International Consensus Statement on Obstructive Sleep Apnea

Large studies failed to detect significant differences in symptoms or clinical outcomes between REM-related and non-REM-related OSA.

“large studies failed to detect significant differences in symptoms or clinical outcomes between patients with REM-related (or REM-predominant) versus non-REM-related OSA.50,60–62”

• cites [50] sourceNot checked — no full text of the source
• cites [60] sourceDisputed

  The Discussion (page unavailable) reports different findings for REM and non-REM sleep-disordered breathing: REM-related breathing was not associated with sleepiness or poorer quality of life, whereas non-REM-related breathing was associated with both. It says the two should be distinguished in managing patients with daytime sleepiness, rather than concluding that the large study found no clinical distinction. The paper does not establish whether broader outcomes, such as cardiovascular outcomes, differ. (LLM, conf=0.83)

  Checked by a reasoning model — may be wrong.

  Experimental — accuracy not yet measured
• cites [61] sourcePartly supported

  The Results report no significant variation in sleepiness or most quality-of-life measures across REM-predominant OSA categories, supporting part of the claim. But they also report poorer objectively measured sleep quality and worse self-perceived health at AHI-REM ≥ 30. The study does not establish the claim's broad comparison between patients with REM-related and non-REM-related OSA. (LLM, conf=0.86)

  Experimental — accuracy not yet measured
• cites [62] sourcePartly supported

  The Abstract reports similar baseline symptom and functional-outcome scores in 330 patients with REM-related versus other OSA, as well as similar changes after PAP treatment. But the Results report a higher prevalence of depression and cardiovascular disease in the non-REM-restricted group, so the claim is too broad if it includes all clinical outcomes. (LLM, conf=0.84)

  Experimental — accuracy not yet measured

Some data suggest that a low arousal threshold may be a risk factor for failure of upper airway surgery to achieve a surgical cure.

“Some data suggest that a low arousal threshold may be a risk factor for failure of UA surgery to achieve a surgical cure.”

• cites [363] sourceDisputed

  The “Modeling Response to Surgery” section says low arousal threshold was correlated with better surgical outcome, not failure to achieve a cure. The “Baseline Loop Gain, Arousal Threshold and Surgical Outcome” section likewise reports an association with lower postoperative AHI. Page numbers are unavailable in our copy. (LLM, conf=0.91)

  Checked by a reasoning model — may be wrong.

  Experimental — accuracy not yet measured
• cites [364] sourceCould not verify

  The Summary of the Findings discusses variable surgical success and identifies high loop gain as a possible contributor to poor response. The Results report surgical outcomes, but neither section links a low arousal threshold to failure to achieve a cure; searches for “low arousal threshold” and “arousal threshold” found no such finding in our copy. Page numbers were unavailable. (LLM, conf=0.74)

  Experimental — accuracy not yet measured
• cites [416] sourcePartly supported

  The Introduction (page unavailable) says a low arousal threshold might predispose patients to repetitive apnea rather than stable breathing. That supports the proposed physiological risk, but the paper does not address whether a low threshold predicts failure of upper-airway surgery to achieve a cure. (LLM, conf=0.78)

  Experimental — accuracy not yet measured

In general, increasing frequency of each symptom occurs with increasing AHI and OSA severity.

“In general, increasing frequency of each symptom occurs with increasing AHI and OSA severity.”

• cites [88] sourceNot checked — no full text of the source
• cites [456] sourceSupported
• cites [505] sourceDisputed

  The Clinical Implications section says symptom–severity dissociation is prominent in OSA patients, contrary to the claimed general increase in symptoms with severity. The Database section also says the questionnaire did not measure symptom frequency and the analysis did not subdivide OSA by severity category. Page numbers are unavailable in our copy. (LLM, conf=0.85)

  Checked by a reasoning model — may be wrong.

  Experimental — accuracy not yet measured

There is increasing evidence supporting use of some home-based type III and IV sleep studies to rule-in but not rule-out moderate to severe OSA.

“There is increasing evidence supporting the use of some home-based type III and IV type sleep studies to “rule-in” but not “rule-out”580 moderate to severe OSA.”

• cites [580] sourceDisputed

  The Discussion (page not available) states that home respiratory polygraphy can both exclude and confirm sleep apnoea–hypopnoea syndrome. The Diagnostic efficacy section likewise reports cut-offs that effectively excluded it, including at a polysomnographic AHI threshold of 15. Although the Introduction describes a recommendation to follow a negative home test with polysomnography, the paper’s own findings contradict the claim that these studies can rule in but not rule out disease; it studies type III monitoring, not type IV. (LLM, conf=0.88)

  Checked by a reasoning model — may be wrong.

  Experimental — accuracy not yet measured

The onset of myocardial infarction is more likely to occur during the nighttime.

“In addition, the onset of MI is more likely to be during the nighttime.666”

• cites [666] sourceDisputed

  The Discussion says nighttime MI onset is increased in patients with obstructive sleep apnea compared with patients without it. It also says patients without sleep apnea have a nighttime nadir and a morning peak, like the general population. The unqualified claim that MI onset is more likely at night therefore reverses the paper’s general finding. (LLM, conf=0.88)

  Checked by a reasoning model — may be wrong.

  Experimental — accuracy not yet measured

Subjective sleep quality improves significantly after medical or surgical treatment of chronic rhinosinusitis.

“with significant improvement in subjective sleep quality after medical and surgical treatment.914–916”

• cites [914] sourceNot checked — no full text of the source
• cites [915] sourceNot checked — no full text of the source
• cites [916] sourceDisputed

  The Treatment: Medical vs. Surgical discussion reports significant improvement in subjective sleep-quality scores after surgery, but says scores among patients receiving medical management worsened. Thus the review does not support significant improvement after either treatment. (LLM, conf=0.88)

  Checked by a reasoning model — may be wrong.

  Experimental — accuracy not yet measured

Nocturnal reflux symptoms in obstructive sleep apnea are associated with sleep fragmentation and poor sleep quality.

“which have been associated with sleep fragmentation and overall poor sleep quality.927,930,936”

• cites [927] sourceNot checked — no full text of the source
• cites [930] sourceNot checked — no full text of the source
• cites [936] sourceDisputed

  The Results (page not provided) report that arousal indexes in the 15-, 30-, and 45-minute periods around reflux episodes were lower than the full-night arousal index, contrary to the attributed association with sleep fragmentation. The paper does not report a measure of overall sleep quality, so it cannot establish that part of the claim. (LLM, conf=0.84)

  Checked by a reasoning model — may be wrong.

  Experimental — accuracy not yet measured

Data were insufficient to analyze differences in quality of life or interface-associated side effects.

“Data were insufficient to perform analyses examining differences in QOL or interface-associated side effects.”

• cites [39] sourceDisputed

  The Mask Selection section (page unavailable) reports analyses of both outcomes: studies found no clinically significant QOL difference between intranasal and nasal interfaces, and fewer side effects with nasal than with oronasal or oral interfaces. It therefore contradicts the claim that data were insufficient to analyze QOL or interface-associated side effects. (LLM, conf=0.88)

  Checked by a reasoning model — may be wrong.

  Experimental — accuracy not yet measured

Individuals with OSA who report daytime sleepiness appear to have greater decline in blood pressure compared with those who are not sleepy.

“Individuals with OSA who report daytime sleepiness appear to demonstrate greater decline in BP compared to those who are not sleepy.39,1211”

• cites [39] sourcePartly supported

  In the PAP Therapy section (page unavailable), the guideline reports blood-pressure reductions with PAP in patients with OSA and hypertension. It says sleepy patients may be more accepting of PAP and may receive blood-pressure benefits, but it does not report a greater blood-pressure decline in sleepy than in non-sleepy patients. (LLM, conf=0.84)

  Experimental — accuracy not yet measured
• cites [1211] sourceDisputed

  In the Subgroup Analyses (Table 3; page not provided), the group with lower Epworth Sleepiness Scale scores (<10) had a larger 24-hour diastolic BP reduction with CPAP than the group with higher scores (≥10): −3.24 versus −2.44 mm Hg. This is the opposite of the claimed direction, although the table does not establish a statistically significant difference between the subgroups. (LLM, conf=0.86)

  Checked by a reasoning model — may be wrong.

  Experimental — accuracy not yet measured

A series of calibration equations were published to allow imputation of AHI levels across definitions.

“A series of calibration equations were published to allow imputation of AHI levels across definitions.14”

• cites [14] sourceSupported

In a cross-sectional analysis of the community-based HypnoLaus cohort, adjusted thresholds for defining moderate and severe OSA were shown to associate with both hypertension and diabetes.

“In a cross-sectional analysis of the community-based general population, HypnoLaus cohort, adjusted thresholds for defining moderate and severe OSA were shown to associate with both hypertension and diabetes.”

• cites [16] sourceSupported

In the Sleep Heart Health Study, application of the current AASM recommended definition (3% desaturation or arousal) resulted in an approximately doubling of the number of individuals classified with moderate OSA (AHI > 15).

“In the Sleep Heart Health Study (SHHS), the application of the current AASM recommended definition (3% desaturation or arousal) resulted in an approximately doubling of the number of individuals classified with moderate OSA (AHI > 15).”

• cites [17] sourceSupported

This study suggested that thresholds for defining disease severity need to be reduced by approximately one half for hypopnea definitions using a 4% desaturation criteria compared to the AASM hypopnea definition.

“Specifically, this study suggested that the thresholds for defining disease severity need to be reduced by approximately one half for hypopnea definitions using a 4% desaturation criteria compared to the AASM hypopnea definition.16”

• cites [16] sourcePartly supported

  The Scoring of polysomnography section identifies the stricter AASM2007 definition as requiring at least 4% oxygen desaturation. The Discussion says AASM2012 criteria produce higher AHI values and approximately twice as many OSA diagnoses, while Statistical analysis says the authors estimated equivalent AHI thresholds. The passages retrieved do not establish that disease-severity thresholds under the 4% definition should be reduced by approximately one half. (LLM, conf=0.75)

  Experimental — accuracy not yet measured

In the unattended PSGs in the Sleep Heart Health Study, events were more reliably scored when desaturation criteria alone were used in comparison to inclusion of the arousal criteria.

“For example, in the unattended PSGs in the SHHS, events were more reliably scored when desaturation criteria alone were used in comparison to inclusion of the arousal criteria.23”

• cites [23] sourceSupported

With aging, women may experience increased rates of death and incident heart failure.

“but with aging, may experience increased rates of death and incident heart failure (HF).19”

• cites [19] sourcePartly supported

  The Results (page unavailable) report that, over 13.6 years, greater obstructive sleep apnea severity was associated with the combined outcome of incident heart failure or death in women. The paper does not establish that aging itself increases those rates, and it reports the two outcomes as a composite rather than demonstrating an increase in each separately. (LLM, conf=0.85)

  Experimental — accuracy not yet measured

In REM sleep, genioglossus muscle activity is lower than in NREM sleep.

“In REM, the genioglossus muscle activity is lower,47”

• cites [47] sourceSupported

More hypoxemia as measured by T90 and lower SaO2 nadir in those with similar AHI is associated with more inflammation (measured by C-reactive protein, platelet count, and endothelial stiffness).

“It has been shown that more hypoxemia as measured by T90 and lower SaO2 nadir in those with similar AHI is associated with more inflammation (measured by C-reactive protein [CRP], platelet count, and endothelial stiffness).28–30”

• cites [28] sourceNot checked — no full text of the source
• cites [29] sourcePartly supported

  The Discussion reports an association between nocturnal intermittent hypoxia, measured by 3% oxygen desaturation index, and higher hs-CRP in middle-aged community residents. The paper does not establish the claimed association using T90 or SaO2 nadir among people with similar AHI, and it does not report platelet count or endothelial stiffness as inflammation outcomes. Page numbers were not available in our copy. (LLM, conf=0.88)

  Experimental — accuracy not yet measured
• cites [30] sourceNot checked — no full text of the source

The ODI may be particularly useful to identify hypoxia-related stresses, which have been related to metabolic disease.

