Abstract
Background
Impaired daytime vigilance is an important consequence of OSA, but several studies have reported no association between objective measurements of vigilance and the apnea-hypopnea index (AHI). Notably, the AHI does not quantify the degree of flow limitation, that is, the extent to which ventilation fails to meet intended ventilation (ventilatory drive).
Research Question
Is flow limitation during sleep associated with daytime vigilance in OSA?
Study Design and Methods
Nine hundred ninety-eight participants with suspected OSA completed a 10-min psychomotor vigilance task (PVT) before same-night in-laboratory polysomnography. Flow limitation frequency (percent of flow-limited breaths) during sleep was quantified using airflow shapes (eg, fluttering and scooping) from nasal pressure airflow. Multivariable regression assessed the association between flow limitation frequency and the number of lapses (response times > 500 ms, primary outcome), adjusting for age, sex, BMI, total sleep time, depression, and smoking status.
Results
Increased flow limitation frequency was associated with decreased vigilance: a 1-SD (35.3%) increase was associated with 2.1 additional PVT lapses (95% CI, 0.7-3.7; P = .003). This magnitude was similar to that for age, where a 1-SD increase (13.5 years) was associated with 1.9 additional lapses. Results were similar after adjusting for AHI, hypoxemia severity, and arousal severity. The AHI was not associated with PVT lapses (P = .20). In secondary exploratory analysis, flow limitation frequency was associated with mean response speed (P = .012), median response time (P = .029), fastest 10% response time (P = .041), slowest 10% response time (P = .018), and slowest 10% response speed (P = .005).
Interpretation
Increased flow limitation during sleep was associated with decreased daytime vigilance in individuals with suspected OSA, independent of the AHI. Flow limitation may complement standard clinical metrics in identifying individuals whose vigilance impairment most likely is explained by OSA.
Key Words: apnea-hypopnea index, arousal severity, automated, desaturation severity, flow limitation, lapses, OSA, polysomnography, respiratory events, sleep and breathing, sleepiness, upper airway, vigilance
FOR EDITORIAL COMMENT, SEE PAGE 769
Take-home Points.
Study Question: Is flow limitation during sleep associated with daytime vigilance in individuals with suspected OSA?
Results: Increased flow limitation frequency during sleep was associated with decreased daytime vigilance, independent of the apnea-hypopnea index and other potential confounders and mediators.
Interpretation: Flow limitation could be used alongside apnea-hypopnea index and patient symptoms to guide decisions on who should and should not be treated for impaired vigilance.
It is estimated that worldwide almost 1 billion adults aged 30 to 69 years have OSA, and of these, almost 425 million have moderate to severe OSA, for which treatment generally is recommended.1 Key symptoms of OSA are excessive daytime sleepiness and impaired vigilance.2 Vigilance is the ability to maintain focused attention over a prolonged period, and impaired vigilance can be the cause of major accidents resulting in loss of life.3 Although substantial evidence exists that patients with OSA have impaired daytime vigilance,2,4,5 the standard clinical measurement of OSA severity, the apnea-hypopnea index (AHI), is poorly associated with objective measurements of vigilance such as psychomotor vigilance task (PVT) performance.
Clinical diagnosis and treatment of OSA are based on the AHI, which quantifies the total number of apneas and hypopneas per hour during sleep.6 Only two studies4,5 have reported a statistically significant association between the AHI and PVT performance (Pearson r = 0.42 and 0.14, respectively), whereas several studies found no significant association.7, 8, 9, 10, 11, 12 One explanation for this is that although the AHI quantifies the frequency of apneas and hypopneas, it does not quantify the severity of obstruction and desaturation events13 or the degree of airway flow limitation, that is, the extent to which ventilation fails to meet intended ventilation (ventilatory drive).6 Previous work demonstrated that the presence of sustained inspiratory flow limitation is associated with decreased PVT performance14 and that elimination of flow limitation improves vigilance.15 However, no studies to date have quantified the relationship between flow limitation as a continuous variable and PVT-quantified vigilance in OSA, nor have any disentangled its effect from hypoxemia severity or arousal severity.
The aim of this study was to investigate and quantify the association between overnight flow limitation and daytime vigilance in a cohort of 998 participants with suspected OSA who were studied with PVT and diagnostic polysomnography. We used two recently published variations of a tool that quantifies the frequency16 and severity17 of flow limitation on a breath-by-breath basis. Our primary hypothesis was that the overall burden of flow limitation during sleep is associated with vigilance impairment assessed using PVT lapse count (response times > 500 ms) and that this relationship is independent of the AHI and other domains of OSA pathologic characteristics such as hypoxemia severity and arousal severity. Secondary exploratory analyses were performed (1) to examine the associations between overnight flow limitation and other PVT outcome metrics and Epworth Sleepiness Scale (ESS) score18 and (2) to determine whether flow limitation during respiratory events and other descriptors of respiratory event severity are associated more strongly with PVT performance.
Study Design and Methods
Study Design and Participants
This was a retrospective cross-sectional analysis of 1,292 consecutive participants with suspected OSA who completed a PVT, ESS questionnaire, and in-laboratory diagnostic polysomnography at the Sleep Disorders Centre at the Princess Alexandra Hospital in Brisbane, Australia, from 2015 through 2018. This study was approved by the Metro South Health Human Research Ethics Committee (Identifiers: HREC/16/QPAH/021 and LNR/2019/QMS/54313). Participants were excluded if they had fewer than 3 h of total sleep time, if their sleep study was missing data or contained poor-quality data, or if they had a neuromuscular disorder or visual impairment (Fig 1). Nine hundred ninety-eight participants remained for analysis.
Figure 1.
Flow diagram showing exclusion criteria applied to the diagnostic sleep studies performed at the Princess Alexandra Hospital. PVT = psychomotor vigilance task.
Polysomnographic Recordings
The Grael acquisition system and ProFusion PSG 4.0 software (Compumedics) were used to record and score the polysomnograms. The following signals were used in this analysis: nasal pressure (direct-coupled amplified and unfiltered, 128 Hz, absence of filtering confirmed by expert visual inspection), pulse oximetry (16 Hz), and EEG (1,024 Hz) using C3-M2 and C4-M1. Polysomnograms were scored according to the American Academy of Sleep Medicine 2012 guidelines.19 Hypopneas were scored based on the following criteria: ≥ 30% drop in nasal pressure amplitude for ≥ 10 s and accompanied with a ≥ 3% oxygen desaturation or arousal. AHI was defined as the average number of apneas and hypopneas per hour during sleep (events/h). For split night CPAP studies, only the period before the onset of titration was used.
