Advertisement

The individual and joint associations of depression and sleep duration with cardiometabolic diseases and mortality: A prospective cohort study

  • Xunjie Cheng
    Correspondence
    Corresponding author. Department of Geriatric Medicine, Center of Coronary Circulation, Xiangya Hospital, Central South University, Xiangya Road 87#, Changsha, 410008, China.
    Affiliations
    Department of Geriatric Medicine, Center of Coronary Circulation, Xiangya Hospital, Central South University, Changsha, China

    National Clinical Research Center for Geriatric Disorders, Xiangya Hospital, Central South University, Changsha, China
    Search for articles by this author
  • Feiyun Ouyang
    Affiliations
    Department of Social Medicine and Health Management, Xiangya School of Public Health, Central South University, Changsha, China
    Search for articles by this author
  • Tianqi Ma
    Affiliations
    Department of Geriatric Medicine, Center of Coronary Circulation, Xiangya Hospital, Central South University, Changsha, China
    Search for articles by this author
  • Lingfang He
    Affiliations
    Department of Geriatric Medicine, Center of Coronary Circulation, Xiangya Hospital, Central South University, Changsha, China
    Search for articles by this author
  • Li Gong
    Affiliations
    National Clinical Research Center for Geriatric Disorders, Xiangya Hospital, Central South University, Changsha, China

    Department of Cardiovascular Surgery, Xiangya Hospital, Central South University, Changsha, China
    Search for articles by this author
  • Jinghua Yin
    Affiliations
    Department of Geriatric Medicine, Center of Coronary Circulation, Xiangya Hospital, Central South University, Changsha, China
    Search for articles by this author
  • Guogang Zhang
    Affiliations
    Department of Cardiovascular Medicine, Xiangya Hospital, Central South University, Changsha, China

    Department of Cardiovascular Medicine, The Third Xiangya Hospital, Central South University, Changsha, China
    Search for articles by this author
  • Yongping Bai
    Correspondence
    Corresponding author. Department of Geriatric Medicine, Center of Coronary Circulation, Xiangya Hospital, Central South University, Xiangya Road 87#, Changsha, 410008, China.
    Affiliations
    Department of Geriatric Medicine, Center of Coronary Circulation, Xiangya Hospital, Central South University, Changsha, China

    National Clinical Research Center for Geriatric Disorders, Xiangya Hospital, Central South University, Changsha, China
    Search for articles by this author

      Highlights

      • Depression and sleep duration were mutually adjusted in only a few studies.
      • Depression and short or long sleep duration were independently associated with increased risk of cardiovascular health.
      • Additive interactions were found between depression and short sleep duration.

      Abstract

      Background and aims

      Depression and sleep duration were only mutually adjusted in a few studies, and it is unknown whether these two factors are independent or overlapping risk factors for cardiometabolic diseases (CMDs) and mortality. This study aimed to evaluate the individual and joint associations of depression and sleep duration with CMDs and mortality.

      Methods

      A total of 261,297 participants who were free of CMD at baseline were included. Sleep duration was divided into three groups (short: <7 h/day, referent: ages 39–64 years: 7–9 h/day; ages 65+ years: 7–8 h/day, and long: ages 39–64 years: >9 h/day; ages 65+ years: >8 h/day). The main outcomes were hypertension, stroke, CHD, DM, all-cause mortality, and cardiovascular mortality.

      Results

      Among the 261,297 participants, depression and short or long sleep duration were independently associated with increased risk of CMDs and mortality (hazard ratio [HR], 1.10–1.38) when they were mutually adjusted, except for the association between short sleep duration and stroke (HR, 1.03; 95% confidence interval [CI], 0.97–1.10). We documented significant additive interactions between depression and short sleep duration in relation to all-cause mortality (relative excess risk due to interaction [RERI], 0.19; 95% CI, 0.02–0.37) and CHD (RERI, 0.30; 95% CI, 0.11–0.48).

      Conclusions

      In this study, depression and short or long sleep duration were independently associated with an increased risk of CMDs and mortality. We also observed significant additive interactions between depression and short sleep duration in relation to all-cause mortality and CHD.

