A Validation Study of Four Different Cluster Analyses of OSA and the Incidence of Cardiovascular Mortality in a Hispanic Population

doi:10.1016/j.chest.2021.06.047

Chest

Volume 160, Issue 6, December 2021, Pages 2266-2274

https://doi.org/10.1016/j.chest.2021.06.047 Get rights and content

Background

Previous studies reported a strong association between sleepiness-related symptoms and comorbidities with poor cardiovascular outcomes among patients with moderate to severe OSA (msOSA). However, the validation of these associations in the Hispanic population from South America and the ability to predict incident cardiovascular disease remain unclear.

Research Question

In Hispanic patients with msOSA, are four different cluster analyses reproducible and able to predict incident cardiovascular mortality?

Study Design and Methods

Using the SantOSA cohort, we reproduced four cluster analyses (Sleep Heart Health Study [SHHS], Icelandic Sleep Apnea Cohort [ISAC], Sleep Apnea Cardiovascular Endpoints [SAVE], and The Institute de Recherche en Sante Respiratoire des Pays de la Loire [IRSR] cohorts) following a cluster analysis similar to each training dataset. The incidence of cardiovascular mortality was constructed using a Kaplan-Meier (log-rank) model, and Cox proportional hazards models were adjusted by confounders.

Results

Among 780 patients with msOSA in our cohort, two previous cluster analyses (SHHS and ISAC) were reproducible. The SAVE and IRSR cluster analyses were not reproducible in our sample. We identified the following subtypes for SHHS: “minimally symptomatic,” “disturbed sleep,” “moderate sleepiness,” and “severe sleepiness.” For ISAC, three different subtypes (“minimally symptomatic,” “disturbed sleep,” and “excessive sleepiness”) were similar to the original dataset. Compared with “minimally symptomatic,” we found a significant association between “excessive sleepiness” and cardiovascular mortality after 5 years of follow-up in SantOSA, hazard ratio (HR), 5.47; 95% CI, 1.74-8.29; P < .01; and HR, 3.23; 95% CI, 1.21-8.63; P = .02, using the SHHS and ISAC cluster analyses, respectively.

Interpretation

Among patients with msOSA, a symptom-based approach can validate different OSA patient subtypes, and those with excessive sleepiness have an increased risk of incident cardiovascular mortality in the Hispanic population from South America.

Graphical Abstract

Section snippets

Description of the OSA in Santiago Cohort

This study followed the current recommendation from the Strengthening the Reporting of Observational Studies in Epidemiology statement.¹¹ OSA in Santiago (SantOSA) is a prospective, clinic-based study of the consequences of OSA in the real world derived from the clinical evaluation of OSA in a tertiary center. The protocol and study registration information is available at www.isrctn.com, study registration no: ISRCTN62293645. This study was approved by the Institutional Review Board of Clinica

Results

A total of 2,018 participants were included in the SantOSA cohort, and 1,113 (55.1%) were excluded by apnea-hypopnea index < 15 events/h. A total of 905 participants reported msOSA, and 125 (8.6%) were excluded because of insufficient or missing data caused by loss of follow-up in the national register of mortality by incorrect ID or foreign (non-Chilean) ID. Finally, the sample included 780 participants with msOSA. The study flow chart and the number of participants included in each cluster

Discussion

The main findings of this study are as follows: (1) Using a clinic-based cohort of patients from a South American country (Chile), we identified different phenotypes of OSA patients using sleep-related symptoms similar to previous studies including both clinic-based and population-based cohorts. (2) Based on the previous cluster analysis, the cardiovascular and multicomponent clusters proposed by Quan et al²³ and Gagnadoux et al²⁴ were not replicable in our cohort. (3) Excessive sleepiness was

Interpretation

Among patients with moderate-to-severe OSA in Chile, a symptom-based approach can validate different OSA patient subtypes, and those with the “excessive sleepiness” subtype have an increased risk of incident cardiovascular mortality according to two different solutions. This subtype should be included in clinical practice and further research.

Take-home Points

Study Question: In Latino patients with moderate to severe OSA, are four different cluster analyses reproducible and able to predict

Acknowledgments

Author contributions: G. L. is the guarantor of the paper. Study design and ISRCTN register: G. L., J. J.; data analysis: J. D., F. L., C. S.; Interpretation of the results: G. L., J. D., J. J.; Manuscript preparation: G. L., J. J., J. D.; Manuscript revision, editing, and approval: G. L., J. D., F. L., C. S., J. J. All authors approved the final manuscript.