Psychomotor Vigilance Task
The PVT is an objective measurement of vigilance and is simple, brief, and free of learning effects with repeated administration.20,21 The PVT measures sustained attention and requires participants to press a button when they see a visual stimulus.22 The Psychology Experiment Building Language 10-min PVT23 was administered between 7 pm and 9 pm to participants before the commencement of polysomnography using an ASUS Transformer Pad with keyboard. Participants were instructed to monitor the screen and press the response button with the thumb or index finger of their dominant hand as soon as they saw the stimulus. One hundred twenty-one single trial stimuli were presented, and the time between successive stimuli varied randomly between 2 and 10 s. The first trial was considered instructional and was excluded. No maximum time limit per trial was used; however, response times were capped at 30 s as described previously.22 Responses of < 100 ms were considered false starts and were excluded. The number of lapses in attention (response times > 500 ms) was the primary outcome measure as recommended.22 Secondary outcome measures were mean response speed, median response time, fastest 10% response time, slowest 10% response time, and slowest 10% response speed.
Quantification of Flow Limitation
We used two validated algorithms to assess the degree of flow limitation objectively on a breath-by-breath basis using the nasal airflow signal. The primary exposure variable was flow limitation frequency, defined as the percentage of breaths during sleep classified as being certain flow-limited.16 This classification was based on multiple flow shape characteristics (eg, fluttering and scooping) that describe nonrounded inspiratory and expiratory flow.16 We also assessed flow limitation severity, which quantifies the mismatch between measured ventilation and intended ventilation (ventilatory drive).17 Specifically, flow limitation severity = (1 – (ventilation / drive)) × 100. A score of 50% indicates that ventilation was one-half of the intended level, and a score of 0% indicates a perfectly patent airway. Figure 2 demonstrates these quantifications in an example participant. The median value of flow limitation severity provided a single summary metric.17 All breaths during sleep, including those that occurred within scored respiratory events, were included in analyses. Breaths within obstructive apneas—scored apneas plus any unscored periods of ventilation of < 10% of the local mean value (7-min sliding window)—were imputed at the median respiratory rate and were attributed a certain flow-limited label and a flow limitation severity of 99.9%. Breaths were excluded if they were within arousals or longer than Q3 + 1.5 × IQR (where Q3 is quartile 3 and IQR is the interquartile range of a participant’s breath durations).
Figure 2.
Demonstration of the two validated algorithms used to quantify airway flow limitation on a breath-by-breath basis. A-C, Flow limitation throughout the polysomnography for one participant (A) demonstrating an unobstructed airway when awake (B) and flow limitation when asleep (C). These two periods are surrounded by dashed boxes, and vertical lines denote the start of breaths. Note the characteristically different airflow shapes being used to estimate flow limitation using continuous (flow limitation severity) and discrete (flow limitation certainty) scales. FL = flow limited; NREM = non-rapid eye movement; REM = rapid eye movement.
Other Quantifications of OSA Severity
To quantify the likely downstream physiologic sequelae of flow limitation, we also quantified hypoxemia and arousal aspects of sleep-disordered breathing. Hypoxemia severity was quantified using the desaturation severity parameter, which is the cumulative area of decreased oxygen saturation signal during scored desaturation events normalized by total sleep time.12 The units of desaturation severity are percent; for example, 2% desaturation severity represents an average oxygen saturation reduction of 2% less than baseline during sleep. Desaturations scored in wake epochs except sleep-wake and wake-sleep transition epochs were excluded.
Arousal severity was quantified by incorporating the intensity24 and cumulative duration25 of arousals into a single metric. Arousal intensity uses C3-M2 and C4-M1 EEG signals to quantify scored arousals using a scale of 0 to 9, where 9 is most intense.24 The intensity and duration of arousals were multiplied, summed together, then normalized by total sleep time. The units of arousal severity are arousal scale; for example, an arousal severity of 1 represents an average arousal intensity of 1 arousal scale during sleep. Raw signal and power spectral density plots of C3-M2 and C4-M1 signals were inspected visually for signal quality. One hundred four participants (10.4%) had one poor-quality signal and 103 participants (10.3%) had two poor-quality signals and were excluded from computing arousal intensity. Refer to e-Appendix 1 for more information.
An additional exploratory analysis was performed to investigate the relationship between PVT performance and several measures of respiratory event severity: (1) standard clinical event frequency (hypopnea index and apnea index); (2) average proportion of certain flow-limited breaths and average flow limitation severity during scored respiratory events; (3) respiratory event duration, depth, and lost ventilation using minute ventilation as a fraction of eupnea26; (4) change in heart rate resulting from respiratory events27; (5) respiratory event-related desaturation duration, depth, area, and severity; and (6) respiratory event-related arousal severity and average arousal intensity. Refer to e-Appendix 1 and e-Figure 1 for more information.
Statistical Analyses
Data processing and statistical analyses were performed in MATLAB version R2023a software (Mathworks) and R version 4.2.2 software (R Foundation for Statistical Computing). All continuous variables (including PVT lapses) were transformed to be normally distributed using the two-parameter Box-Cox transformation. Variables then were standardized between 0 and 1 by subtracting the minimum value then dividing by the range. Multivariable linear regression models were developed to quantify the relationship between the number of PVT lapses and predictor variables. Models were adjusted for age, sex, and BMI (model 0, underadjusted); were adjusted further for total sleep time, smoking status, and depression (model 1, primary model); and were adjusted further for hypertension, COPD, emphysema, heart failure, stroke, type 2 diabetes, and ESS score (model 2, overadjusted). The significance level threshold was set at P < .05. Plots of model residuals were inspected visually for normality, zero mean, and constant variance. Refer to e-Appendix 1 for more information.
Results
Participant characteristics are summarized in Table 1.
Table 1.
