Sleep Medicine 76 (2020) 26e32

Contents lists available at ScienceDirect

Sleep Medicine
journal homepage: www.elsevier.com/locate/sleep

Original Article

Beneficial effects of weekend catch-up sleep on metabolic syndrome

in chronic short sleepers
Dae Jin Kim a, Sue Jean Mun b, Jeong Su Choi d, Jiyoung Kim c, Gha-hyun Lee c,
Hyun-Woo Kim a, Min-Gyu Park a, Jae Wook Cho a, *
a
Department of Neurology, Pusan National University Yangsan Hospital, Yangsan, South Korea
b
Department of Otorhinolaryngology, Pusan National University Yangsan Hospital, Yangsan, South Korea
c
Department of Neurology, Pusan National University Hospital, Busan, South Korea
d
Department of Health and Safety Convergence Science, Graduate School, Korea University, Seoul, South Korea

a r t i c l e i n f o a b s t r a c t

Article history: Objectives: Short sleep duration has been known to be related to metabolic syndrome (MetS) . The aim of
Received 2 June 2020 this study was to investigate the beneficial effects of weekend catch-up sleep (WCUS) on MetS in the
Received in revised form Korean middle-aged population.
20 September 2020
Methods: For this cross-sectional study, 1,812 participants aged 35-60 years were selected from the 2016-
Accepted 22 September 2020
Available online 30 September 2020
2018 Korean National Health and Nutrition Examination Survey (mean age 46.94 years, 49% male). Short
sleep duration was defined as <6hrs on weekdays, and participants were divided into two groups: WCUS
group and no weekend catch-up sleep group.
Keywords:
Metabolic syndrome
Multiple logistic regression was performed to determine the association between WCUS and MetS
Short sleep duration prevalence. The covariates included age, sex, education, income, occupation, smoking, alcohol con-
Catch-up sleep sumption, and physical activity.
Sleep restriction Results: WCUS was significantly associated with lower MetS prevalence in the unadjusted model and in
Middle-aged the model adjusted for socioeconomic and health behavior factors.
Korea Conclusion: These results support the beneficial effects of WCUS on lowering the risk of MetS among
middle-aged chronic short sleepers.
© 2020 Elsevier B.V. All rights reserved.

1. Introduction have MetS [5], and the prevalence of MetS was 11.9e49% in the
AsiaePacific region including South Korea [6,7]. Numerous studies
Adequate sleep is important for mental, physical, and social have found an association between short sleep duration and
well-being. However, sleep habits are changing with rapidly metabolic diseases such as diabetes, hypertension, hyperlipidemia,
changing lifestyles. Although sleep experts recommend 7e9 h per obesity, and MetS [3,8e10]. In a cross-sectional analysis in Canada,
night as the optimal sleep duration for adults [1,2], sleep duration participants had higher risk of MetS when they slept <6 h per night
has been decreasing in the modernized society over the last 50 compared to those who slept 7e8 h per night (adjusted odds ratio
years [3]. This trend has become more significant recently with an 1.76, 1.08e2.84) [11]. In the National Health and Nutrition Exami-
increasing prevalence of people with sleep deficiency due to eco- nation Survey, 2737 participants provided their self-reported sleep
nomic, social, and technological changes [4]. duration and blood test details, and the results indicated that those
Simultaneously, metabolic syndrome (MetS) has secular trends with short sleep duration had a higher risk of MetS. In addition,
of increasing prevalence worldwide and has become a major public short sleep durations have been associated with higher scores of
health concern. Recent studies demonstrated that nearly 35% of all MetS severity in both men and women [12]. A meta-analysis study
adults and 50% of those older than 60 years in the United States including 18 studies demonstrated that those who slept less than
5 h had 1.5-times higher odds of MetS [13]. Another meta-analysis
also reported that short sleep duration is significantly associated
* Corresponding author. Department of Neurology, Pusan National University with MetS risk [14].
Yangsan Hospital, 20 Geumo-ro, Mulgeum-eup, Yangsan, 50612, South Korea. Lifestyle modifications such as engaging in aerobic exercise, diet
Fax: þ82 55 360 2152.
E-mail address: [email protected] (J.W. Cho).
changes, and stress control are considered important in MetS

https://doi.org/10.1016/j.sleep.2020.09.025
1389-9457/© 2020 Elsevier B.V. All rights reserved.
D.J. Kim, S.J. Mun, J.S. Choi et al. Sleep Medicine 76 (2020) 26e32

management [15,16]. Given the importance of sleep insufficiency in A questionnaire was also used to investigate health behavior
MetS, sleep duration could also be a modifiable lifestyle behavior. factors such as smoking status (nonsmoker, ex-smoker, current
Many studies have already shown that catch-up sleep (CUS) is good smoker), alcohol consumption (none, once a month, once a
for health, quality of life, and well-being [17,18]. There is a further week), and level of physical activity (none, moderate activity,
need to explore whether weekend catch-up sleep (WCUS) can play intense activity). Intense physical activity was defined as at least
a role in preventing MetS among people with chronic sleep defi- 20 min of activities that caused shortness of breath or involved
ciency and whether it would be beneficial in reducing MetS risk. strenuous movements.
Therefore, we investigated the positive effects of WCUS on MetS Height, weight, blood pressure, and waist circumference were
among middle-aged people with sleep deficiency during weekdays measured, and body mass index (BMI) was calculated using height
using data from the Korean National Health and Nutrition Exami- (m2) and body weight (kg). Blood samples were taken after more
nation Survey (KNHANES). than 8 h of fasting.

