Seven-year survey of sleep timing in Russian children and adolescents: chronic 1-h forward transition of social clock is associated with increased social jetlag and winter pattern of mood seasonality

Abstract

Previous studies indicate that solar clock (daily changes in the Earth’s surface illumination) is a main zeitgeber for human circadian system. It has been shown that human biological clock is weakly adjusted to such changes in social clock as daylight saving time (DST). There are two changes of social clock in Russian Federation: on 25 March 2011, DST has been replaced by permanent DST (DSTp), which was subsequently revoked on 26 October 2014 (non-DSTp). These manipulations with social clock may lead to prolonged disturbances of human circadian system. Our hypothesis is that during period of DSTp, the dissociation between social and biological clocks was greatest as compared with DST and non-DSTp periods. Here, we examine the effects of DSTp on the sleep timing, social jetlag (SJL), academic performance, and winter and summer seasonality of mood and behavior of 10–24-year-old inhabitants of European North of Russia. A cross-sectional retrospective analysis of questionnaire data (n = 7968) was performed using chi squared-test and analysis of covariance. Our findings indicate that SJL (F2,7967 = 31.9; p < 0.0001; η2 = 0.009), and winter pattern of mood seasonality (χ2 2,7967 = 10.5; p < 0.01) were increased in adolescents during the period of DSTp as compared with DST and non-DSTp periods. The largest increase in SJL was occurred in ages between 10 and 17-year-olds. The finding suggests that increase in SJL can be attributed to a later rise time on free days (F2,7967 = 44.9; p < 0.0001; η2 = 0.012). Similar changes were observed in three subsamples obtained in Syktyvkar, Petrozavodsk, and Vorkuta. Effect sizes of studied relationships were small or very small. The greatest effect sizes (η2 ∼ 0.05) were observed in Arctic city of Vorkuta indicating that in polar region, solar clock is still stronger zeitgeber for human circadian system, than the social clock. In conclusion, we have shown for the first time that there is a greatest dissociation between social and biological clocks during the period of DSTp which potentially exerts a negative influence on adolescents’ sleep habits, mood, and behavior. Our data indicate that “non-DSTp” social clock system most suitable for prevention dissociation between social and biological clocks.

Full text

Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=nbrr20
Download by: [95.128.138.59] Date: 06 February 2017, At: 12:17
Biological Rhythm Research
ISSN: 0929-1016 (Print) 1744-4179 (Online) Journal homepage: http://www.tandfonline.com/loi/nbrr20
Seven-year survey of sleep timing in Russian
children and adolescents: chronic 1-h forward
transition of social clock is associated with
increased social jetlag and winter pattern of mood
seasonality
Mikhail F. Borisenkov, Tatyana A. Tserne, Alexander S. Panev, Ekaterina S.
Kuznetsova, Natalia B. Petrova, Vladimir D. Timonin, Sergey N. Kolomeichuk,
Irina A. Vinogradova, Maria S. Kovyazina, Nikita A. Khokhlov, Anna L. Kosova
& Olga N. Kasyanova
To cite this article: Mikhail F. Borisenkov, Tatyana A. Tserne, Alexander S. Panev, Ekaterina