“Conversely, the ODI may be particularly useful to identify hypoxia-related stresses, which have been related to metabolic disease.16”

• cites [16] sourcePartly supported

  The Introduction and Scoring of polysomnography sections link hypopnea definitions involving oxygen desaturation to the study of diabetes and metabolic syndrome. The ODI table defines ODI as the oxygen desaturation index, but the passage evaluates its association with sleepiness, not metabolic outcomes. Our copy therefore supports the desaturation and metabolic-disease connection only indirectly; it does not establish that ODI is particularly useful for identifying metabolic-disease-related stresses. (LLM, conf=0.70)

  Experimental — accuracy not yet measured

Daytime impairment in functioning tends to be worse in UARS patients than in those with mild OSA.

“Daytime impairment in functioning tends to be worse in patients with UARS compared to those with mild OSA.74,76”

• cites [74] sourceNot checked — no full text of the source
• cites [76] sourceSupported

There is a very high prevalence of OSA in those with coronary artery disease, hypertension, heart failure, arrhythmias, and especially atrial fibrillation.

“There is a very high prevalence of OSA in those with CAD, hypertension, HF, arrhythmias, and especially AF.109–112”

• cites [109] sourceNot checked — no full text of the source
• cites [110] sourceNot checked — no full text of the source
• cites [111] sourcePartly supported

  The Abstract and Introduction (page numbers unavailable) say OSA is common in patients with cardiovascular disease and highly prevalent among cardiology patients; they also link it to coronary disease, heart failure, and atrial fibrillation. They do not establish high OSA prevalence separately in every listed group, particularly people with hypertension or arrhythmias, or say prevalence is especially high in atrial fibrillation. (LLM, conf=0.84)

  Experimental — accuracy not yet measured
• cites [112] sourceNot checked — no full text of the source

A more recent population-based study in Switzerland found a very high prevalence: 23% in women and 50% in men.

“A more recent population-based study in Switzerland found a very high prevalence: 23% in women and 50% in men.97”

• cites [97] sourceSupported

A recent meta-analysis estimated that 49% of subjects hospitalized with cardiac disease (heart failure, acute coronary syndrome, and arrhythmias) had previously undiagnosed OSA.

“A recent MA estimated 49% of subjects hospitalized with cardiac disease (HF, acute coronary syndrome, and arrhythmias) had previously undiagnosed OSA.108”

• cites [108] sourcePartly supported

  Results §3.2 reports a pooled OSA prevalence of 48% among hospitalized cardiac patients; 49% applies specifically to the acute coronary syndrome subgroup, not to all the cardiac conditions listed. Results §3.1 includes one study of patients with cardiovascular disease such as new arrhythmias, but gives no separate arrhythmia estimate. Discussion §5.1 says the pooled figure may estimate previously undiagnosed OSA because diagnosis occurred after admission, rather than establishing that all cases were previously undiagnosed. (LLM, conf=0.90)

  Experimental — accuracy not yet measured

The 4% desaturation criteria for hypopneas may be associated with a more cohesive epidemiologic group-typically older, male, obese, with underlying cardiac disease or diabetes.

“The 4% desaturation criteria for hypopneas (≥4% oxygen desaturation) may be associated with a more cohesive group epidemiologically – typically older, male, obese, and with underlying cardiac disease (e.g., HF, coronary artery disease [CAD], and atrial fibrillation [AF]) or diabetes.”

• cites [36] sourceCould not verify

  The Subjects section says participants were otherwise healthy, and Table 1 in Patients with OSA versus Control Subjects reports age, sex, and BMI. Neither passage links those characteristics—or cardiac disease or diabetes—to a ≥4% desaturation criterion for hypopneas. Searches for the criterion and the listed comorbidities found no passage addressing the claimed epidemiologic group; page numbers were unavailable in our copy. (LLM, conf=0.75)

  Experimental — accuracy not yet measured

In obese individuals referred to sleep clinics, OHS prevalence increases to 10%-20%.

“However, in obese individuals referred to sleep clinics, prevalence rate increases to 10%–20%.84,85”

• cites [84] sourcePartly supported

  The Discussion (page unavailable) reports that previous studies found OHS prevalence of 9–20% among patients with obstructive sleep apnea, broadly matching the quoted range. But that evidence concerns patients with OSA, not all obese individuals referred to sleep clinics; this study itself found 8.5% among its clinic patients with OSA. (LLM, conf=0.86)

  Experimental — accuracy not yet measured
• cites [85] sourcePartly supported

  The Introduction reports a 17% aggregate OHS prevalence among patients with obstructive sleep apnea (OSA) referred to a sleep disorder center, consistent with the claimed 10%–20% range. The estimate is for patients with OSA, however, not all obese individuals referred to sleep clinics; the reported prevalence across studies also ranges from 4% to 50%. (LLM, conf=0.84)

  Experimental — accuracy not yet measured

A smaller subset, 3%-7% in men and 2%-5% in women, had OSAS based on AHI ≥5 with excessive daytime sleepiness.

“A smaller subset, 3%–7% in men and 2%–5% in women, had OSAS based on AHI ≥ 5 with excessive daytime sleepiness.88,90,92”

• cites [88] sourceNot checked — no full text of the source
• cites [90] sourcePartly supported

  The Occurrence of Hypersomnolence section (p. 3) estimates that 4% of men and 2% of women had both an apnea-hypopnea score of at least 5 and self-reported hypersomnolence. The paper supports the criterion and estimates within the quoted ranges, but it does not report the ranges of 3%–7% for men and 2%–5% for women. (LLM, conf=0.88)

  Experimental — accuracy not yet measured
• cites [92]Not checked — no full text of the source

The Wisconsin Sleep Cohort and SHHS found a general prevalence of 24% in men and 9% in women using the alternative hypopnea definition of ≥4% oxygen desaturation.

“These studies found a general prevalence of 24% in men, 9% in women using the alternative definition of hypopnea (≥4% oxygen desaturation).”

• cites [90] sourcePartly supported

  The Occurrence of Hypersomnolence (p. 3) gives prevalence estimates of 24% for men and 9% for women with an apnea-hypopnea score of at least 5. Interpretation of Polysomnographic Data (p. 2) defines hypopnea using a decrease of at least 4% in oxyhemoglobin saturation. These are estimates for middle-aged adults from the Wisconsin Sleep Cohort; this paper does not report an SHHS finding or establish prevalence for the general population. (LLM, conf=0.87)

  Experimental — accuracy not yet measured
• cites [91] sourceCould not verify

  The Abstract reports OSA prevalence of 4% in men and 2% in women, not the claimed 24% and 9%; those figures may concern a different definition, so this is not a direct contradiction. The Home polysomnography section describes SHHS hypopnea scoring but does not report prevalence using a ≥4% oxygen-desaturation threshold. Searches for the claimed percentages and the Wisconsin cohort did not locate that finding in our copy. (LLM, conf=0.76)

  Experimental — accuracy not yet measured

Several groups reported increased frequency of non-specific or functional somatic complaints in UARS patients compared with OSA patients, including irritable bowel, insomnia, difficulty concentrating, cognitive problems, depression, and poor psychomotor performance.

“Several groups have reported an increased frequency of non-specific or functional somatic complaints in patients with UARS, compared to OSA patients including irritable bowel, insomnia, difficulty concentrating, cognitive problems, depression, and poor psychomotor performance.77–79”

• cites [77] sourceNot checked — no full text of the source
• cites [78] sourceNot checked — no full text of the source
• cites [79] sourcePartly supported

  The Introduction reports prior studies finding more insomnia, symptoms resembling functional somatic syndromes, and impaired psychomotor performance in UARS than OSAS. The Discussion reports more somatic symptoms and higher depression scores in UARS. Our copy does not specifically establish the claimed comparisons for irritable bowel, difficulty concentrating, or cognitive problems. (LLM, conf=0.88)

  Experimental — accuracy not yet measured

Residing in less favorable walking environments has been associated with more severe sleep apnea, especially in male and obese individuals.

“Residing in less favorable walking environments has been associated with more severe sleep apnea especially in male and obese individuals.124”

• cites [124] sourceSupported

Trials enrolling obese diabetics found that 86% have some degree of OSA.

“Trials enrolling obese diabetics have found that 86% have some degree of OSA.122”

• cites [122] sourceSupported

Stroke has been consistently identified as a devastating outcome of untreated severe OSA, especially in middle-aged adults.

“Stroke has been consistently identified as a devastating outcome of untreated severe OSA, especially in middle aged adults.120”

• cites [120] sourceSupported

Neighborhood disadvantage, living in poor residential areas, has been associated with childhood sleep apnea in several epidemiological studies.

“Neighborhood disadvantage, living in poor residential areas, has been associated with childhood sleep apnea in several epidemiological studies.125,126”

• cites [125] sourceNot checked — no full text of the source
• cites [126] sourceSupported

Guest et al. estimated that cost savings exceeded the costs of treating OSA after 13 years.

“For example, Guest et al. estimated that cost savings exceeded the costs of treating OSA after 13 years.141”

• cites [141] sourceSupported

Most population-based studies show a two- to threefold greater prevalence of OSA among men compared to women.

“Most population-based studies show a two- to threefold greater prevalence of OSA among men compared to women.18,45,90,93,96,99,104,144–156”

• cites [18] sourceNot checked — no full text of the source
• cites [45] sourceNot checked — no full text of the source
• cites [90] sourceSupported

Some of the cost advantage with use of home sleep testing ($167 vs $782 for lab study) was offset by the need for additional testing in some patients who failed or had non-diagnostic home studies.

“Some of the cost advantage with use of home sleep testing ($167 vs. $782 for lab study) was offset by the need for additional testing in some patients who have failed or had non-diagnostic home studies.138”

• cites [138] sourceSupported

Post-menopausal women, particularly those not on hormone-replacement therapy, are at increased risk for OSA compared to pre-menopausal women, even when correcting for age.

“Post-menopausal women, and particularly those not on hormone-replacement therapy, are at increased risk for OSA compared to pre-menopausal women, even when correcting for age.99,144,147,154”

• cites [99] sourceNot checked — no full text of the source
• cites [144] sourceNot checked — no full text of the source
• cites [147] sourcePartly supported

  The paper’s introductory passage reports that OSA incidence increases after menopause and says some, but not all, studies suggest an inverse association with hormone therapy. Its own analysis, reported in the Abstract and “Type of menopause and OSA risk” section, compares surgical with natural menopause among postmenopausal women and finds a stronger association in women who never used hormone therapy. It does not establish the claimed age-adjusted comparison of postmenopausal with premenopausal women; no page numbers are provided in our copy. (LLM, conf=0.86)

  Experimental — accuracy not yet measured

Sands et al. showed that obese patients with OSA compared to those without OSA have worse pharyngeal collapsibility with less active pharyngeal muscle response.

“A study by Sands et al. showed that obese patients with OSA compared to those without OSA have worse pharyngeal collapsibility with less active pharyngeal muscle response.204”

• cites [204] sourceSupported

Obesity, specifically central adiposity, increases pharyngeal collapsibility that causes recurrent episodes of upper airway obstruction in sleep apnea through mechanical and neuromuscular effects.

