Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Nov 22.
Published in final edited form as: Sleep Med. 2020 May 11;74:92–98. doi: 10.1016/j.sleep.2020.05.001

Symptom subtypes and cognitive function in a clinic-based OSA cohort: a multi-centre Canadian study

AJ Hirsch Allen a,i, Andrew E Beaudin b,c,i, Nurit Fox a, Jill K Raneri d, Robert P Skomro e,i, Patrick J Hanly c,d,f,i, Diego R Mazzotti g, Brendan T Keenan g, Eric E Smith b,c, Sebastian D Goodfellow h, Najib T Ayas a,i,*
PMCID: PMC9680684  NIHMSID: NIHMS1847810  PMID: 32841852

Abstract

Background:

Distinct symptom subtypes are found in patients with OSA. The association between these subtypes and neurocognitive function is unclear.

Objective:

The purposes of this study were to assess whether OSA symptom subtypes are present in a cohort of Canadian patients with suspected OSA and evaluate the relationship between subtypes and neurocognitive function.

Methods:

Patients with suspected OSA who completed a symptom questionnaire and underwent testing for OSA were included. Symptom subtypes were identified using latent class analysis. Associations between subtypes and neurocognitive outcomes (Montreal Cognitive Assessment [MoCA], Rey Auditory Verbal Learning Test [RAVLT], Wechsler Adult Intelligence Scale [WAIS-IV], Digit-Symbol Coding subtest [DSC]) were assessed using analysis of covariance (ANCOVA), controlling for relevant covariates.

Results:

Four symptom subtypes were identified in patients with OSA (oxygen desaturation index ≥5 events/hour). Three were similar to prior studies, including the Excessively Sleepy (N=405), Disturbed Sleep (N=382) and Minimally Symptomatic (N=280), and one was a novel subtype in our sample defined as Excessively Sleepy with Disturbed Sleep (N=247). After covariate adjustment, statistically significant differences among subtypes (p=0.037) and among subtypes and patients without OSA (p=0.044) were observed in DSC scores; the Minimally Symptomatic subtype had evidence of higher DSC scores than all other groups, including non-OSA patients. No differences were seen in MoCA or RAVLT.

Conclusions

Results support the existence of previously identified OSA symptom subtypes of excessively sleepy, disturbed sleep and minimally symptomatic in a clinical sample from Canada. Subtypes were not consistently associated with neurocognitive function across multiple instruments.

Keywords: Sleep apnea, Symptom subtypes, Neurocognitive outcomes, Sleepiness, Cluster analysis

1. Introduction

Obstructive Sleep Apnea (OSA) is a common yet underdiagnosed respiratory disorder characterized by recurrent upper airway obstruction during sleep [1,2]. OSA results in sleep fragmentation and repetitive hypoxemia [3], and is associated with many adverse consequences, including excessive daytime sleepiness, reduced quality of life, decreased learning skills, and neurocognitive impairment [4,5].

It is increasingly recognized that distinct subtypes based on self-reported symptoms are found in patients with OSA [6–10]. Studies using latent class analysis, a statistical procedure for grouping individuals into a set of mutually exclusive clusters based on a collection of categorical measurements [11,12], have identified between three and five primary symptom subtypes of moderate-severe OSA [7–10]. These subtypes have been identified in both clinical [7,8] and population-based cohorts [9,10], and consistently include subtypes defined by severity of sleepiness, presence of disturbed sleep (eg, insomnia) or a lack of traditional symptoms (eg, minimally symptomatic).

These symptom subtypes are associated with adverse outcomes and a differential response to treatment. For example, in a cohort from Iceland, symptom subtypes had a different symptomatic response after two years of treatment with continuous positive airway pressure (CPAP) [13]. Specifically, the excessively sleepy subtype showed a wide range of symptom improvements, whereas symptoms were more resistant to CPAP therapy in the disturbed sleep group. Additionally, in the Sleep Heart Health Study (SHHS), the excessively sleepy subtype was at greater risk of prevalent and incident cardiovascular disease (CVD) [10].

However, the association between these symptom subtypes and neurocognitive function has not been evaluated. Neurocognitive deficits associated with OSA include impaired episodic memory, executive function, attention and visuospatial cognitive functions [14,15]. Longitudinal studies in the general population have found OSA to be an independent risk factor for development of mild cognitive impairment (MCI) [16,17] and dementia [18–20]. In a recent study, Sharma and colleagues found that OSA was associated with markers of increased amyloid burden over a two-year follow-up period in a cohort of cognitively normal elderly men [20]. However, it is unclear whether these neurocognitive effects are mediated by symptoms of OSA (eg, disturbed sleep or sleepiness) or are a consequence of direct neural damage mediated by nocturnal hypoxemia [21,22].

The purpose of this study was: (1) to assess whether previously identified OSA symptom subtypes are present in a cohort of patients with suspected OSA recruited from multiple sleep centers across Western Canada, and (2) to determine the relationship between symptom subtypes and neurocognitive function in this cohort.