2.3. Definition of MetS

2. Materials and methods
MetS was defined as the presence of three or more of the
2.1. Study population and data collection
following factors according to the National Cholesterol Education
Program, Adult Treatment Program III criteria [19,20]: i) central
This study was based on data from KNHANES VII (2016e2018), a
obesity, ii) high fasting plasma glucose 100 mg/dl, iii) high fasting
cross-sectional, nationally representative, population-based survey
serum triglyceride 150 mg/dl, iv) low serum high density lipo-
regarding health and nutritional status of Korean residents estab-
protein cholesterol <40 mg/dl in men and <50 mg/dl in women,
lished by the Korea Centers for Disease Control and Prevention.
and v) high blood pressure with systolic or diastolic blood pressure
Participants were chosen using systematic sampling with multi-
130/85 mmHg.
stage clusters based on age group, geographical area, and sex
We defined central obesity using the AsiaePacific obesity
recorded in household registries. Trained interviewers conducted
criteria based on waist circumference as defined by the World
surveys and administered questionnaires regarding socioeconomic
Health Organization [21]. The waist circumference threshold for
characteristics, health-related behavior, and medical history. Blood
central obesity is  90 cm in men and 85 cm in women according
samples were taken. A total of 24,269 individuals participated in
to the Korean Society for the Study of Obesity [22].
the 2016e2018 KNHANES. We selected the first 9323 participants
aged 35e60 years, and excluded 1106 participants with missing
2.4. Statistical analysis
data on physical examination, questionnaires, and other blood test
variables; 1546 participants who took medication for hypertension,
Data were analyzed using SAS version 9.4 (SAS Institute, Cary,
diabetes, and hyperlipidemia were also excluded. Of the remaining
NC, USA) with complex sample design, and p-values less than 0.05
6671 participants, we enrolled 1825 participants who slept less
for two-tailed tests were considered statistically significant. After
than 6 h on weekdays as people with short sleep duration (Fig. 1).
participants were divided into two groups based on the presence of
Finally, we divided these participants into two groups depending
WCUS, we compared demographic characteristics, sleep duration,
on the presence of WCUS. The WCUS group was defined as par-
and MetS components. An independent t-test or Wilcoxon rank-
ticipants who slept more on weekends than weekdays. The control
sum test with weighted values was used to compare differences
group (no WCUS, nCUS) involved others who slept for the same
of continuous variables, and the c2 test was used for categorical
duration on weekdays and weekends or slept less on weekends.
variables.
All the survey data used for this study were from KNHANES VII,
We calculated odds ratios and 95% confidence intervals (CI)
and publicly available from the Korea Centers for Disease Control
using a logistic regression model to estimate the association be-
and Prevention. We got review exemption from our institutional
tween WCUS and MetS. A final multivariate analysis was adjusted
review board (IRB No. 05-2020-109).
progressively for potential confounders such as age, gender, soci-
odemographic factors (education level, household income, occu-
2.2. Main variables pation), and health behaviors (smoking status, alcohol
consumption, physical activity).
To assess sleep duration on weekdays and weekends, partici- Receiver operating characteristic (ROC) curves were plotted to
pants were asked the following questions: “What time do you go to assess the most suitable WCUS duration for MetS. Cutoff values
bed on weekdays?” “What time do you wake up on weekdays?” were obtained both from the point on the ROC curve and from
“What time do you go to bed on weekends?” and “What time do Youden's index.
you wake up on weekends?” Answers were based on a clock time. Additionally, we performed multi-variable generalized additive
Sleep durations on weekdays and weekends were calculated using model (GAM) analyses in three ways to assess the smooth rela-
these sleep and waking times. WCUS duration was calculated as tionship between metabolic syndrome/metabolic syndrome
weekend sleep duration minus weekday sleep duration. severity score/the number of present components, and catch-up
We collected data on sociodemographic variables (age, sex, sleep duration (detailed methods in suppl data).
educational level, state of employment, occupation, average work
hours per week, household income, marital status, and work 3. Results
schedule type) using a self-reported questionnaire as follows: ed-
ucation level (middle school or lower, high school, or college or 3.1. Characteristics of study population
higher), household income ($1000, $1001e$2000, $2001e$3000,
or $3001 a month), marital status (single, separated/divorced/ Table 1 gives the descriptive characteristics with weighted
widowed, or married), state of employment (employed or unem- values of the study population according to the presence or absence
ployed), occupation (wage and salary worker, self-ownership, or of WCUS (n ¼ 1825). The mean age of participants in the WCUS
unpaid family worker), and work schedule type (day work, eve- group (n ¼ 1121) was 46.07 ± 0.24 years, and lower than that of the
ning/night work, shiftwork, and others). nCUS group (48.39 ± 0.29 years). There was no difference in sex
27
D.J. Kim, S.J. Mun, J.S. Choi et al. Sleep Medicine 76 (2020) 26e32