“Obesity, specifically central adiposity, increases pharyngeal collapsibility that causes recurrent episodes of UA obstruction in sleep apnea through both mechanical and neuromuscular effects.199–203”

• cites [199] sourceNot checked — no full text of the source
• cites [200] sourceNot checked — no full text of the source
• cites [201] sourcePartly supported

  The Introduction and Discussion (page not specified) describe pharyngeal collapse in sleep apnea as involving airway anatomy and dilator-muscle activity, supporting the mechanical and neuromuscular portion of the claim. The Methods compare patients with obese, size-matched controls, but the paper does not establish that obesity—specifically central adiposity—increases collapsibility or causes recurrent obstruction. (LLM, conf=0.86)

  Experimental — accuracy not yet measured
• cites [202] sourceNot checked — no full text of the source
• cites [203] sourceNot checked — no full text of the source

Obesity and fat distribution differentially affect men and women's propensity for OSA.

“Obesity174 and fat distribution175,176 differentially affect men and women’s propensity for OSA.”

• cites [174] sourceSupported
• cites [175] sourceNot checked — no full text of the source
• cites [176] sourceNot checked — no full text of the source

An increase in BMI by 1 standard deviation was associated with a threefold increase in risk of OSA.

“Another study showed an increase in BMI by 1 standard deviation (SD) was associated with a threefold increase in risk of OSA.90”

• cites [90] sourcePartly supported

  The Risk Factors discussion (p. 5) describes a threefold increase for a one-SD increment in any measure of body habitus. But Table 5 (p. 5) gives the BMI-specific odds ratio as 4.17, not threefold, for sleep-disordered breathing defined by an apnea-hypopnea score of at least 5. (LLM, conf=0.88)

  Experimental — accuracy not yet measured

Obesity is one of the most significant risk factors for SDB.

“Obesity is one of the most significant risk factors for SDB.95,156,189,190”

• cites [95] sourceSupported
• cites [156] sourceNot checked — no full text of the source
• cites [189] sourceNot checked — no full text of the source

Facial anatomical distinctions, such as mandible position, impart different degrees of OSA risk in men and women.

“Facial anatomical distinctions, such as mandible position, impart different degrees of risk in men and women.168,169”

• cites [168] sourceSupported
• cites [169] sourcePartly supported

  The Results report that mandibular plane angle had a different association with OSA in women and men, supporting a sex-related difference in craniofacial features associated with OSA. But the Discussion says mandibular depth angle—more directly a measure of mandibular position—was associated with OSA regardless of sex. The study reports associations, not that these features cause different degrees of risk. (LLM, conf=0.84)

  Experimental — accuracy not yet measured

Characteristics of the sleep study in OSA such as AHI, event duration, and measures of oxygen desaturation during sleep are heritable.

“Moreover, characteristics of the sleep study in OSA such as AHI, event duration, and measures of oxygen desaturation during sleep are heritable.252”

• cites [252] sourceSupported

Obesity plays a major role in OSA.

“Obesity plays a major role.90”

• cites [90] sourceSupported

Cardiopulmonary coupling, a measure of sleep depth, is heritable.

“as is cardiopulmonary coupling,254 a measure of sleep depth.255”

• cites [254] sourceSupported
• cites [255] sourcePartly supported

  The cited paper supports the description of high-frequency cardiopulmonary coupling as a sleep-related measure: its Abstract reports a correlation with EEG delta power and calls it a biomarker of stable sleep. It does not establish cardiopulmonary coupling generally as a measure of sleep depth. The heritability statement is attached to a different citation and is not judged here. (LLM, conf=0.82)

  Experimental — accuracy not yet measured

Extreme phenotypes of OSA have been demonstrated.

“Extreme phenotypes have been demonstrated.256”

• cites [256] sourceSupported

Parapharyngeal fat deposition in the subglosso-supraglottic area is associated with increased OSA severity independent of BMI or neck circumference.

“They have also shown that parapharyngeal fat deposition in the subglosso-supraglottic area is associated with increased OSA severity independent of BMI or neck circumference.209,210”

• cites [209] sourceNot checked — no full text of the source
• cites [210] sourceSupported

OSA aggregates in families.

“It has been known for over two decades that OSA aggregates in families.”

• cites [245] sourceNot checked — no full text of the source
• cites [246] sourceNot checked — no full text of the source
• cites [247] sourceSupported

Heart rate response to arousal is heritable.

“Heart rate response to arousal is also heritable253”

• cites [253] sourceSupported

Further evidence that obesity is pathogenic in OSA for some individuals is that weight loss results in marked improvement of OSA.

“Further evidence that obesity is pathogenic in OSA for some individuals, is that weight loss results in marked improvement of OSA.122,214”

• cites [122] sourceSupported
• cites [214] sourceNot checked — no full text of the source

Open bite, overbite, overjet, and proclination of the mandibular incisors are dental findings that may suggest underlying craniofacial deficiency in OSA.

“Dental findings may also be suggestive of underlying craniofacial deficiency and include: open bite, overbite, overjet, and proclination of the mandibular incisors.219,234,237”

• cites [219] sourceNot checked — no full text of the source
• cites [234] sourceNot checked — no full text of the source
• cites [237] sourcePartly supported

  The Discussion (page unavailable) reports that deep overbite and severe overjet were more prevalent in patients with OSAS than in the general Japanese population, and suggests a role for abnormal craniofacial anatomy. It does not identify open bite or proclination of the mandibular incisors as findings; its mention of protruding teeth concerns the maxillary anterior teeth. (LLM, conf=0.88)

  Experimental — accuracy not yet measured

Multiple physiological traits determine risk for OSA.

“There are also multiple physiological traits that determine risk for the disorder.36,257,258”

• cites [36] sourceSupported
• cites [257] sourceSupported
• cites [258] sourceNot checked — no full text of the source

Genome-wide studies have largely focused on quantitative variables from sleep studies that were shown to be heritable.

“Genome-wide studies263–265 have largely focused on quantitative variables obtained from the sleep study to characterize SDB that were shown to be heritable.252”

• cites [263] sourceSupported
• cites [264] sourceSupported
• cites [265] sourceSupported
• cites [252] sourceSupported

Variants in the interleukin 18 receptor and hexokinase genes are associated with oxygen desaturation measures during sleep.

“Variants in the interleukin 18 receptor and hexokinase genes were associated with oxygen desaturation measures during sleep.269”

• cites [269] sourceSupported

A genome-wide study examining common and rare variants by whole genome sequencing identified novel rare variant associations with measures of desaturation: ARMCX3, MRPS33, and C16orf90.

“A genome-wide study that specifically examined both common and rare variants by whole genome sequencing264 identified novel rare variants associations with measures of desaturation, that is, ARMCX3, MRPS33, and C16orf90.”

• cites [264] sourceSupported

The role of nasal obstruction as a contributing factor in the pathogenesis of OSA and SDB is controversial.

“but the role of nasal obstruction as a contributing factor in the pathogenesis of OSA and SDB is controversial.276–278”

• cites [276] sourceSupported
• cites [277] sourceSupported
• cites [278] sourceNot checked — no full text of the source

Nasal obstruction from septal deviation, inferior turbinate hypertrophy, nasal valve collapse, polyposis, and/or inflammation is common in patients with SDB.

“Several recent reviews indicate that nasal obstruction from septal deviation, inferior turbinate hypertrophy, nasal valve collapse, polyposis, and/or inflammation are common in patients with SDB”

• cites [276] sourcePartly supported

  The Abstract and Discussion report nasal obstruction in 70% of the studied OSA patients and identify septal deviation, inferior turbinate hypertrophy, and inflammation as possible contributors. The paper does not establish nasal valve collapse or polyposis as common causes, and its study population was patients with OSA rather than the broader SDB population. It is a prospective study, not a review. (LLM, conf=0.88)

  Experimental — accuracy not yet measured
• cites [277] sourcePartly supported

  The section “Clinical and experimental evidence of correlation between nasal pathology and SDB” (page unavailable) lists septal deviation, enlarged turbinates, nasal valve collapse, rhinitis, and disease with or without nasal polyps as causes of nasal obstruction. It also associates nasal obstruction with snoring and mild SDB, but does not establish that these causes of obstruction are common among patients with SDB. (LLM, conf=0.84)

  Experimental — accuracy not yet measured
• cites [278] sourceNot checked — no full text of the source

Novel variants were associated with AHI and oxygen desaturation measures in admixture mapping.

“Novel variants were associated with AHI as well as oxygen desaturation measures.”

• cites [270] sourceSupported

For most previously described gene variants, there was no evidence of association with OSA diagnosis or with quantitative traits from sleep studies in community-based cohorts with standardized sleep apnea phenotyping.

“For most of the previously described gene variants, there was no evidence of association with OSA diagnosis or with quantitative traits from the sleep study with findings from community-based cohorts which underwent standardized sleep apnea phenotyping.”

• cites [272] sourcePartly supported

  The results sections report that only three of 40 tested candidate variants validated for EHR-derived OSA diagnosis, and three were associated with quantitative traits from sleep-study reports. However, the Study populations and EHR-Derived OSA phenotypes sections describe health-system EHR cohorts, with sleep-study measures available only for a Geisinger subset—not community-based cohorts that all underwent standardized sleep-apnea phenotyping. Page numbers are unavailable in our copy. (LLM, conf=0.88)

  Experimental — accuracy not yet measured

Recently, large biobanks have begun analyzing genetic associations with snoring and ICD-based OSA diagnosis.

“Recently, large biobanks have begun analyzing genetic associations with snoring and ICD-based OSA diagnosis.274”

• cites [274] sourceSupported

Progress in OSA genetics has been made using genome-wide association studies or a special SNP panel.

“More recently, progress has been made using genome-wide association studies263–265 or use of a special SNP panel.266”

• cites [263] sourceSupported
  Experimental — accuracy not yet measured
• cites [264] sourceSupported
  Experimental — accuracy not yet measured
• cites [265] sourceSupported
  Experimental — accuracy not yet measured
• cites [266] sourceSupported
  Experimental — accuracy not yet measured

A variant in GPR83 was found to be associated with AHI, based primarily on the Hispanic Community Health Study.

“Based on primarily the Hispanic Community Health Study, a variant in GPR83 was found to be associated with AHI.263”

• cites [263] sourcePartly supported

  The GWAS Results report a significant association between AHI and a locus spanning GPR83; the study sample was predominantly from HCHS/SOL (11,317 of 12,558 participants with AHI data). However, the Figure 1 caption places the lead variant rs116791765 in an intergenic region, not in GPR83 itself, so the claim overstates the variant’s location. (LLM, conf=0.85)

  Experimental — accuracy not yet measured

A region on chromosome 18q21 including 20 variants on NARS/FECH was associated with both AHI and percentage time SaO2 < 90%, with 12 associations replicated in independent cohorts.

“A region on chromosome 18q21 that included 20 variants on NARS/FECH was associated with both AHI and percentage time SaO2 < 90%, with 12 associations that replicated in independent cohorts.”

• cites [270] sourceSupported

Evidence supports a causal role in OSA for iron metabolism, based on increased FECH expression with lower nocturnal SaO2.

“Evidence for a causal role in OSA was supported by finding that FECH expression was increased in association with lower nocturnal SaO2. These findings suggested a novel for iron metabolism in pathway influencing OSA phenotypes.”

• cites [270] sourcePartly supported

  The Discussion reports that increased FECH expression was associated with lower nocturnal SaO₂ and suggests that iron/heme metabolism influences OSA traits. It does not establish a causal role; the authors say future research is needed to determine the mechanism. (LLM, conf=0.88)

  Experimental — accuracy not yet measured

Mouth breathing during sleep has been associated with up to 2.5 times higher airway resistance.