Financial/nonfinancial disclosures: None declared.

References (40)

A.V. Benjafield et al.
Estimation of the global prevalence and burden of obstructive sleep apnoea: a literature-based analysis
Lancet Respir Med
(2019)
A. Malhotra et al.
Obstructive sleep apnoea
Lancet
(2002)
M.A. Martinez-Garcia et al.
Precision medicine in obstructive sleep apnoea
Lancet Respir Med
(2019)
A. Zinchuk et al.
Phenotypic subtypes of OSA: a challenge and opportunity for precision medicine
Chest
(2020)
G. Bakris et al.
ACC/AHA versus ESC/ESH on hypertension guidelines: JACC guideline comparison
J Am Coll Cardiol
(2019)
W. Quan et al.
High risk characteristics for recurrent cardiovascular events among patients with obstructive sleep apnoea in the SAVE study
EClinicalMedicine
(2018)
F. Gagnadoux et al.
Relationship between OSA clinical phenotypes and CPAP treatment outcomes
Chest
(2016)
M. Sánchez-de-la-Torre et al.
Effect of obstructive sleep apnoea and its treatment with continuous positive airway pressure on the prevalence of cardiovascular events in patients with acute coronary syndrome (ISAACC study): a randomised controlled trial
Lancet Respir Med
(2020)
G. Labarca et al.
Efficacy of continuous positive airway pressure (CPAP) in the prevention of cardiovascular events in patients with obstructive sleep apnea: systematic review and meta-analysis
Sleep Med Rev
(2020)
S. Javaheri et al.
CPAP treatment and cardiovascular prevention: we need to change the design and implementation of our trials
Chest
(2019)

G. Labarca et al.

A clinic-based cluster analysis in patients with moderate-severe obstructive sleep apnea (OSA) in Chile

Sleep Med

(2020)

J.M. Marin et al.

Long-term cardiovascular outcomes in men with obstructive sleep apnoea-hypopnoea with or without treatment with continuous positive airway pressure: an observational study

Lancet

(2005)

J.C. Meurice

Predictive factors of long-term compliance with nasal continuous positive airway pressure treatment in sleep apnea syndrome

Chest

(1994)

E.A. Phillipson

Sleep apnea: a major public health problem

N Engl J Med

(1993)

S.C. Veasey et al.

Obstructive sleep apnea in adults

N Engl J Med

(2019)

W.J. Randerath et al.

Evaluation of a multicomponent grading system for obstructive sleep apnoea: the Baveno classification

ERJ Open Res

(2021)

D. Wraith et al.

Classifying patients by their characteristics and clinical presentations: the use of latent class analysis

Respirology

(2014)

S. Redline et al.

Sleep-disordered breathing in Hispanic/Latino individuals of diverse backgrounds. The Hispanic Community Health Study/Study of Latinos

Am J Respir Crit Care Med

(2014)

B.T. Keenan et al.

Recognizable clinical subtypes of obstructive sleep apnea across international sleep centers: a cluster analysis

Sleep

(2018)

E. von Elm et al.

The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies

PLoS Med

(2007)

Cited by (35)

