recommended for adolescents is the PDSS scale (see [24]). This self-administered Likert-scale questionnaire format consisted of eight questions that mark the propensity of dozing off in different situations in daily life. Each question gives option of scoring in a range of 0–3, where score 0 represents “no chance of dozing” and score 3 represents a “higher chance of dozing”. Based on the total score range from 0 to 24, the subjects were grouped into 4 categories: absence of daytime sleepiness (score 0–6), average daytime sleepiness (score 7–10), excessive daytime sleepiness (score 11–15), excessive daytime sleepiness with need of medical attention (score 16–24). The higher score corresponds to a higher degree of daytime sleepiness [25].

Pittsburg sleep quality index

Pittsburg Sleep Quality Index (PSQI) presented in a questionnaire format was used to show the sleep quality. This self-administered questionnaire represents a combination of Likert-scale and subjective information. In 19 questions, it collects information about the sleep timings and sleep disturbances experienced by an individual during the previous month. The collected information is used to assess seven components of sleep, which when summed up gives a global score 0–21. The individuals with score ≤ 5 are categorised as ‘good sleepers', whilst those with score > 5 are categorised as ‘poor sleepers’ [26].

Fatigue severity scale

The Fatigue Severity Scale (FSS) was used to calculate the level of physical fatigue. This Likert-scale questionnaire enquires about the level of fatigue experienced in nine different situations during the day with score options ranging from 0 to 7. The score 0 represents the complete agreement, whilst score 7 represents complete disagreement with the given statement. Hence, the final score ranges may total from 0 to 63. In this score, a score ≤ 36 marks the absence of fatigue, whilst a score > 36 marks the physical fatigue [27].

Depression, anxiety and stress scale

We used a depression, anxiety and stress scale (DASS-21) to assess the mood of volunteer subjects. This self-administered scale consisted of 21 questions in three subsets of seven questions each related to the depression, anxiety or stress symptoms. This is a Likert-scale questionnaire where score option varies from 0 to 3; a score 0 indicates that the subject does not experience whilst score 3 indicates experience with regard to mood being asked. Thus, the total score ranges from 0 to 21 for each subset and the cut-off score for the normal, mild, moderate and extreme level is different for each subset. On the basis of score, the individuals are divided into four categories for each subset which are as follows. For depression: 0–9: normal; 10–13 mild: 14–20: moderate; 21–27: severe; 28 + extremely severe; for anxiety: 0–7: normal; 8–9 mild: 10–14: moderate; 15–19: severe; 20 + extremely severe and for stress: 0–14: normal; 15–18 mild: 19–25: moderate; 26–33: severe; 34 + extremely severe. The higher score corresponds to a more severe symptom of depression, anxiety or stress [28].

In addition, we also used an internet addiction scale to assessed the level of internet addiction. The self-administered questionnaire consisted of 20 questions related to disturbances in social life, personal life and emotional vulnerability. This is a Likert-scale questionnaire with option score ranging from 0 to 5. The score 0 represents no addiction, whilst score 5 represents an extreme addiction level. The total score thus ranges from 0 to 100. On the basis of the total score, the individuals are divided into four categories: no addiction (score 0–30), mild level of addiction (score 31–50), moderate level of addiction (score 51–70) and extreme level of addiction (score 71–100) [29].

Statistics

Statistical analyses were performed using IBM SPSS version 20 and Graph Pad Prism version 5.0, as appropriate. We also compared data on BMI and chronotype distribution and grades of urban and rural school-going students using chi-square test. The univariate General linear model (GLM) analysed different sleep variables, with social settings (urban and rural) and days of the week (school days and free days) as the fixed factor, and days of the week x urban/rural as the interaction factor. Student’s unpaired t test compared the scores of social jetlag, fatigue, daytime sleepiness, sleep quality, and the symptoms relation to depression, anxiety and stress as well as internet addiction between urban and rural students. Finally, Spearman rank correlation test determined the relationship of chronotype with internet addiction, sleep health and symptoms related to depression, anxiety, and stress. For statistical significance, alpha was set at 0.05.

