TABLE 3.
Studies on rapid eye movement (REM) obstructive sleep apnoea (OSA) in clinical samples
| First author [ref.], year | Study type; sample | Diagnostic criteria | Outcomes | Results |
| O’Connor [43], 2000 | Cross-sectional retrospective in 838 OSA patients (mean age: men 48.6 years; women 50.8 years; women 24.6%) | OSAREM: overall AHI between 10 and 25 events·h−1, AHIREM/AHINREM >2 and AHINREM <10 events·h−1 |
Sex-related differences in OSA | Prevalence of OSAREM: 24% in men, 62% in women. Lower overall AHI in women secondary to low AHINREM, AHIREM not different between sexes. Largest difference between men and women in the mild–moderate OSA range. No relationship with age in both men and women. |
| Resta [44], 2005 | Observational; 45 severely obese OSA patients (20 women age- and weight-matched to 25 men), mean age 44 years, BMI 40 kg·m−2 | OSAREM: overall AHI between 10 and 25 events·h−1, AHIREM/AHINREM >2 events·h−1 and AHINREM <10 events·h−1 | Sex-related differences in sleep and OSA in severely obese patients | Prevalence of OSAREM: 35% in women and 4% in men. Lower OSA severity and sleep efficiency and higher number of awakenings in patients with OSAREM. |
| Haba-Rubio [45], 2005 | Cross-sectional retrospective; 415 OSA patients (mean age 54.1 years, 73% men) | AHIREM/AHINREM >2 Exclusion criteria: 1) AHI ≤5 events·h−1 of TST; 2) previous treatment for SDB; 3) REM sleep <15% of TST during nocturnal recording |
Frequency and clinical characteristics of OSAREM including subjective (ESS) and objective (MWT, n=228) sleepiness | SDBREM in 36.4% of the sample (women: 46.4%, men 53.6%). No difference in symptoms or sleepiness between SDBREM and SDBNREM groups, but OSA more severe in patients with SDBNREM. Declining frequency of SDBREM from mild to severe OSA. SDBREM associated with higher BMI in mild–moderate OSA and more frequent in women except in severe OSA cases. |
| Koo [55], 2008 | Observational; 2486 OSA patients (mean age 50.8 years, 67.1% men) | At least 10 min of REM sleep at PSG OSAREM criteria: 1) overall AHI ≥5 events·h−1; 2) AHINREM <15 events·h−1; and 3) AHIREM/AHINREM ≥2 |
Sex-related differences in OSA | Prevalence of OSAREM: 21% in men, 40.8% in women. Patients with OSAREM were younger and showed longer REM sleep duration. Prevalence of OSAREM decreased with age and increasing BMI in both men and women. No difference in positional AHI was found for any sleep stage between sexes. |
| Koo [56], 2008 | Cross-sectional retrospective; 221 patients with OSAREM (mean age 50 years, 33.5% men) | OSAREM criteria: 1) age ≥18 years; 2) overall AHI ≥5 events·h−1; 3) AHINREM <15 events·h−1; 4) AHIREM/AHINREM >2; and 5) time spent in REM sleep >10 min | Effects of age and sex in OSAREM | Prevalence of OSAREM in the entire sample of OSA patients (n=1540): 14.4% (24.5% of women, 7.9% of men in the entire sample). Comorbidities: depression in 41.2%, at least one cardiovascular risk factor in 67.7% and EDS in 68.1%. In both men and women, OSAREM more frequent in patients younger than 55 years of age and directly related to BMI in women. |
| Pamidi [53], 2011 | Cross-sectional; 931 consecutive OSA patients (mean age 50 years, 44.4% men) | Overall AHI ≥5 events·h−1, with AHIREM/ AHINREM ≥2, AHINREM< 8 events·h−1 and REM duration >10.5 min | Association of OSAREM with subjective sleepiness (ESS) and QoL by SF-12 | Prevalence of OSAREM: 13.5%. OSAREM patients were younger and more often women (76.2 versus 52.4%) compared to nonstage-specific OSA patients. AHIREM did not predict sleepiness or QoL. Depressive symptoms and BMI predicted ESS and QoL in the OSAREM group. |
| Conwell [57], 2012 | Cross-sectional; 931 consecutive OSA patients (mean age 50 years, 44.4% men) | Definition 1: overall AHI ≥5 events·h−1 and AHIREM/AHINREM ≥2 Definition 2: overall AHI ≥5 events·h−1, AHIREM/ AHINREM ≥2 and AHINREM <15 events·h−1 Definition 3: overall AHI ≥5 events·h−1, AHIREM/AHINREM ≥2, AHINREM <8 events·h−1 and at least 10.5 min of REM sleep duration |
Prevalence of OSAREM according to different definitions | Prevalence of OSAREM varied according to the definition used (1: 36.7%; 2: 24.4%; 3: 13.5%). OSAREM more prevalent in women (78 versus 48% in men), younger individuals (mean age 45 versus 52 years) and African Americans. Similar degrees of obesity and sleepiness, better sleep quality, and lower prevalence of diabetes and HT in OSAREM compared to patients with nonsleep stage-dependent OSA. |
