Review article
The prevalence and predictors of obstructive sleep apnea in major depressive disorder, bipolar disorder and schizophrenia: A systematic review and meta-analysis

https://doi.org/10.1016/j.jad.2016.02.060Get rights and content

Highlights

  • Approximately one quarter of patients with serious mental illness had obstructive sleep apnea.

  • OSA appears more prevalent in people with MDD (36%) compared to BD (24%) & Schizophrenia (15%).

  • Increasing age & BMI were moderators of OSA in patients with SMI.

Abstract

Background

Obstructive sleep apnea (OSA) is a health hazard since it is associated with neurocognitive dysfunction and cardio-metabolic diseases. The prevalence of OSA among people with serious mental illness (SMI) is unclear.

Method

We searched major electronic databases from inception till 06/2015. Articles were included that reported the prevalence of OSA determined by polysomnography (PSG) or an apnea-hypopnea index (AHI) >5 events/hr, in people with major depressive disorder (MDD), bipolar disorder (BD) or schizophrenia. A random effects meta-analysis calculating the pooled prevalence of OSA and meta-regression of potential moderators were performed.

Results

Twelve articles were included representing 570,121 participants with SMI (mean age=38.3 years (SD=7.5)), 45.8% male (range=32–80.4) and mean body mass index (BMI) 25.9 (SD=3.7). The prevalence of OSA in SMI in clinical studies was 25.7% (95% CI 13.9 to 42.4%, n=1,535). Higher frequencies of OSA were seen in MDD (36.3%, 19.4–57.4%, n=525) than in BD (24.5%, 95% CI 10.6–47.1, n=681) and schizophrenia (15.4%, 95% CI 5.3–37.1%, n=329). The prevalence of OSA in 568,586 people with SMI from population cohort studies was 10.7% (95% CI 2.4–37.0%) and 19.8% (95% CI 2.5–70.0%) in 358,853 people with MDD. Increasing age (β=0.063, 95% CI 0.0005–0.126, p=0.04, N=10) and BMI predicted increased prevalence of OSA (β=0.1642, 95% CI 0.004–0.3701, p=0.04, N=9).

Conclusion

People with SMI (particularly MDD) have a high prevalence of OSA. Screening for and interventions to manage OSA in SMI including those focused on reducing BMI are warranted.

Introduction

Obstructive sleep apnea (OSA) is an important cause for a number of medical comorbidities and premature mortality (Punjabi, 2008). OSA is characterised by recurrent episodes of pharyngeal obstructions during sleep resulting in a reduction in airflow which can cause derangements in gaseous exchange and disturbed sleep (Punjabi, 2008, Sharafkhaneh et al., 2005, Sharafkhaneh et al., 2004). As a result OSA is associated with a range of adverse outcomes including daytime sleepiness (Johns, 1993) and cognitive dysfunction (Olaithe and Bucks, 2013). It may result in the development of hypertension (Peppard et al., 2000), cardiovascular disease (Peker et al., 2006) and abnormalities in glucose metabolism (Punjabi and Polotsky, 2005). Moreover, OSA often has a deleterious impact on quality of life (Punjabi, 2008) and is associated with symptoms of anxiety and depression (Rezaeitalab et al., 2014).

People with serious mental illness (SMI) (defined as major depressive disorder (MDD), bipolar disorder (BD) and schizophrenia spectrum disorders) have significantly poorer physical health than the general population and are at increased risk of comorbidities (Smith et al., 2013a, Smith et al., 2013b, Stubbs et al., 2014). Of particular concern is the high prevalence of obesity (Mitchell et al., 2013a, Mitchell et al., 2013b) diabetes (Stubbs et al., in press); Vancampfort et al., in press) and metabolic syndrome (Mitchell et al., 2013a, Mitchell et al., 2013b).

