Abstract
Rationale: Sleep-disordered breathing (SDB) occurring primarily during rapid eye movement (REM) sleep is a common clinical problem. The natural history of REM-related SDB and the associated cardiovascular sequelae of disease progression remain to be determined.
Objectives: The objective of the current study was to describe the natural history of REM-related SDB, ascertain predictors of progression, and determine whether the evolution of REM-related SDB into non-REM (NREM) sleep is associated with incident cardiovascular events.
Methods: Participants from the Sleep Heart Health Study with a baseline NREM apnea–hypopnea index (NREM-AHI) of <5 events/h and data from a follow-up sleep study along with information on incident cardiovascular disease were included in the study. Bivariate logistic regression was used to jointly model the predictors of disease progression based on the presence or absence of SDB during NREM and REM sleep using a cut-point of 5 events/h. Explanatory variables such as age, race, body mass index (BMI), change in BMI, and baseline REM-AHI were considered. Proportional hazards regression was then used to establish whether the development of SDB during NREM sleep was associated with incident cardiovascular disease.
Results: The majority of the 1,908 participants included in the study did not develop SDB during NREM sleep. The likelihood of progression of SDB into NREM sleep did increase with higher baseline REM-AHI. BMI and an increase in BMI predicted progression of SDB in both NREM and REM sleep in men but not in women. There was a strong interdependence between developing a NREM-AHI of ≥5 events/h and worsening REM-AHI at follow-up with odds ratios of 6.01 and 4.47, in women and men, respectively. Moreover, the relative risk for incident cardiovascular events among those who developed a NREM-AHI of ≥5 events/h at the follow-up visit was elevated only in women with REM-related SDB at baseline.
Conclusions: SDB during REM sleep is a relatively stable condition and does not progress in the majority of individuals. Progression of SDB into NREM sleep is associated with sex, weight, and age. SDB during REM and NREM sleep tends to develop concurrently. Finally, the development of SDB during NREM sleep is associated with incident cardiovascular events, but only in women with REM-related SDB at baseline.
Keywords: sleep apnea, REM sleep, cardiovascular disease
Sleep-disordered breathing (SDB) occurring only during rapid eye movement (REM) sleep is a common clinical occurrence that can present a therapeutic challenge. The physiological milieu of REM sleep, with its marked decrease in pharyngeal dilator muscle tone and suppression of genioglossus muscle activity, results in a greater propensity for upper-airway collapse during REM sleep than during non-REM (NREM) sleep. Apneas and hypopneas that occur during REM sleep are typically longer, more frequent, and have more severe oxyhemoglobin desaturation than those that occur during NREM sleep (1–3). Patients with SDB that occurs only during REM sleep can have a normal overall apnea–hypopnea index (AHI) despite having frequent SDB events during REM sleep. Even with the obvious differences in the hypoxemic burden of SDB during NREM and REM sleep, the benefit of treating patients with SDB only during REM sleep is unknown.
Recently published longitudinal data from several epidemiological cohorts indicate that untreated SDB during REM sleep is independently associated with adverse cardiovascular outcomes, including incident hypertension (4, 5) and recurrent cardiovascular events (6). A possible explanation for the increased cardiovascular burden is that SDB events during REM sleep are associated with a greater sympathetic response than SDB events during NREM sleep, and thus pose a greater health risk despite the fact that the amount of REM sleep is less compared to NREM sleep. Although observational data on the clinical sequelae of SDB during REM sleep have recently become available, the knowledge gap regarding the natural history of SDB that occurs only during REM sleep remains to be addressed. It has been suggested that SDB isolated to REM sleep may simply be an early manifestation of SDB that will progress and eventually manifest during both NREM and REM sleep. If so, REM-related SDB would not be considered a separate clinical entity that is associated with cardiovascular disease but rather just an early manifestation of SDB that will ultimately manifest across the sleep period. To better understand the eventual development of SDB during NREM and REM sleep, and the associated risk for cardiovascular disease, the natural history of REM-related SDB needs to be defined. The primary objectives of the current study were to characterize the natural history of REM-related SDB and determine the role of demographic and anthropometric factors as determinants of disease progression. An additional goal was to assess whether the progression of SDB into NREM sleep in individuals with REM-related SDB at baseline was associated with incident cardiovascular disease.