Fig. 1. Sample construction and data collected.

ratio, BMI, state of employment, and marital status between the Table 3 shows the difference in sleep duration between partic-
two groups. Lower educational level and household income were ipants with and without MetS. Participants without MetS had
more prevalent in the nCUS group. The WCUS group had more similar weekday sleep durations as those with MetS (329.66 ± 1.22
salaried workers; on the contrary, self-ownership and unpaid vs. 324.65 ± 2.58 min); however, they had a significantly higher
family workers were more prevalent in the nCUS. Regarding health weekend sleep duration (409.95 ± 2.72 vs. 388.11 ± 4.70 for those
behavior factors, both groups had similar patterns of smoking and with MetS), meaning they had more WCUS.
physical activity. The WCUS group showed higher prevalence of The crude and adjusted analyses for the assessment of re-
moderate alcohol consumption (1 per month), and the nCUS lationships between WCUS of short sleepers and MetS are shown in
group showed more heavy alcohol use (1 per week). Table 4. In the crude model, among participants who slept for less
Weekday sleep durations of both groups were similar than 6 h on weekdays, the risk of MetS was 0.655 times lower (95%
(326.99 ± 1.40 min for WCUS and 331.05 ± 1.78 min for nCUS), but CI ¼ 0.514e0.835) in the WCUS group compared to the nCUS group.
weekend sleep durations were notably different. The nCUS group Even after adjusting for age, sex, education level, monthly income,
had similar sleep duration on weekdays and weekends occupation (Model 1), and smoking status, alcohol consumption,
(327.02 ± 1.88 min), while the WCUS group had an average sleep and physical activity (Model 2), the relationship persisted, but the
duration of 452.24 ± 2.56 min on weekends, which was magnitude of the odds ratio was attenuated slightly.
124.24 ± 2.61 min higher than weekdays. The optimal cutoffs for WCUS and weekend sleep duration for
MetS were 40 min [Area under the ROC curve (AUC): 0.564,
3.2. WCUS and MetS sensitivity: 52.8%, Specificity: 59.0%] and 390 min (AUC: 0.574,
sensitivity: 61.0%, Specificity: 50.7%) respectively.
Table 2 demonstrates the prevalence of MetS components The GAM analyses also showed negative relationship between
classified by WCUS. Overall, 22.73% of the sample had MetS (3 MetS and WCUS among the subjects of this study (detailed results
components). The nCUS group had higher prevalence of each MetS in suppl data).
component, especially elevated glucose, abdominal obesity, and
elevated total cholesterol. More MetS components were associated 4. Discussion
with nCUS compared to WCUS. In addition, the nCUS group had
significantly more participants who met the diagnostic criteria of In this study using nationally representative data, we found an
MetS with 27.46% vs. 19.87% for WCUS. association between WCUS and MetS among middle-aged people
28
D.J. Kim, S.J. Mun, J.S. Choi et al. Sleep Medicine 76 (2020) 26e32

Table 1
Descriptive characteristics of study subjects divided by the presence of weekend catch up sleep.

Total nCUS WCUS p-value

(N ¼ 1825) (N ¼ 704) (N ¼ 1121)

Mean (SE) Mean (SE) Mean (SE)

or % (SE) or % (SE) or % (SE)