      Graphical abstract

      Keywords

      1. Introduction

      Cardiometabolic diseases (CMDs), including hypertension, stroke, coronary heart disease (CHD), and diabetes mellitus (DM), are leading causes of death worldwide. In 2019, approximately 20 (35%) million deaths were caused by CMDs globally [
      • Roth G.A.
      • Mensah G.A.
      • Johnson C.O.
      • et al.
      Global burden of cardiovascular diseases and risk factors, 1990-2019: update from the GBD 2019 study [published correction appears in J Am coll cardiol. 2021 apr 20;77(15):1958-1959].
      ,
      Global Burden of Disease Collaborative Network
      Global Burden of Disease Study 2019 (GBD 2019) Results.
      ]. With the increasing epidemic of depression and sleep disorders [
      • Sheehan C.M.
      • Frochen S.E.
      • Walsemann K.M.
      • Ailshire J.A.
      Are U.S. adults reporting less sleep?: findings from sleep duration trends in the National Health Interview Survey.
      ,
      • Arias-de la Torre J.
      • Vilagut G.
      • Ronaldson A.
      • et al.
      Prevalence and variability of current depressive disorder in 27 European countries: a population-based study.
      ,
      • Daly M.
      Prevalence of depression among adolescents in the U.S. From 2009 to 2019: analysis of trends by sex, race/ethnicity, and income.
      ,
      • Otsuka Y.
      • Kaneita Y.
      • Spira A.P.
      • et al.
      Trends in sleep problems and patterns among Japanese adolescents: 2004 to 2017.
      ], many previous studies have explored whether depression and short or long sleep duration are associated with increased risk of CMDs and mortality [
      • Ai S.
      • Zhang J.
      • Zhao G.
      • et al.
      Causal associations of short and long sleep durations with 12 cardiovascular diseases: linear and nonlinear Mendelian randomization analyses in UK Biobank.
      ,
      • Krittanawong C.
      • Tunhasiriwet A.
      • Wang Z.
      • et al.
      Association between short and long sleep durations and cardiovascular outcomes: a systematic review and meta-analysis.
      ,
      • Harshfield E.L.
      • Pennells L.
      • Schwartz J.E.
      • et al.
      Association between depressive symptoms and incident cardiovascular diseases.
      ,
      • Miloyan B.
      • Fried E.
      A reassessment of the relationship between depression and all-cause mortality in 3,604,005 participants from 293 studies.
      ]. However, the results varied widely in prior studies. For example, some studies found that short sleep duration was associated with an increased risk of cardiovascular events and all-cause mortality [
      • Ai S.
      • Zhang J.
      • Zhao G.
      • et al.
      Causal associations of short and long sleep durations with 12 cardiovascular diseases: linear and nonlinear Mendelian randomization analyses in UK Biobank.
      ,
      • Itani O.
      • Jike M.
      • Watanabe N.
      • Kaneita Y.
      Short sleep duration and health outcomes: a systematic review, meta-analysis, and meta-regression.
      ], but results from other studies were not statistically significant [
      • García-Perdomo H.A.
      • Zapata-Copete J.
      • Rojas-Cerón C.A.
      Sleep duration and risk of all-cause mortality: a systematic review and meta-analysis.
      ,
      • Zhou L.
      • Yu K.
      • Yang L.
      • et al.
      Sleep duration, midday napping, and sleep quality and incident stroke: the Dongfeng-Tongji cohort.
      ].
      The heterogeneity in previous studies could be explained by several possible reasons, such as differences in sample size, follow-up period, adjustment of covariables, sex and regional differences in population, and variation in the definition of depression and sleep duration. To overcome these limitations, some studies pooled individual data from cohort studies [
      • Harshfield E.L.
      • Pennells L.
      • Schwartz J.E.
      • et al.
      Association between depressive symptoms and incident cardiovascular diseases.
      ,
      • Wang C.
      • Bangdiwala S.I.
      • Rangarajan S.
      • et al.
      Association of estimated sleep duration and naps with mortality and cardiovascular events: a study of 116 632 people from 21 countries.
      ,
      • Svensson T.
      • Saito E.
      • Svensson A.K.
      • et al.
      Association of sleep duration with all- and major-cause mortality among adults in Japan, China, Singapore, and korea.
      ], and others provided systematic reviews and meta-analyses of previous studies [
      • Krittanawong C.
      • Tunhasiriwet A.
      • Wang Z.
      • et al.
      Association between short and long sleep durations and cardiovascular outcomes: a systematic review and meta-analysis.
      ,
      • Miloyan B.
      • Fried E.
      A reassessment of the relationship between depression and all-cause mortality in 3,604,005 participants from 293 studies.
      ,
      • Yin J.
      • Jin X.
      • Shan Z.
      • et al.
      Relationship of sleep duration with all-cause mortality and cardiovascular events: a systematic review and dose-response meta-analysis of prospective cohort studies.
      ]. However, one crucial limitation remains unresolved: depression and sleep duration were mutually adjusted in only a few previous studies [
      • Mezick E.J.
      • Hall M.
      • Matthews K.A.
      Are sleep and depression independent or overlapping risk factors for cardiometabolic disease?.
      ]. In fact, a study indicated that approximately 10% of patients with short or long sleep duration also presented with depression [
      • Wang C.
      • Bangdiwala S.I.
      • Rangarajan S.
      • et al.
      Association of estimated sleep duration and naps with mortality and cardiovascular events: a study of 116 632 people from 21 countries.
      ], and the bidirectional association between the two factors has been well studied [
      • Riemann D.
      • Krone L.B.
      • Wulff K.
      • Nissen C.
      Sleep, insomnia, and depression.
      ,
      • Plante D.T.
      The evolving nexus of sleep and depression.
      ]. Moreover, some studies have indicated that the associations of short or long sleep duration with CMDs and mortality were not statistically significant when further adjusted for depression and other confounding variables [
      • Cappuccio F.P.
      • Stranges S.
      • Kandala N.B.
      • et al.
      Gender-specific associations of short sleep duration with prevalent and incident hypertension: the Whitehall II Study.
      ,
      • Lan T.Y.
      • Lan T.H.
      • Wen C.P.
      • Lin Y.H.
      • Chuang Y.L.
      Nighttime sleep, Chinese afternoon nap, and mortality in the elderly.
      ].
      A clear understanding of the effects of depression and sleep duration is critical for identifying high-risk groups for CMDs and developing prevention and treatment strategies. Therefore, to clarify whether depression and sleep duration are independent or overlapping risk factors for CMDs and mortality, this study evaluated the individual and joint associations of depression and sleep duration with CMDs and mortality based on participants from the UK Biobank (UKB).

      2. Patients and methods

      2.1 Study population

      All the data used in this study were derived from the UKB (Approved ID: 76118). As detailed elsewhere [
      • Sudlow C.
      • Gallacher J.
      • Allen N.
      • et al.
      UK biobank: an open access resource for identifying the causes of a wide range of complex diseases of middle and old age.
      ], the UKB enrolled approximately 0.5 million middle-aged and elderly people from 22 centers across the United Kingdom between 2006 and 2010. A variety of information, including clinical, socioeconomic, behavioral, and biological samples, was collected at baseline. Health outcomes were also recorded through linkage to routinely available national datasets during follow-up. Written informed consent was obtained from each participant, and the UKB was approved by the West Multi-centre Research Ethics Committee.
      A total of 502,414 participants were included in this study. We excluded participants with CMD at baseline (166,702) and participants with missing information on sleep duration (2,392) and other variables (74,109, mainly physical activity [60,927]). Participants who died within 1 year of enrolment (306) were also excluded from the main analyses. Ultimately, 261,297 participants were included in this study. Detailed information is provided in Supplementary Fig. 1. Multiple information sources were used to define CMDs in this study, including self-reported history and relevant age of disease onset, medication, operation and electronic health records derived from health institutions [
      • Said M.A.
      • Verweij N.
      • van der Harst P.
      Associations of combined genetic and lifestyle risks with incident cardiovascular disease and diabetes in the UK biobank study.
      ,
      • Cheng X.
      • Ouyang F.
      • Ma T.
      • et al.
      Association of healthy lifestyle and life expectancy in patients with cardiometabolic multimorbidity: a prospective cohort study of UK biobank.
      ]. Detailed information is provided in Supplementary Table 1. A participant was considered as having CMD at baseline if the available earliest date of CMD diagnosis was prior to the recruitment date.

      2.2 Ascertainment of main variables

      We defined participants with depression according to three sources of data: self-reported history of depression, antidepressant use, and depression-related hospitalization records (International Classification of Diseases, Tenth Revision [ICD-10]: F32 and F33) [
      • Glanville K.P.
      • Coleman J.R.I.
      • Howard D.M.
      • et al.
      Multiple measures of depression to enhance validity of major depressive disorder in the UK Biobank.
      ]. Detailed information is provided in Supplementary Table 1 and Supplementary Text 1. Self-reported information about sleep duration per 24 h was obtained at baseline. Participants were classified into three groups according to sleep duration in this study: short sleep duration (<7 h/day), referent sleep duration (7–8 h/day for people aged 65+ years and 7–9 h/day for people aged 39–64 years), and long sleep duration (>8 h/day for people aged 65+ years and >9 h/day for people aged 39–64 years) [
      • Hirshkowitz M.
      • Whiton K.
      • Albert S.M.
      • et al.
      National Sleep Foundation's updated sleep duration recommendations: final report.
      ,
      • Duncan M.J.
      • Oftedal S.
      • Kline C.E.
      • Plotnikoff R.C.
      • Holliday E.G.
      Associations between aerobic and muscle-strengthening physical activity, sleep duration, and risk of all-cause mortality: a prospective cohort study of 282,473 U.S. adults [published online ahead of print, 2022 Jul 21].
      ].
      Other confounding variables included in this study were age, sex, race, Townsend deprivation index, family history of CMDs, other mental disorders, body mass index (BMI), smoking status, physical activity, sedentary time, alcohol consumption, and diet. Detailed information is provided in Supplementary Table 2.