2. Methods

Participants were recruited between July 2016 and February 2019 and enrolled in the multi-center Canadian Sleep and Circadian Network–s (CSCN, Website: https://www.cscnweb.ca) adult OSA database. The database includes individuals recruited from the University of Calgary, the University of Saskatchewan, and the University of British Columbia. Participants were eligible for recruitment if they were 18–80 years old, able to complete an English language survey, and referred for suspected OSA. OSA was diagnosed by unattended home sleep apnea testing (HSAT) or in-laboratory attended polysomnography (PSG).

This study was performed according to the Declaration of Helsinki and approved by the Conjoint Health Research Ethics Board of the University of Calgary (ID: REB16-0211), the Biomedical Research Ethics Board of the University of Saskatchewan (ID: 16–106), and the University of British Columbia Clinical Research Ethics Board (ID: H16-00422). Patients were informed of study requirements prior to providing written informed consent.

2.1. Study protocol

At the University of Calgary, OSA was diagnosed by unattended HSAT, whereas at the Universities of Saskatchewan and British Columbia OSA was diagnosed by in-lab attended PSG. OSA severity was defined based upon the oxygen desaturation index (ODI) as Non-OSA (ODI<5), mild OSA (5≤ODI<15), moderate OSA (15≤ODI≤30) and severe OSA (ODI>30 events/hour). All patients completed a standardized sleep questionnaire and cognitive testing prior to any treatment for OSA. Additional study protocol details are provided in the online data supplement.

2.2. PSG and HSAT

2.2.1. Polysomnography (PSG)

Sleep and its various stages were documented using standard electroencephalographic (EEG), electrooculographic (EOG) and electromyographic (EMG) criteria (Sandman 10.1 software). EEG was recorded with electrodes applied at C3-A2 and C4-A1 (according to the International 10–20 system). EMG activity was recorded from the submental muscles and anterior tibialis muscles. Airflow was detected by recording of nasal pressure and a thermocouple. Snoring was measured using a microphone that was attached to the middle of the respitrace band at the level of third intercostal space. The output of the microphone was fed into a sound meter (Model SL120, Pacer Industries, Toronto, Canada) that was calibrated in the range of 40–110 dB using a 1 KHz signal. A single ECG (modified V2) was monitored in order to detect cardiac arrhythmias. Arterial oxygen saturation (SaO2) was monitored continuously with a pulse oximeter (Model N-100, Nellcor Inc., Hayward, CA) attached to the index finger. Chest wall movement was monitored by a respiratory inductive plethysmograph (Respitrace, Ambulatory Monitoring Equipment, Ardsley, NY). The entire record was manually scored for sleep stage, apnea type, and duration by registered polysomnographic technologists (RPSGT) and reviewed by sleep medicine physician blinded to the results of cognitive testing. Respiratory events were scored using standard AASM criteria [23]. Apnea was defined as cessation of airflow for >10 s and hypopnea were defined as a 30% decrease in airflow for >10 s associated with arousal or a 4% desaturation. The oxygen desaturation index (ODI) was based upon a 4% desaturation with mean Spo2 and the duration that Spo2 was less than 90% indexed to the total recording time (TRT; time between “lights off” and “lights on”).

2.2.2. Home sleep apnea testing (HSAT)

HSAT was performed with either the Remmers Sleep Recorder (Sagatech, Calgary, AB, Canada), ApneaLink Air (Resmed, San Diego, CA, USA) or the Apnea Risk Evaluation System (ARES, SleepMed, Kennesaw, GA, USA). All three HSAT systems have been validated against PSG [24–26] and record respiratory airflow via a nasal cannula pressure transducer, snoring via a microphone and sleep position (supine/not supine) from an accelerometer and arterial oxyhemoglobin saturation using pulse oximetry (Spo2). The Spo2 signal was recorded at 1 Hz or higher and analyzed using proprietary algorithms. For all HSATs, the ODI was calculated as the number of times Spo2 decreased by ≥4% divided by the total time of oximetry recording. Similarly, mean Spo2 and the duration of Spo2<90% were indexed to the total recording time. All HSATs were interpreted by a sleep physician. In an effort to harmonize the OSA severity measures, we indexed them to the total recording time of the PSGs, the same denominator used to calculate these indices in HSAT.

2.3. Sleep questionnaire

The same comprehensive survey instrument was used by all participating centers. The questionnaire included questions related to demographics, medical history, lifestyle comorbidities, medications, family history of medical disease, sleep schedule, sleep related symptoms and co-exiting sleep disorders (including restless legs syndrome and insomnia). Daytime sleepiness was assessed with the Epworth Sleepiness Scale (ESS) [27]. The Pittsburgh Sleep Quality Index (PSQI) was used to assess sleep quality [28]. Additional questionnaire details including the RLS and insomnia criteria are provided in the online data supplement [29,30]. The specific questions that were used to determine patient symptom subtype are also provided in the online data supplement. The selection of these questions was made by matching questions from our questionnaire to those of previous publications identifying OSA symptom subtypes (Table 1)[7–10].

Table 1.

Questions from the CSCN Questionnaire that were used to produce the symptom subtypes using LCA methods.