Demographic, Polysomnographic, and PVT Characteristics of All Participants and Participants Stratified by Sex
| Variable | All Participants | Male | Female |
|---|---|---|---|
| Participants, male | 998 (52.9) | 528 | 470 |
| Age, y | 54.4 ± 13.5 | 55.2 ± 13.7 | 53.6 ± 13.3 |
| BMI, kg/m2 | 34.0 ± 9.8 | 32.8 ± 8.2 | 35.5 ± 11.4a |
| Active tobacco use | 151 (15.1) | 80 (15.2) | 71 (15.1) |
| Depression | 204 (20.4) | 75 (14.2) | 129 (27.4)b |
| Hypertension | 394 (39.5) | 206 (39.0) | 188 (40.0) |
| COPD | 93 (9.3) | 50 (9.5) | 43 (9.1) |
| Type 2 diabetes | 214 (21.4) | 113 (21.4) | 101 (21.5) |
| Emphysema | 1 (0.1) | 1 (0.2) | 0 (0.0) |
| Heart failure | 38 (3.8) | 23 (4.4) | 15 (3.2) |
| Stroke | 34 (3.4) | 19 (3.6) | 15 (3.2) |
| Total sleep time, h | 5.2 ± 1.1 | 5.1 ± 1.1 | 5.4 ± 1.2a |
| Rapid eye movement sleep, % | 16.9 ± 8.4 | 16.8 ± 8.3 | 17.0 ± 8.6 |
| Non-rapid eye movement sleep, % | 82.7 ± 7.6 | 82.8 ± 7.5 | 82.5 ± 7.7 |
| Apnea-hypopnea index, events/h | 17.5 ± 29.0 | 22.5 ± 32.5 | 12.9 ± 23.0a |
| Flow limitation frequency, % | 22.8 ± 35.3 | 28.6 ± 37.2 | 17.2 ± 30.6a |
| Flow limitation severity, % | 33.6 ± 18.1 | 36.6 ± 20.0 | 30.4 ± 15.3a |
| Desaturation severity, % | 0.3 ± 1.5 | 0.5 ± 2.0 | 0.2 ± 0.9a |
| Arousal severity (arousal scale) | 0.2 ± 0.2 | 0.3 ± 0.2 | 0.2 ± 0.1a |
| Epworth sleepiness scale score | 9.3 ± 6.5 | 9.0 ± 6.2 | 9.6 ± 6.8 |
| PVT | |||
| Duration, min | 16.4 ± 0.8 | 16.3 ± 0.8 | 16.5 ± 0.8a |
| Lapses,c No. | 14.8 ± 33.9 | 10.3 ± 26.4 | 21.4 ± 39.5a |
| Mean response speed, 1/s | 2.6 ± 0.4 | 2.7 ± 0.4 | 2.4 ± 0.4a |
| Median response time, ms | 382.1 ± 103.8 | 363.1 ± 81.4 | 407.5 ± 122.6a |
| Fastest 10% response time, ms | 299.8 ± 54.8 | 290.3 ± 43.5 | 312.0 ± 65.5a |
| Slowest 10% response time, ms | 673.4 ± 463.6 | 627.0 ± 377.9 | 734.9 ± 557.3a |
| Slowest 10% response speed, 1/s | 1.6 ± 0.5 | 1.7 ± 0.5 | 1.4 ± 0.5a |
Data are presented as No., No. (%), or mean ± SD for back-transformed variables. PVT = psychomotor vigilance task.
P < .05 between male and female groups using Welch t test.
P < .05 between male and female groups using the χ2 test.
Response times > 500 ms.
Relationship Between the Domains of OSA Pathologic Characteristics and OSA Severity
Flow limitation frequency and AHI were correlated moderately (Pearson r = 0.66) (Fig 3). However, it also was observed that two participants with almost identical demographic and polysomnographic characteristics showed substantially different flow limitation frequency and PVT performance. Variables quantifying the three separate domains of OSA pathologic characteristics (flow limitation, hypoxemia severity, and arousal severity) were correlated with each other no more than moderately (Pearson r ≤ 0.68) (Fig 4, e-Fig 2).
Figure 3.
A, Scatterplot illustrating a moderate correlation (Pearson r = 0.66; P < .001) between flow limitation frequency during sleep and the apnea-hypopnea index (AHI) for 998 participants (each participant is represented as a single point, and a regression line is shown in blue). This suggests that flow limitation frequency may capture different pathologic features than the AHI. B, For example, P1 shows a lower AHI and higher proportion of certain and possible flow-limited breaths than P2. C, P1 shows more PVT lapses than P2. Both participants were male, of similar age, and had similar BMI, total sleep time, arousal index, arousal severity, and desaturation severity. Therefore, increased flow limitation may explain why P1 showed worse PVT performance than P2. P1 = participant 1; P2 = participant 2; PVT = psychomotor vigilance task.
Figure 4.
Histograms and scatterplots showing demographic, predictor, and outcome variables (transformed and standardized, with back-transformed tick values). Histograms along the diagonal show the distribution of each variable. Scatterplot regression lines are shown in blue. Squares are colored and numbered based on Pearson correlation coefficients (numbers in boldface represent P < .05). Axes scales are nonlinear because of transformed data being plotted with back-transformed tick marks. Age is in years and BMI is in kilograms per square meter. AHI = apnea-hypopnea index; ArS = arousal severity (arousal scale); DSv = desaturation severity (%); ESS = Epworth Sleepiness Scale; FLF = flow limitation frequency (%); FLS = flow limitation severity (%); NREM = non-rapid eye movement (%); PVT-L = psychomotor vigilance task lapses (No.); TST = total sleep time (h).
Overall Burden of Flow Limitation During Sleep and PVT Lapses
The percentage of flow-limited breaths during sleep was a significant predictor of PVT lapses (Fig 5). Using model 1, a 1-SD (35.3%) increase in flow limitation frequency was associated with 2.1 additional PVT lapses (95% CI, 0.7-3.7; P = .003). These results were similar for flow limitation severity (1.8 lapses; 95% CI, 0.3-3.3; P = .020) and desaturation severity (2.0 lapses; 95% CI, 0.4-3.7; P = .015). A 1-SD increase in age (13.5 years) was associated with 1.9 additional lapses (95% CI, 0.5-3.5; P = .008). AHI and arousal severity were not associated significantly with PVT lapses (P ≥ .20). Relative to models 0 and 1, adjustment for additional covariates in model 2 reduced the effect size of flow limitation frequency and severity.
Figure 5.