S. Kuznetsova, Natalia B. Petrova, Vladimir D. Timonin, Sergey N. Kolomeichuk, Irina
A. Vinogradova, Maria S. Kovyazina, Nikita A. Khokhlov, Anna L. Kosova & Olga N.
Kasyanova (2017) Seven-year survey of sleep timing in Russian children and adolescents:
chronic 1-h forward transition of social clock is associated with increased social jetlag
and winter pattern of mood seasonality, Biological Rhythm Research, 48:1, 3-12, DOI:
10.1080/09291016.2016.1223778
To link to this article: http://dx.doi.org/10.1080/09291016.2016.1223778
Published online: 31 Aug 2016. Submit your article to this journal
Article views: 19 View related articles
View Crossmark data

BIOLOGICAL RHYTHM RESEARCH, 2017
VOL. 48, NO. 1, 312
http://dx.doi.org/10.1080/09291016.2016.1223778
Seven-year survey of sleep timing in Russian children and
adolescents: chronic 1-h forward transition of social clock is
associated with increased social jetlag and winter pattern of
mood seasonality
Mikhail F. Borisenkova, Tatyana A. Tsernea, Alexander S. Paneva,b,
Ekaterina S. Kuznetsovaa,c, Natalia B. Petrovab, Vladimir D. Timoninb,
Sergey N. Kolomeichukd, Irina A. Vinogradovae, Maria S. Kovyazinaf,
Nikita A. Khokhlovf, Anna L. Kosovag and Olga N. Kasyanovah
aInstitute of Physiology, Komi Science Centre, Ural Branch, Russian Academy of Science, Syktyvkar, Russia;
bInstitute of Natural Sciences, Pitirim Sorokin Syktyvkar State University, Syktyvkar, Russia; cInstitute of
Pedagogy and Psychology, Pitirim Sorokin Syktyvkar State University, Syktyvkar, Russia; dInstitute of Biology,
Karelian Science Centre, RAS, Petrozavodsk, Russia; eMedical Institute, Petrozavodsk State University,
Petrozavodsk, Russia; fDepartment of Psychology, Lomonosov Moscow State University, Moscow, Russia;
gInstitute of the North Industrial Ecology Problem, Kola Science Centre, RAS, Apatity, Russia; hPsychology
Office, Middle School No. 15, Apatity, Russia
ABSTRACT
Previous studies indicate that solar clock (daily changes in the Earth’s
surface illumination) is a main zeitgeber for human circadian system.
It has been shown that human biological clock is weakly adjusted
to such changes in social clock as daylight saving time (DST). There
are two changes of social clock in Russian Federation: on 25 March
2011, DST has been replaced by permanent DST (DSTp), which
was subsequently revoked on 26 October 2014 (non-DSTp). These
manipulations with social clock may lead to prolonged disturbances
of human circadian system. Our hypothesis is that during period
of DSTp, the dissociation between social and biological clocks was
greatest as compared with DST and non-DSTp periods. Here, we
examine the eects of DSTp on the sleep timing, social jetlag (SJL),
academic performance, and winter and summer seasonality of mood
and behavior of 10–24-year-old inhabitants of European North of
Russia. A cross-sectional retrospective analysis of questionnaire data
(n = 7968) was performed using chi squared-test and analysis of
covariance. Our ndings indicate that SJL (F2,7967=31.9; p<0.0001;
η2= 0.009), and winter pattern of mood seasonality (χ22,7967= 10.5;
p<0.01) were increased in adolescents during the period of DSTp
ascompared with DST and non-DSTp periods. The largest increase in
SJL was occurred in ages between 10 and 17-year-olds. The nding
suggests that increase in SJL can be attributed to a later rise time on
free days (F2,7967=44.9; p<0.0001; η2=0.012). Similar changes were
observed in three subsamples obtained in Syktyvkar, Petrozavodsk,
and Vorkuta. Eect sizes of studied relationships were small or very
small. The greatest eect sizes (η2∼ 0.05) were observed in Arctic
© 2016 Informa UK Limited, trading as Taylor & Francis Group
KEYWORDS
Permanent daylight
saving time; children and
adolescents; social jetlag;
academic performance;
winter pattern of mood
seasonality
ARTICLE HISTORY
Received 6 July 2016
Accepted 8 August 2016
CONTACT Mikhail F. Borisenkov borisenkovm@yandex.ru

4 M. F. BORISENKOV ET AL.
city of Vorkuta indicating that in polar region, solar clock is still
stronger zeitgeber for human circadian system, than the social clock.
In conclusion, we have shown for the rst time that there is a greatest
dissociation between social and biological clocks during the period
of DSTp which potentially exerts a negative inuence on adolescents’
sleep habits, mood, and behavior. Our data indicate that “non-DSTp”
social clock system most suitable for prevention dissociation between
social and biological clocks.
Introduction
Social jetlag (SJL) is a misalignment between the social and the biological clock that is
widespread among residents of industrialized countries (Roenneberg and Merrow 2007).
Social jetlag has been shown apparently associated with poor academic performance
(Haraszti et al. 2014; van der Vinne et al. 2015), depression (Levandovski et al. 2011), and
obesity (Roenneberg et al. 2012). Previously, it was shown that human circadian system is
purely adjusted to daylight saving time (DST) (Kantermann et al. 2007) and the increases in
the sizes of time zones increase the risk of SJL (Borisenkov 2011). Therefore, such changes
in social clock have the potential to adversely aect the cognitive function, well-being, and
health of aected populations.