      2.3 Ascertainment of outcomes

      Six health outcomes, including all-cause mortality, cardiovascular mortality, CHD, stroke, DM, and hypertension, were analyzed in this study. We used the same methods mentioned above to ascertain the occurrence of the four CMDs during follow-up (Supplementary Table 1). A participant was considered as getting CMD during follow-up if the available earliest date of CMD diagnosis was in the aftermath of recruitment date. Data on deaths were obtained from the National Health Service Information Centre and the National Health Service Central Registers and coded using ICD-10 (Data-Field: 41270). Cardiovascular mortality was defined as death recorded using ICD-10 codes I00–I99. The follow-up of the study started after enrolment and ended on December 31, 2019. This date was selected to minimize the effect of the coronavirus disease 2019 outbreak.

      2.4 Statistical analyses

      The baseline characteristics of the participants were described according to both depression symptoms and sleep duration. Continuous variables (age, Townsend deprivation index, BMI, and sedentary time) are presented as medians (interquartile ranges), and categorical variables (sex, race/ethnicity, smoking, alcohol consumption, physical activity, diet, family history of CMD, and other mental problems) are presented as numbers (percentages). We used two steps to evaluate the associations of depression and sleep duration with CMDs and mortality. First, we evaluated the associations of depression and sleep duration with CMDs and mortality, without adjusting for each other. Cox proportional hazard models were used to estimate hazard ratios (HRs) and 95% confidence intervals (CIs). Schoenfeld residuals were used to test the proportional hazards assumption, and no violations were found [
      • Grambsch P.
      • Therneau T.
      Proportional hazards tests and diagnostics based on weighted residuals.
      ]. Three settings of models were analyzed for both depression and sleep duration. Model 1 was adjusted for age and sex. We additionally adjusted for family history of CMDs, other mental disorders, race, and Townsend deprivation index in Model 2. Model 3 was further adjusted for physical activity, sedentary time, smoking status, alcohol consumption, diet, and BMI. To estimate the associations between hourly sleep duration categories and the outcomes, we repeated these analyses by regrouping participants into 6 groups according to sleep duration: ≤5 h (combined as small sample size), 6 h, 7 h (referent group), 8 h, 9 h, and ≥10 h (combined as small sample size) per day. Second, depression and sleep duration were mutually adjusted in the models with the three settings. We also used two steps to assess the joint associations of depression and sleep duration with CMDs and mortality. First, we evaluated both the multiplicative and additive interactions between depression and sleep duration by adding cross-product terms between depression and sleep duration in the fully adjusted model. The likelihood ratio test was used to assess the presence of multiplicative interaction between models with and without interaction terms. We calculated the additive interactions between depression and short and long sleep duration. Three indicators, including relative excess risk due to interaction, proportion of disease attributable to interaction, and synergy index, were used to measure additive interaction. A previous method established by Li et al. was used to calculate the three indicators based on the Cox proportional hazards model [
      • Li R.
      • Chambless L.
      Test for additive interaction in proportional hazards models.
      ]. Second, for outcomes in which interaction existed, we further assessed the joint association between the two factors. We classified the participants into six mutually exclusive groups according to the joint categories of depression and sleep duration (Supplementary Table 3). Cox proportional hazard models were used to estimate HRs and 95% CIs, with participants without depression and with referent sleep duration as the reference group. Schoenfeld residuals were used to test the proportional hazards assumption, and no violations were found. The confounding variables in Model 3 were adjusted in the models.
      We performed five sensitivity analyses to assess the robustness of the results. First, we only used ICD-10 records to define depression. Second, we excluded participants diagnosed with CMD during the first year of follow-up to minimize reverse causality of depression and sleep duration with CMDs. Third, multiple imputation was performed using multivariate imputation by chained equations to deal with missing values in covariables listed above and 332,882 participants were finally included. Fourth, we excluded participants with extremely short (<3 h/day) or long (>12 h/day) sleep duration. Finally, we only used ICD-9 and ICD-10 records to define CMDs events during follow-up. All analyses were performed using R 4.1.1, and two-tailed p < 0.05 was considered statistically significant. p values were adjusted by Benjamini-Hochberg method to account for multiple comparisons.