A, Forest plot showing the association between 1-SD increase in predictor variables and increased number of psychomotor vigilance task (PVT) lapses. Vertical ticks represent coefficient estimates and horizontal bars represent 95% CIs. Linear regression models were adjusted for age, sex, and BMI (model 0, red bars); further adjusted for total sleep time, smoking status, and depression (model 1, blue bars); and further adjusted for hypertension, COPD, emphysema, heart failure, stroke, type 2 diabetes, and ESS score (model 2, gray bars). All analyses had 998 participants except for those with arousal severity, which had 895 participants with useable EEG signals. B, Graph showing PVT lapses as a function of flow limitation frequency for the average participant using estimates from model 1 (shading within dashed lines represents 95% CI). ESS = Epworth Sleepiness Scale.
Flow limitation frequency remained a statistically significant predictor when AHI and arousal severity were included separately as covariates (P ≤ .006) (Fig 6). Flow limitation frequency remained a significant predictor when all of AHI, desaturation severity, and arousal severity were included as covariates (P = .017). Flow limitation frequency was borderline statistically significant when desaturation severity was included as a covariate (P = .051). Of these models, the model with flow limitation frequency and no covariates showed the best fit.
Figure 6.
Forest plot showing the association between 1-SD increase in predictor variables and increased number of psychomotor vigilance task (PVT) lapses for models with combinations of flow limitation frequency and other predictors. Vertical ticks represent coefficient estimates and horizontal bars represent 95% CIs. Linear regression models were adjusted for age, sex, BMI, total sleep time, smoking status, and depression. Based on AICc, the model with flow limitation frequency and no covariates had the best fit, and the model with flow limitation frequency, apnea-hypopnea index, desaturation severity, and arousal severity had the worst fit. All analyses had 998 participants except for those with arousal severity, which had 895 participants with useable EEG signals. All AICc values were calculated using only these 895 participants because of the effect of sample size on AICc. The model with the best fit (lowest AICc) has a ΔAICc of 0, and all other ΔAICc values were calculated based on the difference from its AICc. AICc = corrected Akaike information criterion; ΔAICc = difference in corrected Akaike information criterion from the model with the best fit.
Secondary Exploratory Analyses
Overnight flow limitation frequency was associated significantly with ESS score (P = .007) and mean response speed (P = .012), median response time (P = .029), fastest 10% response time (P = .041), slowest 10% response time (P = .018), and slowest 10% response speed (P = .005) (Table 2). Similar effect sizes were observed with respiratory event-related flow limitation metrics compared with sleep burden flow limitation metrics (Fig 7). Respiratory event-related desaturation severity, depth, and area were associated significantly with PVT lapses (P ≤ .039). No other measures of respiratory event severity were associated significantly with PVT lapses. This includes hypopnea index, apnea index, arousal intensity, event duration, event depth, lost ventilation, change in heart rate, and respiratory event-related arousal severity.
Table 2.
Associations Between Psychomotor Vigilance Task and Epworth Sleepiness Scale Outcome Variables and Predictor Variables
| Variable | Flow Limitation Frequency | Flow Limitation Severity Median | Apnea-Hypopnea Index | Desaturation Severity | Arousal Severity |
|---|---|---|---|---|---|
| No. of lapses | |||||
| P value | .003 | .020 | .202 | .015 | .706 |
| Δa | 2.143 | 1.746 | 0.957 | 1.981 | –0.279 |
| 95% CI of change | 0.704 to 3.676 | 0.271 to 3.325 | –0.508 to 2.515 | 0.373 to 3.709 | –1.789 to 1.195 |
| Mean response speed, 1/s | |||||
| P value | .012 | .067 | .338 | .025 | .661 |
| Δa | –0.039 | –0.030 | –0.016 | –0.040 | 0.007 |
| 95% CI of change | –0.069 to –0.009 | –0.062 to 0.002 | –0.049 to 0.017 | –0.074 to –0.005 | –0.025 to 0.040 |
| Median response time, ms | |||||
| P value | .029 | .106 | .371 | .029 | .675 |
| Δa | 5.316 | 4.106 | 2.301 | 6.019 | –1.064 |
| 95% CI of change | 0.523 to 10.294 | –0.857 to 9.263 | –2.747 to 7.493 | 0.594 to 11.684 | –6.162 to 3.957 |
| Fastest 10% response time, ms | |||||
| P value | .040 | .370 | .524 | .098 | .285 |
| Δa | 2.798 | 1.266 | 0.916 | 2.547 | –1.535 |
| 95% CI of change | 0.121 to 5.566 | –1.506 to 4.108 | –1.916 to 3.804 | –0.464 to 5.671 | –4.426 to 1.276 |
| Slowest 10% response time, ms | |||||
| P value | .018 | .027 | .781 | .086 | .879 |
| Δa | 22.079 | 21.504 | 2.687 | 18.012 | –1.463 |
| 95% CI of change | 3.679 to 41.559 | 2.347 to 41.835 | –16.626 to 22.308 | –2.479 to 39.838 | –20.901 to 17.803 |
| Slowest 10% response speed, 1/s | |||||
| P value | .004 | .013 | .476 | .026 | .861 |
| Δa | –0.052 | –0.047 | –0.014 | –0.046 | 0.003 |
| 95% CI of change | –0.087 to –0.017 | –0.083 to –0.010 | –0.052 to 0.024 | –0.085 to –0.005 | –0.034 to 0.041 |
| Epworth sleepiness scale score | |||||
| P value | .007 | .018 | .008 | .022 | .049 |
| Δa | 0.533 | 0.491 | 0.567 | 0.515 | 0.412 |
| 95% CI of change | 0.142 to 0.931 | 0.082 to 0.905 | 0.150 to 0.991 | 0.073 to 0.964 | 0.002 to 0.830 |
Linear regression models were adjusted for age, sex, BMI, total sleep time, smoking status, and depression (model 1). Statistically significant associations (P < .05) appear in boldface. All analyses had 998 participants except for those with arousal severity, which had 895 participants with useable EEG signals.
Association between 1-SD increase in predictor variable and change in outcome variable.
Figure 7.
Forest plot showing the association between 1-SD increase in respiratory (resp.) event-related predictor variables and increased number of psychomotor vigilance task (PVT) lapses. Vertical ticks represent coefficient estimates and horizontal bars represent 95% CIs. Linear regression models were adjusted for age, sex, BMI, total sleep time, smoking status, and depression. All analyses had 998 participants except for those based on arousal intensity, which had 895 participants with useable EEG signals.