| Sakao [42], 2015 | Cross-sectional; 468 patients with suspected OSA (mean age 54.9 years, 22.9% women) | Three definitions of OSAREM: I: overall AHI ≥5 events·h−1 and AHIREM/AHINREM ≥2 II: AHINREM <15 events·h−1 in addition to I III: AHINREM <8 events·h−1 and at least 10.5 min of REM sleep in addition to I |
Prevalence and features of OSAREM in Japanese subjects | Prevalence of OSAREM: 24.8% (I), 17.6% (II) and 11% (III). In women, prevalence of OSAREM increased from 33.6 to 40.1% from definition I to III and was higher than the prevalence of OSANREM. Subjects with OSAREM showed lower BMI and HbA1c levels than subjects with OSANREM. |
| Lee [46], 2016 | Cross-sectional retrospective; 1281 Korean OSA patients (mean age 54 years, 18% women) | Overall AHI ≥5 events·h−1, AHINREM <15 events·h−1 and AHIREM to AHINREM ratio >2 | Association of OSAREM with sleepiness (ESS), depressive symptoms (BDI) and health-related QoL (SF-36) | Prevalence of OSAREM: 18% (32.6% in women, 14.1% in men). OSAREM more frequent in mild–moderate than severe OSA. Significant association of OSAREM with depressive symptoms only in men. Sleepiness or QoL similar in OSAREM and non-REM related OSA groups. |
| Al Oweidat [54], 2018 | Cross-sectional; 478 Jordanian patients with OSA (mean age 55.3 years, 44.6% women) | Overall AHI ≥5 events·h−1 Broad definition of OSAREM: AHIREM/AHINREM ≥2; strict definition: AHINREM<5 events·h−1, AHIREM >5 events·h−1 and at least 30 min of REM sleep; OSANREM: AHIREM/AHINREM <2 |
Differences in demographic and PSG features between REM- and NREM-related OSA | Severe OSA in 72% of the sample. Prevalence of OSAREM: 18% (31% in women, 7.5% in men) according to the broad definition, 2.7% (5.2% in women, 0.8% in men) according to the strict definition. Higher arousal index and time spent at SpO2<90% and lower SpO2nadir in patients with OSANREM compared to patients with OSAREM for both broad and strict definitions. No differences in BMI, ESS and snoring between the two groups. |
| Mano [58], 2019 | Retrospective cross-sectional; 3234 Japanese OSA patients (mean age: 52.5 years, 14.5% women) | Three definitions of OSAREM: I: overall AHI ≥5 events·h−1 and AHIREM/AHINREM ≥2 II: AHINREM <15 events·h−1 in addition to I III: AHINREM <8 events·h−1 and at least 10.5 min of REM sleep in addition to I |
Effect of sex and age on OSAREM In women, analysis of the effects of menopause, defined as age >50 years | Overall prevalence of OSAREM 24.6%, 18.6% and 12.2% according to definitions I–III. In men, prevalence of OSAREM decreased with age, from 22.8 in men under 50 years to 19.1% in men over 50 years (definition I). Corresponding values in women were 44.3% under 50 years and 47.7% over 50 years (definition I). In multivariate analysis, adjustment for BMI and CT90 slightly decreased significance, whereas further adjustment for AHINREM strongly reduced the difference between sexes, both below and above age 50. |
| Bahammam [47], 2020 | Prospective observational; 2169 OSA patients (mean age: 46.7 years, 38% women) | Predominant OSAREM: overall obstructive AHI ≥5 events·h−1, AHINREM <15 events·h−1 and AHIREM/AHINREM ≥2. OSANSS: AHI ≥5 events·h−1, criteria for OSAREM not fulfilled | Analysis of clinical and sleep features, and of comorbidities | Prevalence of OSAREM: 17% (25% in women, 12% in men). OSAREM was more frequent at younger age but was unrelated to menopause at multivariate analysis. In OSAREM, frequent nocturnal chest pain, headache at awakening, nocturnal awakening with palpitations and higher prevalence of bronchial asthma, while snoring and overall prevalence of HT and ischaemic heart disease were lower than in OSANSS. In men, OSAREM was independently associated with younger age, HT, bronchial asthma, high sleep efficiency, low amount of NREM sleep stage 1 and lower AHI and SpO2 nadir. In women, OSAREM was independently associated with younger age, higher BMI, less HT, hypothyroidism, less sleepiness, low amount of NREM sleep stage 1 and lower amount of time spent at SpO2: <90%. |
| Sasai-Sakuma [48], 2021 | Retrospective cross-sectional; 1458 Japanese OSA patients (median age 48 years, 9.7% women) | Three definitions of OSAREM: I: overall AHI ≥5 events·h−1 and AHIREM/AHINREM ≥2 II: AHINREM <15 events·h−1 in addition to I III: AHINREM <8 events·h−1 and at least 10.5 min of REM sleep in addition to I |
