Abstract
Study Objectives:
We aimed to identify the relationship between duration of categorized catch-up sleep on free days (CUS) and measured body mass index (BMI) in adults using the data from the seventh Korean National Health and Nutrition Examination Survey (KNHANES VII), 2016.
Methods:
CUS duration was classified as ≤ 0, > 0–1, > 1–2, and > 2 hours. Being overweight or obese was defined as having a BMI ≥ 25.0 kg/m2 or ≥ 30.0 kg/m2, respectively.
Results:
Of 6,382 participants aged 19–80 years in the KNHANES VII survey of 2016, 201 and 583 participants were excluded because of shift-working and insufficient data, respectively. Of 5,598 participants, CUS was observed in 2,274 (44.9%) participants, of which 3,324 (55.1%), 1,043 (19.4%), 724 (14.7%), and 507 (10.8%) had CUS of ≤ 0, > 0–1, > 1–2, and > 2 hours, respectively; the prevalence of obesity was 5.6%, 5.6%, 4.8%, and 6.1%, respectively. The association between BMI and CUS duration showed a significant negative association in the CUS ≤ 0 hours group (beta [95% confidence interval], −0.394 [−0.646, −0.143], P = .002); however, other CUS groups did not show any significant association with BMI (CUS > 0–1 hours: −0.196 [−1.258, 0.865], P = .716; CUS > 1–2 hours, −0.542 [−1.625, 0.541], P = .325; CUS > 2 hours, −0.113 [−0.459, 0.233], P = .519).
Conclusions:
Our findings provide an understanding of the relationship between CUS and BMI and can serve as an instructive basis for the management of BMI.
Citation:
Lee HJ, Cho S, Lee SH, Kim SJ, Kim KM, Chu MK. Catch-up sleep on free days and body mass index: results from the seventh Korea National Health and Nutrition Examination Survey, 2016. J Clin Sleep Med. 2024;20(1):39–47.
Keywords: body mass index, obesity, chronobiology, sleep, epidemiology
BRIEF SUMMARY
Current Knowledge/Study Rationale: Previous studies demonstrated contradictory results on the relationship between catch-up sleep on free days and body mass index. We performed a nationwide population study based on the actual measurement.
Study Impact: There was no significant difference in the risk of being obese or overweight among groups with duration of catch-up sleep on free days of ≤ 0, > 0–1, > 1–2, and > 2 hours. A significant negative linear association between body mass index and catch-up sleep on free days was observed in the group with catch-up sleep on free days with duration ≤ 0 hours.
INTRODUCTION
Obesity is a major global public health issue and a well-known risk factor for cardiovascular, neurodegenerative, psychiatric, and musculoskeletal disorders1 and has significant adverse effects on individuals, societies, and countries. Nevertheless, the prevalence of obesity has increased rapidly.
Sleep-related factors, such as quantity, quality, and duration, play crucial roles in the occurrence of obesity. Short sleep duration has been consistently expressed as an important factor in the development of obesity.2 Poor quality of sleep and evening chronotype are considerably associated with obesity.3,4 Therefore, identifying the sleep-related risk factors of obesity is important because they are modifiable and can be used to prevent or reduce obesity.
Habitual sleep restriction on workdays is often inevitable owing to the lifestyle and work schedules of modern society. Catch-up sleep on free days (CUS), a coping method for workday sleep restriction that entails a sleep deprivation pattern on workdays and compensatory sleep extension on free days, can affect health significantly. Potential associations of CUS with depression, hyperlipidemia, poor glycemic control, hypertension, and life quality have been proposed.5–9 Studies have reported contradictory results regarding the association between CUS duration and body mass index (BMI). Ad libitum sleep expansion on free days in response to sleep restriction for 9 nights cannot sufficiently clear the sleep debt and can only partially mitigate body weight gain.10 A population-based study revealed that individuals with CUS had a lower BMI than those without CUS.6 The relationship between CUS and BMI is dose-dependent.
The aim of this study was to investigate the association between CUS duration and BMI using the data obtained from a nationwide population-based study. We hypothesized that CUS duration is significantly related to BMI independent of sleep duration and chronotype. We also analyzed the correlation between CUS duration and the risk of obesity or being overweight.
METHODS
Study population
The data were obtained from the seventh Korean National Health and Nutrition Examination Survey (KNHANES VII), 2016. The KNHANES is a cross-sectional epidemiological survey that is conducted annually by the Korea National Institute of Health.11 The data obtained using a multistage stratified sampling design were used for this study through an operational definition. Detailed information pertaining to KNHANES has been described previously.11
Assessment of sleep-related factors and CUS duration
Information regarding each participant’s sleep duration on weekdays and free days was collected using questionnaires. The weighted weekly average sleep duration (hours per night) was calculated as ([workday sleep duration × 5] + [free day sleep duration × 2]/7). The chronotype was determined by the corrected midsleep duration on free days (MSFsc) after correcting for accumulated sleep debts over workdays. Midsleep duration on free days (MSF) was calculated as follows: sleep onset + sleep duration/2,12 and MSFsc was calculated using the following equation: MSF – (sleep duration on free days – weighted weekly average sleep duration)/2. Social jetlag (SJL) status was estimated as the difference (absolute value in hours) in the midsleep duration between workdays and free days (|MSF − MSW|). CUS was calculated using the following equation: average sleep duration on free days – average sleep duration on workdays. Accordingly, we categorized participants into those having CUS duration ≤ 0, > 0–1, > 1–2, and > 2 hours.