Results

Table 1 summarises the socio-demographic variables. Overall, there appears homoscedasticity (homogeneity of variance) in the study sample, as revealed by the data from BMI (21.0 ± 0.14 kg/m²) and menstrual health, and self-reported academic performance (Table 1).

Table 1.

Socio-demographic variables of school-going female adolescents

Variables	Urban		Rural		p value
	N	%	N	%
Number of students	1033	59	719	41
BMI (kg/m2)
Underweight	89	9	76	11	0.6
Normal	897	86	635	88	0.87
Overweight	48	5	8	1	0.1
Menstrual health
Regular	899	87	134	13	0.82
Irregular	644	90	71	10	0.53
Grade
IX	316	31	179	25	0.42
X	252	24	193	27	0.67
XI	255	25	178	25	0.8
XII	210	20	169	23	0.6
Self-reported academic performance
Poor	12	2	11	2	0.6
Average	233	22	65	10	0.5
Good	104	10	99	13	0.5
Excellent	684	66	544	75	0.4

Alpha value < 0.05 was considered significant

Chronotype distribution and sleep-related parameters

There was a significant difference in the chronotype distribution between rural and urban student cohorts. The rural students reported a higher percentage of morning chronotype (59%), and lower percentage of the evening chronotype (14%). Conversely, the urban students reported a lower percentage of morning type (21%) and higher percentage of the evening type (44%). Neither the chronotype percentage was almost similar in both student cohorts (rural = 27%; urban = 35%; see Fig. 2).

Fig. 2.

Differences in the percentage distribution of the chronotype between urban and rural school-going female students as assessed using Munich Chronotype Questionnaire. The empty white bar represents the rural student cohort and filled grey bar represents the urban student cohort. Asterisk (*) indicates a significant difference (p < 0.05) as determined by Chi-square test

As compared to rural cohort, the urban student cohort reported a higher sleep latency (F = 110.13; p < 0.001), later sleep onset (F = 567.23; p < 0.001), later sleep end (F = 104.88; p < 0.001), higher sleep inertia (F = 25.72; p < 0.001), shorter sleep duration (F = 89.15; p < 0.001), and shorter total time spent in bed (F = 36.66; p < 0.001; Fig. 3, Table 2; GLM test). When assessed further the impact of a social schedule, we found an earlier sleep onset (F = 46.50; p < 0.001) and sleep end (F = 1204.42; p < 0.001), and lower sleep inertia (F = 71.08; p < 0.001), shorter sleep duration (F = 583.56; p < 0.001) and shorter total time in bed (F = 584.72; p < 0.001) on school days, compared to those on the free days of the week. The differences in sleep variables between school and free days were also dependent on the social settings, i.e. the urban and rural environment. This was revealed by a significant effect of interaction (environmental setting × school/free days) on sleep onset (F = 12.84; p < 0.001), sleep offset (F = 219.46; p < 0.001), sleep inertia (F = 18.25; p < 0.001), sleep duration (F = 55.066; p < 0.001) and total time spent in bed (F = 67.65; p < 0.001; Fig. 3, Table 2; GLM).

Fig. 3.

Box plot of sleep measures during the school day and free day in urban and rural school-going female students as measured by Munich Chronotype Questionnaire. a Sleep latency, b sleep onset, c sleep duration, d sleep inertia, e sleep end, and f total time in bed. The empty white bar represents the school days and a filled grey bar represents free days of both urban and rural cohorts. Asterisk (*) indicates a significant difference (p < 0.05) as determine by General Linear Model (GLM) test

Table 2.