More recently, research has considered the relationship between SMI and respiratory diseases. For instance, in a population cohort study (Partti et al., 2015) demonstrated that schizophrenia is associated with increased odds of having chronic obstructive pulmonary disease (COPD, OR=4.2) and chronic bronchitis (OR=3.8). Furthermore, reduced lung function (Vancampfort et al., 2014) can negatively influence participation in daily life among patients with schizophrenia. One of the most prevalent respiratory disorder is OSA. Key risk factors for OSA include obesity, craniofacial abnormalities (Sutherland et al., 2012), allergic rhinitis (Farabollini et al., 2009) and cardiovascular disease (Young et al., 2004). In addition, lifestyle factors such as smoking and alcohol misuse increase an individual’s risk of OSA (Punjabi, 2008). These comorbidities and lifestyle factors are particularly prevalent among people with SMI and therefore it is reasonable to assume that this group may be at increased risk of OSA. As such, the identification and management of OSA in people with SMI may be important in preventing or attenuating cardio-metabolic comorbidities associated with SMI (Punjabi, 2008, Young et al., 2004). The relationship between OSA and cardiovascular disease, stroke, hypertension and altered glucose metabolism is complex and appears to be bidirectional (Punjabi, 2008, Young et al., 2004). Nevertheless, understanding the prevalence of OSA in people with SMI is important.

Previously, a number of narrative systematic reviews have considered the prevalence of OSA in schizophrenia (Kalucy et al., 2013), and SMI more generally (Gupta and Simpson, 2015). Whilst not focussing on OSA, another review (Plante and Winkelman, 2008) established the critical relationship between sleep disturbances generally and bipolar disorder. Although these reviews have advanced the field, a number of limitations persist. First, none have attempted to conduct a meta-analysis and the prevalence of OSA was limited to considering individual studies as single entities or calculating arithmetic means or medians of the prevalence of OSA. Meta-analyses enable the pooling of data across studies in order to provide a more accurate effect size than when individual studies are considered separately (Ioannidis, 2009). The recent reviews have also considered differing inclusion criteria when reviewing the literature on OSA. For instance, (Gupta and Simpson, 2015) only considered studies measuring OSA confirmed by polysomnography (PSG) the gold standard recognised by the International Classification of Sleep Disorders 2nd edition (ICSD-2) (Medicine, 2005) criteria. However, (Kalucy et al., 2013) considered ‘sleep disordered breathing’ including studies that report OSA from screening questionnaires.

We aimed to conduct a meta-analysis investigating the prevalence of OSA in SMI and investigate diagnostic subgroup differences among people with MDD, BD and schizophrenia when determined according to standardised OSA measures. Where possible, we anticipated conducting a comparative meta-analysis comparing the risk of OSA in people with SMI versus the general population. Moreover, we sought to investigate differences reported in clinical and population cohort studies and conduct meta-regression analyses on potential moderator variables (if reported in at least 4 studies).

Section snippets

Method

This systematic review was conducted in accordance with the Meta-analysis of Observational Studies in Epidemiology guidelines (Stroup et al., 2000) and reported in accordance with the PRISMA statement (Moher et al., 2009) following a predetermined but unpublished protocol.

Search results and study selection

The initial search identified 2,215 publications. After removal of duplicates, 221 abstracts and titles were screened (Fig. 1). At the full text review stage, 62 articles were considered and 50 articles were excluded with reasons which are listed in Fig. 1. Thus, 12 articles met the eligibility criteria and were included in the review (Ancoli-Israel et al., 1999, Carney et al., 2006, Deldin et al., 2006, Hattori et al., 2009, Hrubos-Strøm et al., 2012, Kelly et al., 2013, Levine et al., 2001,

Prevalence of OSA among people with SMI

Data were pooled from 1,535 unique participants with SMI resulting in a pooled prevalence of OSA of 25.7% (95% CI 13.9 to 42.4, Q=156 (df=13) p<0.01) (Fig. 2). The Begg (−0.9, p=0.59) and Egger tests (−0.45, p=0.79) did not demonstrate any evidence of publication bias in the overall pooled analyses.