BMI, overweight, and obesity
BMI was calculated using the following formula: [weight (kilograms)/height (meters squared)]. Having a BMI ≥ 25.0 kg/m2 or ≥ 30.0 kg/m2 were defined as being overweight or obese, respectively.
Lifestyle covariates
Data pertaining to demographic factors (sex and age) and lifestyle-related variables (current employment status, alcohol consumption, smoking habit, highest academic background, physical activity, marital status, and household income) were obtained from responses to KNHANES VII surveys.
Some variables were split into groups for comparison: current employment status (employed vs unemployed), alcohol consumption (< 2 times/week vs ≥ 2 times/week), highest educational level (≥ 12 years vs < 12 years), smoking habit (current vs never or former), marital status (married vs never married), and household income (< 2 million won vs ≥ 2 million won) (1,000 won = 0.85 US dollar). Physical activity was assessed using the Global Physical Activity Questionnaire score, and we categorized the participants according to whether their physical activity fulfilled the World Health Organization’s recommendation (metabolic equivalents minutes per week ≥ 600 vs metabolic equivalents minutes per week < 600).
Assessment of depression
The Patient Health Questionnaire-9 was used to screen depression. The Patient Health Questionnaire-9 consists of 9 items; each item is scored from 0–3 according to the experience in the last 2 weeks.13 Participants who scored 10 or higher on this questionnaire were diagnosed to have depression. The Korean-validated Patient Health Questionnaire-9 has a sensitivity of 81.8% and a specificity of 89.9% for the diagnosis of depression.
Ethics statement
This study was approved by the Institutional Review Board of Severance Hospital, Yonsei University (approval no. 4-2022-0911). The design and all aspects of this study followed the principles of the Declaration of Helsinki and the KNHANES usage guidelines.14
Statistical analyses
Weighted chi-squared test and weighted one-way analysis of variance were conducted for categorical variables and continuous variables for characteristics analysis, respectively. Categorical variables are depicted as numbers with percentages, and continuous variables are shown as means ± standard errors.
We calculated the odds ratios with 95% confidence intervals (CIs) for the occurrence of being overweight (BMI ≥ 25.0 kg/m2) or obese (BMI ≥ 30.0 kg/m2) in different CUS duration groups and their subgroups through multivariable logistic regression analysis.
In multivariable logistic regression analysis, the association between CUS groups (independent variable) and the risk of being overweight (yes/no; dependent variable) or being obese (yes/no; dependent variable) was determined, with demographic (sex and age), lifestyle-related (current job status, alcohol intake, smoking habit, highest educational level, physical activity, marital status, and household income), and sleep-related (weekly average sleep duration, chronotype, and SJL) variables as covariates.
We used 4 models for multivariable logistic regression analysis. In model 1, adjustments were made for demographic variables (sex and age [years], continuous variables). Model 2 included lifestyle-related variables (current job status, alcohol intake, highest educational level, smoking habit, marital status, and household income) in addition to the demographic variables in model 1. Model 3 incorporated sleep-related variables (weekly average sleep period, chronotype [MSFsc], and SJL) in addition to variables in model 1. The final model, model 4, included lifestyle-related variables (current job status, alcohol intake, highest educational level, smoking habit, marital status, house income, and physical activity) with variables in model 3. Additional analyses excluding extreme chronotype (2.5% MSFsc) were conducted using all 4 models (models 1–4) for the risk of being overweight or obese. We also conducted a multivariable linear regression analysis to assess whether CUS duration was associated with continuous BMI after covariates adjustment. The IBM Statistical Package for Social Sciences Statistics for Windows, version 23.0 (IBM Corp., Armonk, New York) was used for most statistical analyses. Statistically significant differences were defined as two-tailed P < .05. Furthermore, Bonferroni’s correction for post hoc analyses was conducted for comparing the 4 CUS duration groups.
RESULTS
Study population
The flow chart of participant selection is summarized in Figure 1. Of the 8,150 participants, we excluded 1,768, 201, and 583 because of age < 19 years, being shift-workers, or incomplete data, respectively. Finally, 5,598 participants were included in this study.
Figure 1. Flow chart depicting the participant selection in the seventh Korea National Health and Nutrition Examination Survey (KNHANES VII-1), 2016.
CUS = catch-up sleep on free days.
Demographic, sleep-related, and lifestyle-related parameters and BMI according to CUS groups
The group with CUS duration > 0 hours included a greater number of females than the group with CUS duration ≤ 0 hours (P = .001). The mean age of participants was lower in the group with CUS duration > 0 hours than that in the group with CUS duration ≤ 0 hours (P < .001). The group with CUS duration > 0 hours exhibited a greater number of participants having the highest academic background of ≥ 12 years and a job than the group with CUS duration ≤ 0 hours (P < .001). Married individuals were more common in the CUS duration ≤ 0 hours group (P < .001). The incidence of alcohol intake and smoking habit were not significantly different among groups. The prevalence of being overweight (BMI ≥ 25.0 kg/m2) in participants with CUS duration ≤ 0 hours (37.6%) and > 0–1 hours (34.0%) was higher than that in participants with CUS duration > 1–2 hours (31.5%) and > 2 hours (32.2%) (P = .008 after Bonferroni’s correction; Table 1).
Table 1.