“Mouth breathing during sleep has been associated with up to 2.5 times higher airway resistance.287”

• cites [287] sourceSupported

A study by Miljeteig et al. divided 683 patients referred for PSG into three groups based on severity of nasal resistance and found no difference in apnea or snoring severity between groups.

“Conversely, a study by Miljeteig et al.298 divided 683 patients referred for PSG into three groups based on severity of nasal resistance and found no difference in apnea or snoring severity between groups.”

• cites [298] sourceSupported

A study of 541 snorers who underwent posterior rhinomanometry concluded that nasal obstruction is an independent risk factor for OSA.

“A study of 541 snorers who underwent posterior rhinomanometry concluded that nasal obstruction is an independent risk factor for OSA.297”

• cites [297] sourcePartly supported

  The Abstract says 541 snorers were included to undergo posterior rhinomanometry, but only 528 completed the measurement. It explicitly concludes that daytime nasal obstruction is an independent risk factor for obstructive sleep apnoea syndrome; the claim overstates the number who underwent the test. (LLM, conf=0.90)

  Experimental — accuracy not yet measured

Lateral pharyngeal wall edema and enlarged lateral pharyngeal fat pads have been identified via tissue specimens and imaging in OSA patients.

“Lateral pharyngeal wall edema and enlarged lateral pharyngeal fat pads have been identified via tissue specimens and imaging in OSA patients.200,205,320–322”

• cites [200] sourceNot checked — no full text of the source
• cites [205] sourceNot checked — no full text of the source
• cites [320] sourceNot checked — no full text of the source
• cites [321] sourceNot checked — no full text of the source
• cites [322] sourcePartly supported

  The ‘Sizes of fat deposits and distribution about the palatopharyngeal airspace’ section reports significantly more postero-lateral fat in patients with OSA than in weight-matched controls, based on magnetic resonance imaging. The paper does not identify lateral pharyngeal wall edema or report findings from tissue specimens; its discussion of possible oedema concerns the soft palate. (LLM, conf=0.88)

  Experimental — accuracy not yet measured

70% of patients with OSA had nasal obstruction confirmed by clinical examination and rhinomanometry.

“A recent study by Magliulo et al.276 demonstrated that 70% of patients with OSA had nasal obstruction confirmed by clinical examination and rhinomanometry.”

• cites [276] sourceSupported

The lumen of snorers and OSA patients is circular or elliptical, with the long axis in the anterior-posterior dimension.

“In contrast, the lumen of snorers and OSA patients is circular or elliptical, with the long axis in the anterior–posterior dimension.314”

• cites [314] sourceSupported

Increasing tonsil grade is significantly associated with increasing OSA severity, with each increase in Brodsky grade correlating to an increase in AHI of approximately 14 events/h.

“The most recent study by Jara and Weaver studied an adult heterogeneous OSA population in the US and found that increasing tonsil grade was significantly associated with increasing OSA severity with adjusted data showing that every increase in tonsil grade as measured by the Brodsky classification system correlated to an increase in AHI of approximately 14 events/h.327”

• cites [327] sourceSupported

A normal pharyngeal lumen is elliptical in shape, with the long axis in the lateral dimension.

“A normal pharyngeal lumen is elliptical in shape, with the long axis in the lateral dimension.314”

• cites [314] sourceSupported

Given the variability in OSA pathogenesis, the strategy to improve pharyngeal anatomy/collapsibility surgically may be prone to failure in the subset of patients in whom OSA is primarily caused by abnormal ventilatory control.

“Given the variability in OSA pathogenesis, the strategy to improve pharyngeal anatomy/collapsibility surgically may be prone to failure in the subset of patients in whom OSA is primarily caused by abnormal ventilatory control.36”

• cites [36] sourceSupported

Elevated loop gain has been found to predict failure of sleep apnea surgeries, at least in some cases.

“From a surgical perspective, elevated loop gain may be important since it has been found to predict failure of sleep apnea surgeries, at least in some cases.363–365”

• cites [363] sourcePartly supported

  The section “The Value of Estimates of Nonanatomical Traits for Upper Airway Surgery” says elevated baseline loop gain was associated with poor surgical outcome, and “Baseline Loop Gain, Arousal Threshold and Surgical Outcome” reports a correlation with postoperative AHI. However, the Abstract says only preoperative AHI was predictive in multivariate regression, so the paper does not establish elevated loop gain as an independent predictor of surgical failure. (LLM, conf=0.85)

  Experimental — accuracy not yet measured
• cites [364] sourcePartly supported

  The Summary of the Findings (page not available in our copy) identifies high loop gain as a potential reason some patients respond poorly to upper airway surgery. But it presents this as a possible explanation, not a finding that elevated loop gain predicts surgical failure. The Results report an association between changes in loop gain and improvements in apnea severity, rather than a test of preoperative loop gain as a predictor of failure. (LLM, conf=0.83)

  Experimental — accuracy not yet measured
• cites [365] sourceSupported

Tongue fat volume calculated from MRI is greater among obese OSA patients versus obese subjects without apnea.

“They showed that tongue fat volume calculated from the MRI was greater among the obese OSA patients versus obese subjects without apnea.208”

• cites [208] sourceSupported

High loop gain due to a high controller gain can be seen in clinical situations such as high altitude or heart failure.

“High loop gain due to a high controller gain (i.e., overly sensitive thermostat in the analogy above) can be seen in clinical situations such as high altitude or heart failure.351,352”

• cites [351] sourceNot checked — no full text of the source
• cites [352] sourceSupported

Control of breathing is a negative feedback system in which chemoreceptors and the lung try to maintain PaCO2 at roughly 40 mmHg.

“Control of breathing is a negative feedback system in which chemoreceptors (e.g., in the carotid body; “controller”) and the lung (“plant”) try to maintain a PaCO2 at roughly 40 mmHg.”

• cites [180] sourcePartly supported

  The Methods section, “Protocol: LG measurement” (page unavailable), describes a respiratory feedback circuit involving gas exchange and chemoreceptors. The paper does not identify the chemoreceptors as the “controller” or the lung as the “plant,” or state that the system maintains PaCO2 at roughly 40 mmHg. (LLM, conf=0.83)

  Experimental — accuracy not yet measured
• cites [344] sourceNot checked — no full text of the source
• cites [345] sourceNot checked — no full text of the source
• cites [346] sourceNot checked — no full text of the source
• cites [347] sourcePartly supported

  The paper’s description of loop gain (section labelled Abstract; page unavailable) describes breathing as a chemical-feedback system involving chemoreceptors and a plant that includes the lungs. Table 1 reports a mean baseline PCO2 of 40 mmHg in the healthy elderly subjects, but this is an end-tidal measurement during sleep on CPAP, not a stated PaCO2 target maintained by the system. The cited passages also do not identify the carotid body specifically. (LLM, conf=0.86)

  Experimental — accuracy not yet measured

Greater Pcrit values are associated with an inferiorly and posteriorly positioned hyoid.

“Genta et al. studied 34 Japanese-Brazilian men with OSA and also found an association between greater Pcrit values with an inferiorly and posteriorly positioned hyoid.338”

• cites [338] sourcePartly supported

  The Results (page not provided) report that higher Pcrit correlated with greater mandibular-plane-to-hyoid distance, a measure of inferior hyoid position. The paper does not report an association with posterior hyoid position. It also states that 11 of the 34 Japanese-Brazilian men did not have OSA, contrary to the citing sentence's description of all 34 as having OSA. (LLM, conf=0.90)

  Experimental — accuracy not yet measured

Among obese rats, there is greater deposition of fat within the tongue compared to non-obese rats.

“Brennick et al. showed that among obese rats, there was greater deposition of fat within the tongue compared to non-obese rats.342”

• cites [342] sourceSupported
  Experimental — accuracy not yet measured

The negative pressure reflex is regulated by cholinergic systems in the brainstem, including the peri-obex region.

“The NPR has been mapped using neurochemical techniques and is thought to be regulated by cholinergic systems in the brainstem, for example, the peri-obex region which is heavily cholinergic.”

• cites [391] sourcePartly supported

  The Results and Discussion identify the brainstem periobex region as important to the negative pressure reflex: bilateral inhibition of its neurons abolished genioglossus responses to negative pressure in rats. But the paper does not establish that the region is heavily cholinergic or that cholinergic systems regulate the reflex; its intervention used muscimol, a GABA-A receptor agonist. No page numbers are available in our copy. (LLM, conf=0.88)

  Experimental — accuracy not yet measured

Manipulating the cholinergic system can influence the negative pressure reflex.

“Pharmacological studies indicate manipulation of the cholinergic system may influence this reflex, recognizing the complex role of acetylcholine throughout the brain and systemically.”

• cites [392] sourceSupported

Through protective upper airway dilator muscle reflexes, muscle activity is increased during wakefulness in OSA compared to matched controls.

“Through protective UA dilator muscle reflexes, the activity of the muscles is increased during wakefulness in OSA compared to matched controls.201”

• cites [201] sourcePartly supported

  The Abstract reports significantly greater waking genioglossus activity in OSA patients than in age-matched controls, with the difference persisting in size-matched subsets. The Introduction and later discussion propose a protective reflex response to negative airway pressure, but describe that mechanism as theoretical or speculative rather than established; the study measured the genioglossus, not upper-airway dilator muscles generally. No page number is provided for these passages. (LLM, conf=0.88)

  Experimental — accuracy not yet measured

It may be possible to quantify loop gain directly from in-lab or home sleep studies.

“It may be possible to quantify loop gain directly from inlab or home sleep studies.369–371”

• cites [369] sourceSupported
• cites [370] sourcePartly supported

  The Abstract and Discussion describe estimating loop gain from ventilation signals collected during routine polysomnography, supporting the in-lab portion of the claim. The Methods describe overnight polysomnography, but our copy does not establish that the approach works with home sleep studies. (LLM, conf=0.83)

  Experimental — accuracy not yet measured
• cites [371] sourcePartly supported

  The Discussion (page unavailable) says loop gain can be quantified from routine clinical polysomnography, and Methodological considerations specifies that the study used data from an in-laboratory clinical environment. Our copy does not establish that the method works with home sleep studies, so the home-study part of the claim is unsupported by this source. (LLM, conf=0.86)

  Experimental — accuracy not yet measured

The genioglossus has both tonic activity and phasic activity.

“Additionally, it is a complex muscle which has both tonic activity (i.e., present throughout the respiratory cycle) as well as phasic activity (i.e., bursts with each inspiration).377,378”

• cites [377] sourceSupported
• cites [378] sourceCould not verify

  The References section names a separate study on tonic and phasic respiratory drives to genioglossus motoneurons, but this paper does not itself state that the muscle has both activity patterns. Searches for tonic, phasic, and inspiratory activity found descriptions of motor-unit recordings during breathing, not a passage establishing the claimed distinction. No page numbers are available in our copy. (LLM, conf=0.70)

  Experimental — accuracy not yet measured

A pilot study found that combining oxybutynin and atomoxetine may improve the apnea-hypopnea index.

“A recent pilot study showed potential benefits of the combination of oxybutynin and atomoxetine on the AHI.”

• cites [393] sourceSupported

Monosynaptic projections from several brain structures to hypoglossal motor control have been shown in animal models.

“Monosynaptic projections from various brain structures have been shown in animal models including locus coeruleus (adrenergic), lateral dorsal/pedunculopontine tegmentum (cholinergic), and hypothalamic (orexinergic) and raphe neurons (serotonergic) among others.”