Prevalence of OSA among different diagnostic subgroups

Data was pooled from 6 clinical studies from 522 people with MDD to establish a pooled prevalence of OSA of 36.3% (95% CI 19.4 to 57.4, Q=82.83, (df=6) p<0.01). The pooled prevalence of OSA among 681

Prevalence of OSA in population cohort studies

Data were pooled from 4 samples utilising two studies including 359,378 people with MDD, 72,043 with BD and 138,700 people with schizophrenia. As reported in Fig. 3, the pooled prevalence of OSA among a total of 568,586 people with SMI was 10.7% (95% CI 2.4 to 37.0, 1368.3 (df=3), p<0.01). The pooled prevalence of OSA in MDD, calculated from pooled data from two unique studies was 19.8% (95% CI 2.5 to 70.0).

General findings

Results our meta-analysis demonstrates in clinical studies that approximately a quarter of people with SMI meet the criteria for OSA. People with MDD appear to have a higher prevalence of OSA compared to people with BD or schizophrenia. However, after the removal of two potentially biased studies the prevalence of OSA in MDD appears similar to BD. When we considered population cohort data only, the pooled prevalence of OSA was lower at 10.7%, although pooled data from the two MDD cohorts

Conclusion

Our meta-analysis established that 25.7% of people with SMI have OSA. Higher frequencies were found in MDD (36.3%) compared to BD (24.5%) and schizophrenia (15.4%) although for BD and schizophrenia our findings are unclear due to the small number of the studies included for these conditions. Clearly people with SMI are at risk of OSA and this may negatively impact a range of health outcomes and therefore warrants closer attention in clinical practice. Moreover, future representative research is

Conflict of interest

BS, SR, PM, DV and MF declare no conflict of interest.

FG has received honoraria for advisory work and lectures from Roche, BMS, Lundbeck, and Sunovion and has a family member with professional links to Lilly and GSK.

Dr De Hert reported being a paid consultant for, receiving grant or research support and honoraria from, and serving on the speakers’ bureaus or advisory boards of Janssen-Cilag, Lundbeck, and Takeda.

Funding

None.

References (64)

  • A. Sharafkhaneh et al.

    Sleep apnea in a high risk population: a study of Veterans Health Administration beneficiaries

    Sleep Med.

    (2004)
  • D. Vancampfort et al.

    Associations between expiratory spirometry parameters and limitations in daily life activities in patients with schizophrenia

    Gen. Hosp. Psychiatry

    (2014)
  • N.T. Vozoris

    Sleep apnea-plus: prevalence, risk factors, and association with cardiovascular diseases using United States population-level data

    Sleep Med.

    (2012)
  • American Academy of Sleep Medicine

    The International Classification of Sleep Disorders, Diagnostic and Coding Manual

    (2005)
  • American Psychiatric Association

    Diagnostic and Statistical Manual of Mental Disorders – DSM-IV-TR

    (2000)
  • C.B. Begg et al.

    Operating characteristics of a rank correlation test for publication bias

    Biometrics

    (1994)
  • R.S. Bucks et al.

    Neurocognitive function in obstructive sleep apnoea: a meta-review

    Respirology

    (2013)
  • R.M. Carney et al.

    Depression and obstructive sleep apnea in patients with coronary heart disease

    Psychosom. Med.

    (2006)
  • S.R. Coughlin et al.

    Obstructive sleep apnoea is independently associated with an increased prevalence of metabolic syndrome

    Eur. Heart J.

    (2004)
  • M. De Hert et al.

    Physical illness in patients with severe mental disorders. I. Prevalence, impact of medications and disparities in health care

    World Psychiatry: Off. J. World Psychiatric Assoc. (WPA)

    (2011)
  • R. DerSimonian et al.