The PAP-RES algorithm: Defining who, why and how to use positive airway pressure therapy for OSA
2024, Sleep Medicine Reviews
Obstructive sleep apnea (OSA) is a common condition that is increasing in prevalence worldwide. Untreated OSA has a negative impact on health-related quality of life and is an independent risk factor for cardiovascular diseases. Despite available data suggesting that cardiovascular risk might differ according to clinical phenotypes and comorbidities, current approaches to OSA treatment usually take a “one size fits all” approach. Identification of cardiovascular vulnerability biomarkers and clinical phenotypes associated with response to positive airway pressure (PAP) therapy could help to redefine the standard treatment paradigm. The new PAP-RES (PAP-RESponsive) algorithm is based on the identification of OSA phenotypes that are likely to impact therapeutic goals and modalities. The paradigm shift is to propose a simplified approach that defines therapeutic goals based on OSA phenotype: from a predominantly “symptomatic phenotype” (individuals with high symptom burden that negatively impacts on daily life and/or accident risk or clinically significant insomnia) to a “vulnerable cardiovascular phenotype” (individuals with comorbidities [serious cardiovascular or respiratory disease or obesity] that have a negative impact on cardiovascular prognosis or a biomarker of hypoxic burden and/or autonomic nervous system dysfunction). Each phenotype requires a different PAP therapy care pathway based on differing health issues and treatment objectives.
Pitolisant: A Game Changer for Managing Sleepiness in OSA?
2024, Chest
Incident Coronary Calcium Score in Patients With OSA With and Without Excessive Sleepiness: Brazilian Longitudinal Study of Adult Health
2024, Chest
Uncertainty exists about the impact of OSA and its phenotypes on cardiovascular disease.
Are OSA and clinical features such as daytime sleepiness associated with incident subclinical coronary atherosclerosis?
In this prospective community-based cohort study, we administered a sleepiness questionnaire, actigraphy, and home sleep studies at baseline. Coronary artery calcium (CAC; 64-slice multidetector CT scan imaging) was measured at two different time points throughout the study (baseline, between 2010 and 2014, and follow-up, between 2016 and 2018). Incidence of subclinical atherosclerosis was defined as baseline CAC of 0 followed by CAC of > 0 at a 5-year follow-up visit. The association of incident CAC outcome was assessed using logistic regression. Stratified analyses based on excessive daytime sleepiness (EDS) were performed.
We analyzed 1,956 participants with available CAC scores at baseline (mean age, 49 ± 8 years; 57.9% female; 32.4% with OSA). In covariate-adjusted analyses (n = 1,247; mean follow-up, 5.1 ± 0.9 years), we found a significant association between OSA and incidence of subclinical atherosclerosis (OR, 1.26; 95% CI, 1.06-1.48), with stronger effects among those reporting EDS (OR, 1.66; 95% CI, 1.30-2.12; P = .028 for interaction). Interestingly, EDS per se was not associated with any CAC outcome. An exploratory analysis of the square root of CAC progression (baseline CAC > 0 followed by a numerical increase in scores at follow-up; n = 319) showed a positive association for both OSA (β = 1.084; 95% CI, 0.032-2.136; P = .043) and OSA with EDS (β = 1.651; 95% CI, 0.208-3.094; P = .025).
OSA, particularly with EDS, predicts the incidence and progression of CAC. These results support biological plausibility for the increased cardiovascular risk observed among patients with OSA with excessive sleepiness.
Excessive daytime sleepiness, morning tiredness, and prognostic biomarkers in patients with chronic coronary syndrome
2024, International Journal of Cardiology
Sleep-related breathing disorders (SRBD) are related to cardiovascular outcomes in patients with chronic coronary syndrome (CCS). Whether SRBD-related symptoms are associated with prognostic biomarkers in patients with CCS is not established.
Associations between frequency (never/rarely, sometimes, often, always) of self-reported SRBD-related symptoms (excessive daytime sleepiness [EDS]; morning tiredness [MT]; loud snoring; multiple awakenings/night; gasping, choking, or apnea when asleep) and levels of biomarkers related to cardiovascular prognosis (high-sensitivity C-reactive protein [hs-CRP], interleukin 6 [IL-6], high-sensitivity cardiac troponin T [hs-cTnT], N-terminal pro B-type natriuretic peptide [NT-proBNP], cystatin C, growth differentiation factor 15 [GDF-15] and lipoprotein-associated phospholipase A₂ activity) were assessed at baseline in 15,640 patients with CCS on optimal secondary preventive therapy in the STABILITY trial. Cross-sectional associations were assessed by adjusted linear regression models testing for trends with the never/rarely category serving as reference.
EDS was associated (geometric mean ratio, 95% confidence interval) with increased levels of IL-6 (often 1.07 [1.03–1.10], always 1.15 [1.10–1.21]), GDF-15 (often 1.03 [1.01–1.06], always 1.07 [1.03–1.11]), NT-proBNP (always 1.22 [1.12–1.33]), and hs-cTnT (always 1.07 [1.01–1.12]). MT was associated with increased levels of IL-6 (often 1.05 [1.01–1.09], always 1.09 [1.04–1.15]), and GDF-15 (always 1.06 [1.03–1.10]). All symptoms were to some degree associated with higher levels of hs-CRP and loud snoring was also associated with decreased levels of NT-proBNP and hs-cTnT.
In patients with CCS, stepwise increased frequency of SRBD-related symptoms, such as EDS and MT, were associated with gradually higher levels of IL-6 and GDF-15, each reflecting distinct pathophysiological pathways.
Associations between sleep-related symptoms, obesity, cardiometabolic conditions, brain structural alterations and cognition in the UK biobank
2023, Sleep Medicine
Sleep disturbances are increasingly recognized as adversely affecting brain health in aging. Our aim was to investigate interrelations between subjective sleep-related symptoms, obesity, cardiometabolic disorders, brain structure and cognitive decline in a population-based aging sample.
Data were extracted from the UK Biobank for anthropometric and demographic information, self-reported sleep behaviours, cardiometabolic measures, structural brain magnetic resonance imaging and cognitive test scores. “Sleep-related symptoms” (SRS) were measured using four questionnaire items: loud snoring, daytime sleepiness, likelihood to nap and difficulty getting up in the morning. Associations were tested using a structural equation model (SEM), adjusted for confounders. Further, multiple regression analysis was used to test for direct relationships between SRS and specific cognitive domains.
Among 36,468 participants with an average age of 63.6 (SD 7.5) years and 46.7% male, we found that SRS were associated with obesity and several pre-existing cardiometabolic disturbances. In turn, cardiometabolic disorders were associated with increased white matter hyperintensities and cortical thinning, which were related to cognitive dysfunction. SRS were also directly related to several structural brain changes and to cognitive dysfunction. Regression analyses showed that SRS were directly associated with slower reaction times, and lower scores in fluid intelligence, working memory and executive function.
Self-reported sleep-related symptoms were associated with cognitive dysfunction directly and through pre-existing cardiometabolic disorders and brain structural alterations. These findings provide evidence that symptoms of sleep disturbances, here defined primarily by hypersomnolence and snoring, are important risk factors or markers for cognitive dysfunction in an aging population.
Heritability and genetic correlations for sleep apnea, insomnia, and hypersomnia in a large clinical biobank
2023, Sleep Health
Comorbid insomnia and sleep apnea is reported to have worse outcomes than either condition alone. The local genetic correlations of these disorders are unknown.
To identify local genomic regions with heritability for clinically diagnosed sleep apnea and insomnia, and to identify local genetic correlations between these disorders and/or hypersomnia.
Fifty thousand two hundred seventeen patients of European ancestry were examined. Global and local heritability and genetic correlations for independent regions were calculated, adjusting for obesity and other covariates.
Sleep apnea and insomnia were significantly globally heritable and had 118 and 168 genetic regions with local heritability p-values <.05, respectively. One region had a significant genetic correlation for sleep apnea and hypersomnia (p-value = 9.85 × 10⁻⁴).
Clinically diagnosed sleep apnea and insomnia have minimal shared genetic architecture, supporting genetically distinct comorbid insomnia and sleep apnea components. However, additional correlated regions may be identified with additional sample size and methodological improvements.