Demographic, lifestyle-related, and sleep-related characteristics and BMI of individuals according to CUS duration.
| CUS ≤ 0 h | 0 < CUS ≤ 1 h | 1 < CUS ≤ 2 h | CUS > 2 h | P | |
|---|---|---|---|---|---|
| (n = 3,324) | (n = 1,043) | (n = 724) | (n = 507) | ||
| Sex, female, n (%) [95% CI] | 1,859 (49.3) [47.7, 50.9] | 661 (56.6) [53.2, 59.9] | 441 (54.0) [50.2, 57.7] | 296 (52.0) [47.0, 57.1] | .001 |
| Age (years) mean ± SE | 51.19 ± 0.45 | 43.46 ± 0.60 | 41.28 ± 0.63 | 38.79 ± 0.72 | <.001 |
| Chronotype | |||||
| Early chronotype, n (%) [95% CI] | 1,125 (27.0) [24.9, 29.2] | 104 (7.3) [5.7, 9.4] | 62 (7.0) [5.2, 9.4] | 65 (10.6) [8.0, 14.1] | <.001 |
| Intermediate chronotype, n (%) [95% CI] | 1,911 (60.6) [58.3, 62.8] | 800 (75.2) [71.5, 78.6] | 550 (76.0) [71.9, 79.7] | 336 (65.4) [60.1, 70.3] | |
| Late chronotype, n (%) [95% CI] | 288 (12.4) [11.0, 14.0] | 139 (17.5) [14.2, 21.3] | 112 (17.0) [13.6, 20.9] | 106 (24.3) [19.1, 29.7] | |
| MSFsc, mean ± SE | 3.25 ± 0.04 | 3.82 ± 0.06 | 3.89 ± 0.08 | 3.93 ± 0.12 | <.001 |
| Highest educational level ≥ 12 years, n (%) [95% CI] | 1,918 (66.4) [63.8, 68.9] | 822 (84.0) [81.2, 86.4] | 608 (86.8) [84.2, 89.1] | 442 (89.6) [86.5, 92.1] | <.001 |
| Alcohol intake (≥ 2 times/wk), n (%) [95% CI] | 713 (24.3) [22.5, 26.2] | 209 (21.3) [18.5, 24.4] | 157 (23.0) [19.6, 26.8] | 123 (25.3) [20.9, 30.3] | .319 |
| Smoking habit (current smoking), n (%) [95% CI] | 587 (22.1) [20.3, 24.1] | 166 (19.2) [16.4, 22.3] | 130 (20.9) [17.5, 24.8] | 116 (25.6) [21.0, 30.8] | .152 |
| Current job status (having a current job), n (%) [95% CI] | 1696 (55.7) [53.3, 58.0] | 675 (65.9) [62.3, 69.2] | 495 (68.6) [64.0, 72.8] | 373 (73.9) [69.2, 78.2] | <.001 |
| Physical activity (≥ 600 MET), n (%) [95% CI] | 1,348 (43.9) [41.5, 46.3] | 520 (53.9) [50.3, 57.5] | 372 (52.6) [48.3, 56.9] | 250 (49.3) [44.0, 54.6] | <.001 |
| Marital status (married), n (%) [95% CI] | 2,951 (82.2) [80.3, 84.0] | 869 (75.8) [71.5, 79.6] | 566 (72.7) [68.2, 76.8] | 343 (60.8) [55.2, 66.1] | <.001 |
| House income (≥ 2 million won), n (%) [95% CI] | 1,721 (46.3) [42.9, 49.7] | 352 (31.3) [27.3, 35.6] | 242 (33.7) [29.0, 38.6] | 172 (33.4) [28.3, 38.9] | <.001 |
| Workday sleep duration per day (h), mean ± SE | 7.26 ± 0.03 | 7.03 ± 0.04 | 6.67 ± 0.05 | 6.17 ± 0.07 | <.001 |
| Free day sleep duration per day (h), mean ± SE | 7.05 ± 0.03 | 7.85 ± 0.04 | 8.42 ± 0.05 | 9.42 ± 0.07 | <.001 |
| Weekly average sleep duration (h), mean ± SE | 7.20 ± 0.03 | 7.26 ± 0.04 | 7.17 ± 0.05 | 7.10 ± 0.06 | <.001 |
| Social jetlag (h), mean ± SE | 0.18 ± 0.02 | 0.80 ± 0.04 | 1.13 ± 0.06 | 1.66 ± 0.07 | <.001 |
| PHQ-9 score, mean ± SE | 2.84 ± 0.10 | 2.40 ± 0.13 | 2.42 ± 0.12 | 2.76 ± 0.19 | <.001 |
| PHQ-9 score > 10, n (%) [95% CI] | 244 (7.0) [6.0, 8.2] | 47 (4.2) [3.0, 5.8] | 25 (2.9) [1.9, 4.5] | 30 (6.0) [3.9, 8.9] | <.001 |
| BMI (kg/m2), mean ± SE | 24.19 ± 0.085 | 23.84 ± 0.14 | 23.55 ± 0.18 | 23.80 ± 0.19 | .003 |
| Overweight (BMI ≥ 25.0 kg/m2), n (%) [95% CI] | 1,231 (37.6) [35.5,39.8] | 355 (34.0) [30.6, 37.6] | 227 (31.5) [27.7, 35.6] | 157 (32.2) [27.7, 37.1] | .008 |
| Obesity (BMI ≥ 30.0 kg/m2), n (%) [95% CI] | 180 (5.6) [4.7, 6.7] | 58 (5.6) [4.2, 7.4] | 38 (4.8) [3.3, 6.9] | 33 (6.1) [4.2, 8.8] | .829 |
BMI = body mass index, CI = confidence interval, CUS = catch-up sleep on free days, MET = metabolic equivalent, MSFsc = midsleep duration on free days corrected for sleep debt over workdays, PHQ-9 = Patient Health Questionnaire-9, SE = standard error.