During the first half hour after having awakened in the morning, how tired do you feel?
How often do you feel tired or fatigued after your sleep?
Over the past month, how likely are you to doze off or fall asleep when sitting down and talking to someone?
Over the past month, how likely are you to doze off or fall asleep when watching TV?
Have you ever nodded off or fallen asleep while driving a vehicle?
Do you have the following respiratory comorbidities - Nasal Congestion and/or Sinusitis?
Do you have difficulty falling asleep?
Do you have problems waking up too early?
Do you have difficulty staying asleep?
 Do you feel too hot when trying to sleep at night?
 Do you feel you cannot breathe comfortably?
During the last month on how many nights or days per week have you been told you had the following? Your breathing stops or you choke or struggle for breath
 Do you snore?
 Do you have unpleasant sensations (discomfort, creepy crawly, or tingly feeling) in your legs combined with an urge or need to move your legs?
Epworth Sleepiness Scale score.
Open in a new tab

2.4. Cognitive testing

The cognitive test battery assessed global cognitive function, episodic memory and information processing speed. The Montreal Cognitive Assessment Tool [31] was used to assess global cognitive function. The Rey Auditory Verbal Learning Test [32] and Wechsler Adult Intelligence Scale-fourth Edition Digit–Symbol Coding subtest [33] were used to assess episodic memory and information processing speed, respectively. Cognitive tests were administered and scored by trained research personnel using standardized methodology at all participating centers.

2.4.1. The Montreal cognitive Assessment (MoCA)

The MoCA was designed as a clinical screening instrument for mild cognitive dysfunction. It assesses different cognitive domains: attention and concentration, executive functions, memory, language, visuoconstructional skills, conceptual thinking, calculations, and orientation. The MoCA takes approximately 10 min to administer [31].

2.4.2. The Rey Auditory Verbal Learning Test (RAVLT)

The RAVLT is designed to evaluate a wide variety of function, including immediate memory, new verbal learning, susceptibility to interference (proactive and retroactive), retention of information after a period of time, and memory recognition. The RAVLT Delayed Recall component is particularly effective in identifying dementia and Alzheimer’s disease [32]. When completing the RAVLT, participants are given a list of 15 unrelated words repeated over five different trials in which they are asked to repeat them. Another list of 15 unrelated words are given and the participant must again repeat the original list of 15 words and then again after 20 min. The RAVLT takes approximately 10–15 min to administer (not including the 20-min interval for the final delayed recall of the list of 15 words) [32].

2.4.3. The Wechsler Adult Intelligence Scale-fourth edition (WAIS-IV) Digit—Symbol coding subtest (DSC)

The DSC is a validated and sensitive measure of cognitive dysfunction impacted by many domains. Good performance on the DSC requires intact motor speed, attention, and visuoperceptual functions, including scanning and the ability to write or draw (ie, basic manual dexterity). Performance might also be affected by associative learning [34]. The DSC is a paper-and-pencil cognitive test presented on a single sheet of paper that requires a subject to match symbols to numbers according to a key located on the top of the page. The subject copies the symbol into spaces below a row of numbers. The number of correct symbols within the allowed time (120 s) constitutes the score [33].

2.5. Data Analysis

MCI was operationally defined as a total MoCA score<26 as per standard protocol [range: 0–30] [28]. RAVLT delayed free recall and delayed recognition and the DSC scores were expressed as z-scores relative to age-matched norms [35–37].

2.6. Statistical analyses

A latent class analysis (LCA) was performed among patients with any OSA (ODI>5)using 14 symptom questions plus the Epworth Sleepiness Scale (ESS), reflecting questions similar to prior publications on symptom clusters in moderate-severe OSA [7–10]. The 14 questions used to perform the LCA are shown in Table 1. The optimal number of clusters was defined based on the Bayesian information criterion (BIC) value and the clinical relevance of the resulting symptom subtypes. BIC is a criterion for model selection based, in part, on the likelihood function and including a penalty term related to the number of model parameters to avoid overfitting. Lower BIC values indicate better model fit. Thus, the optimal cluster solution was defined as that with the minimum BIC value. We then examined the symptom characteristics of the resulting subtypes. If reasonably distinct clinical interpretations were observed for the optimal solution based on BIC, this was chosen as the final number of subtypes. Otherwise, we examined the symptom characteristics for one fewer and one additional subtype to determine if clearer clinical interpretations emerged.

After identifying the optimal number of subtypes, we evaluated differences among subtypes and between subtypes and a reference group of patients without OSA (ODI<5), for three neurocognitive outcomes – the MoCA total score, normalized RAVLT delayed free recall scores, and DSC scores. Analyses were performed using analysis of variance (ANOVA; unadjusted) or analysis of covariance (ANCOVA; adjusted for potential confounders including age, BMI, sex, ODI, education, hypertension, diabetes, COPD, and prior heart disease) applied via linear regression models, with subtype as a categorical predictor (see Table 4). If a significant difference among subtypes was observed, we performed pairwise comparisons between subtypes or between subtypes and the non-OSA group. Statistical significance was based on a P-value less than 0.05. Statistical analyses were performed using Statistical Analysis System (SAS) software (v9.4, Toronto, ON, Canada).

Table 4.

Multivariate linear regression with DSC Z- score as outcome.