Effects of gender and OSA severity on OSAREM prevalence | Prevalence of OSAREM (definition II) was 0% in severe OSA, 18.9% in moderate and 47% in mild OSA. Compared with OSANSS, patients with OSAREM showed higher BMI and female predominance, lower AHI and ODI, and higher sleep efficiency, but no difference in prevalence of daytime sleepiness or cardiometabolic comorbidities. |
| Chiu [49], 2022 | Retrospective cross-sectional; 1331 Korean OSA patients aged >20 years (median age 53 years, 20.5% women) | REM sleep: at least 30 min. OSAREM: AHIREM/AHINREM >2 and AHINREM <15 events·h−1; OSANREM: AHIREM/AHINREM ≤2; OSANSS: AHIREM/AHINREM >2 and AHINREM ≥15 events·h−1; isolated OSAREM: AHIREM >5 events·h−1 with AHINREM <5 events·h−1 | Clinical demographics, OSA-related symptoms, PSG results and medical comorbidities in OSAREM, OSANREM and OSANSS | Prevalence of OSAREM: 31.1%, OSANREM: 60.7%, OSANSS: 8.2%. OSAREM more frequent in women (53.1 versus 25.4% in men), mild–moderate than severe OSA based on overall AHI and associated with longer duration of desaturations and lower SpO2 nadir. No difference in ESS score or prevalence of comorbidities compared to the other groups. |
| Lee [50], 2022 | Retrospective cross-sectional; 692 Korean OSA patients (mean age 50.3 years, 28.2% women) | REM sleep: at least 30 min; OSAREM: overall AHI >5 events·h−1, AHIREM/AHINREM ≥2 and AHINREM <15 events·h−1 | Prevalence and clinical characteristics of OSAREM | Prevalence of OSAREM: 20.2% (38.4% of women, 13% of men). In the OSAREM group, females represented 53.6%. OSAREM present in 69.3% of patients with mild OSA and 30% of patients with moderate OSA. Prevalence of HT and diabetes lower in patients with OSAREM than in patients with OSANREM. |
| Sattaratpaijit [51], 2022 | Retrospective cross-sectional; 408 Thai OSA patients (mean age 49.7 years, 39.9% women) | At least 10.5 min of REM sleep; OSAREM: overall AHI ≥5 events·h−1, AHIREM/AHINREM >2 and AHINREM <15 events·h−1 | Prevalence and clinical characteristics of OSAREM | Prevalence of OSAREM: 21.6% (29.4% of women, 16.3% of men). OSAREM significantly associated with female sex (OR 2.35), age <60 years (OR 2.52) and mild OSA (OR 17.46). |
| Qanash [52], 2023 | Retrospective cross-sectional; 609 Saudi OSA patients (mean age 49 years, 42% women) | Strict definition: AHI ≥5 events·h−1, AHIREM/AHINREM ≥2, AHINREM <15 events·h−1, at least 30 min of REM sleep Intermediate definition AHI ≥5 events·h−1, AHIREM/AHINREM ≥2, AHINREM <15 events·h−1 Lenient definition: AHI ≥5 events·h−1, AHIREM/AHINREM ≥2 |
Prevalence and clinical characteristics of OSAREM | Prevalence of OSAREM: 26% (women 36%, men 18%) according to the strict definition, 33% (women 48.4%, men 21.8%) according to the moderate definition and 52% (women 69.5%, men 35.4%) according to the lenient definition. Milder OSA and desaturation severity in OSAREM compared to OSANREM. No differences in ESS or prevalence of comorbidities between OSAREM and OSANREM. |
| Huang [61], 2023 | Cross-sectional; 4152 subjects with suspected OSA (mean age 49 years, 21% women), Shangai Sleep Health Study cohort | Stratification of the sample according to AHIREM in the no OSA and mild, moderate and severe OSA ranges Analysis in the subgroup with AHINREM <5 events·h−1 |
Association of OSAREM with cardiovascular risk, assessed as FRS and autonomic imbalance, assessed as HRV parameters in both subjects with and without prevalent CVD | Severe OSAREM associated with increased FRS, but not with prevalent CVD, in the multivariate analysis. High LF/HF, high LF and high HF associated with AHIREM in the subgroup with AHINREM<5 events·h−1. HRV parameters mediated the relationship between AHIREM and both prevalence of CVD and high FRS. |
AHI: apnoea–hypopnoea index; BDI: Beck Depression Inventory; BMI: body mass index; CT90: cumulative time spent at SpO2 <90%; CVD: cardiovascular disease; EDS: excessive daytime sleepiness; ESS: Epworth Sleepiness Scale; FRS: Framingham Risk Score; HbA1c: glycosylated haemoglobin; HF: high-frequency spectrum; HRV: heart rate variability; HT: hypertension; LF: low-frequency spectrum; MWT: maintenance of wakefulness test; ODI: oxygen desaturation index; NREM: non-REM; OR: odds ratio; OSANSS: nonstage-specific OSA; PSG: polysomnography; QoL: quality of life; SDB: sleep-disordered breathing; SF-12: Short Form-12; SF-36: Short Form-36; SpO2: oxygen saturation measured by pulse oximetry; TST: total sleep time.