Association Between Weekend Catch-Up Sleep and Depression: Evidence From the 2017–2018 National Health and Nutrition Examination Survey

Article information

Psychiatry Investig. 2025;22(9):997-1005

Publication date (electronic) : 2025 September 1

doi : https://doi.org/10.30773/pi.2024.0252

Wenting Cai, Zhonghong Lai, Shan Huang, Feng Xiao, Xueqin Liu, Yuxu Zhao, Ye Luo

Department of Psychiatry, The Third People’s Hospital of Ganzhou, Ganzhou, China

Correspondence: Ye Luo, MS Department of Psychiatry, The Third People’s Hospital of Ganzhou, No. 10, Zhangjiang North Avenue, Zhanggong District, Ganzhou, Jiangxi 341000, China Tel: +86-13970742331, E-mail: ly791128@163.com

Received 2024 August 25; Revised 2025 February 7; Accepted 2025 May 26.

Abstract

INTRODUCTION

Sleep plays a crucial role in human physiology, encompassing about one-third of our daily life. Despite common misconceptions, sleep is not a passive state but rather an active process that triggers significant physiological changes throughout the body [1]. Hence, ensuring adequate sleep duration and quality is crucial for optimal physical and mental well-being. Unfortunately, sleep disturbances are widespread in contemporary society [2], with sleep deprivation emerging as a prevalent concern [3]. Notably, both inadequate and excessive sleep duration have been linked to elevated risk of various metabolic and cardiovascular conditions, including diabetes [4], hypertension [5], metabolic syndrome [6], cardiovascular disease (CVD) [7], obesity, 8 and psychiatric disorders such as depression [9]. Currently chronic sleep deficiency and sleep disorders are regarded as pressing public health challenges [10].

Depression, a highly prevalent and debilitating disorder, affects approximately 7% of the global population [11]. Individuals with depression often experience impaired functional capacity and significant health-related disability, as evidenced by poor self-rated health status [12]. In 2015, depression ranked as the third leading cause of disability worldwide, due to its substantial prevalence and association with disabilities [13]. Furthermore, depression frequently co-occurs with chronic illnesses may precipitate suicidal behavior [14].

In contemporary society, the duration of sleep is often compromised by social commitments or work schedules, resulting in a population-wide decline in average sleep duration. This sleep curtailment adversely affects both mental well-being and metabolic health. This reduced sleep pattern is linked to a heightened risk of developing depression [15,16]. To compensate for weekday sleep debt, some individuals extend their sleep on weekends, a phenomenon termed weekend catch-up sleep (CUS). This compensatory strategy aims to counteract insufficient weekday sleep through prolonged sleep on free days. Notably, CUS may confer metabolic benefits, including improved insulin sensitivity, lower body mass index (BMI), reduced blood pressure, and a decreased risk of dyslipidemia [17-19]. Among Korean adults, data from the 2016 Korean National Health and Nutrition Examination Survey (NHANES) indicated that those with insufficient sleep who did not practice weekend CUS faced a significantly higher risk of depression compared to those who compensated with weekend CUS [20]. Similarly, a study of Korean high school students suggested that ≥2-hour CUS combined with frequent private tutoring was associated with a lower likelihood of depression [21]. Additionally, research involving a nationally representative Korean sample found that 1–2 hours of weekend CUS was associated with a lower depression risk relative to ≤0 hours of CUS [22]. However, existing evidence on the CUS-depression relationship remains limited and inconsistent. To address this gap, we analyzed data from the 2017–2018 NHANES to assess this association, with subgroup analyses to explore potential variations.

METHODS

RESULTS

Participants

The participant flow is presented in Figure 1. A total of 9,254 participants were sourced from the NHANES database. Among them, 4,804 participants were excluded, comprising 1,150 non-adults, 2,000 individuals with incomplete sleep duration data, 1,064 participants lacking PHQ-9 scores, and 590 participants reporting less than -1 hour of weekend sleep. Consequently, the final sample size for statistical analysis consisted of 4,450 eligible participants.

Figure 1.