Estimate 95% CI P-Value
Excessively Sleepy with Disturbed Sleep Subtype* 0.07 (−0.15,0.29) 0.51
Excessively Sleepy Subtype* 0.08 (−0.13,0.29) 0.47
Minimally Symptomatic Subtype* 0.27 (0.06,0.49) 0.01
Disturbed Sleep Subtype* 0.04 (−0.16,0.25) 0.67
Male −0.37 (−0.48, −0.25) <0.01
Age 0.003 (−0.002, 0.01) 0.21
BMI −0.01 (−0.02, −0.01) <0.01
ODI (4%) 0.001 (−0.002, 0.002) 0.99
≤12 Grade Education −0.44 (−0.56, −0.31) <0.01
Hypertension −0.16 (−0.28, −0.04) 0.01
Diabetes −0.14 (−0.29, 0.001) 0.05
COPD −0.30 (−0.52, −0.08) 0.01
Prior Heart Disease −0.09 (−0.30, 0.11) 0.37
Open in a new tab

Non-OSA group is the reference.

3. Results

3.1. Baseline characteristics

A total of 1499 subjects (185 without OSA and 1314 with OSA) with suspected and available symptom questionnaire data were included. Of these subjects, 1209 had PSGs and the other 290 had HSATs. Furthermore, 526 subjects underwent split night studies in which only the first half of the nights data was used. Characteristics of the patients are shown in Table 2. Subjects had a mean (standard deviation [SD]) age of 54.1 (13.2) years, BMI of 33.5 (8.3) kg/m2, and slightly more than half were men (56.5%). The majority of participants who provided information on education-level had more than a grade 12 education (73.4%). Significant differences existed among symptom subtypes in ODI 4% (events/hour), BMI, age, ESS, sex, self-reported hypertension and diabetes (Table 2).

Table 2.

Sample characteristics according to obstructive sleep apnea symptom subtype.

Total(N = 1499) Non-OSA(N = 185) Excessively Sleepy with Disturbed Sleep(N = 247) Excessively Sleepy(N = 405) Minimally Symptomatic(N = 280) Disturbed Sleep(N = 382) P-Value
ODI 4%(Events/Hr) 23.79 (2.88) 2.10 (5.11) 33.67 (29.17) 30.77 (30.94) 22.89 (23.72) 22.10 (24.24) <0.001
Age (years) 54.09 (13.19) 47.96 (14.47) 52.88 (13.54) 53.72 (12.07) 57.83 (12.75) 55.48 (12.60) <0.001
BMI 33.51 (8.25) 28.79 (6.28) 34.17 (8.58) 35.59 (8.69) 32.88 (6.98) 33.64 (8.32) <0.001
ESS* 9.03 (4.98) 8.90 (5.10) 12.26 (3.86) 13.36 (3.47) 5.12 (2.73) 5.57 (2.74) <0.001
Categorical Variables
Sex Male = 845 (56.52%) Male = 70 (37.63%) Male = 170 (69.11%) Male = 218 (54.50%) Male = 189 (67.50%) Male = 198 (51.70%) <0.01
Education Yes = 957 (75.24%) Yes = 113 (79.58%) Yes = 168 (76.02%) Yes = 262 (74.01%) Yes = 180 (75.00%) Yes = 234 (74.29%) 0.74
Hypertension Yes = 666 (45.31%) Yes = 45 (24.46%) Yes = 109 (44.31%) Yes = 193 (50.92%) Yes = 138 (49.46%) Yes = 181 (47.38%) <0.01
Diabetes Yes = 262 (17.85%) Yes = 16 (6.11%) Yes = 50 (20.24%) Yes = 84 (22.22%) Yes = 44 (15.71%) Yes = 68 (17.85%) <0.01
COPD Yes = 96 (6.56%) Yes = 13 (7.14%) Yes = 16 (6.53%) Yes = 21 (5.59%) Yes = 13 (4.66%) Yes = 33 (8.66%) 0.28
Prior Heart Disease Yes = 135 (9.20%) Yes = 7 (3.83%) Yes = 27 (10.98%) Yes = 35 (9.26%) Yes = 25 (8.99%) Yes = 41 (10.70%) 0.08
Open in a new tab
*

Continuous variables are presented as means and SD, P value from ANOVA, chi-square test comparing variables across subtypes.

*

ESS scores range from 0 to 24.

*

Education: (>12 years of formal education).

*

Prior heart disease was defined as a patient with angina, a myocardial infarction, a stent placement or a coronary artery bypass graft.

3.2. OSA symptom clusters

Clustering analysis among patients with ODI>5 (n = 1314) identified four optimal subtypes based on symptoms. Fig. 1 shows the relative proportion of each symptom and the ESS severity group across the symptom subtypes. Based on the distribution of observed symptoms, we identified three subtypes that were the same as those observed in previous literature [7–10] – the excessively sleepy (n = 405; 30.8%), disturbed sleep (n = 382; 29.1%), minimally symptomatic (n = 280; 21.3%) – and an additional subtype characterized as excessively sleepy with disturbed sleep (n = 382; 29.1%). Thus, results demonstrate the existence of the expected three symptom subtypes, as well as an additional subgroup with both excessive sleepiness and disturbed sleep.