In the Russian Federation, two changes were made to the social time system over a short
period. On 25 March 2011, the country replaced DST with permanent DST (DSTp). This DSTp was
subsequently revoked on 26 October 2014 (non-DSTp). However, for more than 3 years, the
population of the Russian Federation observed DSTp (Table 1; Figure 1). The aim of the present
study was to evaluate the eects of chronic 1-h forward and backward transition of social clock
on sleep timing, academic performance, and mood seasonality of Russian residents.
Materials and methods
Subjects and instruments
This study was approved by the ethical committee of the Institute of Physiology of Komi
Science Centre, the Ural Branch of the Russian Academy of Sciences. A cross-sectional ret-
rospective analysis was conducted on continuous data accumulated over a 7-year period
(from 2009 to 2016). The study participants were enrolled in schools and universities located
in 14 settlements in the European North of Russia. The data were obtained from 7968 par-
ticipants aged 10 to 24 years. For each participant, personal data (i.e., date, place of residence,
age, sex, height, weight, and academic performance) were collected. For details, see Figure 1
and Table 2. Chronotype, the degree of SJL, and sleep characteristics were estimated using
the Munich ChronoType Questionnaire (Roenneberg et al. 2003). A mid-sleep phase shift
between weekdays and weekend of ≥ 1 h and ≥ 2 h were used as the criteria for SJL ≥ 1 h
and SJL ≥ 2 h group assignment, respectively. Winter and summer pattern of seasonality of
mood and behavior were assessed using the Seasonal Pattern Assessment Questionnaire
Table 1.Description of time periods studied.
#Period The social time system Abbreviation
I Before 25 March 2011 Daylight saving time DST
II 26 March 2011–26 October 2014 Permanent DST DSTp
III After 27 October 2014 Permanent non-DST non-DSTp

BIOLOGICAL RHYTHM RESEARCH 5
according to criteria described by Rosen and colleagues (1990). For details, see Table 3 and
our previous publication (Borisenkov et al. 2015).
DST Non-DST
Figure 1.The time and place of the study. Upper panel: Scheme illustrating daylight saving time (DST) and
non-DST social time system. Middle panel: The study was carried out at three periods. I: from 01 January
2009 to 26 March 2011 – during the period DST (i.e. DST acting yearly from April to October); II: from 27
March 2011 to 26 October 2014 – during the period of permanent DST (DST
P
) (i.e. DST acting permanently
throughout the year); III: from 27 October 2014 until the present – when DST was canceled (non-DSTp)
(i.e.non-DST acting permanently throughout the year). Columns indicate the number of persons surveyed.
Lower panel: The study was carried out in 14 settlements (for details see Table 2) located in the northern
European Russia. The sizes of circles are proportional to the number of persons surveyed. Data on sex
ratio are also presented in the scheme.

6 M. F. BORISENKOV ET AL.
Statistical analysis
All the data were divided into three periods as described in Table 1. A chi squared-test was
used for analysis of inuence of periods on non-parametric variables. Two series of one way
analyses of covariance (ANCOVA) were performed for normally distributed variables. Firstly,
we performed the analysis using “period” as xed factor, “anthropometric characteristics”,
“school start time” (SST), “achievement”, and “Global Seasonality Scores” (GSS) as dependent
variables, and “month of survey”, “year of survey”, “settlement population”, “latitude”, “longi-
tude”, “time of sunrise”, “day length”, and “age” as covariates. The variables “BMI” and “settle-
ment population” are non-normally distributed, therefore, we used normalized “lnBMI” and
“ln(settlement population)” in analysis. There is a certain imbalance in age, anthropometric
characteristics and SST in the three groups studied (Table 4) Therefore, the above charac-
teristics were included in the second set of ANCOVAs as covariates. Secondly, one way
ANCOVAs were performed using “period” as xed factor, “chronotype, SJL, and sleep
Table 2.Short characteristics of place of residence and subject of study.
Notes: 1: Moscow; 2: Saint-Petersburg; 3: Petrozavodsk; 4: Apatity; 5: Murmansk; 6: Mutnitsa; 7: Kazhym; 8: Ob’yachevo;
9: Syktyvkar; 10: Ust-Kulom; 11: Timsher; 12: Izhma; 13: Inta; 14: Vorkuta. Lat.: latitude; Long.: longitude; M (SD): mean
(standard deviation); F/M: sex ratio; N: number of persons surveyed.