The model output of GLM univariate analysis showing the effect of environmental setting (urban vs rural) and social schedule (school days vs free days) on sleep–wake parameters

Variables	Factors	F	df	p value
Sleep latency	Rural urban	110.13	1	0.001
	School days/free days	0.50	1	0.479
	Rural urban * days of week	0.64	1	0.426
Sleep onset	Rural urban	567.23	1	0.001
	School days/free days	46.50	1	0.001
	Rural urban * days of week	12.84	1	0.001
Sleep offset	Rural Urban	104.88	1	0.001
	School days/free days	1204.42	1	0.001
	Rural urban * days of week	219.46	1	0.001
Sleep inertia	Rural urban	25.72	1	0.001
	School days/free days	71.08	1	0.001
	Rural urban * days of week	18.25	1	0.001
Sleep duration	Rural urban	89.15	1	0.001
	School days/free days	583.56	1	0.001
	Rural urban * days of week	55.07	1	0.001
Total time on bed	Rural urban	36.66	1	0.001
	School days/free days	584.72	1	0.001
	Rural urban * days of week	67.65	1	0.001

P value in bold indicates a significant (p < 0.05) differences determined by General linear model test

There were also significant differences in other sleep health variables. For example, compared to the rural student cohort, the urban student cohort reported a higher social jetlag (p < 0.05; Student’s unpaired t test), physical fatigue (p < 0.001; Student’s t test), daytime sleepiness (p < 0.001; Student’s unpaired t test) and the overall poor sleep quality (p < 0.001; Student’s unpaired t test; Fig. 4a–d).

Fig. 4.

Mean (± SE) of sleep health (A) and Internet dependence and mood-related parameters (B) of urban and rural school-going female students. a Poor sleep quality as measured by Pittsburgh sleep quality questionnaire, b daytime sleepiness score measured by Epworth sleepiness scale (ESS), c fatigue score measured using fatigue severity scale (FSS), d social jet lag measured by Munich chronotype questionnaire (MCTQ), e internet addiction calculated using internet addiction scale, f depression, g anxiety, h stress symptoms calculated using depression, anxiety and stress scale (DASS-21). An empty bar (white) represents the rural student cohort and filled (grey) bar represents the urban student cohort. Asterisk (*) indicates a significant difference (p < 0.05) as determine by Student’s unpaired t test

Mood-related symptoms

The urban students reported a higher level of stress (p < 0.001; Student’s unpaired t test) and higher depressive-like symptoms (p < 0.001; Student’s unpaired t test), but a lower level of anxiety-related symptoms (p < 0.001; Student’s unpaired t test; Fig. 4f–h). Interestingly, the urban student cohort was also found with a higher level of addiction to the internet (p < 0.001; Student’s unpaired t test; Fig. 4e).

Relationships: chronotype, sleep health measures, and mood-related symptoms

Table 3 describes relationships between chronotype, sleep health measures and mood-related symptoms. There was a significant positive correlation of chronotype with the depressive-like symptoms, internet addiction, stress levels, daytime sleepiness, fatigue, sleep quality and social jet lag. The internet addiction was also positively correlated with chronotype, mood parameters, fatigue and sleep health. Similarly, the mood-related variables showed a significant positive relationship with sleep health.

Table 3.

Correlation matrix showing relationship between chronotype, internet addiction, mood-related parameters and measures of sleep health of school-going female adolescents

	Chronotype	Internet addiction	Depression	Anxiety	Stress	Daytime sleepiness	Fatigue	Sleep quality	Social jet lag
Chronotype
Internet Addiction	*** r = 0.21; p < 0.001
Depression	*r = 0.05; p = 0.04	***r = 0.14; p < 0.001
Anxiety	ns	***r = 0.09; p < 0.001	*** r = 0.49; p < 0.001
Stress	**** r = 0.05; p = 0.03	*** r = 0.18; p < 0.001	*** r = 0.46; p < 0.001	*** r = 0.37; p < 0.001
Daytime Sleepiness	*** r = 0.24; p < 0.001	*** r = 0.07; p = 0.003	ns	*** r = 0.13; p < 0.001	** r = 0.06; p = 0.008
Fatigue	*** r = 0.38; p < 0.001	*** r = 0.24; p < 0.001	*** r = 0.10; p < 0.001	ns	*** r = 0.12; p < 0.001	***r = − 0.43; p < 0.001
Sleep Quality	*** r = 0.20; p < 0.001	*** r = 0.16; p < 0.001	*** r = 0.24; p < 0.001	*** r = 0.19; p < 0.001	*** r = 0.18; p < 0.001	***r = − 0.19; p < 0.001	***r = 0.36; p < 0.001
Social jet lag	*** r = 0.68; p < 0.001	*** r = 0.15; p < 0.001	ns	**r = − 0.07; p = 0.004	ns	***r = − 0.32; p < 0.001	***r = 0.43; p < 0.001	***r = 0.12; p < 0.001