    Meta-analysis in clinical trials

    Controll. Clin. Trials

    (1988)
  • M. Egger et al.

    Bias in Meta-analysis Detected by a Simple, Graphical Test. BMJ

  • B. El-Ad et al.

    Effect of sleep apnea on cognition and mood

    Int. Rev. Psychiatry

    (2005)
  • B. Farabollini et al.

    Allergic rhinitis and associated pathologies: the rationale for steroid options

    Eur. Ann. Allergy Clin. Immunol.

    (2009)
  • P. Gardner-Sood et al.

    Cardiovascular risk factors and metabolic syndrome in people with established psychotic illnesses: baseline data from the IMPaCT RCT study

    Psychol. Med.

    (2016)
  • G.J. Gibson

    Obstructive sleep apnoea syndrome: underestimated and undertreated

    Br. Med. Bull.

    (2005)
  • T.L. Giles et al.

    Continuous positive airways pressure for obstructive sleep apnoea in adults

    Cochrane Database Syst. Rev.

    (2006)
  • M.A. Gupta et al.

    Obstructive sleep apnea and psychiatric disorders: a systematic review

    J. Clin. Sleep Med.: Off. Publ. Am. Acad. Sleep Med.: JCSM

    (2015)
  • M. Hattori et al.

    Risk factors for obstructive sleep apnea syndrome screening in mood disorder patients

    Psychiatry Clin. Neurosci.

    (2009)
  • Higgins, J.P.T., Green, S., 2011. Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 [updated...
  • H. Hrubos-Strøm et al.

    Sleep apnoea, anxiety, depression and somatoform pain: a community-based high-risk sample

    Eur. Respir. J.

    (2012)
  • J.P. Ioannidis

    Integration of evidence from multiple meta-analyses: a primer on umbrella reviews, treatment networks and multiple treatments meta-analyses

    Can. Med. Assoc. J.

    (2009)
  • Cited by (107)

    • Validation of a data collection set for the psychiatric, addiction, sleep and chronobiological assessments of patients with depression: A Delphi study for the SoPsy-depression French national cohort

      2023, Encephale
      Citation Excerpt :

      Four questionnaires characterizing addiction were identified (alcohol, cannabis, tobacco, and other substances), and the steering committee decided to retain additionally the short version of the AUDIT for the mandatory and the full version for the optional set. Furthermore, 8 questionnaires for “sleep” dimension (habits, insomnia, sleep quality, chronotype, sleepiness, hypersomnia, apnea and restless leg syndrome screenings) were initially included, and the steering committee decided to add a questionnaire for chronotype since none was retained despite being an essential marker, and the only chronobiological marker in the mandatory set; the committee further decided to move a necessary screening questionnaire for apnea, hypersomnia and sleepiness from the optional set to the mandatory one because of their high prevalence in patients with depressive disorder [17,48–50]. Finally, 5 questionnaires were included initially examining these dimensions specifically in children and adolescents (habits, sleepiness, chronotype, depressive and ADHD symptoms) with here again the Munich ChronoType Questionnaire (MCTQ) retained in the mandatory set to assess chronotype and social jet-lag syndromes which are especially prevalent in adolescents [51].

    • Multi-dimensional analyses of the associations between depression, nocturnal awakening and asthmatic outcomes

      2023, Journal of Affective Disorders
      Citation Excerpt :

      Thirdly, obstructive sleep apnea played a negative role in asthmatic outcome (Global Initiative for Asthma, 2020). Data involving 6 clinical studies from 522 people showed that individuals with major depressive disorder have a higher prevalence of obstructive sleep apnea characterized by nocturnal hypoxemia and awakening (Stubbs et al., 2016). Nocturnal hypoxemia and awakening secondary to obstructive sleep apnea may lead to a negative impact on health outcomes for individuals with depression, thus should pay closer attention in clinical practice (Stubbs et al., 2016).

    View all citing articles on Scopus
    View full text