Methods
Study Sample
The Sleep Heart Health Study was a prospective cohort study of the cardiovascular consequences of SDB. Details of the study design and cohort follow-up have been reported previously (7). Between 1995 and 1998, participants for the study were recruited from ongoing cohort studies, including the Framingham Offspring and Omni Study, the Atherosclerosis Risk in Communities Study, the Cardiovascular Health Study, the Strong Heart Study, and the cohort studies of respiratory disease in Tucson. Participants were at least 40 years of age and were not on treatment for SDB with positive airway pressure, oral appliance, oxygen, or tracheostomy. All participants provided written consent and the study protocol was approved by the institutional review board of each of the study’s field sites. For the current analysis, the 6,441 participants who completed the baseline examination had to meet the following criteria: 1) a NREM-AHI of <5 events/h on the baseline polysomnogram, 2) a follow-up polysomnogram ∼5 years after the baseline study, and 3) complete data on both visits. Finally, only those participants with more than 15 minutes of REM sleep during the baseline and follow-up sleep studies were included in the analysis to allow for a representative estimate of the REM-AHI. A total of 1,908 participants were included in the current analysis. Sensitivity analyses were also conducted using alternative thresholds for REM-AHI, including 10 and 15 events/h as described in the online supplement.
Data Collection
All study participants completed a baseline examination including a detailed health interview, two unattended home polysomnograms, measurements of blood pressure and anthropometry, and assessments of sleep habits and prescription medication use. Body mass index (BMI) at baseline and follow-up visits were used to calculate the change in BMI over time (ΔBMI). Home polysomnography at both visits was conducted using a portable monitor (P-Series, Compumedics). The following signals were recorded: C3/A1 and C4/A2 electroencephalograms, bilateral electrooculograms, a single bipolar electrocardiogram, a chin electromyogram, oxyhemoglobin saturation by pulse oximetry, chest and abdominal excursion by inductance plethysmography, airflow by an oronasal thermocouple, and body position by a mercury gauge. Details regarding the polysomnographic equipment, hook-up procedures, failure rates, scoring, and quality assurance and control have been published elsewhere (8). Apneas were identified if airflow was absent or nearly absent for at least 10 seconds. Apneas were further classified as obstructive if movement on either the chest or abdominal inductance channels was noted, or as central if no displacement was observed on both of these channels. Hypopneas were scored if there was at least a 30% reduction in airflow or thoracoabdominal movement below baseline values for at least 10 seconds. The AHI was defined as the number of apneas and hypopneas, each associated with at least a 4% decrease in oxygen saturation per hour of sleep. An arousal index was derived as the total number of arousals per hour of sleep according to standard criteria (9). Incident cardiovascular disease was determined as the occurrence of any of the following: myocardial infarction (nonfatal or fatal), percutaneous coronary angioplasty, coronary artery bypass grafting, stroke (nonfatal or fatal), and occurrence of a heart failure episode as previously described (7).
Statistical Analysis
Given that the severity and progression of SDB are associated with sex (10, 11), all analyses of the baseline and change in AHI values were stratified by sex. Sleep stage–related AHI values were categorized using commonly used clinical cut-points: <5.0 (normal), 5.0–14.9 (mild), 15.0–29.9 (moderate), and ≥30.0 events/h (severe). Given the limited sample size in some of the categories, individuals with moderate or severe SDB were collapsed to form a composite group. The primary dependent variables for the current analyses were the follow-up NREM- and REM-AHI values, which were dichotomized using a threshold of 5 events/h for ease of exposition. This resulted in the creation of two distinct binary dependent variables, NREM-AHI ≥ 5 events/h and REM-AHI ≥ 5 events/h, which had to be modeled jointly in a single analysis. If NREM-AHI and REM-AHI were analyzed separately using techniques such as ordinary logistic regression, two regression coefficients for a particular covariate (e.g., age)—one for NREM-AHI ≥ 5 events/h and one for REM-AHI ≥ 5 events/h—would be derived. Estimates obtained from such analyses would not account for the potential dependence of the follow-up states and would preclude a model-based quantification of this interdependence. The strategy of bivariate logistic regression (12) was used to jointly model the four potential scenarios at the follow-up visit: 1) NREM-AHI ≥ 5 events/h, REM-AHI ≥ 5 events/h; 2) NREM-AHI ≥ 5 events/h, REM-AHI < 5 events/h; 3) NREM-AHI <5 events/h, REM-AHI ≥ 5 events/h; and 4) NREM-AHI <5 events/h, REM-AHI <5 events/h.
With this approach, the joint probability for each of the four scenarios was modeled via three components. Two components are the marginal probabilities of NREM-AHI ≥ 5 events/h and REM-AHI ≥ 5 events/h, each modeled as a function of explanatory variables including race, BMI, ΔBMI, and baseline REM-AHI. The third component is the odds ratio ψ, which describes the dependence between the two outcomes at follow-up. This framework provides separate estimates of a predictor’s effect on each outcome, but the estimates are related stochastically through the common odds ratio (ψ). A value of ψ equal to one indicates that the development of NREM-AHI ≥ 5 and REM-AHI ≥ 5 is independent, whereas a value greater than one indicates that the evolution of NREM-AHI and REM-AHI is positively interdependent. Additional details regarding the bivariate logistic regression model are provided in the online supplement. Finally, to determine whether progression of SDB into NREM sleep was associated with incident cardiovascular disease, the proportional hazards model was used. In all multivariable models, the independent variables were considered as categorical variables to simplify the interpretation of the resulting regression coefficients. Multivariable models were also constructed using continuous variables, and the results from these analyses that used continuous predictors are provided in the online supplement. All analyses were conducted using the R studio and Stata 13.0 statistical packages.