View all citing articles on Scopus

: FUNDING/SUPPORT: The authors have reported to CHEST that no funding was received for this study.

View full text

Chest

Sleep: Original ResearchA Validation Study of Four Different Cluster Analyses of OSA and the Incidence of Cardiovascular Mortality in a Hispanic Population

Background

Research Question

Study Design and Methods

Results

Interpretation

Graphical Abstract

Section snippets

Description of the OSA in Santiago Cohort

Results

Discussion

Interpretation

Take-home Points

Acknowledgments

Lancet Respir Med

Lancet

Lancet Respir Med

Chest

J Am Coll Cardiol

EClinicalMedicine

Chest

Lancet Respir Med

Sleep Med Rev

Chest

Sleep Med

Lancet

Chest

Sleep apnea: a major public health problem

N Engl J Med

Obstructive sleep apnea in adults

N Engl J Med

Evaluation of a multicomponent grading system for obstructive sleep apnoea: the Baveno classification

ERJ Open Res

Classifying patients by their characteristics and clinical presentations: the use of latent class analysis

Respirology

Sleep-disordered breathing in Hispanic/Latino individuals of diverse backgrounds. The Hispanic Community Health Study/Study of Latinos

Am J Respir Crit Care Med

Recognizable clinical subtypes of obstructive sleep apnea across international sleep centers: a cluster analysis

Sleep

The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies

PLoS Med

Sleep: Original Research
A Validation Study of Four Different Cluster Analyses of OSA and the Incidence of Cardiovascular Mortality in a Hispanic Population