Discussion
The objective of this study was to investigate and quantify the relationship between nocturnal flow limitation and daytime vigilance in OSA. This was investigated in a retrospective clinical cohort of 998 participants with suspected OSA who completed a PVT and same-night in-laboratory polysomnography. We hypothesized that increased flow limitation during sleep is associated with decreased daytime vigilance. We demonstrate that this association was independent of potential confounders and mediators such as age, sex, BMI, respiratory event frequency (AHI), arousal severity, subjective sleepiness, and key comorbidities. A 1-SD increase in the percentage of flow-limited breaths during sleep (from 22.8% to 58.1%) was associated with 2.1 additional PVT lapses (response times > 500 ms).
Mechanistic Insights
Nocturnal flow limitation may contribute to daytime hypovigilance in OSA via three possible mechanisms. First, flow limitation may cause changes in sleep EEG microstructure and may result in daytime vigilance impairment. Specifically, sustained inspiratory flow limitation resulting in increased K-complex density and decreased slow-wave activity after K complexes is associated with decreased PVT performance.9,14 Additionally, higher frontal EEG slowing ratio in rapid eye movement sleep and lower sleep spindle density have been associated with reduced PVT performance.28 Second, flow limitation may cause hypercapnia29 and acute respiratory acidosis,30 leading to cognitive impairment31,32 that persists throughout the day. Third, flow limitation may cause vigilance impairment through higher upper airway resistance, causing increased work of breathing and leading to fatigue. Increased physical fatigue has been associated with increased PVT lapses.8 Furthermore, patients with OSA can gain weight after CPAP treatment, which could indicate high levels of physical exertion during sleep in untreated OSA. However, this weight gain has been attributed to increased fluid accumulation,33 increased dietary intake, and disordered eating behavior.34 Further interventional studies are necessary to elucidate the physiologic mechanisms by which nocturnal flow limitation contributes to impaired vigilance.
Intermittent hypoxemia and sleep fragmentation resulting from arousals are other mechanisms that may contribute to impaired daytime vigilance in OSA. Although Sforza et al4 reported that arousal count explained 23% of variance in PVT lapses, other studies have reported no association between the arousal index and PVT quantified vigilance.9,12 Furthermore, arousal severity (a combination of arousal intensity, duration, and frequency) was not associated with PVT performance in our study. Intermittent hypoxemia may result in neuronal damage to brain regions involved in attention and motor skills.35 In previous work using data from this cohort, a 10% increase in desaturation severity elevated the odds of increased PVT response time and number of lapses (OR, 1.43-1.52).12 Using linear regression, we found that desaturation severity was associated significantly with the same PVT metrics. This suggests that the severity of desaturations resulting from respiratory events may influence daytime PVT performance, which is consistent with other studies.4,36,37
Because OSA is characterized by respiratory events, we conducted a secondary exploratory analysis to determine whether flow limitation during respiratory events and other descriptors of respiratory event severity are associated more strongly with PVT performance. In this analysis, the only measures of respiratory event severity that were associated significantly with PVT performance were respiratory event-related flow limitation and desaturation severity, area, and depth. PVT performance was associated previously with nocturnal arterial pulsation frequency in this cohort.38 Thus, we expected to observe an association between respiratory event-related change in heart rate and PVT performance, but did not. This is possibly because this metric does not capture changes in heart rate resulting from spontaneous arousals. Because no significant association between PVT performance and the hypopnea index and apnea index was found, it is unlikely that flow limitation frequency and severity metrics are simply differentiating between hypopneas and apneas. Finally, respiratory event-related measures of flow limitation, desaturation severity, and arousal severity are good proxies for sleep burden measures because they are correlated strongly (Pearson r ≥ 0.84) (e-Fig 2).
Clinical Implications
To aid in the interpretation of our findings, we based our analyses on the associations between a 1-SD increase in predictor variables and increased PVT lapses. A 1-SD increase in the percentage of flow-limited breaths during sleep was associated with 2.1 additional lapses (95% CI, 0.7-3.7). It previously was found that older adults are less vigilant than younger adults,39,40 and as a comparison with this known relationship, we note that in our data, the effect size of flow limitation was similar to that for age, where a 1-SD increase (from 54.4 years to 67.9 years) was associated with 1.9 additional lapses. To consider these findings in a real-world context, a vehicle travelling at 100 km/h (62.1 miles/h) will travel 13.9 m (45.6 feet) in 500 ms. Therefore, any lapses in attention may increase accident risk while driving substantially. In a study of commercial truck drivers who completed the PVT, an association was found between lapses and average sleep duration.11 The increase in lapses between the 6- and 7-h group and the < 5-h group (approximately 1-SD difference) was approximately 0.9. Therefore, nocturnal flow limitation may have more than twice the effect on daytime vigilance relative to average sleep duration. These findings highlight the importance of measuring flow limitation when diagnosing OSA. Clinical diagnosis and treatment of OSA relies in part on the AHI; however, in this study, the best model used flow limitation alone, and consistent with six other studies,7, 8, 9, 10, 11, 12 AHI was not associated significantly with vigilance in any model. It also has been reported that patients with upper airway resistance syndrome showed worse PVT performance than patients with OSA.41 After excluding individuals with AHI of > 15 events/h to examine the association between flow limitation and vigilance in individuals with mild OSA, we found that the effect size of flow limitation on PVT lapses increased to 2.7 lapses (95% CI, 0.5-5.0 lapses). This suggests that individuals with mild OSA (AHI < 15 events/h) who exhibit substantial nocturnal flow limitation may benefit from CPAP treatment, particularly those who have impaired vigilance and operate a vehicle.42 Indeed, elimination of flow limitation using CPAP has been demonstrated to improve daytime vigilance.15 Therefore, flow limitation and other measures of OSA severity could be used alongside the AHI and patient symptoms to guide decisions on who should and should not be treated.