Settlement Coordinate Months and periods of survey
Age, M (SD) N
#Population Lat. Long. I II III
1 11980000 55.8 37.6 – 1–5,10–12 2,3,10–12 19.4 (2.2) 311
2 5190000 59.8 30.3 – 10,12 2–4,10–12 17.4 (4.3) 76
3 270000 61.8 34.3 – 1–6,10–12 1,10–12 14.9 (2.5) 2537
4 60000 67.6 33.4 3,10 – 3 15.4 (2.7) 562
5 300000 69.0 33.1 – – 3,4,12 15.2 (2.9) 135
6 800 59.5 49.5 3,5,10 – – 14.5 (2.0) 245
7 1000 60.3 51.5 4 – – 14.1 (2.2) 76
8 6000 60.2 49.4 – 1–4,12 – 14.7 (1.9) 97
9 250000 61.7 50.9 1–5,10–12 1–6,9–12 1–6,9–12 18.0 (3.2) 2127
10 5000 61.7 53.7 3–5 – – 13.2 (2.1) 97
11 1200 61.9 55.1 1,2 – – 14.5 (2.1) 79
12 4000 65.0 53.9 4–6 4,5 – 15.7 (1.8) 126
13 40000 66.0 60.1 1,3,12 – 12 14.0 (1.9) 431
14 87000 67.4 64.1 4 1,4,11,12 1,3,10,12 14.4 (1.7) 1069
Total 15.9 (3.1) 7968
Table 3.Criterions for winter and summer patterns of seasonality assessment (Rosen et al. 1990).
Notes: We assessed the type (winter, summer) and severity (SAD, sub-SAD, or no-SAD) of seasonal affective disorder in study
participants using the responses on the items #11, 12 and 17 of SPAQ. Item#11 – Global Seasonality Score (GSS). Changed
from 0 to 24 scores.
Item#12 – “At what time of year do you feel the worst?” Changed from 1 (Jan.) to 12 (Dec.).
Item#17 – Seasonal changes to be problematic. Changed from: 0 (none), 1 (mild), 2 (moderate), 3 (marked), 4 (severe),
5 (disabling).
Pattern
Item
#11 #17 #12
SADw ≥10 ≥2
sub-SADw ≥10 0, 1 1, 2, 12 (plus any other combination of months excluding 6, 7, 8)
sub-SADw 8, 9 ≥0
SADs ≥10 ≥2
sub-SADs ≥10 0, 1 6, 7, 8 (plus any other combination of months excluding 1, 2, 12)
sub-SADs 8, 9 ≥0
no-SAD Respondents who failed to meet the criteria defined above for SAD or sub-SAD

BIOLOGICAL RHYTHM RESEARCH 7
characteristics” as dependent variables, and “moth of survey”, “year of survey”, “settlement
population”, “latitude”, “longitude”, “time of sunrise”, “day length”, “age”, “weight”, “height”, “BMI”,
and “SST” as covariates. The variables “BMI”, “settlement population”, “sleep onset latency on
week and free days” (SOLW(F)), “sleep inertia” (SlIW(F)), “SJL”, and “chronotype” (MSFsc)
are non-normally distributed, therefore we used normalized “lnBMI”, “ln(settlement popu -
lation)”, “SqrSOLW(F)”, “ SqrSlIW(F)”, “ lnSJL”, and “lnMSFsc” in analysis.
Results
Social jetlag ≥ 1 and 2 h was observed in 86.4 and 59.3% of individuals surveyed, respectively
(Figure 2A). During the period of time corresponding to DSTp, the proportions of people with
SJL ≥ 1 h and 2 h increased by 6.7% (χ22,7967 = 92.3; p < 0.00001) and 16.3% (χ22,7967 = 1241.8;
p < 0.0000001; Figure 2A; Table 5b), respectively. An increase in GSS (F
2,7967
= 48.9; p < 0.0001;
η2 = 0.015; Table 5a) and winter pattern of mood seasonality (χ22,7967 = 10.5; p < 0.01; Table
5b) were also observed during the period of time corresponding to DSTp. There were no
signicant changes in summer pattern of mood seasonality during three period studied
(data not shown). The largest increase in SJL occurred in 10 to 17-year-olds (Figure 2B). This
nding can be primarily attributed to a later rise time on free days (F2,7967 = 44.9; p < 0.0001;
η
2
= 0.012; Figure 3F; Tables 5a and 6). Two analyses of covariance were performed to evaluate
the hypothesis that the changes in SJL can be primarily attributed to a later rise time on free
days (Table 5a): 1: according to the procedure described above; 2: in addition, “get-up time
on free days” was used as covariate. Signicant inuence of “period” on SJL was found only
in rst analysis (F2,7967 = 31.9; p < 0.0001; η2 = 0.009) but not in second one (F2,7967 = .133;
p = 0.875; η2 = 0.000) indicating that our hypothesis is true.