Asterisk (*) indicates significant difference as determined by Spearman rank correlation test

Discussion

Present questionnaire-based study suggests the impact of the social environment on chronotype distribution, sleep variables and mood in school-going female adolescents. We found a higher percentage of morning and evening chronotypes in rural and urban adolescents, respectively, similar to earlier studies in Chinese, Brazilian and Indian populations [30–34]. The prevalence of the morning chronotype amongst rural adolescents is probably because of their longer exposure to direct sunlight [31, 32]. The external light environment can have a direct impact on circadian clock regulated biological functions including the sleep timing. And, a light exposure even at very dim illumination via electronic gadgets during late biological day/early biological night can affect the circadian rhythm and, in turn, the timing of sleep onset to a later hour akin to an evening chronotype [35–37]. We believe that the prevalence of an evening chronotype amongst school-going adolescents is due to limited access of natural light in classrooms during much of the time of day and/ or due exposure to artificial light during much of the early biological night [30–34, 38, 39].

Different environmental settings can act as a determining lifestyle factor [40] and influence the level of physical activity and fitness in both youth and adult [4, 41–43]. This can be even seen in people between urban and rural areas located few kilometres apart in the same geographical area [44]. Limited access of opportunities for physical activity [45] and time spent on electronic devices [3] can also be potent factors contributing to a reduced physical activity.

There were also differences in various sleep variables between school days and free days of the week as well as between urban and rural student cohorts. For example, as compared to school days, we found an increased sleep latency, early sleep onset/end decreased inertia, longer sleep duration and total time in bed on free days. This is perhaps because of differences in the relationship of internal (biological) and external times between two social environments. Understandably, compared to an imposed routine on school days (perhaps leading to a synchronised biological and social timings), free days provide relatively a flexible routine (perhaps “free-running condition); hence subjects may experience each week a ‘social jetlag’ condition [46]. The separation of work days from free days of the week is better discerned in urban than in the rural environment which requires intermittent house-hold working on all days and this probably contributes to differences in the sleep disruption between two social environments. A similar conclusion has been drawn from differences in earlier studies on urban and rural inhabitants of the United States population [33].

We reason that apart from differences in the routine, a higher accessibility to technological advances like the better access of electricity and electronic gadgets (e.g. television, smart phone, etc.) contributes to larger sleep effects in the urban cohort, as also suggested by other studies. It is evident that the use of electronic gadgets was more common amongst urban female students, which possibly led to lesser total time in bed and in turn a shorter sleep duration, consistent with the reported association of reduced sleep duration with screen time in several other previous studies [11, 12, 47]. Interestingly, a late sleeping behaviour with sleep debt attributed to cultural characteristics, climate differences and harmful sleep habits, has also been found amongst Iranian preschool children from an urban area [48]. A Brazilian study also reported later bedtimes and wake-ups on both school days and weekends in adolescent students having electricity at home, as compared to those who did not have it [30]. A similar higher rate of abnormal sleep behaviour (e.g. sleep delays, shorter average sleep duration) was found amongst Chinese urban student population [15]. Interestingly, there can be differences between urban and rural populations (shorter sleep duration, later sleep onset and end times in urban dwellers) even though the subjects residing in both areas had same origins (i.e. geographical location and place of birth) [49].