Flowchart of the screening of participants for the 2017–2018 NHANES database in the United States. NHANE, National Health and Nutrition Examination Survey; CUS, weekend catch-up sleep; PHQ-9, Patient Health Questionnaire-9.

Sleep duration on weekdays and weekend, and weekly average sleep duration

Mean sleep duration on weekdays and weekends were 7.5± 1.6 h and 8.4±1.6 h, respectively. Mean weekly average sleep was 7.7±1.5 h (Table 1).

Table 1.

Baseline characteristics of individuals with CUS duration ≤0 h, 0 h<CUS≤1 h, 1 h<CUS≤2 h, and >2 h

Multivariable logistic regression analyses association between weekend CUS and risk of depression

In Model 1, adjusted for age and sex, the multivariable adjusted odds ratios (ORs) for depression were 0.78 (95% confidence interval [CI] 0.59–1.04), 0.72 (95% CI 0.51–1.02), and 0.91 (95% CI 0.67–1.24) for individuals with different durations of weekend-CUS: >0 to 1 hour, >1 to 2 hours, and >2 hours, respectively, compared to those with ≤0 hours (Table 2). When further adjusting for education, physical activity, smoking, drinking, PIR, and living status in Model 2 in addition to Model 1, the multivariable adjusted ORs were 0.87 (95% CI 0.63–1.20), 0.80 (95% CI 0.59–1.19), and 0.96 (95% CI 0.67–1.36) for individuals with different CUS durations, as detailed in Table 2. Similarly, in Model 3 (adjusted for average sleep duration in addition to Model 1), the multivariable adjusted ORs of experiencing depression were 0.77 (95% CI 0.59–1.02), 0.71 (95% CI 0.51–0.99), and 0.87 (95% CI 0.65–1.18) for individuals with varying CUS durations compared to those with ≤0 hours. Lastly, in Model 4 (adjusted for BMI, race, diabetes, cholesterol, hypertension, and other CVD factors in addition to Models 2 and 3), the multivariable adjusted ORs for depression were 0.89 (95% CI 0.54–1.48), 0.71 (95% CI 0.37–1.37), and 0.97 (95% CI 0.57–1.65) for different CUS duration groups compared to the reference group with ≤0 hours (Table 2).

Table 2.

Multivariable adjusted OR (95% CI) for the association between CUS with depression

Multivariate logistic regression analyses assessing the relationship between weekend CUS and depression under subgroups of weekday sleep duration

Notwithstanding the absence of significant association between weekend CUS and depression in the primary analysis, the potential confounding effects of multiple variables warrant consideration. Specifically, whether weekday sleep duration modulates the potential antidepressant effects of CUS remains indeterminate. To elucidate this interaction, stratified analyses were conducted across prespecified subgroups based on weekday sleep duration patterns. When comparing those with no CUS, those with over 2 hours of CUS had an adjusted OR of 0.58 (95% CI 0.36–0.94), and those with CUS exceeding 2 hours had an OR of 0.47 (95% CI 0.27–0.83), in individuals with weekday sleep duration of 6 hours or less (Table 3). The risk reduction for depression through weekend CUS was notably higher in individuals who slept 6 hours or less on weekdays and had more than 2 hours of CUS on weekends. Weekend CUS resulted in a more significant decline in the risk of depression among those with less than 6 hours of weekday sleep and over 2 hours of CUS on weekends. No statistically significant variances were detected in the association between weekend CUS and the risk of depression in other subgroups.

Table 3.

Adjusted multivariate logistic regression analyses assessing the relationship between weekend CUS and depression under subgroups of weekday sleep duration

Association between CUS duration and severity of depression (PHQ-9 score) among individuals with CUS≤0 h, 0 h<CUS≤1 h, 1 h<CUS≤2 h, and CUS>2 h

We conducted multivariate linear regression analyses to examine the association between the duration of chronic unpredictable stress and the severity of depression as measured by the PHQ-9 score, utilizing four models (Table 4). In Model 1, adjusted for age and sex, a significant negative linear correlation was observed between CUS duration and PHQ-9 scores among individuals with CUS durations of 0 to 1 hour and 1 to 2 hours. Conversely, no significant linear relationship was found between individuals with CUS durations ≤0 hours and those with CUS durations >2 hours. This pattern persisted in Model 3. However, in both Model 2 and Model 4, the groups categorized by CUS durations (CUS≤0 h, 0 h<CUS≤1 h, 1 h<CUS≤2 h, and CUS>2 h) did not display a significant linear association with PHQ-9 scores.