Sleep-related parameters according to CUS groups
Of 5,598 participants, 1,356 (24.2%), 3,597 (64.3%), and 645 (11.5%) were classified as early, intermediate, and late chronotypes, respectively. The distribution of chronotypes of participants was significantly different among the different CUS groups. Sleep duration on workdays and free days, weekly average sleep duration, and SJL were significantly different among the CUS groups (Table 1).
Multivariable logistic regression analyses for association of the risk of being overweight or obese according to CUS groups
In model 1 adjusted for sex and age, the multivariable-adjusted odds ratios for the risk of being overweight (BMI ≥ 25.0 kg/m2) were 0.962 (95% CI, 0.815, 1.134), 0.863 (95% CI, 0.709, 1.051), and 0.906 (95% CI, 0.724, 1.134) for participants with CUS duration > 0–1, > 1–2, and > 2 hours, respectively, compared with those with ≤ 0 hours. Models 2, 3, and 4 showed similar patterns in the group with CUS duration > 1–2 hours of a decreased odds ratio, whereas the groups with CUS duration > 0–1 hours and > 2 hours showed no significant difference with the group with CUS duration ≤ 0 hours (Table 2).
Table 2.
Multivariable adjusted OR (95% CI) for the risk of being overweight (BMI ≥ 25.0 kg/m2) in different CUS groups.
| CUS ≤ 0 h | 0 < CUS ≤ 1 h | 1 < CUS ≤ 2 h | CUS > 2 h | |
|---|---|---|---|---|
| (n = 3,324) | (n = 1,043) | (n = 724) | (n = 507) | |
| Model 1 | REF | 0.962 (0.815, 1.134) | 0.863 (0.709, 1.051) | 0.906 (0.724, 1.134) |
| Model 2 | REF | 0.977 (0.824, 1.157) | 0.857 (0.702, 1.047) | 0.900 (0.713, 1.136) |
| Model 3 | REF | 0.987 (0.837, 1.163) | 0.878 (0.713, 1.081) | 0.918 (0.706, 1.081) |
| Model 4 | REF | 1.001 (0.844, 1.187) | 0.869 (0.704, 1.073) | 0.902 (0.689, 1.181) |
Model 1: adjusted for age and sex. Model 2: adjusted for age, sex, current job status, alcohol intake, highest educational level, smoking habit, marital status, and household income. Model 3: adjusted for age, sex, workday sleep duration, chronotype (midsleep duration on free days corrected for sleep debt over workdays), and social jetlag. Model 4: adjusted for age, sex, workday sleep duration, chronotype (midsleep duration on free days corrected for sleep debt over workdays), social jetlag, current job status, alcohol intake, highest educational level, smoking habit, marital status, and household income. 95% CI = 95% confidence interval, BMI = body mass index, CUS = catch-up sleep on free days, OR = odds ratio, REF = reference.
Multivariable logistic regression analyses for the association between the risk of obesity (BMI ≥ 30.0 kg/m2) in different CUS groups also showed a pattern similar to that of the risk of being overweight, with no significantly different risk between different CUS groups (Table 3).
Table 3.
Multivariable adjusted OR (95% CI) for the risk of obesity (BMI ≥ 30.0 kg/m2) in different CUS groups.
| CUS ≤ 0 h | 0 < CUS ≤ 1 h | 1 < CUS ≤ 2 h | CUS > 2 h | |
|---|---|---|---|---|
| (n = 3,324) | (n = 1,043) | (n = 724) | (n = 507) | |
| Model 1 | REF | 0.954 (0.650, 1.400) | 0.801 (0.516, 1.245) | 1.016 (0.662, 1.558) |
| Model 2 | REF | 1.042 (0.705, 1.539) | 0.843 (0.541, 1.314) | 1.029 (0.660, 1.602) |
| Model 3 | REF | 0.973 (0.665, 1.424) | 0.811 (0.512, 1.284) | 1.023 (0.632, 1.284) |
| Model 4 | REF | 1.062 (0.719, 1.568) | 0.849 (0.532, 1.355) | 1.019 (0.621, 1.671) |
Model 1: adjusted for age and sex. Model 2: adjusted for age, sex, current job status, alcohol intake, highest educational level, smoking habit, marital status, and household income. Model 3: adjusted for age, sex, workday sleep duration, chronotype (midsleep duration on free days corrected for sleep debt over workdays), and social jetlag. Model 4: adjusted for age, sex, workday sleep duration, chronotype (midsleep duration on free days corrected for sleep debt over workdays), social jetlag, current job status, alcohol intake, highest educational level, smoking habit, marital status, and household income. 95% CI = 95% confidence interval, BMI = body mass index, CUS = catch-up sleep on free days, OR = odds ratio, REF = reference.
Additional multivariable logistic regression analyses for the association of the risk of being overweight or obese in different CUS groups after excluding extreme chronotypes
We conducted additional multivariable logistic regression analyses after excluding extreme chronotypes (2.5% MSFsc). All 4 models showed no significant difference in the risk of being overweight (Table S1 (245.2KB, pdf) in the supplemental material) or obese (Table S2 (245.2KB, pdf) ) in different CUS groups.