Methodologic Considerations
This study has several limitations. First, we did not have the medication history for all participants; however, for 446 participants whose medication history data was available, the effect size of flow limitation frequency did not change appreciably after adjusting for medications that act on the CNS and may affect vigilance. Second, flow limitation frequency was borderline statistically significant (P = .051) when desaturation severity was included as a covariate. This is possibly because of collinearity between flow limitation frequency and desaturation severity, which are correlated moderately (Pearson r = 0.63). Third, as an alternative to the Box-Cox transform, we also constructed generalized linear models using the log link function and untransformed PVT lapses. Our results generally were consistent; however, the best model based on Akaike information criterion was the model including flow limitation frequency, AHI, desaturation severity, and arousal severity. Fourth, in terms of being a meaningful measurement of vigilance in everyday life, the PVT is not as realistic as other tasks such as a driving simulator. However, PVT response time correlates with driving simulator performance in patients with OSA.43 Fifth, it might be thought that the PVT should have been completed after polysomnography to examine the direct effect of nocturnal flow limitation on PVT performance. However, the effect of sleeping in a laboratory for the first time (termed the first-night effect) may affect sleep architecture and next-day PVT performance.44,45 Flow limitation during sleep is unlikely to be different between home and laboratory environments, provided that sleep state and posture are unchanged. It should be noted that more time can be spent in the supine position during laboratory polysomnography, which may result in larger estimates of AHI in supine-predominant OSA.46 Further work may involve performing a positional analysis in this cohort and characterizing the night-to-night variability of the flow limitation frequency metric. Sixth, sleep duration was calculated from the scored sleep stages of the diagnostic sleep studies. No assessment of habitual sleep duration was made. This makes the relationship between total sleep time and vigilance difficult to interpret. Seventh, only individuals with suspected OSA were included in this study. Although significant variability in flow limitation was found in the 128 participants with an AHI of < 5 events/h in our study (as might be expected from previous investigation into upper airway resistance syndrome47,48), we cannot exclude that the relationship between flow limitation and vigilance may be different in asymptomatic people, given that healthy people without any symptoms can have 30% inspiratory flow limitation frequency during sleep.49 Eighth, this was a cross-sectional study. Finally, numerous factors may affect PVT performance that were not measured here. Therefore, future work may involve performing a follow-up PVT after CPAP treatment to identify when vigilance impairment is sleep related (and treatable or not treatable with CPAP) vs from other sources.
Interpretation
Increased flow limitation during sleep was associated with decreased daytime vigilance in individuals with suspected OSA, independent of the AHI. Flow limitation may complement standard clinical metrics in identifying individuals whose vigilance impairment most likely is explained by OSA.
Funding/Support
E. S. was supported by the National Health and Medical Research Council of Australia [Grant 2001729] administered by The University of Queensland. T. L. has received research funding from the European Union’s Horizon 2020 Research and Innovation Programme [Decision 965417], NordForsk [NordSleep Project 90458] via Business Finland [Grant 5133/31/2018], and the Research Committee of the Kuopio University Hospital Catchment Area for the State Research Funding [Project 5041794]. J. T. has received research funding from the European Union’s Horizon 2020 Research and Innovation Programme [Decision 965417], NordForsk via Business Finland, and the National Health and Medical Research Council of Australia. S. A. S. was supported by the National Heart, Lung, and Blood Institute of the National Institutes of [Grants R01HL146697 and R01HL168067] and the American Academy of Sleep Medicine Foundation [Grant 228-SR-20]. S. K. was supported by the Finnish Cultural Foundation (Central Fund), Finnish Tampere Tuberculosis Foundation, Kuopio University Hospital [Grant 5041804], Maud Kuistila Memorial Foundation, and Regional Council of Pohjois-Savo European Regional Development Fund. P. I. T. has received research funding from the National Health and Medical Research Council of Australia [Grants 2001729 and 2007001], Queensland Health Metro North, and The University of Queensland. D. L. M. was supported by the National Health and Medical Research Council of Australia [Grants 2001729 and 2007001] and The University of Queensland Research Stimulus Allocation Two (Fellowships). A. A. was supported by the National Institute of Health, the American Heart Association, and the American Academy of Sleep Medicine.
Financial/Nonfinancial Disclosures
The authors have reported to CHEST the following: J. T. has received royalties related to Bittium BrainStatus EEG outside the submitted work. S. A. S. has received personal fees as a consultant for Nox Medical, Merck, Apnimed, Respicardia, Inspire Medical Systems, Eli Lilly, LinguaFlex, and Forepont outside the submitted work; receives grant support from Apnimed, Prosomnus, and Dynaflex for unrelated studies; and holds patents on phenotyping from wearable technology and sleep apnea pharmacotherapy through his institution’s Innovation department. His industry interactions are managed by Brigham and Women’s Hospital and Mass General Brigham in accordance with their conflict of interest policies. P. I. T. has received honoraria from the University of Sydney, the University of Adelaide, and Flinders University outside the submitted work. Outside the submitted work, A. A. has received a grant from Somnifix; consulting fees from Apnimed, Somnifix, LinguaFlex, Respicardia, Eli Lilly, and Inspire; honoraria from the American Academy of Orofacial Pain for a lecture; honoraria from University of Alberta for a grant review; equipment loaned from Philips Respironics; and has a patent pending for phenotyping sleep apnea using wearables. None declared (E. S., T. L., S. K., T. G., D. L. M., B. D.).
Acknowledgments
Author contributions: P. I. T., J. T., T. L., S. A. S., D. L. M., and E. S. contributed to study design. E. S., D. L. M., S. A. S., P. I. T., S. K., A. A., T. L., and J. T. contributed to algorithm development. B. D., E. S., and D. L. M. contributed to data collection, collation, and quality checking. E. S., P. I. T., S. A. S., and D. L. M. contributed to data analysis. All authors contributed to interpretation of results and preparation of the manuscript. E. S. takes responsibility for (is the guarantor of) the content of the manuscript, including the data and analysis.
Role of sponsors: The sponsors had no role in the design of the study, the collection and analysis of the data, or the preparation of the manuscript.
Other contributions: The authors thank the staff and participants who visited the Princess Alexandra Hospital Sleep Disorders Centre.
Data availability: Data and code will be shared on request to the corresponding author with the permission of the Princess Alexandra Hospital in Brisbane, Australia.
Additional information: The e-Appendix, e-Figures, and e-Tables are available online under “Supplementary Data.”
Footnotes
Drs Sands and Terrill contributed equally to this manuscript.