Similar data were observed in three subsamples obtained in Syktyvkar, Petrozavodsk, and
Vorkuta with some features (Table 5a, b; Figure 2C–E). In Petrozavodsk was observed signif-
icant impact of “period” on “academic performance” with maximal values during the non-
DSTp period (F2,2535 = 20.0; p = 0.0001; η2 = 0.009). There were no signicant eect of “period”
on SADw in Petrozavodsk and Vorkuta. “Period” had greatest eect size (η2 ∼ 0.05) on SJL,
GUTF, and GSS in Vorkuta’s subsample (Table 5a). There was a lower percentage of SADw in
children and adolescents living in Vorkuta, as compared with their peers living in other
Table 4.Anthropometric characteristics of persons studied.
Notes: BMI: body mass index; SST: school start time; Data presented as mean±SD; one way analysis of covariance were per-
formed using “period” as fixed factor, “anthropometric characteristics” as dependent variables, and parameters presented
in the Table 2 as covariates; the variables “BMI” and “settlement population” are non-normally distributed, therefore we
used normalized “lnBMI” and “ln(settlement population)” in analysis; F – Fisher tests; P – significance of F-test; η2 – effect
size; differences between values marked with the letters are significant (A > a – P < 0.05; B > b – P < 0.01; C > c – P < 0.001;
D > d – P < 0.0001) (post hoc comparisons, Tukey test).
Parameter
Period
νF P η2
I II III
(n=2499) (n=3033) (n=2436)
Sex (F/M), % 57.0/43.0 57.7/42.3 57.4/42.6
Age, yrs 15.4±2.8a,d 15.6±3.0A,d 16.6±3.2D7967 56.4 0.001 0.014
Weight, kg 54.9±12.5d55.1±12.7d58.9±13.0D7967 23.2 0.001 0.006
Height, cm 165.0±10.6a,c 165.7±10.9A,c 167.7±9.5C7967 8.6 0.001 0.002
BMI, kg/m220.0±3.2c19.9±3.2c20.8±3.4C7967 18.6 0.001 0.005
SST, h:mm 8:35±0:46A8:32±0:37a8:33±1:30 7967 3.2 0.041 0.004

8 M. F. BORISENKOV ET AL.
settlements of Northern European Russia (7.13% vs. 5.14%; χ21,7967 = 8.96; p < 0.005). At the
same time, mean values of GSS were the same in compared groups (9.09 vs. 9.02; t
1,7967
= 0.46;
p > 0.05). This means that children and adolescents from Vorkuta less likely to perceive sea-
sonal changes in their mood and behavior as a serious problem. However, this question is
outside the scope of the study.
Discussion
The main nding of the study is that during the period when DSTp was observed in Russian
Federation, the number of people suering from SJL (i.e. those with chronic sleep deprivation
during workweek nights) was signicantly increased. To compensate sleep deprivation, peo-
ple began to sleep more on weekend nights, mainly due to delay time of awakening.
Figure 2.A: Percentage of SJL≥1h (a) and SJL≥2h (b) in children and adolescents in all population at I–III
preiod studied; B–E: Age-associated changes in SJL in all population (B), Syktyvkar (C), Petrozavodsk (D),
and Vorkuta (E) subsamples; Error bars represent SE.

BIOLOGICAL RHYTHM RESEARCH 9
Previously, it was shown that human circadian system entrained by solar but not social
clocks (Roenneberg et al. 2007). The practice of yearly forward and backward transitions of
social clock reduces the ability of the human biological clock to seasonal adjustment
(Kantermann et al. 2007). Our data indicate that DSTp causes chronic dissociation between
social and biological clock that persists for more than three years. The greatest dissociation
Table 5a.SJL, academic performance, and Global Seasinality Scores in children and adolescents.
&the Russian grading system is coded into five grades: 1–5. A low grade indicates low achievement; SJL: social jetlag; GSS:
global seasonality scores; one way analysis of covariance was performed as described in Materials and Methods.
