Elsevier

Sleep Medicine

Volume 36, August 2017, Pages 109-118
Sleep Medicine

Original Article
The relationship between weight change and daytime sleepiness: the Sleep Heart Health Study

https://doi.org/10.1016/j.sleep.2017.05.004Get rights and content

Highlights

  • Weight gain is associated with worse daytime sleepiness.

  • About one-fifth of the relationship can be explained by obstructive sleep apnoea.

  • About one-sixth of the relationship can be explained by poor physical health.

Abstract

Objective

Through a causal framework, we aim to assess the association between weight change and daytime sleepiness, and the role of obstructive sleep apnoea (OSA) in this relationship.

Methods

From the Sleep Heart Health Study, we selected individuals who were: (1) 40–64 years old, with (2) body mass index (BMI) ≥18.5 kg/m2, (3) no history of stroke, treatment for OSA, and tracheostomy at baseline. We used multiple linear regression to assess the relationship between five-year weight change and daytime sleepiness (assessed through Epworth Sleepiness Scale (ESS)) at five years, adjusting for daytime sleepiness, demographics, diabetes, subjective sleep duration, sleep disturbance, smoking status, weight, and use of antidepressants and benzodiazepines at baseline, in those with complete data (N = 1468). We further assessed the potential mediating role of OSA in this relationship.

Results

At baseline, the study participants were on average 55 years old, 46% males, with mean BMI 28 kg/m2; and 25% had ESS>10. ESS at five years worsened by 0.36 units (95% confidence interval (CI) 0.12–0.61, p = 0.004) with every 10-kg weight gain. When stratified by sex, this relationship was only found in women (0.55, 95% CI 0.25–0.86, p < 0.001; p-interaction = 0.02). Approximately one-fifth of the relationship between weight change and daytime sleepiness was mediated by severity of OSA at five years.

Conclusion

Weight gain has a detrimental effect on daytime sleepiness, mostly through pathways other than OSA. This study provides further evidence and understanding of the relationship between obesity and excessive daytime sleepiness.

Introduction

The association between obesity and excessive daytime sleepiness (EDS) has long been accepted; however, to our knowledge, its causality has not been formally assessed. Whether weight change has a causal effect on daytime sleepiness, and what the pathways are remain largely unexplored. EDS, the irresistible urge to fall asleep despite one's intention to remain awake [1], is highly prevalent in the general population (up to 30%) [2] and is known to affect work performance [3], [4], mental health [5], [6], quality of life [5] and motor-vehicle-related deaths [7], [8].

A recent longitudinal study by Fernandez-Mendoza et al. [9] found that weight gain over 7.5 years was associated with incident and persistent EDS over the same duration; whilst weight loss was associated with its remission. Similarly, in another study by Palm et al. [10], increase in body mass index (BMI) over 10–13 years was associated with incident EDS over the same duration. However, these studies used one-/two-item non-validated questionnaires to assess EDS, and they did not assess potential mediating pathways between weight change and daytime sleepiness, which may provide useful information for interventions targeting EDS remission.

Using the Sleep Heart Health Study (SHHS) dataset, a large population-based multicentre cohort study for assessing the cardiovascular outcomes of sleep apnoea, and a causal framework, we aimed to study the relationship between weight change and daytime sleepiness, measured through the Epworth Sleepiness Scale, a validated [11], and widely used eight-item questionnaire to measure daytime sleepiness. We also performed mediation analyses to assess the likely pathways through which weight change may affect daytime sleepiness. Potential mediators considered include obstructive sleep apnoea, mental health, physical health, and sleep duration, as suggested by a previous review [12].

Section snippets

Data source

We used the SHHS dataset; a large population-based multicentre cohort study in the US, which aimed to assess the cardiovascular outcomes of sleep apnoea in community-dwelling adults. A total of 6441 individuals were recruited from six ongoing, population-based, parent cohorts: Atherosclerosis Risk in Communities Study, Cardiovascular Health Study, Framingham Heart Study, New York Cohorts, Strong Heart Study and Tucson Cohorts. To be recruited into the SHHS, individuals were required to be 40

Baseline characteristics, and levels of exposure and outcome

The study participants were on average 55 years old, 46% were female, the mean baseline BMI was 28 kg/m2, and 25% had EDS. Compared to men, women were more likely to have lower education, weight, BMI, ESS scores, OAHI, AHI, and SF-36 physical and mental component score at baseline. They were also more likely to have longer objective sleep duration, worse sleep disturbance score, insomnia, used benzodiazepines and antidepressants, and less likely to be a current or former smoker at baseline (

Discussion

In this study, through a causal framework, we have consistently found evidence supporting an association between weight gain and worse daytime sleepiness that may be partly mediated by severity of obstructive sleep apnoea. The relationship seemed to be more pronounced in women and those with poorer mental health.