Table 4.

Association between CUS duration and severity of depression (PHQ-9 score) among individuals with CUS≤0 h, 0 h<CUS≤1 h, 1 h<CUS ≤2 h, and CUS>2 h

DISCUSSION

In this cross-sectional study, we found no evidence that weekend CUS increases depression risk compared to no weekend CUS. However, among participants averaging ≤6 hours of weekday sleep, weekend CUS was associated with reduced depression risk—particularly when exceeding 2 hours.

Prior research has established a strong link between sleep disorders and depression [9]. Conditions such as insomnia, restless legs syndrome, and obstructive sleep apnea elevate depression risk. A recent meta-analysis further indicates that both insufficient and excessive sleep duration increase depression susceptibility [15]. Our findings align with this evidence: participants consistently averaging <6 hours of sleep weekly showed higher depression rates than those maintaining 6–8 hours (Supplementary Table 1). Given epidemiological evidence that weekend CUS prevalence declines with age (particularly ≥60 years), we conducted age-stratified analyses. In the minimally adjusted model (Model 1: sex-adjusted), weekend CUS showed a nominally protective association with depression. However, this association attenuated to non-significance in fully adjusted models (Models 2–4), suggesting residual confounding rather than age-specific effects. The role of age in modifying CUS-depression relationships remains inconclusive, necessitating further investigation via prospective cohorts with longitudinal sleep monitoring (Supplementary Table 2).

The negative health impacts of inadequate sleep duration are well-established, with weekend CUS frequently employed as a compensatory strategy. Among Korean adolescents, one study identified CUS as significantly associated with both suicide attempts and self-injury risk, independent of depression [32]. However, this study did not adjust for sleep duration in its analyses, potentially overlooking that weekend CUS may primarily reflect chronic sleep deprivation. A separate Korean study found that >2 hours of weekend CUS significantly reduced depression risk compared to <0 hours of weekend CUS. This research further examined interactions among chronotype, sleep duration, CUS, sleep environment, sleep deprivation, and school-related factors in a multi-district sample of high school students [21]. Notably, both studies focused exclusively on adolescents, leaving the CUS-depression relationship in adults understudied. Adolescents display distinct sleep patterns from adults, including age-related reductions in total sleep time [33] and delayed circadian rhythm phases that progressively shift during maturation [34]. Consistent with these developmental differences, delayed sleep phase disorder is markedly more prevalent in adolescents than adults [35].

A Korean study was the first to investigate the association between weekend CUS and depression in adults [22]. The research, which analyzed a nationally representative sample of Korean adults, found that individuals with 1–2 hours of CUS had a significantly lower depression risk compared to those with no CUS. In contrast, no significant risk differences were observed among those with either no CUS, 0–1 hour of CUS, or >2 hours of CUS. Additionally, the study identified a linear trend between longer CUS (>2 hours) and higher PHQ-9 scores, while no such trend existed for no CUS, 0–1 hour, or 1–2 hour groups. These findings suggest a non-linear relationship between CUS duration and depression risk. The authors hypothesized that specifically 1–2 hours of weekend CUS might serve as a behavioral adaptation that could mitigate depression risk.