• cites [375] sourcePartly supported

  The section headed “The lateral reticular formation as a key state-dependent hub” (page unavailable) describes premotor neurons in the reticular formation that innervate the hypoglossal motor pool, supporting a projection from a brain region. The copy does not establish monosynaptic projections from the cited locus coeruleus, tegmental, hypothalamic, and raphe neurons; its account of tegmental cholinergic activity instead describes intervening circuits. (LLM, conf=0.84)

  Experimental — accuracy not yet measured
• cites [403] sourcePartly supported

  The section on serotonin and pharyngeal motor control describes medullary raphé neurons projecting to the hypoglossal motor nucleus, and the sleep-related suppression section describes inputs from the locus coeruleus complex. But the paper does not establish that these projections are monosynaptic or describe hypothalamic orexinergic projections; its Figure 2 instead places the cholinergic LDT/PPT input at the pontine reticular formation, not directly at hypoglossal motor neurons. No page numbers are available in our copy. (LLM, conf=0.86)

  Experimental — accuracy not yet measured
• cites [404] sourceNot checked — no full text of the source
• cites [405] sourceCould not verify

  The passages on upper-airway muscle function and hypoglossal output (section and page unavailable) discuss their roles in obstructive sleep apnoea, but do not describe monosynaptic projections in animal models. Searches for “monosynaptic” and the named brain structures and neurotransmitter systems found no passage addressing the claim. (LLM, conf=0.76)

  Experimental — accuracy not yet measured

A negative pressure reflex causes strong genioglossus activation when sub-atmospheric inspiratory pressure is applied.

“A negative pressure reflex (NPR) exists whereby a sub-atmospheric (negative inspiratory or suction) pressure leads to a robust activation of the genioglossus muscle.”

• cites [388] sourceNot checked — no full text of the source
• cites [389] sourceNot checked — no full text of the source
• cites [390] sourceSupported

Validated functional status and quality-of-life measures most used in clinical and research settings include the ESS, SAQLI, FOSQ, and PSQI.

“Validated functional status and QOL measures most used in the clinical and research settings include the ESS,436 Sleep Apnea Quality of Life Index (SAQLI),437–441 Functional Outcomes of Sleep Questionnaire (FOSQ),437,441,442 and the Pittsburgh Sleep Quality Index (PSQI).443”

• cites [436] sourceNot checked — no full text of the source
• cites [437] sourcePartly supported

  The Discussion (page unavailable) supports the validity of the FOSQ and SAQLI as patient-reported measures for obstructive sleep apnea, and the paper identifies both instruments. An introductory passage says the FOSQ is widely used in sleep research but the SAQLI is less widely used; it therefore does not establish that both are among the measures most used in clinical and research settings. This citation is attached to the SAQLI and FOSQ, not the ESS or PSQI. (LLM, conf=0.85)

  Experimental — accuracy not yet measured
• cites [438] sourceNot checked — no full text of the source
• cites [439] sourceNot checked — no full text of the source
• cites [440] sourcePartly supported

  The SAQLI and Discriminative properties sections describe the SAQLI as an obstructive-sleep-apnoea quality-of-life measure and report evidence for its validity. The introduction mentions the FOSQ as an emerging sleep-specific questionnaire, but this paper does not validate it or establish that these measures are among the most used in clinical and research settings. The cited paper’s contribution therefore supports the SAQLI portion, not the full claim. (LLM, conf=0.84)

  Experimental — accuracy not yet measured
• cites [441] sourceSupported
  Experimental — accuracy not yet measured
• cites [442] sourceNot checked — no full text of the source

Combinations of therapy may be required to eliminate apnea when using arousal-threshold manipulation.

“Consequently, combinations of therapy may well be required to eliminate apnea using this approach.”

• cites [415] sourceSupported

Among general clinic patients, the Berlin questionnaire has reported sensitivity of 89%.

“and 89% and 88% respectively among general clinic patients.”

• cites [422] sourceNot checked — no full text of the source
• cites [423] sourceCould not verify

  The paper’s Introduction and results discuss STOP-Bang sensitivity, not an 89% sensitivity for the Berlin questionnaire among general clinic patients. A search for the Berlin questionnaire found it in the References, but no passage reporting the claimed figure or population; pages are unavailable in our copy. (LLM, conf=0.76)

  Experimental — accuracy not yet measured

Many patients with a high arousal threshold and sleep apnea experience a lowering of arousal threshold over time on therapy.

“many patients with a high arousal threshold with sleep apnea will experience a lowering of arousal threshold over time on therapy.”

• cites [417] sourceSupported
• cites [418] sourcePartly supported

  The main text (page not supplied) says that CPAP lowers the arousal threshold in patients with obstructive sleep apnoea, supporting the central point. It does not establish that this happens in many patients or specify that the lowering occurs over time on therapy; its reference to change over time concerns how the elevated threshold may develop. (LLM, conf=0.84)

  Experimental — accuracy not yet measured

Among surgical patients, the STOP-BANG questionnaire has reported sensitivity of 91%.

“the Berlin and STOP-BANG questionnaires have reported sensitivities of 79%–82% and 91% respectively among surgical patients”

• cites [422] sourceNot checked — no full text of the source
• cites [423] sourcePartly supported

  The pooled predictive parameters section reports 91% sensitivity for STOP-Bang among surgical patients, but specifically for detecting moderate-to-severe OSA with a score of 3 as the cut-off. The citing sentence omits that qualification; the same passage reports 84% sensitivity for any OSA. (LLM, conf=0.87)

  Experimental — accuracy not yet measured

Existing data suggest that improvements in apnea from sedatives or hypnotics that raise the arousal threshold are relatively modest.

“the existing data suggest that any improvements in apnea which occur with these agents are relatively modest.”

• cites [400] sourcePartly supported

  The Discussion (page unavailable) reports that eszopiclone reduced the apnoea/hypopnoea index by about 45% in patients with a low arousal threshold, but only two of eight reached fewer than 10 events per hour; it says additional strategies may be needed to eliminate apnoea. This supports the point that improvement can be incomplete, but the study does not establish that improvements with sedatives or hypnotics generally are relatively modest. (LLM, conf=0.82)

  Experimental — accuracy not yet measured
• cites [401] sourceNot checked — no full text of the source

The Berlin questionnaire has a sensitivity of 82% for identifying clinically relevant OSA.

“The Berlin and STOP-BANG questionnaires have sensitivities of 82% and 94%, respectively,422,423”

• cites [422] sourceNot checked — no full text of the source
• cites [423] sourceCould not verify

  The Results section reports 94% sensitivity for STOP-Bang in detecting moderate-to-severe OSA in sleep-clinic patients, but does not report 82% sensitivity for the Berlin questionnaire. A search of our copy for Berlin questionnaire sensitivity found only a reference identifying that questionnaire, not the claimed figure. The cited paper supports the STOP-Bang half of the printed sentence, not the Berlin claim assessed here. (LLM, conf=0.76)

  Experimental — accuracy not yet measured

Among general clinic patients, the STOP-BANG questionnaire has reported sensitivity of 88%.

“and 89% and 88% respectively among general clinic patients.”

• cites [422] sourceNot checked — no full text of the source
• cites [423] sourcePartly supported

  The section on predictive parameters in the general population (page not available) reports 88% sensitivity for detecting moderate-to-severe OSA, but its participants were from the Sleep Heart Health Study, not general clinic patients. The same section reports different pooled sensitivities for sleep-clinic patients, so the claimed population is not supported. (LLM, conf=0.88)

  Experimental — accuracy not yet measured

An agent that raises the arousal threshold may cause substantial hypoxemia and hypercapnia that could affect end organ function.

“an agent which raises the arousal threshold may yield substantial hypoxemia and hypercapnia which could impact end organ function.”

• cites [414] sourcePartly supported

  The Discussion says apnea causes hypoxemia that may contribute to cardiovascular complications. It does not address whether an agent that raises the arousal threshold causes substantial hypoxemia or hypercapnia; searches for those terms found no supporting passage. (LLM, conf=0.82)

  Experimental — accuracy not yet measured

The STOP-BANG questionnaire has a sensitivity of 94% for identifying clinically relevant OSA.

“The Berlin and STOP-BANG questionnaires have sensitivities of 82% and 94%, respectively,422,423”

• cites [422] sourceNot checked — no full text of the source
• cites [423] sourcePartly supported

  The ‘Predictive parameters ... in the sleep clinic population’ passage reports 94% pooled sensitivity for moderate-to-severe OSA at a STOP-Bang cutoff of 3. The citing sentence gives 94% without those qualifications; the paper reports different sensitivities for surgical patients and the general population. (LLM, conf=0.88)

  Experimental — accuracy not yet measured

At a cutoff score of 5, the STOP-BANG had the best combination of sensitivity and specificity for severe OSA (AHI ≥ 30) in sleep clinic patients (66%, 66%) and surgical patients (56%, 74%).

“At this cutoff score, the STOP-BANG also has the best combination of sensitivity and specificity for severe OSA (AHI ≥ 30) in sleep clinic patients (66%, 66%) and surgical patients (56%, 74%).”

• cites [423] sourceSupported

The NoSAS questionnaire has a sensitivity of 65%-90% in identifying sleep clinic patients with clinically relevant OSA.

“while the NoSAS has a sensitivity of 65%–90% in identifying sleep clinic patients with clinically relevant OSA.”

• cites [424] sourcePartly supported

  The Results report that NoSAS screened patients referred to a sleep clinic for OSA, including moderate-to-severe OSA. But at a NoSAS cutoff of ≥7, sensitivity was 94.9% for moderate-to-severe OSA—above the claimed 65%–90% range. The claim does not specify a cutoff, and the cited Results passage does not establish that range. (LLM, conf=0.86)

  Experimental — accuracy not yet measured
• cites [425] sourcePartly supported

  The Results report that NoSAS detected 76.8% of sleep-center participants with severe sleep-disordered breathing (AHI ≥30 events/h), supporting a sensitivity within the claimed range. Our copy does not report a 65%–90% sensitivity range for clinically relevant OSA; the separately reported 74.8% is a correct-classification rate, not sensitivity. (LLM, conf=0.83)

  Experimental — accuracy not yet measured
• cites [426] sourceNot checked — no full text of the source
• cites [427] sourcePartly supported

  The Conclusions tables report NoSAS sensitivity in sleep-center patients of 64.9% for AHI ≥15 events/h, approximately the claimed lower bound. They report 59.0% for AHI ≥5 and 64.4% for AHI >30; they do not support a sensitivity as high as 90%. (LLM, conf=0.87)

  Experimental — accuracy not yet measured
• cites [428] sourcePartly supported

  The Abstract reports NoSAS sensitivities of 71.2%, 80.4%, and 83.1% for sleep-clinic patients at three SDB severity cutoffs. These values fall within 65%–90%, but the paper does not report that range for NoSAS; Table 2 attributes sensitivity above 90% to STOP-Bang instead. (LLM, conf=0.88)

  Experimental — accuracy not yet measured
• cites [429] sourceCould not verify

  The Results of accuracy outcomes section reports sensitivities for several screening questionnaires in sleep-disorder patients, but does not report a NoSAS result. Searches for “NoSAS” and “65%–90%” found no passage in our copy supporting the claimed sensitivity range for that questionnaire; the reported ranges concern other questionnaires. (LLM, conf=0.76)

  Experimental — accuracy not yet measured
• cites [430] sourceNot checked — no full text of the source
• cites [431] sourceNot checked — no full text of the source

A STOP-BANG cutoff score of 5 had the best combination of sensitivity and specificity for moderate-severe OSA (AHI ≥ 15) in sleep medicine (60%, 72%) and surgical populations (45%, 56%).

“The ROC curves for the STOP-BANG at different OSA severity classes showed a STOP-BANG cutoff score of 5 had the best combination of sensitivity and specificity for moderate-severe OSA (AHI ≥ 15) in sleep medicine (60%, 72%) and surgical populations (45%, 56%).”