Results
A total study of 1,908 participants with a NREM-AHI of <5 events/h at baseline were included in the current analyses. Demographic and covariate data for the sample, stratified by sex, are presented in Table 1. The mean age and BMI of the overall sample were 60.7 years and 27.9 kg/m2, respectively, with no significant differences between men (n = 684) and women (n = 1,224). Approximately one-third of the overall sample experienced a BMI increase of at least 1 kg/m2 between the baseline and follow-up visits, with 35% of women and 30% of men experiencing a BMI gain of 1 kg/m2 (P = 0.03). The average BMI increase in women and men was 0.19 and 0.21 kg/m2, respectively (P = 0.88). REM–related SDB at baseline was present in 550 of 1,224 women (44.9%) and 315 of 684 men (46.1%).
Table 1.
Study sample characteristics by sex for participants with NREM-AHI < 5 events/h at baseline
| Characteristic | Women (n = 1,224) | Men (n = 684) |
|---|---|---|
| Age, yr | 60.7 (10.3) | 60.9 (10.0) |
| Baseline BMI, kg/m2 | 28.0 (5.4) | 27.6 (3.9) |
| ΔBMI, kg/m2 | 0.19 (2.8) | 0.21 (2.0) |
| Number with ΔBMI ≥ 1 kg/m2, n (%) | 427 (34.9) | 204 (29.8) |
| Race, n (%) | ||
| White | 908 (74.2) | 542 (79.2) |
| Native American | 172 (14.1) | 63 (9.2) |
| Other | 144 (11.8) | 79 (11.6) |
| Overall AHI, events/h | 2.7 (2.9) | 3.0 (2.6) |
| Baseline REM-AHI, events/h | 8.3 (10.8) | 7.9 (9.2) |
| Baseline REM-AHI, n (%) | ||
| <5.0 events/h | 674 (55.1) | 369 (53.9) |
| 5.0–14.9 events/h | 323 (26.4) | 196 (28.7) |
| 15.0–29.9 events/h | 152 (12.4) | 88 (12.9) |
| ≥30 events/h | 75 (6.1) | 31 (4.5) |
Definition of abbreviations: AHI = apnea–hypopnea index; BMI = body mass index; ΔBMI = change in BMI between the first and second visit; NREM-AHI = apnea–hypopnea index during non–rapid eye movement sleep; REM-AHI = apnea–hypopnea index during rapid eye movement sleep.
Table 2 depicts the number of participants (and proportion of the sample) who progressed into the four possible categories of NREM-AHI and REM-AHI for a given baseline REM-AHI. Among those with a baseline REM-AHI of <5 events/h, 10.5% of women and 19.0% of men had a follow-up NREM-AHI and REM-AHI of ≥5 events/h. In the subset with a baseline REM-AHI of 5.0–14.9 events/h, 17.7% of women and 37.8% of men had a follow-up NREM-AHI and REM-AHI of ≥5 events/h. Finally, in the subgroup with a baseline REM-AHI of ≥15.0 events/h, 27.3% and 42.0% of women and men, respectively, progressed to have a NREM-AHI of ≥5.0 events/h while maintaining a REM-AHI of ≥5.0 events/h. Interestingly, in each of the three baseline REM-AHI categories, a majority of the sample maintained a NREM-AHI of <5 events/h during the follow-up. In the three baseline REM-AHI categories (<5.0, 5.0–14.9, and ≥15.0 events/h), 87.2%, 80.8%, and 72.7% of women did not progress to have a NREM of ≥5.0 events/h (P < 0.0001 for trend). For men, the respective proportions that did not demonstrate progression were 75.1%, 58.7%, and 58.0%. Although overall a significant proportion of the sample did not progress, this proportion decreased with increasing REM-AHI at baseline in women and men. The distributions of the change in NREM-AHI and REM-AHI between the follow-up and baseline visits are provided in the online supplement.
Table 2.
Distribution of NREM and REM SDB at follow-up stratified by baseline REM-AHI
| Baseline REM-AHI | Follow-Up NREM-AHI and REM-AHI |
|||
|---|---|---|---|---|
| NREM-AHI | REM-AHI | Women | Men | |
| <5.0 | <5 | <5 | 340 (50.5%) | 159 (43.0%) |
| ≥5 | 248 (36.8%) | 118 (32.0%) | ||
| ≥5 | <5 | 15 (2.2%) | 22 (6.0%) | |
| ≥5 | 71 (10.5%) | 70 (19.0%) | ||
| 5.0–14.9 | <5 | <5 | 78 (24.1%) | 43 (21.9%) |
| ≥5 | 183 (56.7%) | 72 (36.7%) | ||
| ≥5 | <5 | 5 (1.5%) | 7 (3.6%) | |
| ≥5 | 57 (17.7%) | 74 (37.8%) | ||
| ≥15.0 | <5 | <5 | 25 (11.0%) | 4 (3.4%) |
| ≥5 | 140 (61.7%) | 65 (54.6%) | ||
| ≥5 | <5 | 0 (0.0%) | 0 (0.0%) | |
| ≥5 | 62 (27.3%) | 50 (42.0%) | ||
Definition of abbreviations: NREM-AHI = apnea–hypopnea index during non–rapid eye movement sleep; REM-AHI = apnea–hypopnea index during rapid eye movement sleep; SDB = sleep-disordered breathing.