#two analyses were performed to assess the hypothesis that can be primarily attributed to a later rise time on free days 1:
according to the procedure described above; 2: in addition, “get-up time on free days” was used as covariate; ν – degrees
of freedom; F – Fisher tests; P – significance of F-test; η2 – effect size; differences between values marked with the letters
are significant (A > a – P < 0.05; B > b – P < 0.01; C > c – P < 0.001; D > d – P < 0.0001) (post hoc comparisons, Tukey test);
values marked with italics are insignificant, p>0.05.
Parameter Group
Period
νF P η2
I II III
SJL, h#All 2.16 ± 1.29d2.70 ± 1.44D2.13 ± 1.33d7967 1: 31.9 0.0001 0.009
2: .133 0.875 0.000
Syktyvkar 2.18 ± 1.23B2.17 ± 1.17 2.03 ± 1.23b2125 6.5 0.001 0.007
Petrozavodsk – 2.83 ± 1.47C2.52 ± 1.45c2535 11.9 0.001 0.005
Vorkuta 2.46 ± 1.63B2.70 ± 1.48D1.95 ± 1.32b,d 1067 23.4 0.0001 0.045
GUTF, h All 10.24 ± 1.85d11.11 ± 1.96D10.27 ± 1.91d7967 44.9 0.0001 0.012
Syktyvkar 10.22 ± 1.68A10.40 ± 1.69C10.03 ± 1.75a,c 2125 6.6 0.001 0.008
Petrozavodsk – 11.26 ± 1.96C10.84 ± 1.92c2535 16.5 0.0001 0.007
Vorkuta 10.37 ± 2.16a10.92 ± 2.15A,D 9.90 ± 1.99d1067 25.4 0.0001 0.049
GSS, scores All 9.70 ± 4.87C9.54 ± 4.86C8.26 ± 4.92c7967 48.9 0.0001 0.015
Syktyvkar 8.68 ± 4.43 9.51 ± 4.92B8.17 ± 4.84b2125 7.5 0.001 0.011
Petrozavodsk – 9.37 ± 4.92B8.69 ± 4.88b2535 10.9 0.001 0.005
Vorkuta 8.91 ± 4.50c10.73 ± 4.46C,D 8.12 ± 5.04d1067 27.2 0.0001 0.054
Achievement&,
scores
All 4.04 ± 0.61a4.03 ± 0.55b4.09 ± 0.57A,B 7967 .614 0.541 0.000
Syktyvkar 4.05 ± 0.59 4.17 ± 0.51 4.13 ± 0.57 2125 2.9 0.056 0.004
Petrozavodsk – 4.00 ± 0.53b4.07 ± 0.54B2535 20.0 0.0001 0.009
Vorkuta 3.84 ± 0.53b3.98 ± 0.58 4.03 ± 0.55B1067 2.4 0.091 0.005
Table 5b.Percentage of SJL and SADw in children and adolescents.
Notes: S-SADw: sub-syndrome of winter seasonal affective disorder; SADw: winter SAD; χ2 – hi-squared statistics; the rest
abbreviations as in Table 5a.
Parameter Group
Period
νχ2PI II III
SJL ≥ 1 h, % All 84.43d91.10D82.47d7967 92.3 0.00001
Syktyvkar 85.55A85.20 80.86a2125 8.0 0.025
Petrozavodsk – 92.60B88.53b2535 8.8 0.005
Vorkuta 83.33a90.21A,C 78.79c1067 19.8 0.001
SJL ≥ 2 h, % All 53.34d69.70D52.63d7967 1241.8 0.0000001
Syktyvkar 55.35A57.89A49.59a2125 9.3 0.01
Petrozavodsk – 73.04C63.78c2535 16.7 0.001
Vorkuta 60.42A67.28D47.06a,d 1067 37.3 0.0001
S-SADw, % All 10.98c14.82C11.60c7967 19.1 0.001
Syktyvkar 8.11a9.38 13.20A2125 7.4 0.025
Petrozavodsk – 14.63 11.69 2535 2.8 0.1
Vorkuta 8.51a18.21A,D 7.96d1067 23.6 0.001
SADw, % All 8.33 9.33B7.13b7967 10.5 0.01
Syktyvkar 7.03 11.25A6.13a2125 5.9 0.05
Petrozavodsk – 8.90 8.77 2535 0.5 0.5
Vorkuta 6.38 6.17 5.14 1067 3.8 0.25

10 M. F. BORISENKOV ET AL.
Figure 3.Age-associated changes in sleep onset on week- (A) and free (B) days; sleep onset latency on
week- (C) and free (D) days; get-up time on week- (E) and free (F) days; sleep inertia on week- (G), and
free (H) days; chronotype (MSFsc) (I), and sleep duration (J) at three periods studied. All parameters are
expressed in hours. I period II period III period.