Multivariable linear regression analyses for association between CUS duration and BMI in different CUS groups
We performed multivariable linear regression analysis to assess the relationship between CUS duration and BMI using the 4 models. In model 1, participants with CUS duration ≤ 0 hours showed a considerably negative linear relationship between CUS duration and BMI (beta and 95% CI, −0.394 [−0.646, −0.143], P = .002). Individuals with CUS duration > 0–1, > 1–2, and > 2 hours showed no such significant linear relationship. Similar relationships were noticed in models 2, 3, and 4. CUS duration ≤ 0 hours showed a significantly positive linear relationship with BMI, whereas CUS duration > 0–1, > 1–2, and > 2 hours showed no such significant linear relationship (Figure 2 and Table 4). Additional multivariable analyses excluding extreme chronotypes (2.5% chronotype) for the association between CUS duration and BMI showed similar patterns; CUS duration ≤ 0 hours showed a significant negative linear relationship, whereas CUS duration > 0–1, > 1–2, and > 2 hours showed no such significant linear relationship (Table S3 (245.2KB, pdf) ).
Figure 2. Association between CUS duration and body mass index among individuals in different CUS groups.
CUS = catch-up sleep on free days.
Table 4.
Association between CUS duration and BMI among individuals in different CUS groups.
| CUS ≤ 0 h (n = 3,324) | 0 < CUS ≤ 1 h (n = 1,043) | 1 < CUS ≤ 2 h (n = 724) | CUS > 2 h (n = 507) | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| β | 95% CI | P | β | 95% CI | P | β | 95% CI | P | β | 95% CI | P | |
| With CUSa | −0.285 | −0.527, −0.043 | .021 | −0.284 | −1.368, 0.801 | .606 | −0.462 | −1.543, 0.618 | .400 | −0.091 | −0.447, 0.264 | .613 |
| With CUSb | −0.276 | −0.520, −0.032 | .027 | −0.258 | −1.331, 0.815 | .636 | −0.489 | −1.560, 0.582 | .369 | −0.105 | −0.461, 0.250 | .560 |
| With CUSc | −0.404 | −0.652, −0.156 | .002 | −0.225 | −1.296, 0.846 | .679 | −0.502 | −1.593, 0.588 | .364 | −0.101 | −0.446, 0.243 | .562 |
| With CUSd | −0.394 | −0.646, −0.143 | .002 | −0.196 | −1.258, 0.865 | .716 | −0.542 | −1.625, 0.541 | .325 | −0.113 | −0.459, 0.233 | .519 |
aModel 1: adjusted for age and sex. bModel 2: adjusted for age, sex, current job status, alcohol intake, highest educational level, smoking habit, marital status, and household income. cModel 3: adjusted for age, sex, workday sleep duration, chronotype (midsleep duration on free days corrected for sleep debt over workdays), and social jetlag. dModel 4: adjusted for age, sex, workday sleep duration, chronotype (midsleep duration on free days corrected for sleep debt over workdays), social jetlag, current job status, alcohol intake, highest educational level, smoking habit, marital status, and household income. 95% CI = 95% confidence interval, BMI = body mass index, CUS = catch-up sleep on free days.
Multivariable logistic regression analyses for the association of the risk of being overweight or obese with CUS duration according to age and sex
We further analyzed the association of the risk of being overweight or obese with CUS duration according to age and sex. We categorized participants into age groups of 20 years: aged 19–40, 41–60, and 61–80 years. We conducted multivariable logistic regression analyses, adjusting for demographic, lifestyle-related, and sleep-related factors for the risk of being overweight or obese. For the risk of being overweight, all participants aged 40–60 years with CUS duration > 2 hours had a significant negative association, and female participants aged 61–80 years with CUS duration > 0–1 hours had a significant positive association (Table S4 (245.2KB, pdf) ). For the risk of being obese, all participants aged 19–40 years with CUS duration > 2 hours showed a significant positive association (Table S5 (245.2KB, pdf) ). Other CUS groups did not reveal any significant association with the risk of being obese or overweight according to age and sex.
DISCUSSION
The main findings of the present study were as follows: (1) CUS duration > 0 hours was observed in 50.7% of adults aged 19–80 years; (2) there was no significant difference in the risk of being obese or overweight according to CUS groups; and (3) a significant negative linear association between CUS duration and BMI was observed in the group with CUS duration ≤ 0 hours. In contrast, other CUS groups did not show any significant association between BMI and CUS duration. Based on these findings, we could partially reject our hypothesis that CUS duration is related to BMI, because a significant association was observed only in participants with CUS duration ≤ 0 hours. The unfavorable effects of short sleep duration on adult obesity have been reported in several cross-sectional and longitudinal studies,2–15 and negative effects of insufficient sleep duration on obesity have also been observed in children and adolescents.16 Sleep extension on free days, CUS, has been demonstrated to partially ameliorate the deleterious effects of insufficient sleep. CUS has also been associated with a decreased risk of hypertension, depression, and metabolic syndrome.17–19 Nevertheless, other studies showed no or negative effects of CUS on metabolic change and depression.20–23 With regard to the effect of CUS on BMI in a population-based setting, Im et al investigated the effect of CUS on BMI using data from a population-based study including 2,156 adults in Korea and found that individuals with CUS had a significantly lower BMI than those without CUS after adjusting for age, sex, sleep duration, chronotype, mood-related factors, and sociodemographic factors.6 However, the study determined the body weight and height based on participants’ reports rather than actual measurements. In addition, the study did not adjust for SJL and did not categorize the CUS duration. Although BMI evaluation based on self-reported data has been used in several epidemiological studies, such self-reported data have often been shown to be inaccurate because of errors such as underreporting of weight and overreporting of height by the participants.24 Therefore, actual measurements should be used for the evaluation of obesity, and BMI and prevalence of obesity should be estimated based on the actual height and weight.25 SJL is a chronic jetlag-like phenomenon that reflects circadian misalignment between an individual’s internal sleep–wake cycles and actual clock-regulated physiology, most often due to aberrant work schedules.26,27 SJL is reported to be associated with the incidence of depression, metabolic syndrome, atherosclerosis, and increased BMI,28 and a significant positive correlation has been observed between SJL and CUS duration.21 Because the negative effects of SJL may ameliorate the positive effects of CUS, SJL should be included in the evaluation of the effects of CUS on BMI. Moreover, in the present study, CUS duration was categorized as > 0–1, > 1–2, and > 2 hours, but Im et al6 did not categorize CUS duration and only identified the presence of CUS. CUS may have different effects according to CUS duration. A population study showed that individuals with CUS duration > 1–2 hours had a decreased risk of depression compared with those with CUS duration ≤ 0 hours. In contrast, individuals with CUS duration > 0–1 and > 2 hours did not show any significant difference in the risk of depression.21 Therefore, the effects of CUS should be analyzed considering the CUS duration.