• cites [423] sourcePartly supported

  Table 4 (Discussion; page unavailable) reports exactly the stated sensitivity and specificity at a STOP-Bang cutoff of 5 for moderate-to-severe OSA: 60% and 72% in sleep-clinic patients, and 45% and 56% in surgical patients. The paper reports results across cutoffs but does not identify 5 as having the best combination of sensitivity and specificity. (LLM, conf=0.87)

  Experimental — accuracy not yet measured

The arousal threshold is not a fixed trait but changes with treatment.

“the arousal threshold is not a fixed trait but rather a dynamic phenomenon which changes with treatment.”

• cites [417] sourceSupported
• cites [418] sourceSupported

A technique using signal processing of polysomnographic recordings can estimate the arousal threshold from clinically available data.

“Sands et al. developed a technique using signal processing of the polysomnographic recordings which can also estimate the arousal threshold using clinically available data.”

• cites [370] sourceSupported

Among surgical patients, the Berlin questionnaire has reported sensitivity of 79%-82%.

“the Berlin and STOP-BANG questionnaires have reported sensitivities of 79%–82% and 91% respectively among surgical patients”

• cites [422] sourceNot checked — no full text of the source
• cites [423] sourceCould not verify

  The surgical-population results report STOP-Bang sensitivities, including 91% for moderate-to-severe OSA, but do not report a Berlin questionnaire sensitivity. A search for “Berlin questionnaire” found it in the References, without the claimed 79%–82% figure for surgical patients. (LLM, conf=0.76)

  Experimental — accuracy not yet measured

An increased risk of cardiopulmonary disease and mortality has been established for clinically relevant OSA (AHI ≥ 15).

“where an increased risk of cardiopulmonary disease and mortality has been established (true positive)432–434”

• cites [432] sourcePartly supported

  The paper’s Abstract (page unavailable) reports that cohort studies consistently associate OSA with increased cardiovascular mortality, and its section on independent cardiometabolic effects reports an association with cardiovascular events. It defines moderate-to-severe OSA as AHI ≥15, but does not establish that threshold specifically for the reported risks or address pulmonary disease risk. (LLM, conf=0.87)

  Experimental — accuracy not yet measured
• cites [433] sourcePartly supported

  The Abstract (page not provided) reports that moderate-to-severe OSA, defined by RDI ≥15/hr, was independently associated with increased all-cause mortality. The Introduction (page not provided) describes evidence linking OSA to cardiovascular conditions, but this study did not establish an increased risk of cardiopulmonary disease; it also cautions that causation is difficult to establish. (LLM, conf=0.86)

  Experimental — accuracy not yet measured
• cites [434] sourcePartly supported

  The Discussion and Table 6 report significantly increased all-cause and cardiovascular mortality for severe sleep-disordered breathing (AHI ≥30), but not for moderate disease (AHI 15–<30). The Abstract says sleep-disordered breathing is linked to incident cardiovascular disease; this study does not establish increased cardiopulmonary disease risk across the claimed AHI ≥15 group. Page numbers were not supplied in our copy. (LLM, conf=0.88)

  Experimental — accuracy not yet measured

The SAQLI showed moderate correlations (0.3 ≤ r < 0.7) with related domains of general health-status questionnaires such as SF-36 and the Global QOL Scale.

“The SAQLI showed moderate correlations (0.3 ≤ r < 0.7) with related domains of general health-status questionnaires such as the 36-item Short Form (SF-36) and the Global QOL Scale.”

• cites [437] sourcePartly supported

  The Criterion Validity section (page unavailable) reports moderate correlations between SAQLI subscales and the SF-36 social functioning and vitality domains (r = 0.40–0.58). Table 4 also reports some SF-36 correlations below 0.3, and our copy does not report a correlation with the Global QOL Scale (searched: “Global QOL Scale”). (LLM, conf=0.86)

  Experimental — accuracy not yet measured
• cites [438] sourceNot checked — no full text of the source
• cites [439] sourceNot checked — no full text of the source
• cites [440] sourcePartly supported

  The Discriminative properties section reports good correlations between the SAQLI and related measures. Table 4 reports several moderate correlations with SF-36 domains and a global rating of change, but its SF-36 vitality correlations are 0.76 and 0.88—above the claimed range. The paper calls the global measure a rating of change, not a “Global QOL Scale.” (LLM, conf=0.88)

  Experimental — accuracy not yet measured
• cites [441] sourcePartly supported

  The Results report SAQLI correlations of r = 0.52 with the SF-36 mental component and r = 0.42 with its physical component, both within the claimed range. The Methods and Results do not identify or report a correlation with a Global QOL Scale, so that part of the claim is unsupported. (LLM, conf=0.89)

  Experimental — accuracy not yet measured

PSG measures (AHI, minimum O2 saturation) were either poorly or not significantly correlated with any of the questionnaires.

“PSG measures (AHI, minimum O2 saturation) were either poorly or not significantly correlated with any of the questionnaires.”

• cites [436] sourceNot checked — no full text of the source
• cites [437] sourcePartly supported

  The Relation with AHI section reports no correlation between AHI and FOSQ or SAQLI scores, except a weak correlation with SAQLI emotional functioning (r = 0.134). Our copy does not report correlations between minimum O2 saturation and the questionnaires, or establish the claim for every questionnaire used in the study. (LLM, conf=0.86)

  Experimental — accuracy not yet measured
• cites [441] sourcePartly supported

  The Results (page unavailable) say AHI severity was significant in the adjusted models only for the SF-36 physical component score in females with severe OSA, broadly supporting a limited association with questionnaire scores. The paper does not report correlations between minimum O2 saturation and the questionnaires; its reported correlation coefficients compare the questionnaires with one another, so that part of the claim is unsupported. (LLM, conf=0.86)

  Experimental — accuracy not yet measured

The FOSQ and SAQLI demonstrated excellent internal consistency (α ≥ 0.9).

“while the FOSQ and SAQLI demonstrated excellent internal consistency (α ≥ 0.9).437,438,440,442”

• cites [437] sourceSupported
• cites [438] sourceNot checked — no full text of the source
• cites [440] sourcePartly supported

  The Reliability and internal consistency section (page not specified) reports SAQLI Cronbach’s alpha values of 0.94, 0.90, and 0.93 for three domains. The paper mentions the FOSQ but does not report its internal consistency, so it supports only the SAQLI part of the claim. (LLM, conf=0.88)

  Experimental — accuracy not yet measured

In a recent multicenter prospective study of 812 patients with type 2 diabetes, obesity, heart failure, or resistant hypertension, the STOP-BANG questionnaire with high-risk set to 3 points or above had greater sensitivity than the Berlin or OSA50 questionnaires (95%, 75%, and 88%, respectively).

“A recent multicenter prospective study enrolled 812 patients diagnosed with type 2 diabetes, obesity, HF, or resistant hypertension who completed multiple OSA screening questionnaires. Following a diagnostic sleep study, the STOP-BANG questionnaire (with “high-risk” set to 3 points or above) had a greater sensitivity than either the Berlin questionnaire or the OSA50 questionnaire (95%, 75%, and 88”

• cites [455] sourceSupported

The FOSQ showed moderate correlations with related domains of the SF-36 and the Sickness Impact Profile.

“The FOSQ showed moderate correlations with related domains of the SF-36 and the Sickness Impact Profile.”

• cites [437] sourcePartly supported

  The Criterion Validity section reports moderate correlations between FOSQ activity level and SF-36 domains, except bodily pain; the Discussion also describes moderate correlations with SF-36 social and physical role domains. Our copy does not report a correlation between the FOSQ and the Sickness Impact Profile (searched: “Sickness Impact Profile” and “SIP”). Page numbers were not available in the indexed passages. (LLM, conf=0.88)

  Experimental — accuracy not yet measured
• cites [441] sourcePartly supported

  The Results (page not supplied) report correlations between the FOSQ and SF-36 summary scales, but describe them as small (r=.31 and .37), not moderate. The paper does not report a comparison with the Sickness Impact Profile. (LLM, conf=0.88)

  Experimental — accuracy not yet measured
• cites [442] sourceNot checked — no full text of the source

The FOSQ and SAQLI demonstrated excellent test-retest reliability (ICC ≥ 0.9).

“The FOSQ and SAQLI demonstrated excellent test–retest reliability (ICC ≥ 0.9)439,440,442”

• cites [439] sourceNot checked — no full text of the source
• cites [440] sourcePartly supported

  The Reliability and internal consistency section reports excellent SAQLI test–retest reliability, with ICCs of 0.91–0.94 in 14 untreated patients. The paper mentions the FOSQ as another sleep-specific questionnaire but does not report its test–retest reliability, so this source supports only the SAQLI portion of the claim. Page numbers were not available in our copy. (LLM, conf=0.90)

  Experimental — accuracy not yet measured
• cites [442] sourceNot checked — no full text of the source

In a retrospective review of 4770 patients from the SHHS, the STOP-BANG had the highest sensitivity for moderate-to-severe OSA (87%).

“The largest comparative study to date was a retrospective review by Silva et al. that included 4770 patients from the SHHS. For this cohort, the STOP-BANG had the highest sensitivity for moderate-to-severe OSA (87%).”

• cites [456] sourceSupported

The Berlin questionnaire has a sensitivity of 69-87% in surgical patients.

“Its sensitivity is 69–87% in surgical patients.”

• cites [489] sourceSupported

In the same SHHS cohort, the 4-Variable screening tool had the highest specificity.

“The 4-Variable screening tool (sex, BMI, BP, and snoring) had the highest specificity.”

• cites [456] sourceSupported

There is very little evidence on the benefits of widespread screening for OSA in asymptomatic adults without comorbidities.

“There is very little evidence on the benefits of widespread screening for OSA in asymptomatic adults without comorbidities.”

• cites [429] sourceCould not verify

  The Results and Discussion assess questionnaire accuracy in patients with and without known sleep problems, not the benefits of widespread screening in asymptomatic adults without comorbidities. A search for screening benefits in that population found no passage addressing the claim; the copy instead discusses possible benefits in surgical patients. (LLM, conf=0.75)

  Experimental — accuracy not yet measured

After increasing the STOP-BANG high-risk threshold to 5 or above, it had greater specificity (69%) than the Berlin (38%) or OSA50 (21%).

“After increasing the “high-risk” threshold to 5 or above on the STOP-BANG, it also had a greater specificity (69% compared to 38% for the Berlin and 21% for the OSA 50).”

• cites [455] sourceSupported

These complications have increasingly resulted in malpractice lawsuits.

“Moreover, these complications have increasingly resulted in malpractice lawsuits.”

• cites [478] sourceNot checked — no full text of the source
• cites [479] sourcePartly supported

  The Materials and Methods and Injury and Liability sections document malpractice claims associated with postoperative opioid-induced respiratory depression. They do not establish that such lawsuits have become more frequent. The discussion reports a decline for one subgroup of claims, but that does not establish an overall trend. (LLM, conf=0.86)

  Experimental — accuracy not yet measured

The preoperative assessment is a cost-effective opportunity to screen for and select high-risk patients who can then be offered sleep study and treatment for OSA.

“the preoperative assessment is a cost-effective opportunity to screen for and select high risk patients who can then be offered sleep study and treatment for OSA.”

• cites [488] sourcePartly supported

  The Discussion (page not available in our copy) favors preoperative STOP-Bang screening followed by polysomnography for screen-positive patients when costs and benefits are assessed over a lifetime. It also says no screening was favored when only the perioperative period was considered. The claim omits that qualification on cost-effectiveness. (LLM, conf=0.86)

  Experimental — accuracy not yet measured

Reports of other screening instruments have suggested minimal burden and disruption of the clinical workflow.