BIOLOGICAL RHYTHM RESEARCH 11
between social and biological clocks was observed in 10–17-year-olds. One possible expla-
nation for this – higher sensitivity of the circadian system of adolescents to light signals
(Crowley et al. 2015).
Surprisingly, social clock manipulations had the greatest eect on SJL, GUTF, and GSS in
children and adolescents living in the Arctic city of Vorkuta. This means that in polar region
the solar clock is still stronger zeitgeber for human circadian system, than the social clock.
This issue requires more thorough study in special research.
According to Roenneberg et al. (2012), SJL ≥ 1 h occurs in 70% of the population. The
higher SJL index values that we observed can potentially be attributed to the fact that we
focused primarily on young people, who are at greater risk of SJL than adults (Roenneberg
et al. 2004). In addition, our study involved only residents living at high latitudes. These
residents more commonly exhibit late chronotype (Borisenkov 2010) and, accordingly, SJL.
Previous studies have shown that changes in the circadian system are associated with
SJL; SJL ≥ 1 h is associated with a marked decrease in the amplitude of the circadian rhythm
of wrist temperature (Polugrudov et al. 2016), SJL ≥ 2 h is associated with a larger cortisol
awakening response (Rutters et al. 2014) and an increased risk of depression (Levandovski
et al. 2011).
Limitations
The study was carried out retrospectively, therefore it was impossible to organize data col-
lection in accordance with the goal and objectives of the research. All signicant eects
were either small or very small in size (eta squared less than 0.05). We used a cross-sectional
study design, which does not allow to nd out the cause-eect relationships between the
studied parameters.
Table 6.Impact of three periods on chronotype and sleep characteristics.
Notes: SlOW(F): sleep onset on weekdays (free days); SOL W(F): sleep onset latency on weekdays (free days); GUTW(F): get-up
time on weekdays (free days); SlI W(F): sleep inertia on weekdays (free days); MSFSC: chronotype; SlD average weekly sleep
duration; One way analysis of covariance were performed using “period” as fixed factor, “chronotype and sleep characteris-
tics” as dependent variables, and parameters presented in the Tables 2 and 4 as covariates; the variables “BMI”, “settlement
population”, “SOLW(F)”, “SlIW(F)”, “SJL”, and “MSFsc” are non-normally distributed, therefore we used normalized “lnBMI” and
“ln(settlement population)”, “SqrSOLW(F)”, “ SqrSlIW(F)”, “ lnSJL”, and “lnMSFsc” in analysis; F – Fisher tests; P – significance of
F-test; η2 – effect size; differences between values marked with the letters are significant (A > a – P < 0.05; B > b, E > e –
P < 0.01; C > c – P < 0.001; D > d – P < 0.0001) (post hoc comparisons, Tukey test); values marked with italics are
insignificant, p>0.05.
Parameter
Period
F P η2
I II III
Weekdays
SlOw, h 23.73 ± 1.23d,b 24.10 ± 1.31D,E 23.89 ± 2.30e,B 11.26 0.001 0.005
SOLw, h 0.43 ± 0.39A,b 0.48 ± 0.45B,C 0.40 ± 0.44a,c 14.89 0.000 0.004
GUTw, h 7.13 ± 0.78A,B 7.09 ± 0.68a7.05 ± 0.80b9.68 0.000 0.003
SlIw, h 0.14 ± 0.14 0.14 ± 0.16 0.14 ± 0.16 3.40 0.034 0.001
Free days
SlOF, h 24.94 ± 1.73d25.47 ± 1.86D24.85 ± 2.90d25.08 0.000 0.007
SOLF, h 0.37 ± 0.40a0.40 ± 0.44A,C 0.36 ± 0.42c14.26 0.000 0.004
GUTF, h 10.24 ± 1.85d11.11 ± 1.96D10.27 ± 1.91d44.85 0.000 0.012
SlIF, h 0.31 ± 0.29a0.33 ± 0.35A0.32 ± 0.41 5.90 0.003 0.002
Weekdays/free days
MSFsc, h 4.78 ± 1.48c5.17 ± 1.57C4.77 ± 1.54c30.33 0.000 0.008
SlD, h 7.69 ± 1.18C7.40 ± 1.21c7.34 ± 1.31c1.01 0.366 0.000

12 M. F. BORISENKOV ET AL.
Conclusion
We have shown for the rst time an increase in SJL, delay in rise time on free days and
increased percentage of winter pattern of seasonality of mood and behavior in children and
adolescents during the period of time that DSTp was observed in the Russian Federation.