      3. Results

      A total of 261,297 participants were included in this study. Table 1 shows the baseline characteristics of the participants. Compared to participants without a history of depression, those with depression were more likely to be female, have other mental disorders, and drink heavily. Compared to participants with referent sleep duration, those with short sleep duration were more likely to have high Townsend deprivation index and drink heavily, and those with long sleep duration were more likely to be male, have other mental disorders, and drink heavily. No appreciable differences in BMI and diet quality were apparent across the groups.
      Table 1Baseline characteristics of participants.
      CharacteristicsTotalDepression symptomsSleep duration (hours/day)
      Non-depressionDepressionReferent
      Referent sleep duration: 7–8 h/day for people aged 65+ years and 7–9 h/day for people aged 39–64 years.
      ShortLong
      No. (%) of participants261,297235,94825,349195,56759,5416189
      Age, median (IQR
      Interquartile range.
      ), year
      55.8 (48.5, 62.1)55.9 (48.5, 62.2)55.3 (48.4, 61.4)55.6 (48.3, 62)55.7 (49.0, 61.7)65.5 (56.4, 67.6)
      Male, No. (%)113,530 (43.4)106,116 (45.0)7414 (29.2)111,713 (57.1)26,995 (45.3)3508 (56.7)
      White, No. (%)249,715 (95.6)225,114 (95.4)24,601 (97.0)188,283 (96.3)55,519 (93.2)5913 (95.5)
      Townsend Deprivation Index, median (IQR)−2.32 (−3.74, 0.12)−2.35 (−3.76, 0.03)−1.99 (−3.58, 0.79)−2.4 (−3.79, −0.09)−2.03 (−3.58, 0.68)−2.05 (−3.55, 0.77)
      BMI
      Body mass index.
      , median (IQR)
      25.89 (23.53, 28.68)25.85 (23.52, 28.6)26.33 (23.74, 29.47)25.75 (23.45, 28.47)26.32 (23.81, 29.3)26.59 (24.11, 29.35)
      Sedentary time, median (IQR), hour4 (3, 6)4 (3, 6)4.5 (3, 6)4 (3, 5.5)4.5 (3, 6)5 (3.5, 6)
      Never/former smoking, No. (%)234,831 (89.9)213,312 (90.4)21,519 (84.9)177,194 (90.6)52,249 (87.8)5388 (87.1)
      Non/moderate alcohol consumption, No. (%)120,193 (46.0)110,265 (46.7)9928 (39.2)92,092 (47.1)25,598 (43.0)2503 (40.4)
      Healthy diet, No. (%)147,245 (56.4)132907 (56.3)14,338 (56.6)110,704 (56.6)33,107 (55.6)3434 (55.5)
      Regular physical activity, No. (%)216,006 (82.7)196,117 (83.1)19,889 (78.5)162,547 (83.1)48,537 (81.5)4922 (79.5)
      Family history of CMD
      Cardiometabolic diseases, including coronary heart disease, hypertension, diabetes mellitus, and stroke.
      , No. (%)
      197,794 (75.7)178,180 (75.5)19,614 (77.4)147,723 (75.5)45,516 (76.4)4555 (73.6)
      Other mental problems, No. (%)9211 (3.5)4802 (2.0)4409 (17.4)6162 (3.2)2443 (4.1)606 (9.8)
      a Interquartile range.
      b Body mass index.
      c Cardiometabolic diseases, including coronary heart disease, hypertension, diabetes mellitus, and stroke.
      d Referent sleep duration: 7–8 h/day for people aged 65+ years and 7–9 h/day for people aged 39–64 years.
      During a median follow-up of 10.9 years, we documented 9,481, 1,304, 10,814, 5,227, 5,786, and 29,115 incident cases of all-cause mortality, cardiovascular mortality, CHD, stroke, DM, and hypertension, respectively. When depression and sleep duration were not simultaneously adjusted in the models, we observed positive associations of depression and short or long sleep duration with CMDs and mortality (see Fig. 1), except for the association of short sleep duration with cardiovascular mortality and stroke (HR, 1.14; 95% CI, 1.00–1.30 and HR, 1.03; 95% CI, 0.97–1.10) (Fig. 1D). The associations were attenuated but remained significant for all outcomes when depression and sleep duration were included in the models simultaneously, except for the long sleep duration with CHD. Depression, short sleep duration, and long sleep duration were associated with higher risk of DM (HR, 1.36; 95% CI, 1.26–1.47), DM (HR, 1.18; 95% CI, 1.11–1.25), and all-cause mortality (HR, 1.38; 95% CI, 1.27–1.51), respectively (Table 2). Similar results were observed when participants were grouped into 6 groups according to sleep duration. Detailed information is provided in Supplementary Tables 4–6.
      Fig. 1
      Fig. 1Individual associations of depression and sleep duration with cardiometabolic diseases and mortality.
      Three different kinds of model were used in the analyses: a) only sleep duration was included; b) only depression was included; and c) both sleep duration and depression were included. Non-depression and referent sleep duration (Referent sleep duration: 7–8 h/day for aged 65+ years and 7–9 h/day for aged 39–64 years) were used as reference groups, respectively. All models were adjusted for age, sex, race, Townsend deprivation index, family history of cardiometabolic disease, other mental disorders, body mass index, sedentary time, physical activity, smoking status, alcohol consumption, and diet. CI: confidence interval.
      Table 2Associations of depression and sleep duration with cardiometabolic diseases and mortality with adjusting for each other.
      Model 1
      Model 1 was adjusted for age and sex.
      Model 2
      Model 2 was additionally adjusted for race, Townsend deprivation index, family history of CMD, and other mental disorders.
      Model 3
      Model 3 was further adjusted for BMI, sedentary time, physical activity, smoking status, alcohol consumption, and diet.
      HR (95% CI)p
      p values were adjusted by Benjamini-Hochberg method.
      HR (95% CI)p
      p values were adjusted by Benjamini-Hochberg method.
      HR (95% CI)p
      p values were adjusted by Benjamini-Hochberg method.
      All-cause mortality
      Short sleep duration1.16 (1.11, 1.22)<0.0011.14 (1.08, 1.19)<0.0011.10 (1.05, 1.15)<0.001
      Long sleep duration1.53 (1.40, 1.67)<0.0011.44 (1.32, 1.58)<0.0011.38 (1.27, 1.51)<0.001
      Depression1.39 (1.30, 1.48)<0.0011.16 (1.08, 1.24)<0.0011.12 (1.05, 1.20)0.002
      Cardiovascular mortality
      Short sleep duration1.24 (1.09, 1.41)0.0011.19 (1.05, 1.36)0.0091.14 (1.00, 1.29)0.061
      Long sleep duration1.52 (1.20, 1.91)0.0011.42 (1.12, 1.79)0.0051.32 (1.05, 1.67)0.026
      Depression1.72 (1.46, 2.02)<0.0011.39 (1.17, 1.66)<0.0011.33 (1.12, 1.58)0.002
      Coronary heart disease
      Short sleep duration1.19 (1.14, 1.25)<0.0011.17 (1.12, 1.22)<0.0011.12 (1.07, 1.17)<0.001
      Long sleep duration1.19 (1.08, 1.31)0.0011.17 (1.06, 1.28)0.0031.10 (1.00, 1.21)0.059
      Depression1.43 (1.35, 1.52)<0.0011.35 (1.27, 1.43)<0.0011.28 (1.21, 1.37)<0.001
      Stroke