Male sex, n, weighted % 757, 48.95 (1.19) 297, 48.65 (2.09) 460, 49.13 (1.49) 0.853
Age, year 46.94 (0.19) 48.39 (0.29) 46.07 (0.24) <0.0001*
Height, cm 165.58 (0.24) 165.14 (0.39) 165.84 (0.30) 0.153
Weight, kg 66.61 (0.32) 66.81 (0.54) 66.50 (0.41) 0.658
BMI, kg/m2 24.18 (0.09) 24.38 (0.14) 24.06 (0.11) 0.069
Waist circumference, cm 82.67 (0.25) 83.23 (0.41) 82.33 (0.31) 0.084
Systolic blood pressure, mmHg 115.13 (0.41) 116.70 (0.63) 114.17 (0.52) 0.002*
Diastolic blood pressure, mmHg 77.73 (0.30) 78.19 (0.44) 77.44 (0.37) 0.177
Fasting plasma glucose, mg/dl 98.71 (0.55) 99.99 (0.89) 97.94 (0.69) 0.067
Total cholesterol, mg/dl 206.29 (0.93) 208.90 (1.52) 204.72 (1.17) 0.029*
HDL, mg/dl 51.93 (0.33) 51.78 (0.54) 52.02 (0.42) 0.726
TG, mg/dl 146.31 (2.91) 150.30 (4.67) 143.89 (3.80) 0.295
Weekdays sleep duration, min 328.52 (1.10) 331.05 (1.78) 326.99 (1.40) 0.075
Weekends sleep duration, min 404.99 (2.35) 327.02 (1.88) 452.24 (2.56) <0.001*
Difference between weekends and weekdays, min 76.47 (2.33) 4.02 (0.62) 125.25 (2.61) <0.001*
Working hours per week, min 44.24 (0.46) 44.07 (0.93) 44.35 (0.54) 0.805
Work schedule type, weighted % 0.599
Day work 81.27 (1.27) 80.49 (1.95) 81.74 (1.56)
Evening work & etc. 18.73 (1.27) 19.51 (1.95) 18.26 (1.56)
Household income, weighted % <0.001*
100 3.58 (0.51) 5.42 (0.97) 2.46 (0.55)
101-200 7.57 (0.77) 10.16 (1.38) 6.00 (0.86)
201-300 13.26 (0.92) 15.01 (1.56) 12.21 (1.15)
301 75.59 (1.28) 69.41 (2.05) 79.33 (1.51)
Education level, weighted % <0.001*
 Middle school 9.54 (0.79) 13.64 (1.32) 7.06 (0.89)
High school 35.52 (1.42) 39.11 (2.11) 33.34 (1.80)
 College 54.94 (1.56) 47.25 (2.19) 59.60 (1.91)
Marital status, weighted % 0.417
Single 7.67 (0.72) 8.13 (1.14) 7.39 (0.92)
Separated/divorced/widowed 8.10 (0.68) 9.13 (1.12) 7.47 (0.84)
Married 84.23 (1.02) 82.74 (1.53) 85.14 (1.31)
State of employment, weighted % 0.155
Employed 76.14 (1.09) 74.09 (1.86) 77.38 (1.37)
Unemployed 23.86 (1.09) 25.91 (1.86) 22.62 (1.37)
Occupation, weighted % <0.001*
Wage and salary worker 54.46 (1.36) 44.54 (2.19) 60.48 (1.70)
Self-ownership 19.36 (1.10) 25.94 (1.94) 15.37 (1.27)
Unpaid family worker 2.32 (0.38) 3.61 (0.73) 1.54 (0.40)
None of the above 23.86 (1.09) 25.91 (1.86) 22.62 (1.37)
Smoking status, weighted % 0.142
Nonsmoker 53.93 (1.22) 52.01 (2.10) 55.09 (1.61)
Ex-smoker 20.44 (1.05) 19.39 (1.76) 21.08 (1.31)
Current smoker 25.63 (1.18) 28.61 (1.97) 23.83 (1.49)
Alcohol consumption, weighted % 0.009*
None 20.31 (1.06) 21.91 (1.75) 19.34 (1.33)
 once a month 29.40 (1.23) 24.55 (1.88) 32.34 (1.63)
 once a week 50.29 (1.35) 53.54 (2.20) 48.32 (1.78)
Level of physical activity, weighted % 0.077
Intense activity 16.68 (0.99) 15.52 (1.64) 17.39 (1.22)
Moderate activity 21.20 (1.14) 18.74 (1.69) 22.70 (1.44)
None 62.11 (1.34) 65.75 (2.06) 59.91 (1.64)

*p < 0.05.
SE, standard error; BMI, body mass index; HDL, high-density lipoprotein; TG, total cholesterol.

with sleep insufficiency. Among people sleeping less than 6 h on Another study showed that shorter the sleep duration, greater the
weekdays, those who sleep more on weekends than weekdays have effect of WCUS on weight gain [25]. Likewise, it could have similar
significantly lower MetS risk, especially with an extra 40-min sleep. beneficial effects on MetS.
Our study also shows that sleep duration on weekends of partici- There are several hypotheses about how short sleep duration
pants with MetS is generally less than those without MetS. causes negative metabolic changes, one of which is short sleep- and
To the best of our knowledge, this is the first report showing a appetite-related hormonal changes [27]. It is generally accepted
significant relationship between WCUS and MetS, and our results that short sleep reduces leptin, a satiety hormone, and increases
are in line with previous studies on WCUS and metabolic diseases appetite-stimulating ghrelin and cortisol, which causes metabolic
[23]. Hwangbo et al. found that 1-h WCUS was helpful in reducing changes [28e32] and increase in subjective hunger and appetite. In
the risk of hypertension among participants who slept <6 h per addition, some experimental studies have found that short sleep
night [24]. In a study with children and adolescents, WCUS was duration increases energy or caloric intake itself regardless of
associated with a decreased risk of being overweight [25,26]. changes in energy expenditure or leptin or ghrelin levels [33,34].
29
D.J. Kim, S.J. Mun, J.S. Choi et al. Sleep Medicine 76 (2020) 26e32