Key findings from this cross-sectional study suggest that weekend CUS is not associated with an increased risk of depression compared to non-practitioners. However, individuals who typically sleep fewer than 6 hours on weekdays may experience reduced depression risk through CUS, particularly when exceeding 2 hours. However, there are several limitations to this study. Firstly, the study did not exclude individuals with ≤-1 hour of CUS on weekends. The disrupted rhythm of individuals with greater CUS on weekends compared to weekday sleep duration may significantly impact the accuracy of the results. Secondly, the study did not conduct subgroup analyses based on average weekday sleep duration to determine potential differences in the risk of depression across different sleep duration groups. Our study excluded individuals with ≤-1 hour of CUS on weekends and also did not perform subgroup analyses on weekday sleep duration. While our findings did not reveal a significant association between CUS and depression, we observed that CUS on weekends was linked to a reduced risk of depression among those with less than 6 hours of average weekday sleep, with a more pronounced reduction in depression risk observed in individuals with over 2 hours of CUS. Multivariate linear regression analyses were conducted using four models to explore the association between the duration of CUS and the severity of depression, as evaluated by the PHQ-9 score (Table 4). In Model 1, upon adjusting for age and sex, a significant negative linear relationship between CUS duration and PHQ-9 scores was observed in individuals with 0 h<CUS≤1 h and 1 h<CUS≤2 h, while there was no significant linear relationship in those with CUS≤ 0 h and CUS>2 h. A similar pattern was seen in Model 3. However, in both Model 2 and Model 4, the four defined groups based on CUS duration did not exhibit a significant linear relationship with PHQ-9 scores.

There are several possible explanations for the non-linear relationship between CUS and depression. First, CUS may compensate for weekday sleep deprivation, though this effect likely varies across populations. For individuals with severe weekday sleep deprivation, weekend CUS may mitigate its negative impacts [36]. However, the threshold for this protective effect remains unclear and warrants confirmation in future cohort studies. Second, while CUS may offset sleep debt, prolonged exposure could lead to adverse outcomes, including circadian rhythm disruption and social jet lag. These conditions are associated with metabolic dysfunction [37], elevated BMI, atherosclerosis [38], and depression [39]. Other factors may also contribute, but limited data prevented a deeper analysis of circadian and social jet lag effects.

The relationship between CUS duration and depression may be influenced by chronotype—a behaviorally expressed circadian preference that regulates sleep-wake timing and daily activity patterns [40]. A meta-analysis of 36 cross-sectional studies found that later sleep timing (e.g., delayed bedtimes) was significantly associated with more severe depressive symptoms [41]. Additionally, longitudinal evidence supports an association between CUS duration and depressive outcomes [42]. Chronotype may further mediate this relationship, as individuals with later chronotypes are more prone to sleep deprivation and thus more likely to engage in weekend CUS. A large-scale, population-based study in South Korea [17] demonstrated that late chronotypes were overrepresented among individuals who practiced CUS compared to those who did not. Unfortunately, our study was unable to further investigate the impact of chronotype on the relationship between CUS and depression due to a lack of available data on chronotype.

There are several limitations to this study. First, the data were obtained from the NHANES database, which evaluated sleep duration using self-reported questionnaires (weekday/weekend) rather than objective measures like polysomnography or actigraphy. While large-scale objective sleep monitoring remains logistically challenging, most epidemiological studies rely on self-reports, which have demonstrated reasonable agreement with actigraphy-derived sleep duration [43]. Second, as a cross-sectional study, this analysis cannot establish causality between CUS and depression. Additionally, depression was assessed via the PHQ-9 questionnaire instead of clinical interviews aligned with DSM-5 criteria, potentially introducing misclassification bias. However, the PHQ-9 is a validated screening tool with high sensitivity and moderate specificity for depression in population-based studies [44]. Finally, key confounders—such as chronotype, circadian disruption, and social jet lag—could not be adjusted for due to data limitations, restricting further exploration of their role in the CUS-depression relationship.

This study has several notable strengths. First, the data were sourced from NHANES, a nationally representative, cross-sectional database that combines interview-based questionnaires with physical examinations, enhancing its reliability and accessibility.

Second, unlike previous studies, we excluded individuals with CUS durations of ≤-1 hour and conducted subgroup analyses for weekday sleep duration, enabling a more precise assessment of CUS’s effects. Although this study did not establish a direct causal link between CUS and depression risk, a key finding emerged: among participants with <6 hours of weekday sleep, CUS was associated with a significantly reduced depression risk. Given inconsistencies in existing literature, future research should prioritize high-quality cohort studies or randomized controlled trials to validate causality.

Supplementary Materials

Article information Continued