      Short sleep duration1.08 (1.01, 1.15)0.0331.05 (0.99, 1.13)0.1281.03 (0.97, 1.10)0.410
      Long sleep duration1.30 (1.14, 1.47)<0.0011.26 (1.11, 1.43)0.0011.22 (1.07, 1.38)0.004
      Depression1.39 (1.27, 1.51)<0.0011.27 (1.17, 1.39)<0.0011.25 (1.14, 1.36)<0.001
      Diabetes mellitus
      Short sleep duration1.45 (1.37, 1.53)<0.0011.34 (1.27, 1.42)<0.0011.18 (1.11, 1.25)<0.001
      Long sleep duration1.59 (1.40, 1.81)<0.0011.46 (1.28, 1.66)<0.0011.21 (1.07, 1.38)0.005
      Depression1.70 (1.58, 1.84)<0.0011.61 (1.49, 1.74)<0.0011.36 (1.26, 1.47)<0.001
      Hypertension
      Short sleep duration1.22 (1.19, 1.25)<0.0011.19 (1.16, 1.22)<0.0011.13 (1.10, 1.16)<0.001
      Long sleep duration1.23 (1.16, 1.30)<0.0011.20 (1.13, 1.28)<0.0011.12 (1.06, 1.19)<0.001
      Depression1.42 (1.37, 1.47)<0.0011.37 (1.32, 1.42)<0.0011.29 (1.24, 1.33)<0.001
      Participants with non-depression and referent sleep duration were considered as reference groups, respectively. Referent sleep duration: 7–8 h/day for aged 65+ years and 7–9 h/day for aged 40–65 years.
      a Model 1 was adjusted for age and sex.
      b Model 2 was additionally adjusted for race, Townsend deprivation index, family history of CMD, and other mental disorders.
      c Model 3 was further adjusted for BMI, sedentary time, physical activity, smoking status, alcohol consumption, and diet.
      d p values were adjusted by Benjamini-Hochberg method.
      In the assessment of the joint associations, we documented significant additive interactions between depression and short sleep duration in relation to all-cause mortality and CHD (Table 3). Regarding all-cause mortality, the multivariable-adjusted HRs were 1.07 (95% CI, 0.98–1.16) for depression, 1.08 (95% CI, 1.03–1.14) for short sleep duration, and 1.34 (95% CI, 1.20–1.50) for their joint effect (Fig. 2A ), with a relative excess risk due to interaction of 0.19 (95% CI, 0.02–0.37). Additive interaction accounted for 56% (0.19/0.34) of the death events. Regarding CHD, the multivariable-adjusted HRs were 1.21 (95% CI, 1.12–1.31) for depression, 1.10 (95% CI, 1.05–1.15) for short sleep duration, and 1.61 (95% CI, 1.45–1.78) for their joint effect (Fig. 2B), with a relative excess risk due to interaction of 0.30 (95% CI, 0.11–0.48). Additive interaction accounted for 49% (0.30/0.61) of the CHD events (Table 3).
      Table 3Additive and multiplicative interactions between depression and sleep duration.
      OutcomesAdditive interaction (95% CI
      CI: confidence interval.
      )
      Multiplicative interaction
      Depression and short sleep durationp for interaction
      p values were adjusted by Benjamini-Hochberg method.
      RERI
      Relative excess risk due to interaction.
      p
      p values were adjusted by Benjamini-Hochberg method.
      AP
      Proportion of disease attributable to interaction.
      p
      p values were adjusted by Benjamini-Hochberg method.
      S
      Synergy index.
      p
      p values were adjusted by Benjamini-Hochberg method.
      All-cause mortality0.19 (0.02, 0.37)0.0390.14 (0.03, 0.26)0.0232.30 (1.02, 5.16)0.0560.136
      Cardiovascular mortality0.19 (−0.34, 0.72)0.5100.12 (−0.18, 0.42)0.4731.45 (0.53, 3.94)0.5010.707
      Coronary heart disease0.30 (0.11, 0.48)0.0040.18 (0.08, 0.29)0.0011.95 (1.27, 2.99)0.0030.028
      Stroke0.11 (−0.14, 0.36)0.4280.08 (−0.09, 0.26)0.4061.46 (0.62, 3.46)0.4200.719
      Diabetes mellitus−0.02 (−0.25, 0.22)0.896−0.01 (−0.16, 0.14)0.8920.97 (0.64, 1.48)0.8940.804
      Hypertension0.03 (−0.08, 0.14)0.6230.02 (−0.05, 0.09)0.6221.07 (0.84, 1.36)0.6210.181
      OutcomesDepression and long sleep durationp for interaction
      p values were adjusted by Benjamini-Hochberg method.
      RERIp
      p values were adjusted by Benjamini-Hochberg method.
      APp
      p values were adjusted by Benjamini-Hochberg method.
      Sp
      p values were adjusted by Benjamini-Hochberg method.
      All-cause mortality0.08 (−0.23, 0.39)0.6220.05 (−0.14, 0.25)0.6151.19 (0.64, 2.20)0.6160.136
      Cardiovascular mortality−0.15 (−1.00, 0.69)0.740−0.10 (−0.70, 0.49)0.7570.77 (0.17, 3.52)0.7550.707
      Coronary heart disease0.06 (−0.25, 0.37)0.7220.04 (−0.17, 0.26)0.7161.19 (0.51, 2.78)0.7170.028
      Stroke0.14 (−0.30, 0.59)0.5650.09 (−0.17, 0.36)0.5361.34 (0.56, 3.18)0.5400.719
      Diabetes mellitus0.07 (−0.36, 0.49)0.7690.04 (−0.21, 0.29)0.7661.11 (0.58, 2.14)0.7660.804
      Hypertension−0.15 (−0.34, 0.04)0.134−0.11 (−0.27, 0.04)0.1680.67 (0.39, 1.17)0.1870.181
      Referent sleep duration: 7–8 h/day for aged 65+ years and 7–9 h/day for aged 40–65 years.
      a Relative excess risk due to interaction.
      b Proportion of disease attributable to interaction.
      c Synergy index.
      d p values were adjusted by Benjamini-Hochberg method.
      e CI: confidence interval.
      Fig. 2
      Fig. 2Joint associations of depression and sleep duration with all-cause mortality and coronary heart disease.
      Participants were classified into six mutually exclusive groups according to the joint categories of depression (Yes/No) and sleep duration (Referent sleep duration: 7–8 h/day for aged 65+ years and 7–9 h/day for aged 39–64 years). All models were adjusted for age, sex, race, Townsend deprivation index, family history of cardiometabolic disease, other mental disorders, body mass index, sedentary time, physical activity, smoking status, alcohol consumption, and diet. Participants with non-depression and referent sleep duration was considered as reference group. CI: confidence interval.
      In the sensitivity analyses, when we excluded participants diagnosed with CMD during the first year of follow-up and extremely short or long sleep duration, or when we used only ICD-9 and ICD-10 records to define CMDs events, the results were consistent with those of the main analyses. When we used only ICD-10 records to define depression, there were no statistically significant associations of depression with all-cause and cardiovascular mortality after adjustment of sleep duration. We documented marginal additive interaction between depression and short sleep duration in relation to all-cause mortality and CHD. When we used multivariate imputations to address missing covariables, short sleep duration was associated with increased risk of cardiovascular mortality and long sleep duration was associated with increased risk of CHD. Other results were not materially altered. Detailed information is provided in Supplementary Tables 8–14.