Table 2 cardiovascular disease, such as insulin-like growth factor-1 and

Prevalence of the metabolic syndrome components divided by weekend catch up myeloperoxidase [41] and improve recovery process of immune
sleep.
and inflammatory parameters [42]. People, especially those who
Total nWCUS, n (%) WCUS, n (%) p-value slept less than 6 h on weekdays, with WCUS of more than 1 h had
n ¼ 704 n ¼ 1121 low possibility of having high C-reactive protein (CRP) levels [43].
Another study also showed that 1e2 h of WCUS could reduce high
Number of components of the metabolic syndrome
0 90 (12.9) 159 (14.1) 0.508 sensitivity-CRP level, thereby lowering the risk of cardiovascular
1 614 (87.1) 962 (85.9) 0.508 disease [44]. These findings suggest the possible mechanism of CUS
2 373 (54.6) 481 (45.0) 0.000* in attenuating the harmful metabolic impact of sleep deficiency.
3 185 (27.5) 207 (19.9) 0.001*
However, this does not mean that longer sleep duration always
4 75 (11.0) 89 (8.9) 0.205
5 14 (1.8) 15 (1.5) 0.611 has a positive impact on MetS. Several epidemiologic studies pre-
Presence of each components of metabolic syndrome viously demonstrated a U-shaped relationship between sleep
Elevated blood glucose 246 (36.2) 319 (30.4) 0.035* duration and MetS prevalence [11,45,46], reflecting that very long
Abdominal obesity 216 (31.9) 285 (27.0) 0.041* sleep durations had a negative impact on body metabolism. Most of
Elevated blood pressure 196 (28.6) 260 (25.1) 0.156
these studies showed the lowest prevalence of MetS with 7 h of
High triglyceride 241 (35.9) 316 (30.3) 0.023*
Low HDL cholesterol 362 (49.4) 574 (48.3) 0.650 sleep. In the present study, the subjects of WCUS group had short
mean sleep duration (weekdays 5 þ weekends 2/7) of
*p < 0.05.
nWCUS, no weekend catch up sleep; WCUS, weekend catch up sleep; HDL, high- 350.37 ± 1.15 min, which were less than 7 h, even if they had more
density lipoprotein. sleep on weekends. Therefore, our results do not imply that an

Table 3
Analysis of average sleep duration and metabolic syndrome of short sleepers (<6 h).

Total (n ¼ 1825) Normal (N ¼ 1433) Metabolic syndrome (N ¼ 392) p-value

On weekdays, min 328.52 ± 1.10 329.66 ± 1.22 324.65 ± 2.58 0.083

On weekends, min 404.99 ± 2.35 409.95 ± 2.72 388.11 ± 4.70 <0.001*
Difference between weekends and weekdays, min 76.47 ± 2.33 80.30 ± 2.67 63.46 ± 4.42 0.001*

*p < 0.05.
All sleep duration is with weighted value and presented as minutes (mean ± standard error).

Table 4
Adjusted ORs (95% Cis) for the prevalence of metabolic syndrome for short sleepers according to the presence of weekend catch up sleep.

Unadjusted Model 1 Model 2

OR (95% CI) p value OR (95% CI) p value OR (95% CI) p value

nWCUS 1 (Ref.) 1 (Ref.) 1 (Ref.)

WCUS 0.655 (0.514, 0.835) 0.001* 0.692 (0.536, 0.893) 0.005* 0.718 (0.554, 0.931) 0.012*

OR, odds ratio; CI, confidence interval; nWCUS, no weekend catch up sleep; WCUS, weekend catch up sleep.
Model 1: adjusted for age, sex, education level, monthly income, and status of workers.
Model 2: Model 1 þ smoking status, and consumption frequency of alcohol, and physical activity.