A recent study by Fernandez-Mendoza et al. [9] similarly found higher risk of incident EDS with weight gain, and remitted EDS with weight loss. Another recent longitudinal study by

Conclusion

Our study showed, for the first time, that weight gain is associated with daytime sleepiness as assessed through the ESS, and that approximately one-fifth of this relationship occurs through obstructive sleep apnoea, and approximately one-sixth through poor overall physical health. We explicitly described the extent to which the assumptions of ‘no unmeasured confounding’ and ‘well-defined intervention’, may affect causal inference from our study findings. This provides further understanding of

Acknowledgements

WLN is supported by a Monash Graduate Scholarship, a Monash International Post-graduate Research Scholarship and a Baker Bright Sparks Top-Up Scholarship. JES is funded by an NHMRC Senior Research Fellowship. AP is supported by an NHMRC Development Fellowship and is a researcher within a National Health and Medical Research Council, Centre for Research Excellence in Obesity Policy and Food Systems grant (APP1041020).

This study was supported in part by the Victorian Government's Operational

References (37)

  • J. Fernandez-Mendoza et al.

    Natural history of excessive daytime sleepiness: role of obesity, weight loss, depression, and sleep propensity

    Sleep

    (2015)
  • N.M. Punjabi et al.

    Predictors of objective sleep tendency in the general population

    Sleep

    (2003)
  • L.A. Panossian et al.

    Daytime sleepiness in obesity: mechanisms beyond obstructive sleep apnea—a review

    Sleep

    (2012)
  • S.F. Quan et al.

    The Sleep Heart Health Study: design, rationale, and methods

    Sleep

    (1997)
  • National Sleep Research Resource

    Sleep Heart Health Study

    (2016)
  • D.A. Dean et al.

    Scaling up scientific discovery in sleep medicine: the national sleep research resource

    Sleep

    (2016)
  • M.W. Johns

    A new method for measuring daytime sleepiness: the Epworth sleepiness scale

    Sleep

    (1991)
  • M.W. Johns

    Reliability and factor analysis of the Epworth sleepiness scale

    Sleep

    (1992)
  • Cited by (16)

    • Longitudinal study of narcolepsy symptoms in first, second, and third-degree relatives of simplex and multiplex narcolepsy families

      2019, Sleep Medicine
      Citation Excerpt :

      Of note, BMI was among the significant predictors of hypersomnolence at the second evaluation: the heavier the family members were, the greater the risk of being hypersomnolent a few years later. A similar finding was obtained recently in the Sleep Heart Health Study [25]. The authors found that weight gain had a detrimental effect on somnolence that could only be partially explained by the severity of OSA.

    • The relationships between improvements in daytime sleepiness, fatigue and depression and psychomotor vigilance task testing with CPAP use in patients with obstructive sleep apnea

      2018, Sleep Medicine
      Citation Excerpt :

      Only patients who were both adherent to, and adequately treated by, CPAP therapy by the time of study closure were deemed eligible to proceed with post-treatment analysis. Due to the previously described independent associations between obesity and impaired vigilance [36] as well as daytime sleepiness in patients with [44] and without OSA [45], those who had significant weight changes from baseline at the time of the potential post-treatment analysis (change in body mass index [BMI] of 10% or greater from baseline) were excluded from further participation to avoid a potential confounding effect of weight change on variables of interest. Additionally, patients who had changes in their dosages of medications that would affect daytime sleepiness (sedative-hypnotics like benzodiazepines, benzodiazepine-receptor agonists and barbiturates), stimulants (methylphenidate, amphetamines, wakefulness-promoting agents like modafinil and armodafinil) and antidepressants (tricyclic antidepressants, monoaminooxidase inhibitors and selective serotonin- or serotonin/norepinephrine reuptake inhibitors) over the course of the study were dropped from post-treatment analysis.

    • Course, risk factors, and mental health outcomes of excessive daytime sleepiness in rural Chinese adolescents: A one-year prospective study

      2018, Journal of Affective Disorders
      Citation Excerpt :

      First, the current study has only 1-year follow-up period, which is much shorter than these previous studies (5 years or above). Second, the population in the current study was relatively lean with a prevalence of obesity of around 3%, when compared with those figures ( around 30%) reported in previous studies (Ng et al., 2017). Finally, these previous studies were conducted in adult population while the current study was conducted in adolescents.

    View all citing articles on Scopus
    View full text