      4. Discussion

      Based on a large prospective cohort study of the UKB, we investigated the individual and joint associations of depression and sleep duration with CMDs and mortality over a median follow-up of 10.9 years. Overall, depression and short or long sleep duration were independently associated with increased risk of CMDs and mortality, except for the association of short sleep duration with cardiovascular mortality and stroke, and that for long sleep duration and CHD. In addition, we documented significant additive interactions between depression and short sleep duration in relation to all-cause mortality and CHD, and the joint effects could be decomposed to 56% and 49% for the additive interactions, respectively.

      4.1 Comparison with other studies

      This study indicated that both depression and short or long sleep duration were independent rather than overlapping risk factors for CMDs and mortality in the UKB. A few previous publications further adjusted for depression (or sleep duration) and other confounding variables when they evaluated the association of sleep duration (or depression) with specific cause of death and cardiovascular event and reported that the associations were reduced or no longer statistically significant after further adjustment [
      • Gangwisch J.E.
      • Malaspina D.
      • Posner K.
      • et al.
      Insomnia and sleep duration as mediators of the relationship between depression and hypertension incidence.
      ,
      • Azevedo Da Silva M.
      • Singh-Manoux A.
      • Shipley M.J.
      • et al.
      Sleep duration and sleep disturbances partly explain the association between depressive symptoms and cardiovascular mortality: the Whitehall II cohort study.
      ]. For example, the HRs of depression for cardiovascular mortality reduced from 1.63 (95% CI, 1.01–2.64) to 1.53 (95% CI, 0.91–2.57) after adjustment for sleep duration and disturbances in a previous study [
      • Azevedo Da Silva M.
      • Singh-Manoux A.
      • Shipley M.J.
      • et al.
      Sleep duration and sleep disturbances partly explain the association between depressive symptoms and cardiovascular mortality: the Whitehall II cohort study.
      ]. However, it is unclear whether the reduced HRs in previous publications were due to depression or sleep duration because more than one confounding variable was adjusted simultaneously with depression or sleep duration. Our analyses were distinct in two primary ways: (a) We provided a comprehensive estimation of the individual associations of depression and sleep duration with four types of CMDs and two causes of death in the same population, and (b) we specifically evaluated the effects of depression (or sleep duration) on the association of sleep duration (or depression) with CMDs and mortality by comparing models with and without adjustment for each other. Consistent with previous findings, the positive association of depression and sleep duration with CMDs and mortality was attenuated when the two factors were analyzed simultaneously in the models in our analysis. The current study, as well as previous studies, demonstrated that sleep duration may operate as a confounder or mediator between depression and CMDs/mortality and vice versa [
      • Mezick E.J.
      • Hall M.
      • Matthews K.A.
      Are sleep and depression independent or overlapping risk factors for cardiometabolic disease?.
      ]. However, most previous studies on sleep duration and depression have been conducted without adjusting for each other [
      • Harshfield E.L.
      • Pennells L.
      • Schwartz J.E.
      • et al.
      Association between depressive symptoms and incident cardiovascular diseases.
      ,
      • Miloyan B.
      • Fried E.
      A reassessment of the relationship between depression and all-cause mortality in 3,604,005 participants from 293 studies.
      ,
      • Zhou L.
      • Yu K.
      • Yang L.
      • et al.
      Sleep duration, midday napping, and sleep quality and incident stroke: the Dongfeng-Tongji cohort.
      ,
      • Svensson T.
      • Saito E.
      • Svensson A.K.
      • et al.
      Association of sleep duration with all- and major-cause mortality among adults in Japan, China, Singapore, and korea.
      ]. This study indicated that care is required when interpreting the results from these studies, as they may misestimate the effects of depression and sleep duration. In contrast to prior studies, the positive associations of depression and sleep duration with CMDs and mortality remained statistically significant after adjusting for each other in our study. A possible reason for this discrepancy is the population heterogeneity. For example, some studies indicated that short sleep duration increased the risk of all-cause mortality in elderly Europeans [
      • Ferrie J.E.
      • Shipley M.J.
      • Cappuccio F.P.
      • et al.
      A prospective study of change in sleep duration: associations with mortality in the Whitehall II cohort.
      ,
      • Bellavia A.
      • Åkerstedt T.
      • Bottai M.
      • Wolk A.
      • Orsini N.
      Sleep duration and survival percentiles across categories of physical activity.
      ], but not in Asians [
      • Svensson T.
      • Saito E.
      • Svensson A.K.
      • et al.
      Association of sleep duration with all- and major-cause mortality among adults in Japan, China, Singapore, and korea.
      ]. Future studies are warranted to clarify the underlying reasons for this discrepancy.
      The associations of short or long sleep duration with cardiovascular mortality, CHD, and stroke were marginally significant or even not statistically significant when the two factors were mutually adjusted in both main and sensitivity analyses. These results must be carefully interpreted. Regarding cardiovascular mortality, this was possible because of the relatively low cardiovascular mortality rate in the UKB. We observed only 1,304 cardiovascular deaths in this study, which was significantly lower than the other outcomes (more than 5,000 events). Regarding CHD, some previous studies have documented a marginally significant association between long sleep duration and CHD [
      • Krittanawong C.
      • Tunhasiriwet A.
      • Wang Z.
      • et al.
      Association between short and long sleep durations and cardiovascular outcomes: a systematic review and meta-analysis.
      ,
      • Lao X.Q.
      • Liu X.
      • Deng H.B.
      • et al.
      Sleep quality, sleep duration, and the risk of coronary heart disease: a prospective cohort study with 60,586 adults.
      ,
      • Wang D.
      • Li W.
      • Cui X.
      • et al.
      Sleep duration and risk of coronary heart disease: a systematic review and meta-analysis of prospective cohort studies.
      ], which was consistent with our study. A previous meta-analysis of 19 studies found that long sleep duration slightly increased the risk of CHD (HR, 1.12; 95% CI, 1.01–1.24) [
      • Krittanawong C.
      • Tunhasiriwet A.
      • Wang Z.
      • et al.
      Association between short and long sleep durations and cardiovascular outcomes: a systematic review and meta-analysis.
      ]. With respect to stroke, the association between short sleep duration and total stroke, ischemic, and hemorrhage stroke varied widely in previous studies. Some studies have reported no association between short sleep duration and stroke [
      • Ai S.
      • Zhang J.
      • Zhao G.
      • et al.
      Causal associations of short and long sleep durations with 12 cardiovascular diseases: linear and nonlinear Mendelian randomization analyses in UK Biobank.
      ,
      • Titova O.E.
      • Michaëlsson K.
      • Larsson S.C.
      Sleep duration and stroke: prospective cohort study and mendelian randomization analysis.
      ], which is consistent with the result of our study. A recent study based on the Mendelian randomization method suggested no causal association between short sleep duration and ischemic stroke (HR, 1.11; 95% CI, 0.85–1.45) [
      • Ai S.
      • Zhang J.
      • Zhao G.
      • et al.
      Causal associations of short and long sleep durations with 12 cardiovascular diseases: linear and nonlinear Mendelian randomization analyses in UK Biobank.
      ]. However, some studies have found that short sleep duration increased the risk of stroke [
      • Krittanawong C.
      • Tunhasiriwet A.
      • Wang Z.
      • et al.
      Association between short and long sleep durations and cardiovascular outcomes: a systematic review and meta-analysis.
      ,
      • Cappuccio F.P.
      • Cooper D.
      • D'Elia L.
      • Strazzullo P.
      • Miller M.A.
      Sleep duration predicts cardiovascular outcomes: a systematic review and meta-analysis of prospective studies.
      ], which was also observed in participants with sleep duration ≤5 h/day in this study. The possible reasons for these discrepancies are differences in definition of short sleep duration and stroke events, sample size, and adjustment of covariables. Future studies are necessary to clarify the underlying mechanisms. Although the associations between short or long sleep duration and cardiovascular mortality, CHD, and stroke are not clear, they remain independent risk factors for other CMDs and all-cause mortality. Thus, a healthy sleep duration (ages 25–64 years: 7–9 h/day; ages 65+ years: 7–8 h/day) is recommended in daily life. As for depression, we documented statistically significant associations between depression and CMDs but not all-cause and cardiovascular mortality when using only ICD-10 records to define depression in sensitivity analyses. The possible reason was that some participants with depression but no corresponding ICD-10 records were misclassified to non-depression group, which would attenuate the associations of depression with CMDs and mortality (Table 2 and Supplementary Table 8). The effects of misclassification were more significant for all-cause mortality as the HR was relatively smaller when compared with that for CMDs (1.12 vs. 1.25–1.36), and the possible reason for insignificant association between depression and cardiovascular mortality was small number of events (1,304 vs. 5,227–29,115).
      We found a significant additive interaction, which is more relevant to public health measures, between depression and short sleep duration in relation to all-cause mortality and CHD. Our study indicated that 56% and 49% of all-cause mortality and CHD among participants with depression and short sleep duration, respectively, could be attributed to additive interaction between the two factors. These findings indicate that approximately 50% of cases of death and CHD would not occur if either risk factor was well controlled. Meanwhile, these results highlight that the majority of deaths and CHD cases depend on both depression and short sleep duration, cost-effective strategies should be established for the prevention of co-occurrence of depression and short sleep duration in patients who already have one condition.