Supplementary Data
References
- 1.Benjafield A.V., Ayas N.T., Eastwood P.R., et al. Estimation of the global prevalence and burden of obstructive sleep apnoea: a literature-based analysis. Lancet Respir Med. 2019;7(8):687–698. doi: 10.1016/S2213-2600(19)30198-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Mazza S., Pepin J., Naegele B., et al. Most obstructive sleep apnoea patients exhibit vigilance and attention deficits on an extended battery of tests. Eur Respir J. 2005;25(1):75–80. doi: 10.1183/09031936.04.00011204. [DOI] [PubMed] [Google Scholar]
- 3.Langner R., Eickhoff S.B. Sustaining attention to simple tasks: a meta-analytic review of the neural mechanisms of vigilant attention. Psychol Bull. 2013;139(4):870. doi: 10.1037/a0030694. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Sforza E., Haba-Rubio J., De Bilbao F., et al. Performance vigilance task and sleepiness in patients with sleep-disordered breathing. Eur Respir J. 2004;24(2):279–285. doi: 10.1183/09031936.04.00091903. [DOI] [PubMed] [Google Scholar]
- 5.Kim H., Dinges D.F., Young T. Sleep-disordered breathing and psychomotor vigilance in a community-based sample. Sleep. 2007;30(10):1309–1316. doi: 10.1093/sleep/30.10.1309. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Arora N., Meskill G., Guilleminault C. The role of flow limitation as an important diagnostic tool and clinical finding in mild sleep-disordered breathing. Sleep Sci. 2015;8(3):134–142. doi: 10.1016/j.slsci.2015.08.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Batool-Anwar S., Kales S.N., Patel S.R., et al. Obstructive sleep apnea and psychomotor vigilance task performance. Nat Sci Sleep. 2014;6:65–71. doi: 10.2147/NSS.S53721. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Lee I.-S., Bardwell W.A., Ancoli-Israel S., et al. Number of lapses during the psychomotor vigilance task as an objective measure of fatigue. J Clin Sleep Med. 2010;6(2):163–168. [PMC free article] [PubMed] [Google Scholar]
- 9.Parekh A., Mullins A.E., Kam K., et al. Slow-wave activity surrounding stage N2 K-complexes and daytime function measured by psychomotor vigilance test in obstructive sleep apnea. Sleep. 2019;42(3) doi: 10.1093/sleep/zsy256. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Prasad B., Steffen A.D., Van Dongen H.P.A., et al. Determinants of sleepiness in obstructive sleep apnea. Sleep. 2018;41(2) doi: 10.1093/sleep/zsx199. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Pack A.I., Maislin G., Staley B., et al. Impaired performance in commercial drivers: role of sleep apnea and short sleep duration. Am J Respir Crit Care Med. 2006;174(4):446–454. doi: 10.1164/rccm.200408-1146OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Kainulainen S., Duce B., Korkalainen H., et al. Severe desaturations increase psychomotor vigilance task-based median reaction time and number of lapses in obstructive sleep apnoea patients. Eur Respir J. 2020;55(4) doi: 10.1183/13993003.01849-2019. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Kulkas A., Tiihonen P., Julkunen P., et al. Novel parameters indicate significant differences in severity of obstructive sleep apnea with patients having similar apnea–hypopnea index. Med Biol Eng Comput. 2013;51(6):697–708. doi: 10.1007/s11517-013-1039-4. [DOI] [PubMed] [Google Scholar]
- 14.Parekh A., Kam K., Mullins A.E., et al. Altered K-complex morphology during sustained inspiratory airflow limitation is associated with next-day lapses in vigilance in obstructive sleep apnea. Sleep. 2021;44(7) doi: 10.1093/sleep/zsab010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Meurice J., Paquereau J., Denjean A., et al. Influence of correction of flow limitation on continuous positive airway pressure efficiency in sleep apnoea/hypopnoea syndrome. Eur Respir J. 1998;11(5):1121–1127. doi: 10.1183/09031936.98.11051121. [DOI] [PubMed] [Google Scholar]
- 16.Mann D.L., Georgeson T., Landry S.A., et al. Frequency of flow limitation using airflow shape. Sleep. 2021;44(12) doi: 10.1093/sleep/zsab170. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Mann D.L., Terrill P.I., Azarbarzin A., et al. Quantifying the magnitude of pharyngeal obstruction during sleep using airflow shape. Eur Respir J. 2019;54(1) doi: 10.1183/13993003.02262-2018. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Johns M.W. A new method for measuring daytime sleepiness: the Epworth sleepiness scale. Sleep. 1991;14(6):540–545. doi: 10.1093/sleep/14.6.540. [DOI] [PubMed] [Google Scholar]
- 19.Berry R.B., Brooks R., Gamaldo C.E., et al. American Academy of Sleep Medicine; 2012. The AASM Manual for the Scoring of Sleep and Associated Events: Rules, Terminology and Technical Specifications. [Google Scholar]
- 20.Arsintescu L., Kato K.H., Cravalho P.F., et al. Validation of a touchscreen psychomotor vigilance task. Accid Anal Prev. 2019;126:173–176. doi: 10.1016/j.aap.2017.11.041. [DOI] [PubMed] [Google Scholar]
- 21.Basner M., Hermosillo E., Nasrini J., et al. Repeated administration effects on psychomotor vigilance test performance. Sleep. 2018;41(1) doi: 10.1093/sleep/zsx187. [DOI] [PubMed] [Google Scholar]
- 22.Basner M., Dinges D.F. Maximizing sensitivity of the psychomotor vigilance test (PVT) to sleep loss. Sleep. 2011;34(5):581–591. doi: 10.1093/sleep/34.5.581. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Mueller S.T., Piper B.J. The psychology experiment building language (PEBL) and PEBL test battery. J Neurosci Methods. 2014;222:250–259. doi: 10.1016/j.jneumeth.2013.10.024. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Azarbarzin A., Ostrowski M., Hanly P., et al. Relationship between arousal intensity and heart rate response to arousal. Sleep. 2014;37(4):645–653. doi: 10.5665/sleep.3560. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Shahrbabaki S.S., Linz D., Hartmann S., et al. Sleep arousal burden is associated with long-term all-cause and cardiovascular mortality in 8001 community-dwelling older men and women. Eur Heart J. 2021;42(21):2088–2099. doi: 10.1093/eurheartj/ehab151. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Terrill P.I., Edwards B.A., Nemati S., et al. Quantifying the ventilatory control contribution to sleep apnoea using polysomnography. Eur Respir J. 2015;45(2):408–418. doi: 10.1183/09031936.00062914. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Azarbarzin A., Zinchuk A., Wellman A., et al. Cardiovascular benefit of continuous positive airway pressure in adults with coronary artery disease and obstructive sleep apnea without excessive sleepiness. Am J Respir Crit Care Med. 2022;206(6):767–774. doi: 10.1164/rccm.202111-2608OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Mullins A.E., Kim J.W., Wong K.K.H., et al. Sleep EEG microstructure is associated with neurobehavioural impairment after extended wakefulness in obstructive sleep apnea. Sleep Breath. 2021;25(1):347–354. doi: 10.1007/s11325-020-02066-5. [DOI] [PubMed] [Google Scholar]