“reports of other screening instruments have suggested minimal burden and disruption of the clinical workflow.”

• cites [446] sourceNot checked — no full text of the source
• cites [459] sourceNot checked — no full text of the source
• cites [460] sourcePartly supported

  The ‘Opportunistic screening’ section describes a one-minute pulse check during a routine consultation, and Table 2 (‘The screening programme’) records short staff and administration times. These support a limited claim about time burden, but the paper does not report whether screening minimally disrupted clinical workflow. (LLM, conf=0.79)

  Experimental — accuracy not yet measured
• cites [461] sourceNot checked — no full text of the source

In men, there is only an association between caffeinated soda and severe SDB.

“Women with SDB have the strongest correlations between caffeinated soda and OSA. There was only an association between caffeinated soda and severe SDB in men.524 Coffee and tea consumption were not associated with SDB.”

• cites [524] sourceSupported

CPAP treatment improves nocturia in patients with OSA.

“Treatment with CPAP has been shown to improve nocturia in patients with OSA.522”

• cites [522] sourceSupported

Neck circumference or collar size is associated with OSA in adults.

“Neck circumference (NC) or collar size is associated with OSA in adults.184,528–534”

• cites [184] sourceSupported
  Experimental — accuracy not yet measured
• cites [528] sourceSupported
  Experimental — accuracy not yet measured
• cites [529] sourceNot checked — no full text of the source
• cites [530] sourceNot checked — no full text of the source
• cites [531] sourceSupported
• cites [532] sourceNot checked — no full text of the source
• cites [533] sourceSupported
  Experimental — accuracy not yet measured
• cites [534] sourceSupported
  Experimental — accuracy not yet measured

Findings from population-based and retrospective studies have demonstrated that habitual snoring and witnessed apneas are predictors for OSA.

“Findings from several population-based and retrospective studies have demonstrated that habitual snoring and witnessed apneas are predictors for OSA.”

• cites [88] sourceNot checked — no full text of the source
• cites [145] sourceNot checked — no full text of the source
• cites [505] sourcePartly supported

  The Results (page not provided) report that patients with OSA more often reported snoring and witnessed apneas, supporting an association with OSA. But the Database section says the questionnaire did not measure snoring frequency, and the study examined a referred sleep-laboratory cohort—not several population-based and retrospective studies. The Introduction also cautions that classic symptoms are not strongly predictive. (LLM, conf=0.87)

  Experimental — accuracy not yet measured

Men are more likely than women to report snoring (34% vs. 19%).

“Sex differences in reported sleep symptoms have been reported, with men being more likely to report snoring (34% vs. 19%) and women being more likely to report hypersomnolence (22.6% vs. 15.5%).90”

• cites [90] sourcePartly supported

  Table 1 (Collection of Polysomnographic Data, p. 2) supports the direction of the sex difference: 44% of men and 28% of women were classified as habitual snorers. It does not support the attributed 34% versus 19% figures; its definition of habitual snoring also includes reported snorting or breathing pauses. (LLM, conf=0.87)

  Experimental — accuracy not yet measured

A BMI greater than 30 increases risk for OSA.

“Obesity is a well-established risk factor for the development and progression of OSA. A BMI greater than 30 increases risk for OSA350 and weight reduction is a treatment option that can improve OSA severity.189”

• cites [350] sourcePartly supported

  Section III, “Pathogenesis of Sleep Apnea” (page not provided), identifies excess body weight as the dominant risk factor for OSA in the general population. The paper supports the increased-risk point, but the passages found do not establish BMI greater than 30 as a specific risk threshold. (LLM, conf=0.84)

  Experimental — accuracy not yet measured

Many validated questionnaires that include some or all OSA symptoms can help identify patients at risk for OSA.

“Many validated questionnaires445,446,510 include some or all of these symptoms and may help identify patients at risk for OSA.509”

• cites [509] sourceSupported

Women are more likely than men to report hypersomnolence (22.6% vs. 15.5%).

“Sex differences in reported sleep symptoms have been reported, with men being more likely to report snoring (34% vs. 19%) and women being more likely to report hypersomnolence (22.6% vs. 15.5%).90”

• cites [90] sourcePartly supported

  The Occurrence of Hypersomnolence (p. 3) reports that hypersomnolence was more common among women than men and gives the cited 22.6% and 15.5% figures. Those figures apply only to subjects with apnea-hypopnea scores of 5 or higher, a restriction the citing sentence omits. (LLM, conf=0.88)

  Experimental — accuracy not yet measured

Neck circumference is associated with OSA severity.

“Neck circumference is associated with OSA severity.138,528”

• cites [138] sourceCould not verify

  The Study Design passages describe OSA classification by apnea–hypopnea index, and the Economic Analysis passages address testing and costs; neither discusses neck circumference or its relationship to OSA severity. Searches for “neck circumference” and related association terms found no passage addressing the claim. No page number was provided for these passages. (LLM, conf=0.74)

  Experimental — accuracy not yet measured
• cites [528] sourceSupported

Obesity is a well-established risk factor for the development and progression of OSA.

“Obesity is a well-established risk factor for the development and progression of OSA. A BMI greater than 30 increases risk for OSA350 and weight reduction is a treatment option that can improve OSA severity.189”

• cites [350] sourceSupported

Increased caffeine intake is associated with less cognitive impairment in patients with untreated OSA.

“Studies on patients with untreated OSA have demonstrated that increased caffeine intake is associated with less cognitive impairment523 thereby indicating a potential therapeutic effect of caffeine and evidence for self-treatment of OSA.”

• cites [523] sourceSupported

Coffee and tea consumption are not associated with SDB.

“Women with SDB have the strongest correlations between caffeinated soda and OSA. There was only an association between caffeinated soda and severe SDB in men.524 Coffee and tea consumption were not associated with SDB.”

• cites [524] sourceSupported

Neck circumference greater than 40 cm (16 inch) is associated with snoring and OSA.

“Specifically, NC > 40 cm (16 inch) is associated with snoring and OSA.528,529,536–538”

• cites [528] sourcePartly supported

  The Clinical Research and Conclusions sections associate larger neck circumference with more severe obstructive sleep apnea syndrome. The paper does not establish a >40 cm threshold or an association between that threshold and snoring; its Polysomnography section says only that snoring noise was recorded. (LLM, conf=0.88)

  Experimental — accuracy not yet measured
• cites [529] sourceNot checked — no full text of the source
• cites [536] sourcePartly supported

  The Risk factors for habitual snoring section reports an association between wide neck circumference (>40 cm) and habitual snoring, and the Discussion reports an association with sleep-disordered breathing (SDB). The Interview questionnaire section defines SDB by reported symptoms, and the study acknowledges that it did not use polysomnography to diagnose it; the paper therefore does not directly establish an association with diagnosed OSA. Page numbers are unavailable in our copy. (LLM, conf=0.87)

  Experimental — accuracy not yet measured
• cites [537] sourceNot checked — no full text of the source
• cites [538] sourceNot checked — no full text of the source

In the SHHS, women with SDB have the strongest correlations between caffeinated soda and OSA.

“Women with SDB have the strongest correlations between caffeinated soda and OSA. There was only an association between caffeinated soda and severe SDB in men.524 Coffee and tea consumption were not associated with SDB.”

• cites [524] sourceSupported

Cephalometry alone may miss a significant proportion of patients with OSA in low-risk populations.

“Analysis of preoperative clinical screening tests for OSA revealed that cephalometry alone may miss a significant proportion of patients with OSA in low-risk populations.561”

• cites [561] sourcePartly supported

  The Discussion (p. 6) says false-negative rates mean most clinical prediction models may miss a significant proportion of patients with OSA. Study Limitations (p. 11) warns that models tested in high-risk patients can have lower sensitivity in lower-risk populations. The paper does not establish that cephalometry alone has this false-negative rate; it describes a combined clinical-cephalometry model as among the most accurate. (LLM, conf=0.84)

  Experimental — accuracy not yet measured

Low inter-examiner agreement in FTP scoring has been suggested.

“FTP and AHI are positively correlated with higher FTP grade associated with higher AHI547,552,553; however, low inter-examiner agreement in FTP scoring has been suggested.554”

• cites [554] sourceSupported

An 18% false negative rate for HSAT in high-risk patients has been observed.

“An 18% false negative rate for HSAT in high-risk patients has been observed572; therefore, if a single HSAT is negative, inconclusive, or technically inadequate, an in-laboratory attended 573,579 PSG should be performed.”

• cites [572] sourcePartly supported

  The Discussion says HSAT can produce false negatives in patients at increased risk of moderate to severe OSA and recommends PSG after a negative or inconclusive result. It does not report an observed 18% false-negative rate; the numerical estimate in the provided passage is 45 to 230 additional false negatives per 1,000 high-risk patients in a meta-analysis. (LLM, conf=0.88)

  Experimental — accuracy not yet measured

Emerging data from small RCTs show HSAT can be feasibly implemented after stroke or TIA or in a stroke rehabilitation unit.

“However, there is emerging data from small RCTs that HSAT can be feasibly implemented after stroke or transient ischemic attack (TIA) or even patients in a stroke rehabilitation unit.586,587”

• cites [586] sourcePartly supported

  The Methods and analysis sections describe a 55-patient randomised trial in a stroke rehabilitation unit that uses type-III polygraphy and plans to assess its feasibility. But this is a pre-results protocol, not RCT evidence that testing was feasible; it also does not study patients after TIA. (LLM, conf=0.87)

  Experimental — accuracy not yet measured
• cites [587] sourceSupported

MRI may be superior to radiocephalometry for assessment of anatomic measures in OSA patients.

“Because the enhanced soft tissue resolution of MRI affords greater measurement accuracy and allows for the determination of additional airway measurements, which cannot be achieved through cephalometry (e.g., tongue volume, PAS area), MRI may be superior to radiocephalometry for the assessment of anatomic measures in OSA patients.567”

• cites [567] sourceSupported

• AssumptionRests on an assumption: MRI provides superior soft tissue definition without ionizing radiation but offers less precise bony contour definition.

Each increase in tonsillar size correlates to an increase in AHI of approximately 14 events/h in adults with OSA.

“Tonsil size may predict severity of OSA, and a study of adults with OSA demonstrated that each increase in tonsillar size correlates to an increase in AHI of approximately 14 events/h.327”

• cites [327] sourceSupported

Sensitivity is increased for HSAT devices using four channels compared to one or two channels.

“MA comparing one channel, two channel, and four channel HSATs showed that the sensitivity is increased for devices using four channels compared to one or two channels.584,585”

• cites [584] sourceCould not verify

  The Materials section describes one home-testing device, and the Sensitivity and Specificity and Between-Site Effects sections compare that device with polysomnography at different diagnostic thresholds and study sites. Our copy does not report a meta-analysis comparing the sensitivity of one-, two-, and four-channel HSAT devices; searches for those channel comparisons found no supporting passage. The four-versus-three comparison in Table II concerns sites, not channels. (LLM, conf=0.76)

  Experimental — accuracy not yet measured
• cites [585] sourceNot checked — no full text of the source

Technical failure rate is higher for portable monitoring done at home than in the lab.

“The technical failure rate is also higher for portable monitoring done at home as opposed to in the lab.583”

• cites [583] sourceSupported

Evidence on using home sleep apnea testing to diagnose sleep-disordered breathing in COPD patients is limited, with only one Level 1b and one Level 3b study available.