The present study found a linear relationship between CUS duration and BMI in individuals with CUS duration ≤ 0 hours, but other CUS duration groups did not show any such significant linear relationship with BMI. This finding suggests a possibility that CUS duration ≤ 0 hours and more sleep time on workdays than on free days were quantitatively associated with a higher BMI; however, CUS duration > 0 hours was not associated with BMI. The possible mechanisms for the significant negative association between CUS duration and BMI in participants with CUS duration ≤ 0 hours should be considered. One possible explanation is that negative CUS (CUS duration < 0 hours) is associated with more sleep insufficiency than positive CUS (CUS duration ≥ 0 hours), and more sleep insufficiency may lead to an increase in BMI.29–31 Sleep insufficiency, caused by various conditions, including reduced sleep duration, poor sleep quality, and shift-working, has been associated with an increase in BMI. Less sleep duration on workdays than on free days may exacerbate sleep insufficiency by not compensating for sleep insufficiency on free days and lead to a significant increase in BMI. The lack of a linear association between CUS and BMI in positive CUS groups may be attributed to the partial compensation of sleep insufficiency by CUS. No significant difference in the risk of being overweight or obese among different CUS groups may be related to the fact that although a significant association between CUS and BMI exists in specific CUS groups the effect is not adequate to affect the overall risk.
Data from a nationwide population-based study in Korea were used in this study. Because the association between CUS and BMI is inconsistent across different sex, age, regional, and social characteristics, these evenly distributed multistage stratified samples may be especially useful to understand the phenomena. However, the findings of the present study are not applicable to individuals < 19 years of age and those engaged in shiftwork. In addition, socioeconomic status and ethnicity across the countries can have significant effects on both for BMI and sleep.32–35 Although we included socioeconomic status variables (monthly income and educational level) in our analyses, their relationships with BMI may vary according to the level of economic development in each country.36 Therefore, the findings of this study should be interpreted with caution, and further investigations should be conducted in more diverse populations to elucidate the effect of CUS duration and BMI.
The present study had several limitations. First, we evaluated sleep-related parameters such as sleep period on weekdays and free days, sleep onset time, and wake-up time using a questionnaire rather than using objective methods such as actigraphy and polysomnography. Nevertheless, objective assessment of sleep through epidemiological studies is very difficult, and most epidemiological studies have assessed sleep-related parameters based on participants’ reports. Furthermore, self-reported sleep duration has been correlated with values determined using actigraphy.37 Second, this was a cross-sectional study that examined the association between CUS duration and BMI in a population-based setting. Consequently, our study could not determine the cause–effect relationship between CUS duration and BMI. Third, we did not include other potential factors regarding sleep and BMI. Excessive daytime sleepiness and insomnia are sleep-related factors that have been significantly associated with BMI.38,39 Nevertheless, detailed information on excessive daytime sleepiness and insomnia was lacking in the KNHANES VII. Further studies including data pertaining to excessive daytime sleepiness and insomnia in the analyses will provide a better understanding of the association between CUS duration and BMI. Finally, there are certain issues with the generalizability of the results, as mentioned above, that will need to be addressed in future studies.
However, the present study also has strengths such as a population-based sample proportional to population distribution, a large sample size, BMI assessment through actual measurement of height and weight, and analyses that adjusted for various factors related to BMI, including mood, sleep, and lifestyle-related factors.
In conclusion, CUS was observed in 55.1% of individuals aged 19–80 years. Individuals with CUS duration ≤ 0 hours exhibited a significant negative association with BMI. Nevertheless, other CUS groups did not show any such significant association with BMI. The findings of the present study may provide a better understanding of the association of CUS with BMI and may be helpful in the management of obesity.
DISCLOSURE STATEMENT
All authors have seen and approved this manuscript. This research was supported by a grant from the Korea Health Technology R&D Project through the Korea Health Industry Development Institute (KHIDI), funded by the Ministry of Health & Welfare, Republic of Korea (grant no. HV22C0106) and a National Research Foundation of Korea (NRF) grant from the Korean government (MSIT) (2022R1A2C1091767). M.K.C. was a site investigator for a multicenter trial sponsored by Biohaven Pharmaceuticals, Allergan Korea, and the Ildong Pharmaceutical Company. He has received lecture honoraria from Eli Lilly and Company, Handok-Teva, and Ildong Pharmaceutical Company over the past 24 months. He received grants from Yonsei University College of Medicine (6-2021-0229), the Korea Health Industry Development Institute (KHIDI) (HV22C0106), and an NRF grant from the Korean government (MSIT) (2022R1A2C1091767). The other authors declare no conflicts of interest.