In general, the poor quality of sleep correlates with the physical fatigue level, negative mood and depressive-like symptoms [50]. In the present study, urban student cohort with poor sleep quality as suggested by delayed sleep onset and daytime sleepiness had higher fatigue level, similar to the results from other studies [51–53]. These urban students also reported stress-like symptoms, akin to higher stress levels and severe internalising problems reported from a study amongst the United States urban youths, compared to their rural counterparts [54]. At the same time, however, the anxiety levels were high in rural student cohort, perhaps due to fear for their success because of access to relatively limited resource opportunities [55]. Both cohorts though also reported mild depressive-like symptoms, perhaps feeling a similar failure and/ or peer pressure. The mental health issues in young females were positively correlated with poor sleep health, as reported in studies on young urban Indian students [56], Chinese males [57] and undergraduate Singaporean students [58].

The technology advances seem to accelerate these tendencies, as can be speculated by differences between urban and rural student cohorts. Understandably, compared to rural student cohort, the urban cohort in general had better access to electronic gadgets (smart phones, computers at home) as well as 24-h internet connection. We do not discount if the television viewing is a factor affecting sleep; for example, a study showed positive correlation of television at home with delayed bedtime during the school day in a Brazilian community [30]. In Indian context, television viewing, which is still more prevalent in urban areas due to socio-economic reasons, is a family activity during late evening hours when popular programmes are generally broadcasted. We therefore suggest that the sleep behaviour amongst 14–18-year adolescents susceptible to hormonal changes and peer pressure is influenced by the social environment. A recent study found that COVID-19 imposed social restrictions affected both the body clock and sleep behaviour. Very interestingly, a relaxed social time pressure during the pandemic promoted more sleep, less social jet lag, and less use of the alarm clock for daily wake-ups [59].

Although not tested in this study, sleep also influences learning and memory; even a short-term sleep deprivation impairs the memory consolidation, as shown in mice [60, 61]. The sleep debt accumulated over the workdays can have serious health implications; for example, extended sleep-deficit periods in individuals may result in excessive daytime sleepiness, fatigue, irritability, low concentration, mood swings, anxiety, depression and demotivation [62, 63].

Conclusions and limitations

The present study showed relationships between chronotype, measures of sleep health and mood-related parameters. In general, an evening chronotype seemed to have a greater internet addiction, poor sleep quality, daytime sleepiness, increased fatigue, and higher depressive- and stress-like symptoms. There were positive correlations of the internet addiction scores with depression, anxiety and stress scores and poor sleep health. These results provide an insight into how different social environments (e.g. urban and rural settings) even lying within a close geographical location can have an impact on the sleep and mood behaviours amongst adolescents. We speculate that the disruption of good sleep–wake pattern possibly by exposure to artificial light at night or spending time with electronic gadgets (mobile phone internet) negatively impacts sleep health and mood in humans.

The present conclusions may be considered with a few caveats. First, we recognise that the present study was done on town and rural settings, one each, and did not include the information on (i) when and wherefrom the present town settlers have migrated, (ii) the family size and total family income, and (iii) the participation in extracurricular activities. Notably though, as per the 2011 census of the Ministry of Statistics and Programme Implementation, Government of India, a rural–urban divide does exist in adult literacy rate for both sexes as follows: females, rural/urban—50.6%/76.9%; males, 74.1%/88.3%. Second, instead of the generally recommended PDSS scale for adolescents, we used ESS scale. We emphasise, however, that ESS scale has been used successfully in adolescents in previous studies (see above), and we believe that our volunteers in the 14–18-year age group understood the questions included in the ESS questionnaire.

Author Contributions

Conceptualization and supervision: VK; Investigation, methodology, data curation and analysis: NS, NAJ and VK; Visualisation and writing: NAJ, NS and VK; Reviewing and editing of manuscript drafts: VK and NAJ; Final version: VK and NAJ; Revision: VK and NAJ. All authors approved the final version of the manuscript.

Funding

VK is supported currently by a UGC-BSR faculty fellowship award (# F.No.26-13/2020(BSR).

Declarations

Conflict of interest

Authors declare no conflicting interest.