To determine whether REM-related SDB at baseline was independently associated with the occurrence of SDB during NREM sleep at follow-up, bivariate logistic regression analyses, stratified by sex, were used. Table 3 shows the adjusted odds ratios for developing a NREM-AHI of ≥5.0 events/h as a function of mild and moderate-to-severe REM–related SDB, as well for covariates such as BMI, change in BMI (ΔBMI dichotomized as an increase of ≥1 kg/m2), and race. First, age was positively associated with development of SDB during NREM sleep in both women and men. The odds ratios for development of SDB during NREM sleep in women and men per 1-year increase in age were 1.04 (95% confidence interval [CI], 1.02–1.06) and 1.02 (95% CI, 1.00–1.04), respectively, after adjusting for race, BMI, ΔBMI, and baseline REM-AHI. Baseline BMI and ΔBMI ≥ 1 kg/m2 predicted the development of a NREM-AHI of ≥5.0 events/h only in men with adjusted odds ratios of 1.10 (95% CI, 1.05–1.15) and 2.23 (95% CI, 1.55–3.21), respectively. Finally, in women, both mild and moderate-to-severe REM-related SDB were also independent predictors for progressing to a NREM-AHI of ≥5.0 events/h, with a statistically significant difference in the odds ratio between the two baseline REM-AHI categories (P = 0.04). In men, mild and moderate-to-severe REM-related SDB at baseline also predicted the occurrence of NREM-AHI ≥ 5.0 events/h at follow-up. However, the difference in the odds ratios comparing mild and moderate-to-severe REM-AHI groups was not statistically significant (P = 0.51).
Table 3.
Adjusted odds ratio* from the bivariate logistic regression model
| Adjusted Odds Ratios (95% CI) |
|||
|---|---|---|---|
| Women | Men | P Value† | |
| Outcome: NREM-AHI ≥ 5 events/h at follow-up |
|||
| Age, year | 1.04 (1.02–1.06) | 1.02 (1.00–1.04) | 0.19 |
| BMI, kg/m2 | 1.01 (0.98–1.04) | 1.10 (1.05–1.15) | 0.01 |
| ΔBMI ≥ 1 kg/m2 | 1.31 (0.94–1.83) | 2.23 (1.55–3.21) | 0.03 |
| Baseline REM-AHI, events/h | |||
| 5.0–14.9 | 1.46 (1.01–2.11) | 1.98 (1.35–2.91) | 0.26 |
| ≥15.0 | 2.27 (1.50–3.43) | 1.68 (1.06–2.68) | 0.35 |
| Outcome: REM-AHI ≥ 5 events/h at follow-up |
|||
| Age, year | 1.04 (1.03–1.05) | 1.01 (0.99–1.03) | 0.02 |
| BMI, kg/m2 | 1.10 (1.07–1.13) | 1.12 (1.07–1.18) | 0.50 |
| ΔBMI ≥ 1 kg/m2 | 1.78 (1.34–2.36) | 1.89 (1.27–2.82) | 0.81 |
| Baseline REM-AHI, events/h | |||
| 5.0–14.9 | 2.56 (1.88–3.48) | 2.67 (1.80–3.96) | 0.87 |
| ≥15.0 | 5.58 (3.51–8.88) | 20.87 (7.54–57.74) | 0.02 |
| Correlation | |||
| Odds ratio parameter (Ψ) | 6.01 (3.67–9.86) | 4.47 (2.81–7.10) | 0.39 |
Definition of abbreviations: BMI = body mass index; ΔBMI = change in BMI between the first and second visit; CI = confidence interval; NREM-AHI = apnea–hypopnea index during non–rapid eye movement sleep; REM-AHI = apnea–hypopnea index during rapid eye movement sleep.
Adjusted odds ratios for age and BMI are reported per unit increase.
P value for comparing the adjusted odds ratios between men and women.