Fig. 1.

Fig. 1.

Open in a new tab

Symptom profile of the identified obstructive sleep apnea symptom subtypes in the CSCN Cohort. The relative differences in symptom burden among subtypes are shown by the color scale, which represents the standardized (z-score) symptom proportion or Epworth Sleepiness Severity Category across groups. Brighter red indicates higher relative symptom burden. ESS = Epworth Sleepiness Scale.

Table 2 summarizes the clinical characteristics of these symptom subtypes. Both the excessively sleepy with disturbed sleep and the excessively sleepy subtypes were younger and had higher ODI compared with the other subtypes. The excessively sleepy subtype had a higher BMI than the other subtypes. The disturbed sleep had a higher proportion of women and less severe OSA when compared with the other subtypes. Although statistically significant, these differences were relatively small from a clinical standpoint. Patients in all subtypes were obese, were between 52.9 and 57.8 years of age, and had either moderate (the minimally symptomatic and disturbed sleep subtypes) or severe (the excessively sleepy and the excessively sleepy with disturbed sleep subtypes) OSA on average. This underscores the fact that patients with clinically similar disease severity and demographic characteristics present with distinct OSA subtypes.

When compared to the non-OSA group, all four subtypes had higher proportions of hypertension and diabetes. Finally, all of the subtypes except for the minimally symptomatic group had a higher proportion of prior heart disease than the non-OSA group.

3.3. Neurocognitive outcomes

The majority of participants completed the battery of cognitive tests. In total, 1270 individuals completed the MoCA; 599 (47.2%) had MoCA score below 26, indicating a high degree of MCI within our cohort. Table 3 illustrates the cognitive scores among participants by OSA symptom subtype. Cognitive outcomes were compared first between symptom subtypes and then to the non-OSA group. There was no statistically significant difference between symptom subtypes and the continuous MoCA scores after adjusting for the previously described covariates (overall p = 0.399). Results were similar in covariate adjusted analyses including the non-OSA group (overall p = 0.309). Similarly, adjusted models of the RAVLT delayed recall scores indicated no statistically significant differences among subtypes (overall p = 0.723) or when symptom subtypes were compared to the non-OSA group (overall p = 0.839). On the other hand, a statistically significant difference was found when investigating the adjusted relationship between symptom subtypes and DSC scores (overall p = 0.037). In particular, the excessively sleepy with disturbed sleep (p = 0.032), the excessively sleepy (p = 0.021) and the disturbed sleep (p = 0.007) subtypes all had lower DSC scores compared to the minimally symptomatic group. This relationship persisted in analyses including the non-OSA group (overall p = 0.044). Interestingly, the excessively sleepy with disturbed sleep (p = 0.515), the excessively sleepy (p = 0.470) and the disturbed sleep (p = 0.673) subtypes all had similar DSC scores to the non-OSA group, while the minimally symptomatic group scored higher on the DSC test than the non-OSA group (p = 0.015). The adjusted regression results for the normalized DSC scores are presented in Table 4.

Table 3.

Neurocognitive scores by obstructive sleep apnea symptom subtype.

Total(1270) Non-OSA(N = 150) Excessively Sleepy with Disturbed Sleep(N = 233) Excessively Sleepy(N = 405) Minimally Symptomatic(N = 192) Disturbed Sleep(N = 290) P-Value
MoCA (Total) 25.39 (2.88) 25.67 (2.67) 25.50 (2.93) 25.54 (2.65) 24.92 (23.32) 25.28 (2.91) 0.08
RAVLT Delayed Recall Z score −0.23 (1.09) −0.03 (1.05) −0.25 (1.17) −0.25 (1.01) −0.32 (1.01) −0.22 (1.10) 0.20
DSC Z Score −0.58 (0.99) −0.51 (0.98) −0.65 (0.95) −0.65 (0.99) −0.42 (0.98) −0.62 (1.00) 0.047
Open in a new tab
*

ESS scores range from 0 to 24.

*

P value from ANOVA, chi-square test, or Poisson regression comparing variable across subtypes.

4. Discussion

This study demonstrates the existence of four distinct symptom subtypes in Canadian patients with OSA presenting to sleep clinics in three Western Provinces. Three of the four subtypes identified in the CSCN cohort were similar to those identified in multiple prior studies [7–10], including patients characterized by disturbed sleep, minimal symptoms, and excessive sleepiness. In addition, in our cohort we identified a fourth subtype characterized by excessive sleepiness together with disturbed sleep.