- 29.Rimpilä V., Saaresranta T., Huhtala H., et al. Transcutaneous CO(2) plateau as set-point for respiratory drive during upper airway flow-limitation. Respir Physiol Neurobiol. 2014;191:44–51. doi: 10.1016/j.resp.2013.10.014. [DOI] [PubMed] [Google Scholar]
- 30.Mehta V., Vasu T.S., Phillips B., et al. Obstructive sleep apnea and oxygen therapy: a systematic review of the literature and meta-analysis. J Clin Sleep Med. 2013;09(03):271–279. doi: 10.5664/jcsm.2500. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Azuma K., Kagi N., Yanagi U., et al. Effects of low-level inhalation exposure to carbon dioxide in indoor environments: a short review on human health and psychomotor performance. Environ Int. 2018;121:51–56. doi: 10.1016/j.envint.2018.08.059. [DOI] [PubMed] [Google Scholar]
- 32.Olaithe M., Bucks R.S., Hillman D.R., et al. Cognitive deficits in obstructive sleep apnea: insights from a meta-review and comparison with deficits observed in COPD, insomnia, and sleep deprivation. Sleep Med Rev. 2018;38:39–49. doi: 10.1016/j.smrv.2017.03.005. [DOI] [PubMed] [Google Scholar]
- 33.Herculano S., Grad G.F., Drager L.F., et al. Weight gain induced by continuous positive airway pressure in patients with obstructive sleep apnea is mediated by fluid accumulation: a randomized crossover controlled trial. Am J Respir Crit Care Med. 2021;203(1):134–136. doi: 10.1164/rccm.202005-1853LE. [DOI] [PubMed] [Google Scholar]
- 34.Tachikawa R., Ikeda K., Minami T., et al. Changes in energy metabolism after continuous positive airway pressure for obstructive sleep apnea. Am J Respir Crit Care Med. 2016;194(6):729–738. doi: 10.1164/rccm.201511-2314OC. [DOI] [PubMed] [Google Scholar]
- 35.Gagnon K., Baril A.A., Gagnon J.F., et al. Cognitive impairment in obstructive sleep apnea. Pathol Biol (Paris) 2014;62(5):233–240. doi: 10.1016/j.patbio.2014.05.015. [DOI] [PubMed] [Google Scholar]
- 36.Pahari P., Korkalainen H., Karhu T., et al. Obstructive sleep apnea-related intermittent hypoxaemia is associated with impaired vigilance. J Sleep Res. 2022 doi: 10.1111/jsr.13803. [DOI] [PubMed] [Google Scholar]
- 37.Tanno S., Tanigawa T., Maruyama K., et al. Sleep-related intermittent hypoxia is associated with decreased psychomotor vigilance in Japanese community residents. Sleep Med. 2017;29:7–12. doi: 10.1016/j.sleep.2016.08.024. [DOI] [PubMed] [Google Scholar]
- 38.Kainulainen S., Duce B., Korkalainen H., et al. Increased nocturnal arterial pulsation frequencies of obstructive sleep apnoea patients is associated with an increased number of lapses in a psychomotor vigilance task. ERJ Open Res. 2020;6(4):00277–02020. doi: 10.1183/23120541.00277-2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Adam M., Rétey J.V., Khatami R., et al. Age-related changes in the time course of vigilant attention during 40 hours without sleep in men. Sleep. 2006;29(1):55–57. doi: 10.1093/sleep/29.1.55. [DOI] [PubMed] [Google Scholar]
- 40.Parasuraman R., Giambra L. Skill development in vigilance: effects of event rate and age. Psychol Aging. 1991;6(2):155–169. doi: 10.1037//0882-7974.6.2.155. [DOI] [PubMed] [Google Scholar]
- 41.Stoohs R.A., Philip P., Andries D., et al. Reaction time performance in upper airway resistance syndrome versus obstructive sleep apnea syndrome. Sleep Med. 2009;10(9):1000–1004. doi: 10.1016/j.sleep.2008.11.005. [DOI] [PubMed] [Google Scholar]
- 42.Bonsignore M.R., Randerath W., Schiza S., et al. European Respiratory Society statement on sleep apnoea, sleepiness and driving risk. Eur Respir J. 2021;57(2) doi: 10.1183/13993003.01272-2020. [DOI] [PubMed] [Google Scholar]
- 43.Huang Y., Hennig S., Fietze I., et al. The psychomotor vigilance test compared to a divided attention steering simulation in patients with moderate or severe obstructive sleep apnea. Nat Sci Sleep. 2020;12:509. doi: 10.2147/NSS.S256987. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Sprajcer M., Gupta C., Roach G., et al. Can we put the first night effect to bed? An analysis based on a large sample of healthy adults. Chronobiol Int. 2022;39(12):1567–1573. doi: 10.1080/07420528.2022.2133611. [DOI] [PubMed] [Google Scholar]
- 45.Byun J.-H., Kim K.T., Moon H.-J., et al. The first night effect during polysomnography, and patients’ estimates of sleep quality. Psychiatry Res. 2019;274:27–29. doi: 10.1016/j.psychres.2019.02.011. [DOI] [PubMed] [Google Scholar]
- 46.Yo S.W., Joosten S.A., Wimaleswaran H., et al. Body position during laboratory and home polysomnography compared to habitual sleeping position at home. J Clin Sleep Med. 2022;18(9):2103–2111. doi: 10.5664/jcsm.9990. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Guilleminault C., Stoohs R., Clerk A., et al. A cause of excessive daytime sleepiness: the upper airway resistance syndrome. Chest. 1993;104(3):781–787. doi: 10.1378/chest.104.3.781. [DOI] [PubMed] [Google Scholar]
- 48.Gold A.R., Gold M.S. Con: sleep fragmentation causes hypersomnolence in OSA. Sleep Med Rev. 2021;55 doi: 10.1016/j.smrv.2020.101399. [DOI] [PubMed] [Google Scholar]
- 49.Palombini L.O., Tufik S., Rapoport D.M., et al. Inspiratory flow limitation in a normal population of adults in São Paulo, Brazil. Sleep. 2013;36(11):1663–1668. doi: 10.5665/sleep.3122. [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.