Short sleepers tend to eat a large quantity of food without inhibi- extremely long WCUS, which is having more than 8 h of mean sleep
tion of eating behavior [35]. A study using functional magnetic duration, is better.
resonance imaging reported that participants with restricted sleep Various psychosocial and environmental factors may
showed a pronounced response to unhealthy food such as high- contribute to weekday sleep deficiency, including early work start
caloric and high-fat food [36]. This might be due to accumulation time, heavy work burden, and increased social activities. In this
of b amyloid protein plaques and neurofibrillary tangles generating study, there was no significant difference in average work hours
global decrease in brain function related decision making [37]. and work schedule type between the WCUS and nCUS groups, but
These neurotoxins are eliminated during sleep, especially slow- the WCUS group had more participants with high education, high
wave sleep. The release of growth hormone (GH) and GH income, and had more wage workers. Considering these results,
releasing hormone, which are responsible for many metabolic and we inferred that white-collar workers in Korea had an increased
homeostatic reactions, is also disturbed with slow-wave sleep chance on weekends to compensate for the lack of sleep due to
deficiency [27]. Furthermore, insufficient sleep may increase fa- excessive work or social activities, than others. Regarding health
tigue and tiredness during the day and reduce physical activity, behavior factors, physical activity and smoking status were similar
thereby reducing energy consumption and causing weight gain or between groups, but frequent drinking was more prevalent in the
metabolic disorders [35,38]. Another possible mechanism is an nCUS group. However, even after adjustment of all these socio-
increase in sympathetic nerve activity, which causes inhibition of demographic and health behavior factors, WCUS was associated
insulin secretion, increase in insulin resistance, and lipolysis [39]. with lower MetS prevalence. If people cannot avoid sleep defi-
CUS could have an effect of counteracting these bad influences ciency during the week, sleeping an extra 40 min on weekends
produced by sleep deficiency. Spiegel et al. showed that sufficient could help prevent MetS.
“recovery sleep” was effective in stabilizing changes in carbohy- In the present study, we performed weighted and unweighted
drate metabolism and endocrine function caused by lack of sleep analyses and showed weighted results reflecting the characteristics
[40]. Other studies reported that recovery sleep after chronic sleep of the multistage sampling design. There was no significant dif-
restriction produced positive changes in blood biomarkers of ference between those two results.
30
D.J. Kim, S.J. Mun, J.S. Choi et al. Sleep Medicine 76 (2020) 26e32