ABBREVIATIONS
- BMI
body mass index
- CI
confidence interval
- CUS
catch-up sleep on free days
- KNHANES VII
seventh Korean National Health and Nutrition Examination Survey
- MSF
midsleep duration on free days
- MSFsc
midsleep duration on free days corrected for sleep debt
- SJL
social jetlag
REFERENCES
- 1. Blüher M . Obesity: global epidemiology and pathogenesis . Nat Rev Endocrinol. 2019. ; 15 ( 5 ): 288 – 298 . [DOI] [PubMed] [Google Scholar]
- 2. Zhou Q, Zhang M, Hu D . Dose-response association between sleep duration and obesity risk: a systematic review and meta-analysis of prospective cohort studies . Sleep Breath. 2019. ; 23 ( 4 ): 1035 – 1045 . [DOI] [PubMed] [Google Scholar]
- 3. Fatima Y, Doi SA, Mamun AA . Sleep quality and obesity in young subjects: a meta-analysis . Obes Rev. 2016. ; 17 ( 11 ): 1154 – 1166 . [DOI] [PubMed] [Google Scholar]
- 4. Teixeira GP, Guimarães KC, Soares AGNS, et al . Role of chronotype in dietary intake, meal timing, and obesity: a systematic review . Nutr Rev. 2022. ; 81 ( 1 ): 75 – 90 . [DOI] [PubMed] [Google Scholar]
- 5. Oh YH, Kim H, Kong M, Oh B, Moon JH . Association between weekend catch-up sleep and health-related quality of life of Korean adults . Medicine (Baltimore). 2019. ; 98 ( 13 ): e14966 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Im HJ, Baek SH, Chu MK, et al . Association between weekend catch-up sleep and lower body mass: population-based study . Sleep. 2017. ; 40 ( 7 ): zsx089 . [DOI] [PubMed] [Google Scholar]
- 7. Hwangbo Y, Kim WJ, Chu MK, Yun CH, Yang KI . Association between weekend catch-up sleep duration and hypertension in Korean adults . Sleep Med. 2013. ; 14 ( 6 ): 549 – 554 . [DOI] [PubMed] [Google Scholar]
- 8. Kuula L, Pesonen AK, Kajantie E, et al . Sleep and lipid profile during transition from childhood to adolescence . J Pediatr. 2016. ; 177 : 173 – 178.e1 . [DOI] [PubMed] [Google Scholar]
- 9. Ness KM, Strayer SM, Nahmod NG, Chang AM, Buxton OM, Shearer GC . Two nights of recovery sleep restores the dynamic lipemic response, but not the reduction of insulin sensitivity, induced by five nights of sleep restriction . Am J Physiol Regul Integr Comp Physiol. 2019. ; 316 ( 6 ): R697 – R703 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Depner CM, Melanson EL, Eckel RH, et al . Ad libitum weekend recovery sleep fails to prevent metabolic dysregulation during a repeating pattern of insufficient sleep and weekend recovery sleep . Curr Biol. 2019. ; 29 ( 6 ): 957 – 967.e4 . [DOI] [PubMed] [Google Scholar]
- 11. Kweon S, Kim Y, Jang MJ, et al . Data resource profile: the Korea National Health and Nutrition Examination Survey (KNHANES) . Int J Epidemiol. 2014. ; 43 ( 1 ): 69 – 77 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Roenneberg T, Kuehnle T, Juda M, et al . Epidemiology of the human circadian clock . Sleep Med Rev. 2007. ; 11 ( 6 ): 429 – 438 . [DOI] [PubMed] [Google Scholar]
- 13. Choi HS, Choi JH, Park KH, et al . Standardization of the Korean version of Patient Health Questionnaire-9 as a screening instrument for major depressive disorder . J Korean Acad Fam Med. 2007. ; 28 ( 2 ): 114 – 119 . [Google Scholar]
- 14. Lee J, Kim HR . The association between long working hours and high-sensitivity C-reactive protein in older aged individuals: the Korea National Health and Nutrition Examination Survey (KNHANES) 2015 . J Occup Environ Med. 2018. ; 60 ( 9 ): 775 – 780 . [DOI] [PubMed] [Google Scholar]
- 15. Bacaro V, Ballesio A, Cerolini S, et al . Sleep duration and obesity in adulthood: an updated systematic review and meta-analysis . Obes Res Clin Pract. 2020. ; 14 ( 4 ): 301 – 309 . [DOI] [PubMed] [Google Scholar]
- 16. Cappuccio FP, Taggart FM, Kandala NB, et al . Meta-analysis of short sleep duration and obesity in children and adults . Sleep. 2008. ; 31 ( 5 ): 619 – 626 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Koo DL, Yang KI, Kim JH, et al . Association between morningness-eveningness, sleep duration, weekend catch-up sleep and depression among Korean high-school students . J Sleep Res. 2021. ; 30 ( 1 ): e13063 . [DOI] [PubMed] [Google Scholar]
- 18. Son SM, Park EJ, Cho YH, et al . Association between weekend catch-up sleep and metabolic syndrome with sleep restriction in Korean Adults: a cross-sectional study using KNHANES . Diabetes Metab Syndr Obes. 2020. ; 13 : 1465 – 1471 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Dewald-Kaufmann JF, Oort FJ, Meijer AM . The effects of sleep extension and sleep hygiene advice on sleep and depressive symptoms in adolescents: a randomized controlled trial . J Child Psychol Psychiatry. 2014. ; 55 ( 3 ): 273 – 283 . [DOI] [PubMed] [Google Scholar]