The associations between having REM-related SDB at baseline and a REM-AHI of ≥5.0 events/h at follow-up are also shown in Table 3. Baseline BMI, ΔBMI, and severity of baseline REM-AHI were all associated with having a follow-up REM-AHI of ≥5.0 events/h in women and men, whereas increasing age was significantly associated with a follow-up REM-AHI ≥5.0 events/h only for women. Finally, the interdependence between having a NREM-AHI of ≥5.0 events/h and a REM-AHI of ≥5.0 events/h at follow-up was highly significant, with an odds ratio (Ψ) of 6.01 (95% CI, 3.67–9.86) and 4.47 (95% CI, 2.81–7.10) for women and men, respectively. Therefore, individuals are more likely to develop both a NREM-AHI of ≥5 events/h and a REM-AHI of ≥5 events/h than to develop either one alone. In all of the models constructed, race was not a predictor of either NREM-AHI ≥ 5.0 events/h or REM-AHI ≥ 5.0 events/h at follow-up in either men or women. Although all of the above analyses focused on the progression of REM-related SDB, the resolution of REM-related SDB was also examined. Stratified analyses showed that younger age, lower BMI at baseline, and a decrease in BMI between the follow-up and baseline visits were independently associated with improvement from REM-related SDB at baseline to no SDB at the follow-up visit (see Table E1 in the online supplement).
Because the above analyses required a minimum REM sleep time of 15 minutes, we conducted sensitivity analyses using a 30-minute threshold. The analyses summarized in Tables E2 and E3 replicate those shown in Tables 2 and 3, respectively, but required a minimum of 30 minutes for REM sleep time at baseline. These analyses showed that there were no significant differences in the variables associated with the evolution of NREM- and REM-related SDB comparing the 15- and 30-minute thresholds in REM sleep time. Additional sensitivity analyses examined whether alternative cut-points for dichotomizing NREM-AHI and REM-AHI, such as 10 and 15 events/h, would influence the predictors associated with the SDB progression. Irrespective of the cut-point chosen, the magnitude of the associations remained relatively consistent, indicating that although using a threshold of 5 events/h to define REM-related SDB may be prone to misclassification, the correlates of SDB progression are relatively consistent. Finally, given that differences in body position could also influence the observed differences in SDB severity between the two visits, the distribution of body position during REM sleep was also examined and noted to be comparable. The average time spent in REM sleep for the baseline and follow-up visits was 80.2 and 83.1 minutes, and 23.4 and 27.2 of these minutes, respectively, were spent in the supine position. No statistically significant differences in the overall amount of REM time or the time in the supine REM position between the two visits were found.
To determine whether the development of SDB during NREM sleep was associated with incident cardiovascular disease, multivariable proportional hazards models stratified by sex were used. In these models (Table 4), the final category of NREM-AHI and REM-AHI was used as the independent variable while accounting for confounders such as age, baseline BMI, race, smoking status, and prevalent hypertension, cardiovascular disease, and type 2 diabetes mellitus. In women with a baseline REM-AHI of ≥5 events/h, developing a NREM-AHI of ≥5 events/h at follow-up was independently associated with incident cardiovascular disease. In contrast, no association with incident cardiovascular disease was observed in women with a REM-AHI of <5 events/h and in men with any value of baseline REM-AHI.
Table 4.
Adjusted hazard ratio* for incident cardiovascular disease
| Baseline REM-AHI | Follow-Up NREM-AHI and REM-AHI |
|||
|---|---|---|---|---|
| NREM-AHI | REM-AHI | Women | Men | |
| <5.0 | <5 | <5 | 1.00 (Reference) | 1.00 (Reference) |
| ≥5 | 1.15 (0.69–1.90) | 0.56 (0.32–0.98) | ||
| ≥5 | Any | 1.36 (0.73–2.50) | 0.98 (0.58–1.66) | |
| 5.0–14.9 | <5 | <5 | 0.69 (0.31–1.53) | 0.66 (0.27–1.59) |
| ≥5 | 1.60 (0.93–2.75) | 1.27 (0.73–2.20) | ||
| ≥5 | Any | 2.11 (1.14–3.90) | 1.17 (0.66–2.09) | |
| ≥15.0 | <5 | <5 | 1.00 (0.31–3.32) | 1.84 (0.42–8.03) |
| ≥5 | 1.02 (0.56–1.89) | 1.20 (0.68–2.12) | ||
| ≥5 | Any | 2.27 (1.17–4.43) | 0.53 (0.25–1.12) | |
Definition of abbreviations: NREM-AHI = apnea–hypopnea index during non–rapid eye movement sleep; REM-AHI = apnea–hypopnea index during rapid eye movement sleep.
Adjusted for age, body mass index, change in body mass index, race, smoking status, and prevalent hypertension, type 2 diabetes, or cardiovascular disease.