All prior studies, as well as the present analysis, support the existence of three core OSA symptom subtypes related to excessive sleepiness, insomnia and a lack of reported symptoms. However, additional subtypes have also been identified in specific cohorts, as in our sample. For example, the study by Keenan and colleagues [8], using an international clinic-based sample, identified five optimal subtypes, including the core three (labelled disturbed sleep, minimal symptoms, and upper airway symptoms with sleepiness) and two additional subtypes characterized as less symptomatic (labelled upper airway symptoms dominant and sleepiness dominant). Similarly, the recent study by Mazzotti and colleagues [10], using the SHHS, identified the three subtypes of disturbed sleep, minimally symptomatic, and excessively sleepy, as well as a fourth subtype labeled moderately sleepy, which has similarities to the sleepiness dominant OSA subtype identified by Keenan et al, [8]. Thus, our observation of four OSA subtypes, including the three identified in prior studies and one additional subtype defined by both excessive sleepiness and disturbed sleep, is consistent with prior studies. Ultimately, differences in the optimal subtypes identified across different studies is likely due to specific differences in the cohorts. For example, our CSCN cohort was a sleep clinic-based sample that included patients with mild OSA, whereas prior studies identifying symptom subtypes have been restricted to patients with at least moderate OSA. Moreover, the SHHS cohort was a community-based study that included patients >40 years-old recruited from both rural and urban geographical regions; as a result, the SHHS patients were significantly older than our CSCN cohort. Ultimately, the observed consistency in the three core symptom subtypes in the current study and prior research, which includes cohort-specific questionnaires and both clinic-based and population-based samples from throughout the world, further supports the claim that these subtypes represent a robust trait in patients with OSA.

In addition to confirming the existence of the OSA subtypes in another clinical cohort, we sought to understand the neurocognitive differences among subtypes. Ultimately, symptom subtypes were not significantly associated with neurocognitive impairment as measured by the total MoCA or normalized RAVLT delayed free recall scores. There were significant differences in DSC scores by symptom subtype, with the minimally symptomatic group scoring significantly higher than both the non-OSA group and the other three subtypes. Why patients with obstructive sleep apnea without self-reported symptoms demonstrate better cognitive performance than both apneics in other symptom subtypes and non-OSA patients in our sample remains to be determined.

Several prior publications investigating community-dwelling volunteers have reported OSA to be an independent risk factor for the development of MCI and dementia [16,18–20]. Several studies in Chinese OSA patients found decrements in the visuospatial/executive function, attention, and delayed recall components of the MoCA were associated with increasing OSA severity [38,39]. In addition, a previous study from our group found a significant relationship between MCI (characterized as a MoCA total score <26) and OSA severity according to ODI. Therefore, the present results suggest that the relationship between OSA and cognitive impairment is likely mediated by other pathophysiological characteristics of OSA, such as overall hypoxemia and sleep fragmentation, as opposed to being mediated by certain OSA symptom subtypes. Nevertheless, it is also possible that the instruments used in this study to assess cognition are not sensitive enough to characterize variability in normal cognitive function, especially in subclinical cognitive impairment. Thus, further studies using more sensitive tests for cognitive ability are warranted.

This study has several limitations. First, not all patients in our cohort were diagnosed with OSA using the same methods. Patients were diagnosed by either unattended HSAT or PSG. In an effort to harmonize the OSA severity measures, we indexed them to the total recording time of the PSGs, the same denominator used to calculate these indices in HSAT. However, given total recording time is longer than total sleep time, OSA severity may be underestimated in our cohort. Second, we used a clinic-based cohort, and referral bias is a potential issue. For example, truly asymptomatic patients would be less likely to be referred to our clinic than patients who were either symptomatic or had clinical features predicting more severe disease (ie, high BMI or comorbidities). However, advantages of a clinic-based cohort include the relatively greater severity of OSA and presence of symptoms. Third, the MoCA has reduced sensitivity and specificity for detecting MCI in clinical populations apart from that on which it was validated [40,41]. Fourth, the use of both PSG and HSAT to diagnose OSA prevented the possibility of using important PSG measures to delineate OSA subtypes because several of these measures including sleep efficiency and arousal index were only available on patients who underwent PSGs [42]. Lastly, the inclusion of patients who underwent HSATs in order to diagnose or rule out OSA prompted us to use the ODI rather than the AHI as a measure of sleep apnea severity. This was a limitation of our study for two reasons. First, there was nonuniform testing means for patients in our study. Second, previous studies using similar questions to distinguish between symptom subtypes used AHI rather than ODI as a measure of OSA severity, which to some degree limits the comparability of our results.

In our study in a multi-center clinic-based cohort from Canada, we identified four distinct symptom subtypes, three of which were similar to those identified in previous studies. While our study supports them as robust traits of patients with OSA, these subtypes were not associated with neurocognitive function as assessed by a broad range of tests designed to detect mild cognitive impairment. This suggests that the neurocognitive deficits associated with OSA may not be attributable to the symptoms of disturbed sleep or sleepiness, but could bedue to neuronal damage from nocturnal hypoxemia or sleep fragmentation not translated into symptoms. Future studies are needed to determine whether these symptom clusters are associated with future adverse health events, including cognitive decline.

Supplementary Material

supplementary

Support:

CIHR (Sleep Disordered Breathing Team Grant), BC Lung Association Operating Grant, Canadian Sleep and Circadian Network CSCN.