- 20. van Leeuwen WM, Hublin C, Sallinen M, Härmä M, Hirvonen A, Porkka-Heiskanen T . Prolonged sleep restriction affects glucose metabolism in healthy young men . Int J Endocrinol. 2010. ; 2010 : 108641 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Kim KM, Han SM, Min IK, Heo K, Kim WJ, Chu MK . Weekend catch-up sleep and depression: results from a nationally representative sample in Korea . Sleep Med. 2021. ; 87 : 62 – 68 . [DOI] [PubMed] [Google Scholar]
- 22. McGrath ER, Espie CA, Power A, et al . Sleep to lower elevated blood pressure: a randomized controlled trial (SLEPT) . Am J Hypertens. 2017. ; 30 ( 3 ): 319 – 327 . [DOI] [PubMed] [Google Scholar]
- 23. So-Ngern A, Chirakalwasan N, Saetung S, Chanprasertyothin S, Thakkinstian A, Reutrakul S . Effects of two-week sleep extension on glucose metabolism in chronically sleep-deprived individuals . J Clin Sleep Med. 2019. ; 15 ( 5 ): 711 – 718 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Lu S, Su J, Xiang Q, Zhou J, Wu M . Accuracy of self-reported height, weight, and waist circumference in a general adult Chinese population . Popul Health Metr. 2016. ; 14 ( 1 ): 30 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Jeong J-y, Kim D-H, Kim K-Y, Ryu SY, Lee S-Y, Park YS . Accuracy of self-reported height, weight and body mass index in Community Health Survey in South Korea . J Health Inf Stat. 2017. ; 42 ( 3 ): 241 – 249 . [Google Scholar]
- 26. Roenneberg T, Allebrandt KV, Merrow M, Vetter C . Social jetlag and obesity . Curr Biol. 2012. ; 22 ( 10 ): 939 – 943 . [DOI] [PubMed] [Google Scholar]
- 27. Wittmann M, Dinich J, Merrow M, Roenneberg T . Social jetlag: misalignment of biological and social time . Chronobiol Int. 2006. ; 23 ( 1-2 ): 497 – 509 . [DOI] [PubMed] [Google Scholar]
- 28. Islam Z, Hu H, Akter S, et al . Social jetlag is associated with an increased likelihood of having depressive symptoms among the Japanese working population: the Furukawa Nutrition and Health Study . Sleep. 2020. ; 43 ( 1 ): zsz204 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Altman NG, Izci-Balserak B, Schopfer E, et al . Sleep duration versus sleep insufficiency as predictors of cardiometabolic health outcomes . Sleep Med. 2012. ; 13 ( 10 ): 1261 – 1270 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Chang WP, Yang CM . Influence of sleep-wake cycle on body mass index in female shift-working nurses with sleep quality as mediating variable . Ind Health. 2020. ; 58 ( 2 ): 161 – 169 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Ming X, Koransky R, Kang V, Buchman S, Sarris CE, Wagner GC . Sleep insufficiency, sleep health problems and performance in high school students . Clin Med Insights Circ Respir Pulm Med. 2011. ; 5 : 71 – 79 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Vieira LS, Bierhals IO, Vaz JDS, Meller FO, Wehrmeister FC, Assunção MCF . Socioeconomic status throughout life and body mass index: a systematic review and meta-analysis . Cad Saude Publica. 2019. ; 35 ( 10 ): e00125518 . [DOI] [PubMed] [Google Scholar]
- 33. McLaren L . Socioeconomic status and obesity . Epidemiol Rev. 2007. ; 29 ( 1 ): 29 – 48 . [DOI] [PubMed] [Google Scholar]
- 34. Grandner MA, Patel NP, Gehrman PR, et al . Who gets the best sleep? Ethnic and socioeconomic factors related to sleep complaints . Sleep Med. 2010. ; 11 ( 5 ): 470 – 478 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Jehan S, Myers AK, Zizi F, et al . Sleep health disparity: the putative role of race, ethnicity and socioeconomic status . Sleep Med Disord. 2018. ; 2 ( 5 ): 127 – 133 . [PMC free article] [PubMed] [Google Scholar]
- 36. Dinsa GD, Goryakin Y, Fumagalli E, Suhrcke M . Obesity and socioeconomic status in developing countries: a systematic review . Obes Rev. 2012. ; 13 ( 11 ): 1067 – 1079 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Lockley SW, Skene DJ, Arendt J . Comparison between subjective and actigraphic measurement of sleep and sleep rhythms . J Sleep Res. 1999. ; 8 ( 3 ): 175 – 183 . [DOI] [PubMed] [Google Scholar]
- 38. Mokhber S, Zargham Ravanbakhsh P, Jesmi F, Pishgahroudsari M, Ghanbari Jolfaei A, Pazouki A . Comparing the excessive daytime sleepiness of obese and non-obese patients . Iran Red Crescent Med J. 2016. ; 18 ( 7 ): e21964 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Chan WS, Levsen MP, McCrae CS . A meta-analysis of associations between obesity and insomnia diagnosis and symptoms . Sleep Med Rev. 2018. ; 40 : 170 – 182 . [DOI] [PubMed] [Google Scholar]