Discussion
The current study yielded several important findings regarding the natural history of REM-predominant SDB in a community-based cohort of middle-aged and older adults. First, irrespective of the baseline REM-AHI, a significant proportion of both men and women did not go on to develop a NREM-AHI of ≥5.0 events/h over a median follow-up period of 5.3 years. The likelihood for progression did significantly increase with worsening baseline REM-AHI. In both sexes, age was a predictor of developing a follow-up NREM-AHI of ≥5.0 events/h, and in women, age also predicted the progression of REM-AHI. Second, as expected, in men the baseline BMI and increase in BMI were significant factors for the occurrence of SDB during NREM and REM sleep. In women, the baseline BMI and increase in BMI were only significant predictors of the occurrence of SDB during REM sleep. Third, there was a strong interdependence in the incident occurrence of a NREM- and REM-AHI of ≥5.0 events/h at follow-up, indicating that progression of SDB during NREM sleep is contemporaneous with progression during REM sleep. Finally, a baseline REM-AHI of ≥5 events/h and the development of a NREM-AHI of ≥5 events/h at follow-up was associated with cardiovascular disease in women but not in men.
Progression of SDB at both individual and population levels has been previously reported in the literature. Longitudinal data from the Wisconsin Sleep Cohort demonstrated that prevalence estimates of SDB have increased significantly over the last two decades, largely owing to temporal increases in BMI (13). Although there is some conflicting evidence (14–16), many studies suggest that at an individual level, SDB seems to advance over time with an increase in the AHI on subsequent testing, especially in individuals with snoring or mild-to-moderate SDB at baseline (17–21). Although increasing age has been a consistent predictor of SDB progression across studies, the association between weight gain and SDB progression is a topic of debate. Whereas some studies have described clear worsening or improvement of SDB occurring in tandem with increasing or decreasing BMI, respectively (14, 17, 22, 23), others have not reported such associations (19–21). The results of the current investigation corroborate that both baseline BMI and an increase in BMI, particularly in men, contribute to the progression of SDB in general, with some sex-related differences in the relative importance of the two factors.
Determining the significance of REM-related SDB on health outcomes has become a topic of great interest. Initial studies had mixed results in demonstrating an association between REM-related SDB and sleepiness and quality of life (24–27). Conversely, evidence showing the significance of REM-related SDB in cardiometabolic outcomes has been consistent. Cross-sectional analyses of both clinic- and community-based samples have shown an association between REM-AHI and glycosylated hemoglobin (28), as well as insulin resistance (29). Even more intriguing is the independent, longitudinal association of REM-related SDB with hypertension that was noted in sizeable cohort studies. REM-related SDB was associated with incident hypertension in both the Wisconsin Sleep Cohort and the MAILES (Men Androgens Inflammation Lifestyle Environment and Stress) study (4, 5). In addition, incident nondipping of nocturnal blood pressure was observed in subjects with severe REM-related SDB in the Wisconsin Sleep Cohort (30). Finally, severe REM-related SDB was found to be associated with recurrent cardiovascular events in men from a large community-based cohort (6). The collective findings from previous studies have considerable clinical implications. A major deficiency is the lack of information regarding the trajectory of REM-related SDB over time. Although there are longitudinal data on key cardiometabolic outcomes in the setting of REM-related SDB, there is a glaring absence of information regarding the impact of REM-related SDB over time. Specifically, there are no reports on the progression of REM-related SDB to more “general” SDB that includes NREM sleep stages. It is plausible that studies linking REM-related SDB to cardiometabolic outcomes are simply capturing SDB earlier along the spectrum of SDB severity, and progression to sleep state-nonspecific SDB is likely to occur in at least some individuals with REM-related SDB. Therefore, the positive associations noted between REM-related SDB and cardiometabolic outcomes may actually just reflect the well-established association between sleep state-nonspecific SDB and adverse health outcomes. The current investigation addresses this important knowledge gap and demonstrates a sex-specific difference in incident cardiovascular disease when SDB that occurs primarily in REM sleep progresses to NREM sleep stages. Women with REM-related SDB at baseline differed from men with regard to cardiovascular outcomes when there was progression of SDB to NREM. Finally, the results presented herein underscore the strong interdependence seen in the longitudinal occurrence of SDB during NREM and REM sleep.
Several strengths and weaknesses of the presented work are worth noting. Its strengths include the large sample size with strong representation of both sexes. In addition, the cohort included people across the severity spectrum of REM-related SDB at baseline. Finally, the generalizability of the findings is compelling given that the sample was community based. Weaknesses include the relatively short time interval between sleep studies. However, the time needed to advance from REM-related SDB to sleep state-nonspecific SDB is not clear and is likely to be quite variable. Despite the large sample size in the current study, there were a limited number of subjects with severe REM-related SDB breathing at baseline. Condensing these moderate and severe REM-related SDB categories precludes estimation of separate effects along the severity spectrum. Nonetheless, the findings presented fill an important gap in the current understanding of the natural history of SDB during REM sleep, and suggest a need to examine not only the long-term health consequences of the baseline severity of REM-related SDB but also its longitudinal trajectory. In conclusion, findings from the current study suggest that REM-related SDB is a relatively stable condition in the majority of individuals. Predictors of disease progression (e.g., age, BMI, and ΔBMI) vary by sex. Finally, incidence cardiovascular disease is most notable in women with REM-related SDB at baseline who progress to develop SDB during NREM sleep.