Chronic 1-h forward transition of social clock potentially exerts a negative inuence on
adolescents’ sleep habits, mood, and behavior.
Disclosure statement
No potential conict of interest was reported by the authors.
Funding
The study was supported in part by Program of UD RAS, project # 15–3-4–50 (MFB) and by the grant
# 15-16-10001 a(p) from the Russian Foundation for Humanities (SNK).
References
Borisenkov MF. 2010. Human chronotypes in the North. Hum Physiol. 36:117–122.
Borisenkov MF. 2011. The pattern of entrainment of the human sleep-wake rhythm by the natural
photoperiod in the North. Chronobiol Int. 28:921–929.
Borisenkov MF, Petrova NB, Timonin VD, Fradkova LI, Kolomeichuk SN, Kosova AL, Kasyanova ON. 2015.
Sleep characteristics, chronotype, and winter depression in 10–20-year-olds in northern European
Russia. J Sleep Res. 24:288–295.
Crowley SJ, Cain SW, Burns AC, Acebo C, Carskadon MA. 2015. Increased sensitivity of the circadian
system to light in early/mid-puberty. J Clin Endocrinol Metab. 100:4067–4073.
Haraszti RÁ, Ella K, Gyöngyösi N, Roenneberg T, Káldi K. 2014. Social jetlag negatively correlates with
academic performance in undergraduates. Chronobiol Int. 31:603–612.
Kantermann T, Juda M, Merrow M, Roenneberg T. 2007. The human circadian clock’s seasonal adjustment
is disrupted by daylight saving time. Curr Biol. 17:1996–2000.
Levandovski R, Dantas G, Fernandes LC, Caumo W, Torres I, Roenneberg T, Hidalgo MP, Allebrandt KV.
2011. Depression scores associate with chronotype and social jetlag in a rural population. Chronobiol
Int. 28:771–778.
Polugrudov AS, Panev AS, Smirnov VV, Paderin NM, Borisenkov MF, Popov SV. 2016. Wrist temperature
and cortisol awakening response in humans with social jetlag in the North. Chronobiol Int. 33:802–
809.
Roenneberg T, Merrow M. 2007. Entrainment of the human circadian clock. Cold Spring Harb Symp
Quant Biol. 72:293–299.
Roenneberg T, Wirz-Justice A, Merrow M. 2003. Life between clocks: daily temporal patterns of human
chronotypes. J Biol Rhythms. 18:80–90.
Roenneberg T, Kuehnle T, Pramstaller PP, Ricken J, Havel M, Guth A, Merrow M. 2004. A marker for the
end of adolescence. Curr Biol. 14:R1038–R1039.
Roenneberg T, Kumar CJ, Merrow M. 2007. The human circadian clock entrains to sun time. Curr Biol.
17:R44–R45.
Roenneberg T, Allebrandt KV, Merrow M, Vetter C. 2012. Social jetlag and obesity. Curr Biol. 22:939–943.
Rosen LN, Targum SD, Terman M, Bryant MJ, Homan H, Kasper SF, Hamovit JR, Docherty PJ, Welch
B, Rosenthal NE. 1990. Prevalence of seasonal aective disorder at four latitudes. Psychiatry Res.
31:131–144.
Rutters F, Lemmens SG, Adam TC, Bremmer MA, Elders PJ, Nijpels G, Dekker JM. 2014. Is social jetlag
associated with an adverse endocrine, behavioral, and cardiovascular risk prole? J Biol Rhythms.
29:377–383.
van der Vinne V, Zerbini G, Siersema A, Pieper A, Merrow M, Hut RA, Roenneberg T, Kantermann T.
2015. Timing of examinations aects school performance dierently in early and late chronotypes.
J Biol Rhythms. 30:53–60.