Despite its strengths, our study has some limitations. First, References
sleep duration based on survey questionnaires does not stand for
actual sleep duration. This limitation has frequently emerged in [1] Consensus Conference P, Watson NF, Badr MS, et al. Recommended amount of
sleep for a healthy adult: a joint consensus statement of the American acad-
population-based studies like ours. Several studies have reported emy of sleep medicine and sleep research society. J Clin Sleep Med 2015;11:
that self-assessed sleep duration overestimates or underestimates 591e2.
objectively measured sleep [47,48]. However, the validity of self- [2] Hirshkowitz M, Whiton K, Albert SM, et al. National Sleep Foundation's sleep
time duration recommendations: Methodology and results summary. Sleep
reported sleep duration has been supported by many studies Health 2015;1:40e3.
[49,50], and substantial research has been conducted using data [3] Larcher S, Benhamou PY, Pepin JL, et al. Sleep habits and diabetes. Diabetes
from large-scale questionnaire surveys. Second, this survey did Metab 2015 Sep;41(4):263e71.
[4] Sheehan CM, Frochen SE, Walsemann KM, et al. Are U.S. Adults reporting less
not seek data on subjective sleep insufficiency. Sleeping less than sleep?: findings from sleep duration trends in the national health interview
6 h does not always mean the participants have sleep deficiency survey, 2004-2017. Sleep 2019 Feb 1;42(2):zsy221.
because they might be natural short sleepers. There is possibility [5] Aguilar M, Bhuket T, Torres S, et al. Prevalence of the metabolic syndrome in
the United States, 2003-2012. JAMA 2015;313:1973e4.
of having completed the study with questions about sleep quality
[6] Ranasinghe P, Mathangasinghe Y, Jayawardena R, et al. Prevalence and trends
or EDS scales. Based on the literature, there is no consensus of metabolic syndrome among adults in the Asia-pacific region: a systematic
definition of “sleep debt” or “inadequate sleep,” which can reflect review. BMC Publ Health 2017 Jan 21;17:101.
[7] Lee SH, Tao S, Kim HS. The prevalence of metabolic syndrome and its related
short sleep duration or subjectively insufficient sleep. Therefore,
risk complications among Koreans. Nutrients 2019 Jul 30;11:1755.
the question about subjective sleepiness and/or sleep quality [8] Choi KM, Lee JS, Park HS, et al. Relationship between sleep duration and the
scales could be meaningful in establishing the definition and metabolic syndrome: Korean national health and nutrition survey 2001. Int J
elucidating the negative effects of sleep deficiency in future Obes (Lond) 2008;32:1091e7.
[9] Fatima Y, Doi SAR, Mamun AA. Longitudinal impact of sleep on overweight
research. In a previous study with similar participants sleeping and obesity in children and adolescents: a systematic review and bias-
<6 h, those with subjective sleep insufficiency showed a more adjusted meta-analysis. Obes Rev 2015;16:137e49.
significant association of an hour of WCUS with decreased risk for [10] Ning X, Lv J, Guo Y, et al. Association of sleep duration with weight gain and
general and central obesity risk in Chinese adults: a prospective study.
hypertension [24]. Third, this is a cross-sectional study using Obesity (Silver Spring) 2020 Feb;28:468e74.
nationwide survey dataset suggesting potential association within [11] Chaput JP, McNeil J, Despre s JP, et al. Seven to eight hours of sleep a night is
an arbitrary timeslot. Inferring causality could be possible in the associated with a lower prevalence of the metabolic syndrome and reduced
overall cardiometabolic risk in adults. PloS One 2013;8:e72832.
future with a longitudinal study design. Finally, because this study [12] Smiley A, King D, Bidulescu A. The association between sleep duration and
was conducted on only Koreans, these findings cannot be gener- metabolic syndrome: the NHANES 2013/2014. Nutrients 2019 Oct 26;11(11):
alized to other racial and ethnic groups. More studies conducting 2582.
[13] Iftikhar Imran H, Donley Meredith A. Jesse mindel, adam pleister, sheryll
on various racial groups would be helpful to support the rela-
soriano, ulysses J. Magalang. Sleep duration and metabolic syndrome. An
tionship between WCUS and MetS. updated dose-risk metaanalysis. Ann Am Thorac Soc 2015;12(9):1364e72.
In conclusion, this survey study provides evidence that short [14] Xi B, He D, Zhang M, et al. Short sleep duration predicts risk of metabolic
syndrome: a systematic review and meta-analysis. Sleep Med Rev 2014;18:
weekday sleep and extra compensatory WCUS are independently
293e7.
associated with lower MetS prevalence among Korean middle-aged [15] Nishizawa H, Shimomura I. Population approaches targeting metabolic syn-
people. For those who sleep <6 h on weekdays, WCUS of at least drome focusing on Japanese trials. Nutrients 2019;11:1430.
40 min could be expected to lower MetS risk. Further research is [16] Frere NO, Soliman SSA, Foda MA, et al. Effect of patient education on meta-
bolic syndrome components among females in Zagazig University outpatient
required to investigate the biologic mechanisms linking CUS and clinics, Egypt: an intervention study. Diabetes Metab Syndr 2019;13:
MetS. 1897e900.
[17] Oh YH, Kim H, Kong M, et al. Association between weekend catch-up sleep
and health-related quality of life of Korean adults. Medicine (Baltim) 2019;98:
e14966.
CRediT authorship contribution statement [18] Kang SG, Lee YJ, Kim SJ, et al. Weekend catch-up sleep is independently
associated with suicide attempts and self-injury in Korean adolescents. Compr
Dae Jin Kim: Conceptualization, Writing - original draft. Sue Psychiatr 2014;55:319e25.
[19] Expert panel on detection, evaluation, and treatment of high blood cholesterol
Jean Mun: Methodology. Jeong Su Choi: Data curation. Jiyoung in adults. JAMA 2001;285:2486e97.
Kim: Methodology. Gha-hyun Lee: Formal analysis. Hyun-Woo [20] Grundy SM. Definition of metabolic syndrome. Circulation 2004;109:433e8.
Kim: Software, Validation. Min-Gyu Park: Supervision. Jae Wook [21] Health communications Australia Pty limit. Sydney, Australia. 2000.
[22] Lee SY. Appropriate waist circumference cutoff points for central obesity in
Cho: Writing - review & editing, Supervision.
Korean adults. Diabetes Res Clin Pract 2007;75(1):72e80.
[23] Im HJ, Baek SH, Chu MK, et al. Association between weekend catch-up sleep
and lower body mass: population-based study. Sleep 2017;40.
Acknowledgement [24] Hwangbo Y, Kim W, Chu MK, et al. Association between weekend catch-up
sleep duration and hypertension in Korean adults. Sleep Med 2013;14:
549e54.
We thank the Department of Biostatistics, Biomedical Research [25] Kim CW. Weekend catch-up sleep is associated with decreased risk of being
Institute, Pusan National University Hospital. overweight among fifth-grade students with short sleep duration. J Sleep Res
2012;21:546e51.
[26] Sunwoo JS, Yang KI, Kim JH, et al. Sleep duration rather than sleep timing is
associated with obesity in adolescents. Sleep Med 2020 Apr;68:184e9.
Appendix A. Supplementary data [27] Smiley A, Wolter S, Nissan D. Mechanisms of association of sleep and meta-
bolic syndrome. J Med Clin Res Rev 2019;3:1e9.
Supplementary data to this article can be found online at [28] Chaput JP, Despre s JP, Bouchard C, et al. Short sleep duration is associated
with reduced leptin levels and increased adiposity: results from the Quebec
https://doi.org/10.1016/j.sleep.2020.09.025. family study. Obesity 2007;15:253e61.
[29] Taheri S, Lin L, Austin D, et al. Short sleep duration is associated with reduced
leptin, elevated ghrelin, and increased body mass index. PLoS Med 2004
Conflict of interest Dec;1:62.
[30] Spiegel K, Tasali E, Penev P, et al. Brief communication: sleep curtailment in
healthy young men is associated with decreased leptin levels, elevated ghrelin
None. levels, and increased hunger and appetite. Ann Intern Med 2004;141:846e50.
The ICMJE Uniform Disclosure Form for Potential Conflicts of [31] Spiegel K, Leproult R, L’hermite-Bale riaux M, et al. Leptin levels are dependent
on sleep duration: relationships with sympathovagal balance, carbohydrate
Interest associated with this article can be viewed by clicking on the
following link: https://doi.org/10.1016/j.sleep.2020.09.025.
31
D.J. Kim, S.J. Mun, J.S. Choi et al. Sleep Medicine 76 (2020) 26e32