Supplementary Material
Footnotes
Supported by National Institutes of Health grants HL118414, HL117167, HL086862, and HL146709.
Author Contributions: N.M.P. takes responsibility for the integrity of the work as a whole from inception to publication. R.N.A., E.J.M., and N.M.P. were responsible for data analysis and interpretation. R.N.A. and N.M.P. were responsible for project inception. All authors contributed to the development of the final manuscript.
This article has an online supplement, which is accessible from this issue’s table of contents online at www.atsjournals.org
Author disclosures are available with the text of this article at www.atsjournals.org.
References
- 1.Grace KP, Hughes SW, Horner RL. Identification of the mechanism mediating genioglossus muscle suppression in REM sleep. Am J Respir Crit Care Med. 2013;187:311–319. doi: 10.1164/rccm.201209-1654OC. [DOI] [PubMed] [Google Scholar]
- 2.McSharry DG, Saboisky JP, Deyoung P, Jordan AS, Trinder J, Smales E, et al. Physiological mechanisms of upper airway hypotonia during REM sleep. Sleep (Basel) 2014;37:561–569. doi: 10.5665/sleep.3498. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Penzel T, Kantelhardt JW, Bartsch RP, Riedl M, Kraemer JF, Wessel N, et al. Modulations of heart rate, ECG, and cardio-respiratory coupling observed in polysomnography. Front Physiol. 2016;7:460. doi: 10.3389/fphys.2016.00460. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Appleton SL, Vakulin A, Martin SA, Lang CJ, Wittert GA, Taylor AW, et al. Hypertension is associated with undiagnosed OSA during rapid eye movement sleep. Chest. 2016;150:495–505. doi: 10.1016/j.chest.2016.03.010. [DOI] [PubMed] [Google Scholar]
- 5.Mokhlesi B, Finn LA, Hagen EW, Young T, Hla KM, Van Cauter E, et al. Obstructive sleep apnea during REM sleep and hypertension. results of the Wisconsin Sleep Cohort. Am J Respir Crit Care Med. 2014;190:1158–1167. doi: 10.1164/rccm.201406-1136OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Aurora RN, Crainiceanu C, Gottlieb DJ, Kim JS, Punjabi NM. Obstructive sleep apnea during REM sleep and cardiovascular disease. Am J Respir Crit Care Med. 2018;197:653–660. doi: 10.1164/rccm.201706-1112OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Quan SF, Howard BV, Iber C, Kiley JP, Nieto FJ, O’Connor GT, et al. The Sleep Heart Health Study: design, rationale, and methods. Sleep. 1997;20:1077–1085. [PubMed] [Google Scholar]
- 8.Redline S, Sanders MH, Lind BK, Quan SF, Iber C, Gottlieb DJ, et al. Sleep Heart Health Research Group. Methods for obtaining and analyzing unattended polysomnography data for a multicenter study. Sleep. 1998;21:759–767. [PubMed] [Google Scholar]
- 9.Berry RB, Budhiraja R, Gottlieb DJ, Gozal D, Iber C, Kapur VK, et al. American Academy of Sleep Medicine; Deliberations of the Sleep Apnea Definitions Task Force of the American Academy of Sleep Medicine. Rules for scoring respiratory events in sleep: update of the 2007 AASM Manual for the Scoring of Sleep and Associated Events. J Clin Sleep Med. 2012;8:597–619. doi: 10.5664/jcsm.2172. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Fietze I, Laharnar N, Obst A, Ewert R, Felix SB, Garcia C, et al. Prevalence and association analysis of obstructive sleep apnea with gender and age differences—results of SHIP-Trend. J Sleep Res. 2019;28:e12770. doi: 10.1111/jsr.12770. [DOI] [PubMed] [Google Scholar]
- 11.Theorell-Haglöw J, Miller CB, Bartlett DJ, Yee BJ, Openshaw HD, Grunstein RR. Gender differences in obstructive sleep apnoea, insomnia and restless legs syndrome in adults—what do we know? A clinical update. Sleep Med Rev. 2018;38:28–38. doi: 10.1016/j.smrv.2017.03.003. [DOI] [PubMed] [Google Scholar]
- 12.le Cessie S, van Houwelingen JC. Logistic regression for correlated binary data. J R Stat Soc Ser C Appl Stat. 1994;43:95–108. [Google Scholar]
- 13.Peppard PE, Young T, Barnet JH, Palta M, Hagen EW, Hla KM. Increased prevalence of sleep-disordered breathing in adults. Am J Epidemiol. 2013;177:1006–1014. doi: 10.1093/aje/kws342. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Fisher D, Pillar G, Malhotra A, Peled N, Lavie P. Long-term follow-up of untreated patients with sleep apnoea syndrome. Respir Med. 2002;96:337–343. doi: 10.1053/rmed.2001.1277. [DOI] [PubMed] [Google Scholar]