Footnotes

CRediT authorship contribution statement

AJ Hirsch Allen: Conceptualization, Data curation, Formal analysis, Writing - original draft. Andrew E. Beaudin: Writing - original draft, Formal analysis. Nurit Fox: Data curation, Writing - review & editing, Project administration. Jill K. Raneri: Data curation, Project administration, Writing - review & editing. Robert P. Skomro: Conceptualization, Methodology, Writing - review & editing. Patrick J. Hanly: Conceptualization, Methodology, Writing - review & editing. Diego R. Mazzotti: Conceptualization, Methodology, Writing - review & editing. Brendan T. Keenan: Conceptualization, Formal analysis, Methodology, Writing - review & editing. Eric E. Smith: Conceptualization, Methodology. Sebastian D. Goodfellow: Methodology, Writing - review & editing. Najib T. Ayas: Conceptualization, Methodology, Writing - review & editing.

Conflict of interest

None.

The ICMJE Uniform Disclosure Form for Potential Conflicts of Interest associated with this article can be viewed by clicking on the following link: https://doi.org/10.1016/j.sleep.2020.05.001.

References

  • [1].Young T, Palta M, Dempsey J, et al. The occurrence of sleep-disordered breathing among middle-aged adults. N Engl J Med 1993;328(17):1230–5. [DOI] [PubMed] [Google Scholar]
  • [2].Gurubhagavatula I Consequences of obstructive sleep apnoea. Indian J Med Res 2010;131:188–95. [PubMed] [Google Scholar]
  • [3].American Academy of Sleep Medicine. International classification of sleep disorders. 3rd ed. Darien: American academy of sleep medicine; 2014. [Google Scholar]
  • [4].Engleman H, Joffe D. Neuropsychological function in obstructive sleep apnoea. Sleep Med Rev 1999;3(1):59–78. [DOI] [PubMed] [Google Scholar]
  • [5].Morsy NE, Farrag NS, Zaki NF, et al. Obstructive sleep apnea: personal, societal, public health, and legal implications. Rev Environ Health 2019;34(2):153–69. [DOI] [PubMed] [Google Scholar]
  • [6].Zinchuk AV, Jeon S, Koo BB, et al. Polysomnographic phenotypes and their cardiovascular implications in obstructive sleep apnoea. Thorax 2018;73:472–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [7].Ye L, Pien GW, Ratcliffe SJ, et al. The different clinical faces of obstructive sleep apnoea: a cluster analysis. Eur Respir J 2014;44:1600–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [8].Keenan BT, Kim J, Singh B, et al. Recognizable clinical subtypes of obstructive sleep apnea across international sleep centers: a cluster analysis. Sleep 2018;41(3):zsx214. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [9].Kim J, Keenan BT, Lim DC, et al. Symptom-based subgroups of Koreans with obstructive sleep apnea. J Clin Sleep Med 2018;14:437–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [10].Mazzotti DR, Keenan BT, Lim DC, et al. Symptom subtypes of obstructive sleep apnea predict incidence of cardiovascular outcomes. Am J Respir Crit Care Med 2019;200(4):493–506. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [11].McLachlan GJ, Peel D. Finite mixture models. New York: John Wiley & Sons, Inc.; 2000. [Google Scholar]
  • [12].12a Vermunt JK, Magidson J. Latent class cluster analysis. In: Hagenaars JA, McCutcheon AL, editors. Applied latent class Analysis. Cambridge: Cambridge University Press; 2002. p. 89–106. [Google Scholar]; 12b Pien GW, Ye L, Keenan BT, et al. Changing faces of obstructive sleep apnea: treatment effects by cluster designation in the Icelandic sleep apnea cohort. Sleep 2018;41(3):zsx201. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [13].Bucks RS, Olaithe M, Eastwood P. Neurocognitive function in obstructive sleep apnoea: a meta-review. Respirology 2013;18:61–70. [DOI] [PubMed] [Google Scholar]
  • [14].Gagnon K, Baril AA, Gagnon JF, et al. Cognitive impairment in obstructive sleep apnea. Pathol Biol 2014;62:233–40. [DOI] [PubMed] [Google Scholar]
  • [15].Yaffe K, Laffan AM, Harrison SL, et al. Sleep-disordered breathing, hypoxia, and risk of mild cognitive impairment and dementia in older women. J Am Med Assoc 2011;306:613–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [16].Albert MS, DeKosky ST, Dickson D, et al. The diagnosis of mild cognitive impairment due to Alzheimer–s disease: recommendations from the National Institute on Aging-Alzheimer–s Association workgroups on diagnostic guidelines for Alzheimer–s disease. Alzheimers Dement 2011;7:270–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [17].Chang WP, Liu ME, Chang WC, et al. Sleep apnea and the risk of dementia: a population-based 5-year follow-up study in Taiwan. PloS One 2013;8:e78655. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [18].Leng Y, McEvoy CT, Allen IE, et al. Association of sleep-disordered breathing with cognitive function and risk of cognitive impairment: a systematic review and meta-analysis. JAMA Neurology 2017;74:1237–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [19].Sharma RA, Varga AW, Bubu OM, et al. Obstructive sleep apnea severity affects amyloid burden in cognitively normal elderly. A longitudinal study. Am J Respir Crit Care Med 2018;197:933–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [20].Sforza E, Roche F. Sleep apnea syndrome and cognition. Front Neurol 2012;3:87. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [21].Quan SF, Chan CS, Dement WC, et al. The association between obstructive sleep apnea and neurocognitive performance–the Apnea Positive Pressure Long-term Efficacy Study (APPLES). Sleep 2011;34:303–314B. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [22].Berry R, Budhiraja R, Gottlieb DJ, et al. Rules for scoring respiratory events in sleep: update of the 2007 AASM manual for the scoring of sleep and associated events. Deliberations of the Sleep Apnea Definitions Task Force of the American Academy of Sleep Medicine. J Clin Sleep Med 2012;8(5):597–619. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [23].Issa FG, Morrison D, Hadjuk E, et al. Digital monitoring of sleep-disordered breathing using snoring sound and arterial oxygen saturation. Am Rev Respir Dis 1993;148:1023–9. [DOI] [PubMed] [Google Scholar]
  • [24].Westbrook PR, Levendowski DJ, Cvetinovic M, et al. Description and validation of the apnea risk evaluation system: a novel method to diagnose sleep apneahypopnea in the home. Chest 2005;128:2166–75. [DOI] [PubMed] [Google Scholar]
  • [25].Erman MK, Stewart D, Einhorn D, et al. Validation of the ApneaLink™ for the screening of sleep apnea: a novel and simple single-channel recording device. J Clin Sleep Med 2007;3:387–92. [PMC free article] [PubMed] [Google Scholar]
  • [26].Johns MW. A new method for measuring daytime sleepiness: the Epworth sleepiness scale. Sleep 1991;14:540–5. [DOI] [PubMed] [Google Scholar]
  • [27].Buysse DJ, Reynolds CF 3rd, Monk TH, et al. The Pittsburgh Sleep Quality Index: a new instrument for psychiatric practice and research. Psychiatr Res 1989;28:193–213. [DOI] [PubMed] [Google Scholar]
  • [28].Allen RP, Picchietti D, Hening WA, et al. Restless legs syndrome: diagnostic criteria, special considerations, and epidemiology. Sleep Med 2003;4:101–19. [DOI] [PubMed] [Google Scholar]
  • [29].Bastien CH, Vallières A, Morin CM. Validation of the Insomnia Severity Index as an outcome measure for insomnia research. Sleep Med Clin 2001;2:297–307. [DOI] [PubMed] [Google Scholar]
  • [30].Nasreddine ZS, Phillips NA, Bédirian V, et al. The montreal cognitive assessment, MoCA: a brief screening Tool for mild cognitive impairment. J Am Geriatr Soc 2005;53:695–9. [DOI] [PubMed] [Google Scholar]
  • [31].Rey A L–examen psychologique dans les cas d–encéphalopathie traumatique. (Les problems.). [The psychological examination in cases of traumatic encephalopathy. Problems.]vol. 28. Archives de Psychologie; 1941. p. 215–85. [Google Scholar]
  • [32].Wechsler D Wechsler adult intelligence scale –. 4th ed. San Antonio, TX: Pearson; 2008. [Google Scholar]
  • [33].Judith Jaeger. Digit symbol substitution test: the case for sensitivity over specificity in neuropsychological testing. J Clin Psychopharmacol 2018;38:513–9. 5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [34].Strauss E, Sherman EMS, Spreen O. A compendium of neuropsychological tests: administration, norms, and commentary. 3rd ed. New York, NY, US: Oxford University Press; 2006. [Google Scholar]
  • [35].Gale SD, Baxter L, Connor DJ, et al. Sex differences on the Rey auditory verbal learning test and the brief visuospatial memory test-revised in the elderly: normative data in 172 participants. J Clin Exp Neuropsychol 2007;29:561–7. [DOI] [PubMed] [Google Scholar]
  • [36].Smith EE, O–Donnell M, Dagenais G, et al. Early cerebral small vessel disease and brain volume, cognition, and gait. Ann Neurol 2015;77:251–61. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [37].Chen R, Xiong KP, Huang JY, et al. Neurocognitive impairment in Chinese patients with obstructive sleep apnoea hypopnoea syndrome. Respirology 2011;16:842–8. [DOI] [PubMed] [Google Scholar]
  • [38].Wen X, Wang N, Liu J, et al. Detection of cognitive impairment in patients with obstructive sleep apnea hypopnea syndrome using mismatch negativity. Neural regeneration research 2012;7:1591–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [39].Gagnon K, Baril A-A, Montplaisir J, et al. Detection of mild cognitive impairment in middle-aged and older adults with obstructive sleep apnoea. Eur Respir J 2018;52. [DOI] [PubMed] [Google Scholar]
  • [40].Villeneuve S, Pepin V, Rahayel S, et al. Mild cognitive impairment in moderate to severe COPD: a preliminary study. Chest 2012;142:1516–23. [DOI] [PubMed] [Google Scholar]
  • [41].Biundo R, Weis L, Bostantjopoulou S, et al. MMSE and MoCA in Parkinson–s disease and dementia with Lewy bodies: a multicenter 1-year follow-up study. J Neural Transm 2016;123:431–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [42].Zinchuk A, Yaggi HK. Phenotypic subtypes of OSA. Chest 2019;157(2):403–20. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

supplementary
NIHMS1847810-supplement-supplementary.docx (28.5KB, docx)

RESOURCES