regulation, cortisol, and thyrotropin. J Clin Endocrinol Metab 2004;89: sleep restriction and recovery in healthy young men. PloS One 2011;6:
5762e71. e28230.
[32] Reynolds AC, Dorrian J, Liu PY, et al. Impact of five nights of sleep restriction [42] Faraut B, Boudjeltia KZ, Vanhamme L, et al. Immune, inflammatory and car-
on glucose metabolism, leptin and testosterone in young adult men. PloS One diovascular consequences of sleep restriction and recovery. Sleep Med Rev
2012;7:e41218. 2012;16:137e49.
[33] Calvin AD, Carter RE, Adachi T, et al. Effects of experimental sleep restric- [43] Han KM, Lee HJ, Kim L, et al. Association between weekend catch-up sleep and
tion on caloric intake and activity energy expenditure. Chest 2013;144: high-sensitivity C-reactive protein levels in adults: a population-based study.
79e86. Sleep 2020. https://doi.org/10.1093/sleep/zsaa010.
[34] St-Onge MP1, Roberts AL, Chen J, et al. Short sleep duration increases energy [44] Jung SW, Lee K, Lee J. Does weekend catch-up sleep affect high-sensitivity C-
intakes but does not change energy expenditure in normal-weight in- reactive protein levels among Korean workers?: a cross-sectional study using
dividuals. Am J Clin Nutr 2011;94:410e6. KNHANES. J Occup Environ Med 2019 Sep;61(9):e367e73.
[35] Chaput JP, Despres JP, Bouchard C, et al. The association between short sleep [45] Fan L, Hao Z, Gao L, et al. Non-linear relationship between sleep duration and
duration and weight gain is dependent on disinhibited eating behavior in metabolic syndrome: a population-based study. Medicine (Baltim) 2020;99:
adults. Sleep 2011;34:1291e7. e18753.
[36] St-Onge MP1, Wolfe S1, Sy M2, et al. Sleep restriction increases the neuronal [46] Smiley A. Association of sleep and metabolic syndrome/metabolic syndrome
response to unhealthy food in normal-weight individuals. Int J Obes (Lond) components among participants in jackson heart study (JHS) and the reasons
2014;38:411e6. for geographic and racial differences in stroke (REGARDS). Indiana University,
[37] Killgore WDS, Balkin TJ, Wesensten NJ. Impaired decision making following 49 ProQuest Dissertations Publishing; 2019. p. 13886336.
h of sleep deprivation. J Sleep Res 2006;15:7e13. [47] Jackson CL, Patel SR, Jackson WB, et al. Agreement between self-reported and
[38] Schmid SM, Hallschmid M, Jauch-Chara K, et al. Short-term sleep loss de- objectively measured sleep duration among white, black, Hispanic, and Chi-
creases physical activity under free-living conditions but does not increase nese adults in the United States: multi-Ethnic Study of Atherosclerosis. Sleep
food intake under time-deprived laboratory conditions in healthy men. Am J 2018;41.
Clin Nutr 2009;90:1476e82. [48] Jackson CL, Ward JB, Johnson DA, et al. Concordance between self-reported
[39] Hucking K, Hamilton-Wessler M, Ellmerer M, et al. Burst-like control of lipolysis and actigraphy-assessed sleep duration among African-American adults:
by the sympathetic nervoussystem in vivo. J Clin Invest 2003;111:257e64. findings from the Jackson Heart Sleep Study. Sleep 2020;12:43.
[40] Spiegel Impact of sleep debt on metabolic and endocrine function. Lancet [49] Hauri PJ, Wisbey J. Wrist actigraphy in insomnia. Sleep 1992;15:293e301.
1999;354:1435e9. [50] Lockley SW, Skene DJ, Arendt J. Comparison between subjective and acti-
[41] Boudjeltia KZ, Faraut B, Esposito MJ, et al. Temporal dissociation between graphic measurement of sleep and sleep rhythms. J Sleep Res 1999;8:
myeloperoxidase (MPO)-modified LDL and MPO elevations during chronic 175e83.

32