- 15.Quan SF. Evolution of OSA. Thorax. 1998;53:532. doi: 10.1136/thx.53.6.532. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Sforza E, Addati G, Cirignotta F, Lugaresi E. Natural evolution of sleep apnoea syndrome: a five year longitudinal study. Eur Respir J. 1994;7:1765–1770. doi: 10.1183/09031936.94.07101765. [DOI] [PubMed] [Google Scholar]
- 17.Berger G, Berger R, Oksenberg A. Progression of snoring and obstructive sleep apnoea: the role of increasing weight and time. Eur Respir J. 2009;33:338–345. doi: 10.1183/09031936.00075408. [DOI] [PubMed] [Google Scholar]
- 18.Inoue Y, Nanba K, Hazama G, Takata K, Arai H. Long-term follow-up study on patients with sleep apnea syndrome. Psychiatry Clin Neurosci. 2001;55:245–246. doi: 10.1046/j.1440-1819.2001.00843.x. [DOI] [PubMed] [Google Scholar]
- 19.Lindberg E, Elmasry A, Gislason T, Janson C, Bengtsson H, Hetta J, et al. Evolution of sleep apnea syndrome in sleepy snorers: a population-based prospective study. Am J Respir Crit Care Med. 1999;159:2024–2027. doi: 10.1164/ajrccm.159.6.9805070. [DOI] [PubMed] [Google Scholar]
- 20.Pendlebury ST, Pépin JL, Veale D, Lévy P. Natural evolution of moderate sleep apnoea syndrome: significant progression over a mean of 17 months. Thorax. 1997;52:872–878. doi: 10.1136/thx.52.10.872. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Sahlman J, Pukkila M, Seppä J, Tuomilehto H. Evolution of mild obstructive sleep apnea after different treatments. Laryngoscope. 2007;117:1107–1111. doi: 10.1097/MLG.0b013e3180514d08. [DOI] [PubMed] [Google Scholar]
- 22.Svanborg E, Larsson H. Natural evolution of obstructive sleep apnea syndrome. Sleep. 1993;16(8) Suppl:S124–S125. doi: 10.1093/sleep/16.suppl_8.s124. [DOI] [PubMed] [Google Scholar]
- 23.Svanborg E, Larsson H. Development of nocturnal respiratory disturbance in untreated patients with obstructive sleep apnea syndrome. Chest. 1993;104:340–343. doi: 10.1378/chest.104.2.340. [DOI] [PubMed] [Google Scholar]
- 24.Chami HA, Baldwin CM, Silverman A, Zhang Y, Rapoport D, Punjabi NM, et al. Sleepiness, quality of life, and sleep maintenance in REM versus non-REM sleep-disordered breathing. Am J Respir Crit Care Med. 2010;181:997–1002. doi: 10.1164/rccm.200908-1304OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Kass JE, Akers SM, Bartter TC, Pratter MR. Rapid-eye-movement-specific sleep-disordered breathing: a possible cause of excessive daytime sleepiness. Am J Respir Crit Care Med. 1996;154:167–169. doi: 10.1164/ajrccm.154.1.8680674. [DOI] [PubMed] [Google Scholar]
- 26.Khan A, Harrison SL, Kezirian EJ, Ancoli-Israel S, O’Hearn D, Orwoll E, et al. Osteoporotic Fractures in Men (MrOS) Study Research Group. Obstructive sleep apnea during rapid eye movement sleep, daytime sleepiness, and quality of life in older men in Osteoporotic Fractures in Men (MrOS) Sleep Study. J Clin Sleep Med. 2013;9:191–198. doi: 10.5664/jcsm.2474. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Su CS, Liu KT, Panjapornpon K, Andrews N, Foldvary-Schaefer N. Functional outcomes in patients with REM-related obstructive sleep apnea treated with positive airway pressure therapy. J Clin Sleep Med. 2012;8:243–247. doi: 10.5664/jcsm.1902. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Grimaldi D, Beccuti G, Touma C, Van Cauter E, Mokhlesi B. Association of obstructive sleep apnea in rapid eye movement sleep with reduced glycemic control in type 2 diabetes: therapeutic implications. Diabetes Care. 2014;37:355–363. doi: 10.2337/dc13-0933. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Chami HA, Gottlieb DJ, Redline S, Punjabi NM. Association between glucose metabolism and sleep-disordered breathing during REM sleep. Am J Respir Crit Care Med. 2015;192:1118–1126. doi: 10.1164/rccm.201501-0046OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Mokhlesi B, Hagen EW, Finn LA, Hla KM, Carter JR, Peppard PE. Obstructive sleep apnoea during REM sleep and incident non-dipping of nocturnal blood pressure: a longitudinal analysis of the Wisconsin Sleep Cohort. Thorax. 2015;70:1062–1069. doi: 10.1136/thoraxjnl-2015-207231. [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.