      4.2 Strengths and limitations

      This study has several strengths. First, we comprehensively evaluated the associations of depression and sleep duration with CMDs and mortality in models with and without adjustment for each other, which confirmed that the two factors were independently associated with increased risk of CMDs and mortality in this study. Second, we analyzed both multiplicative and additive interactions between depression and sleep duration and CMDs and mortality, which provided valuable information for identifying high-risk groups. Finally, the large sample size, long follow-up period, and large number of covariates of the UKB enabled us to establish three settings of models and to perform five sensitivity analyses to strengthen our results.
      This study has some limitations. First, reverse causality may have existed in our study. To minimize this risk, we excluded participants diagnosed with CMD during the first year of follow-up, and the results were consistent with the main analysis. Second, sleep duration and part of patients with depression were defined according to self-reported medical conditions, and there may be recall and reporting bias. Furthermore, use of antidepressants is also linked to other conditions other than depression, such as anxiety, insomnia and pain, this may also be prone to bias. When using only ICD-10 records to define depression, the associations between depression and all-cause and cardiovascular mortality were no longer significant. Future studies based on more accurate measurements of sleep duration (e.g., polysomnography) and depression are required. Third, all variables used in this study were collected at baseline and assumed to be unchanged during the follow-up. Further studies investigating changes in some variables over time will further elucidate the individual and joint association of depression and sleep duration with CMDs and mortality. Finally, most of the participants from the UKB were British people aged above 40 years, the generalizability of our findings to other populations may be limited, and future studies are required to replicate our findings in populations from other countries or young people.
      In conclusion, our findings showed that depression and short or long sleep duration were independent, rather than overlapping, risk factors for CMDs and mortality in the UKB. Furthermore, we documented a significant additive interaction between depression and short sleep duration in relation to all-cause mortality and CHD. Our findings suggest that both depression and short or long sleep duration should be well controlled to reduce the burden of CMDs and premature death. Further studies are required to validate our results in other populations.

      Financial support

      This study was funded by the Science and Technology Innovation Program of Hunan Province (No. 2021RC2014 and No. 2020RC4006 ), the National Natural Science Foundation of China (No. 81822004 ), the National Key Research and Development Program of China (No. 2020YFC2008002 ), and the Project of Innovation-driven Plan in Central South University (No. 2020CX017) .

      CRediT authorship contribution statement

      Xunjie Cheng: Conceptualization, Methodology, Software, Formal analysis, Data curation, Writing – original draft. Feiyun Ouyang: Data curation, Validation, Writing – review & editing. Tianqi Ma: Data curation, Validation, Writing – review & editing. Lingfang He: Data curation, Validation, Writing – review & editing. Li Gong: Data curation, Validation, Writing – review & editing. Jinghua Yin: Data curation, Validation, Writing – review & editing. Guogang Zhang: Resources, Supervision, Writing – review & editing. Yongping Bai: Resources, Supervision, Funding acquisition, Writing – review & editing.

      Declaration of competing interest

      The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

      Appendix A. Supplementary data

      The following is the Supplementary data to this article:

      References

        • Roth G.A.
        • Mensah G.A.
        • Johnson C.O.
        • et al.
        Global burden of cardiovascular diseases and risk factors, 1990-2019: update from the GBD 2019 study [published correction appears in J Am coll cardiol. 2021 apr 20;77(15):1958-1959].
        J. Am. Coll. Cardiol. 2020; 76: 2982-3021https://doi.org/10.1016/j.jacc.2020.11.010
        • Global Burden of Disease Collaborative Network
        Global Burden of Disease Study 2019 (GBD 2019) Results.
        Institute for Health Metrics and Evaluation (IHME), Seattle, United States2020 (Accessed 2022 01 18)
        • Sheehan C.M.
        • Frochen S.E.
        • Walsemann K.M.
        • Ailshire J.A.
        Are U.S. adults reporting less sleep?: findings from sleep duration trends in the National Health Interview Survey.
        Sleep. 2019; 42 (zsy221): 2004-2017https://doi.org/10.1093/sleep/zsy221
        • Arias-de la Torre J.
        • Vilagut G.
        • Ronaldson A.
        • et al.
        Prevalence and variability of current depressive disorder in 27 European countries: a population-based study.
        Lancet Public Health. 2021; 6 (e729-e738)https://doi.org/10.1016/S2468-2667(21)00047-5
        • Daly M.
        Prevalence of depression among adolescents in the U.S. From 2009 to 2019: analysis of trends by sex, race/ethnicity, and income.
        J. Adolesc. Health. 2021; (S1054-139X(21)00443-2)https://doi.org/10.1016/j.jadohealth.2021.08.026
        • Otsuka Y.
        • Kaneita Y.
        • Spira A.P.
        • et al.
        Trends in sleep problems and patterns among Japanese adolescents: 2004 to 2017.
        Lancet Reg Health West Pac. 2021; 9100107https://doi.org/10.1016/j.lanwpc.2021.100107
        • Ai S.
        • Zhang J.
        • Zhao G.
        • et al.
        Causal associations of short and long sleep durations with 12 cardiovascular diseases: linear and nonlinear Mendelian randomization analyses in UK Biobank.
        Eur. Heart J. 2021; 42: 3349-3357https://doi.org/10.1093/eurheartj/ehab170
        • Krittanawong C.
        • Tunhasiriwet A.
        • Wang Z.
        • et al.
        Association between short and long sleep durations and cardiovascular outcomes: a systematic review and meta-analysis.
        Eur. Heart J. Acute Cardiovasc. Care. 2019; 8: 762-770https://doi.org/10.1177/2048872617741733
        • Harshfield E.L.
        • Pennells L.
        • Schwartz J.E.
        • et al.
        Association between depressive symptoms and incident cardiovascular diseases.
        JAMA. 2020; 324: 2396-2405https://doi.org/10.1001/jama.2020.23068
        • Miloyan B.
        • Fried E.
        A reassessment of the relationship between depression and all-cause mortality in 3,604,005 participants from 293 studies.
        World Psychiatr. 2017; 16: 219-220https://doi.org/10.1002/wps.20439
        • Itani O.
        • Jike M.
        • Watanabe N.
        • Kaneita Y.
        Short sleep duration and health outcomes: a systematic review, meta-analysis, and meta-regression.
        Sleep Med. 2017; 32: 246-256https://doi.org/10.1016/j.sleep.2016.08.006
        • García-Perdomo H.A.
        • Zapata-Copete J.
        • Rojas-Cerón C.A.
        Sleep duration and risk of all-cause mortality: a systematic review and meta-analysis.
        Epidemiol. Psychiatr. Sci. 2019; 28: 578-588https://doi.org/10.1017/S2045796018000379
        • Zhou L.
        • Yu K.
        • Yang L.
        • et al.
        Sleep duration, midday napping, and sleep quality and incident stroke: the Dongfeng-Tongji cohort.
        Neurology. 2020; 94: e345-e356https://doi.org/10.1212/WNL.0000000000008739
        • Wang C.
        • Bangdiwala S.I.
        • Rangarajan S.
        • et al.
        Association of estimated sleep duration and naps with mortality and cardiovascular events: a study of 116 632 people from 21 countries.
        Eur. Heart J. 2019; 40: 1620-1629https://doi.org/10.1093/eurheartj/ehy695
        • Svensson T.
        • Saito E.
        • Svensson A.K.
        • et al.
        Association of sleep duration with all- and major-cause mortality among adults in Japan, China, Singapore, and korea.
        JAMA Netw. Open. 2021; 4e2122837https://doi.org/10.1001/jamanetworkopen.2021.22837
        • Yin J.
        • Jin X.
        • Shan Z.
        • et al.
        Relationship of sleep duration with all-cause mortality and cardiovascular events: a systematic review and dose-response meta-analysis of prospective cohort studies.
        J. Am. Heart Assoc. 2017; 6e005947https://doi.org/10.1161/JAHA.117.005947
        • Mezick E.J.
        • Hall M.
        • Matthews K.A.
        Are sleep and depression independent or overlapping risk factors for cardiometabolic disease?.
        Sleep Med. Rev. 2011; 15: 51-63https://doi.org/10.1016/j.smrv.2010.03.001
        • Riemann D.
        • Krone L.B.
        • Wulff K.
        • Nissen C.
        Sleep, insomnia, and depression.
        Neuropsychopharmacology. 2020; 45: 74-89https://doi.org/10.1038/s41386-019-0411-y
        • Plante D.T.
        The evolving nexus of sleep and depression.
        Am. J. Psychiatr. 2021; 178: 896-902https://doi.org/10.1176/appi.ajp.2021.21080821
        • Cappuccio F.P.
        • Stranges S.
        • Kandala N.B.
        • et al.
        Gender-specific associations of short sleep duration with prevalent and incident hypertension: the Whitehall II Study.
        Publish. Correct. Appear Hypertension. 2007 Nov; 50 (Hypertension. 2007;50(4):693-700): e170https://doi.org/10.1161/HYPERTENSIONAHA.107.095471
        • Lan T.Y.
        • Lan T.H.
        • Wen C.P.
        • Lin Y.H.
        • Chuang Y.L.
        Nighttime sleep, Chinese afternoon nap, and mortality in the elderly.
        Sleep. 2007; 30: 1105-1110https://doi.org/10.1093/sleep/30.9.1105
        • Sudlow C.
        • Gallacher J.
        • Allen N.
        • et al.
        UK biobank: an open access resource for identifying the causes of a wide range of complex diseases of middle and old age.
        PLoS Med. 2015; 12e1001779https://doi.org/10.1371/journal.pmed.1001779
        • Said M.A.
        • Verweij N.
        • van der Harst P.
        Associations of combined genetic and lifestyle risks with incident cardiovascular disease and diabetes in the UK biobank study.
        JAMA Cardiol. 2018; 3: 693-702https://doi.org/10.1001/jamacardio.2018.1717
        • Cheng X.
        • Ouyang F.
        • Ma T.
        • et al.
        Association of healthy lifestyle and life expectancy in patients with cardiometabolic multimorbidity: a prospective cohort study of UK biobank.
        Front Cardiovasc. Med. 2022; 9830319https://doi.org/10.3389/fcvm.2022.830319
        • Glanville K.P.
        • Coleman J.R.I.
        • Howard D.M.
        • et al.
        Multiple measures of depression to enhance validity of major depressive disorder in the UK Biobank.
        BJPsych Open. 2021; 7 (. Published 2021 Feb 5): e44https://doi.org/10.1192/bjo.2020.145
        • Hirshkowitz M.
        • Whiton K.
        • Albert S.M.
        • et al.
        National Sleep Foundation's updated sleep duration recommendations: final report.
        Sleep Health. 2015; 1: 233-243https://doi.org/10.1016/j.sleh.2015.10.004
        • Duncan M.J.
        • Oftedal S.
        • Kline C.E.
        • Plotnikoff R.C.
        • Holliday E.G.
        Associations between aerobic and muscle-strengthening physical activity, sleep duration, and risk of all-cause mortality: a prospective cohort study of 282,473 U.S. adults [published online ahead of print, 2022 Jul 21].
        J Sport Health Sci. 2022; S2095–2546 (00077-1)https://doi.org/10.1016/j.jshs.2022.07.003
        • Grambsch P.
        • Therneau T.
        Proportional hazards tests and diagnostics based on weighted residuals.
        Biometrika. 1994; 81: 515-526
        • Li R.
        • Chambless L.
        Test for additive interaction in proportional hazards models.
        Ann. Epidemiol. 2007; 17: 227-236https://doi.org/10.1016/j.annepidem.2006.10.009
        • Gangwisch J.E.
        • Malaspina D.
        • Posner K.
        • et al.
        Insomnia and sleep duration as mediators of the relationship between depression and hypertension incidence.
        Am. J. Hypertens. 2010; 23: 62-69https://doi.org/10.1038/ajh.2009.202
        • Azevedo Da Silva M.
        • Singh-Manoux A.
        • Shipley M.J.
        • et al.
        Sleep duration and sleep disturbances partly explain the association between depressive symptoms and cardiovascular mortality: the Whitehall II cohort study.
        J. Sleep Res. 2014; 23: 94-97https://doi.org/10.1111/jsr.12077
        • Ferrie J.E.
        • Shipley M.J.
        • Cappuccio F.P.
        • et al.
        A prospective study of change in sleep duration: associations with mortality in the Whitehall II cohort.
        Sleep. 2007; 30: 1659-1666https://doi.org/10.1093/sleep/30.12.1659
        • Bellavia A.
        • Åkerstedt T.
        • Bottai M.
        • Wolk A.
        • Orsini N.
        Sleep duration and survival percentiles across categories of physical activity.
        Am. J. Epidemiol. 2014; 179: 484-491https://doi.org/10.1093/aje/kwt280
        • Lao X.Q.
        • Liu X.
        • Deng H.B.
        • et al.
        Sleep quality, sleep duration, and the risk of coronary heart disease: a prospective cohort study with 60,586 adults.
        J. Clin. Sleep Med. 2018; 14: 109-117https://doi.org/10.5664/jcsm.6894
        • Wang D.
        • Li W.
        • Cui X.
        • et al.
        Sleep duration and risk of coronary heart disease: a systematic review and meta-analysis of prospective cohort studies.
        Int. J. Cardiol. 2016; 219: 231-239https://doi.org/10.1016/j.ijcard.2016.06.027
        • Titova O.E.
        • Michaëlsson K.
        • Larsson S.C.
        Sleep duration and stroke: prospective cohort study and mendelian randomization analysis.
        Publish. Correct Appear Stroke. 2020; 51 (Stroke. 2020;51(11):3279-3285): e347https://doi.org/10.1161/STROKEAHA.120.029902
        • Cappuccio F.P.
        • Cooper D.
        • D'Elia L.
        • Strazzullo P.
        • Miller M.A.
        Sleep duration predicts cardiovascular outcomes: a systematic review and meta-analysis of prospective studies.
        Eur. Heart J. 2011; 32: 1484-1492https://doi.org/10.1093/eurheartj/ehr007