“Limited data is available to assess the utility of HSATs in the diagnosis of SDB in patients with COPD (one Level 1b study and one Level 3b study).611,612”

• cites [611] sourceSupported
• cites [612] sourcePartly supported

  The Introduction (page unavailable) identifies WatchPAT as a home sleep apnea testing device and says previous WatchPAT studies excluded patients with COPD; this paper therefore provides an example of the limited evidence the citing sentence describes. The paper does not establish that only two studies were available or assign them Level 1b and Level 3b classifications. (LLM, conf=0.82)

  Experimental — accuracy not yet measured

• AssumptionRests on an assumption: Patients with COPD are at risk for OSA, hypoventilation, and nocturnal hypoxemia.

In the general population, oximetry ODI correlates reasonably well with PSG AHI, primarily when 4% desaturation criteria are used to score respiratory events.

“In the general population, two Level 1b studies,499,616 three Level 2b studies,617–619 and one Level 3b620 found reasonably good correlations between oximetry ODI and PSG AHIs, though primarily when the 4% desaturation criteria was used to score respiratory events.”

• cites [499] sourceNot checked — no full text of the source
• cites [616] sourcePartly supported

  The Results report a strong correlation between home-oximetry ODI and PSG AHI (r = 0.811), and Materials and methods specify a ≥4% desaturation threshold. However, the participants were patients visiting a sleep clinic for sleep problems, not a general-population sample. The paper also did not compare ODI–AHI correlations under 4% and other scoring thresholds, so it does not establish that the correlation holds primarily with the 4% criterion. (LLM, conf=0.88)

  Experimental — accuracy not yet measured
• cites [617] sourceNot checked — no full text of the source
• cites [618] sourcePartly supported

  The Selection of ODI parameters section reports strong correlations between PSG AHI and oximetry ODI, with the highest R² for the 4% ODI measure (0.878). The PSG examination section also specifies a 4% desaturation criterion for scoring hypopneas. However, the study enrolled patients with suspected OSA at a hospital sleep center, not a general-population sample. (LLM, conf=0.86)

  Experimental — accuracy not yet measured
• cites [619] sourcePartly supported

  The Results (page not available) report a good correlation (Spearman coefficient 0.76) between oximetry ODI and respiratory event index, and the Methods specify 4% desaturation criteria. But the study involved patients hospitalized with heart failure and compared ODI with portable-monitoring REI, not PSG AHI in the general population. (LLM, conf=0.90)

  Experimental — accuracy not yet measured
• cites [620] sourcePartly supported

  Table 2 (p. 22) reports reasonably good correlations between oximetry ODI and PSG AHI (r = 0.603–0.617). But the study enrolled patients referred for suspected OSA, not a general-population sample (Subjects, p. 5). Its PSG hypopnea definition used a 3% desaturation threshold, and ODI3 correlated slightly more strongly with AHI than ODI4; it therefore does not support the claim that the correlation was primarily associated with 4% criteria (Polysomnography, p. 5; Table 2, p. 22). (LLM, conf=0.89)

  Experimental — accuracy not yet measured

Among high pretest probability patients with a normal HSAT who later had PSG, OSA was found in 24%.

“and 24% of the patients with normal HSATs.615”

• cites [615] sourceSupported

In 173 CHF patients, at an ODI3% cutoff >7.5 desaturations/h, sensitivity was 97%, specificity 32%, negative likelihood ratio 0.08, and positive likelihood ratio 1.42.

“The results demonstrated that at the cutoff of >7.5 desaturations/h, the ODI3% had a sensitivity 97%, a specificity 32%, a negative likelihood ratio 0.08, and a positive likelihood ratio 1.42.”

• cites [623] sourcePartly supported

  The Diagnostic accuracy of overnight pulse oximetry section reports all four figures at the >7.5 desaturations/h cutoff. The Abstract distinguishes the 173 patients with adequate sleep-study data from the 171 whose ODI could be analysed, so the claim's sample size for these test results is imprecise. (LLM, conf=0.90)

  Experimental — accuracy not yet measured

High-resolution pulse oximetry could not reliably differentiate central from obstructive respiratory events in acute CHF.

“However, HRPO was unable to reliably differentiate between central and obstructive respiratory events.”

• cites [619] sourceSupported

In the Ward et al. CHF study, diagnostic accuracy increased at an ODI3% cutoff of 12.5 desaturations/h, with sensitivity 93% and specificity 73%.

“The diagnostic accuracy increased using a cutoff of 12.5 desaturations/h, with a sensitivity 93% and a specificity 73%.”

• cites [623] sourceSupported

In the Ward et al. CHF study, ODI3% had an AUC of 0.92 (95% CI 0.88-0.96) for detecting SDB in CHF at a cutoff >7.5 desaturations/h.

“The ODI3% had an AUC of 0.92 (95% C.I. 0.88–0.96) for the detection of SDB in CHF, at the cutoff of >7.5 desaturations/h.”

• cites [623] sourceSupported

Non-Level 1 studies in the hFrEF group were downgraded due to small numbers and/or antiquated criteria used to define respiratory events.

“The non-Level 1 studies in this group were downgraded due to small numbers and/or antiquated criteria used to define respiratory events.584,607”

• cites [584] sourceSupported
• cites [607] sourceNot checked — no full text of the source

Erdenebayar et al. found moderate accuracy between in-lab PSG AHI and the piezoelectric sensor snoring index when worn simultaneously with PSG.

“In an exploratory study by Erdenebayar et al., there was moderate accuracy between in-lab PSG AHI and the piezoelectric sensor snoring index when worn at the same time as the PSG.621”

• cites [621] sourcePartly supported

  The Data analysis section compares piezo-electric sensor recordings with PSG annotations, and the Results report 74.3% overall OSA-classification accuracy. But the reported accuracy is for a method combining the snoring index with pulse-rate-variability features, not for agreement between PSG AHI and the snoring index alone; the reported AHI correlation is likewise between PSG AHI and the method’s estimated AHI. (LLM, conf=0.86)

  Experimental — accuracy not yet measured

In stable COPD outpatients with OSA symptoms, agreement between HSAT and PSG is best when both are performed on the same night, with 4% desaturation criteria and an AHI cutoff of 15.

“The best agreement was found when the studies were done on the same night, used a 4% desaturation criteria for respiratory events, and used an AHI cutoff of 15 to define SDB.”

• cites [611] sourcePartly supported

  The Results report closer agreement with PSG for the simultaneous in-laboratory portable-monitor recording than for HSAT performed on a different night. They used a ≥4% desaturation rule and evaluated an AHI cutoff of ≥15, but do not establish that this combination gave the best agreement: results with the ≥3% rule were similar, and the ≥15 cutoff increased specificity while reducing sensitivity. The same-night comparison was with an in-laboratory monitor, not HSAT. (LLM, conf=0.87)

  Experimental — accuracy not yet measured

Among high pretest probability patients with a technically inadequate HSAT who later had PSG, OSA was found in 71%.

“found OSA in 71% of the patients with technically inadequate studies”

• cites [615] sourceSupported

The main limitation of the Chernyshev stroke study was that it included only 21 participants.

“The main limitation to the study was that it included only 21 participants.”

• cites [614] sourceSupported

Patients who have had a stroke have roughly a 60%-70% risk of having OSA.

“Patients who have had a stroke have a roughly 60%–70% risk of having OSA.613”

• cites [613] sourceSupported

In COPD patients, kappa values between HSAT and PSG ranged from 0.48 to 0.93 depending on whether nights differed, desaturation criteria, and AHI cutoff.

“Kappa values ranged from as low as 0.48 (different night, 3% desaturation criteria, AHI cutoff of 5) to 0.93 (same night, 4% desaturation criteria, AHI cutoff of 15).”

• cites [611] sourcePartly supported

  The results text accompanying Table 3 reports kappa agreement between portable-monitor recordings and PSG in patients with COPD, and discusses different AHI cutoffs and 3% versus 4% desaturation scoring. However, the passages available in our copy do not give the claimed endpoints of 0.48 and 0.93 or tie them to the specified night, scoring, and cutoff combinations. (LLM, conf=0.77)

  Experimental — accuracy not yet measured

Possible pathophysiologic mechanisms linking OSA to myocardial ischemia include sympathetic nervous system hyperactivity, hypertension, endothelial dysfunction, metabolic dysregulation, insulin resistance, and a hypercoagulable state.

“A number of possible pathophysiologic mechanisms have been implicated including sympathetic nervous system hyperactivity, hypertension, endothelial dysfunction, metabolic dysregulation, insulin resistance, and hypercoagulable state.641,655–663”

• cites [641] sourcePartly supported

  The Discussion (page unavailable) identifies sympathetic activation, endothelial dysfunction, insulin resistance, hypertension, hyperlipidemia, and diabetes as possible links between OSA and coronary heart disease. It does not identify a hypercoagulable state as one of those mechanisms. (LLM, conf=0.88)

  Experimental — accuracy not yet measured
• cites [655] sourceNot checked — no full text of the source
• cites [656] sourcePartly supported

  The paper's Abstract (page not provided) reports an association between OSA and insulin resistance and describes a metabolic pathway leading to cardiovascular disease. It does not specifically address myocardial ischemia or establish sympathetic hyperactivity, hypertension, endothelial dysfunction, or a hypercoagulable state as mechanisms. With multiple citations attached to the sentence, this source supports its metabolic and insulin-resistance portion, not the full list. (LLM, conf=0.85)

  Experimental — accuracy not yet measured
• cites [657] sourceNot checked — no full text of the source
• cites [658] sourceNot checked — no full text of the source
• cites [659] sourcePartly supported

  The Abstract (page not given) identifies OSA as a cause of hypertension and reports endothelial dysfunction in patients with OSA; the study’s discussion (section 30 (2), page not given) further describes microcirculatory endothelial dysfunction. The copy does not establish sympathetic hyperactivity, metabolic dysregulation, insulin resistance, or a hypercoagulable state as mechanisms linking OSA to myocardial ischemia. (LLM, conf=0.86)

  Experimental — accuracy not yet measured
• cites [660] sourceNot checked — no full text of the source
• cites [661] sourcePartly supported

  The Discussion says hypertension, endothelial dysfunction and increased coagulation in OSA may contribute to cardiovascular events, and it links coagulation markers to myocardial infarction. The cited paper does not address sympathetic nervous system hyperactivity, metabolic dysregulation or insulin resistance as mechanisms, so it supports only part of the list. (LLM, conf=0.86)

  Experimental — accuracy not yet measured
• cites [662] sourceNot checked — no full text of the source
• cites [663] sourceNot checked — no full text of the source

During obstructive apneas, increased adrenergic tone and hypoxemia may increase the risk of myocardial ischemia.

“During obstructive apneas, increased adrenergic tone and hypoxemia may increase the risk of myocardial ischemia.666,667”

• cites [666] sourceSupported
• cites [667] sourceSupported

OSA is highly prevalent (38%-65%) in patients with coronary artery disease compared with the general population.

“Compared to the general population, OSA is highly prevalent (38%–65%) in patients with CAD.635”

• cites [635] sourceSupported

Among adverse cardiovascular consequences, the evidence linking OSA and hypertension is the most robust.

“While OSA has been associated with several adverse CV consequences, the evidence linking OSA and HTN is the most robust.680,681”

• cites [680] sourceSupported
• cites [681] sourceSupported

In a secondary SHHS analysis, after adjustment, OSA significantly predicted incident CAD events only in men aged 70 or younger (HR 1.10 per 10-unit increase in AHI).

“After adjusting for other risk factors, OSA was significantly associated and predicted incident CAD events, defined as MI, revascularization procedure, or coronary heart disease related death, however this was only observed in men aged 70 or younger (HR 1.10, 95% CI = 1.00–1.21 per 10-unit increase in AHI).”

• cites [641] sourcePartly supported

  The Results report the stated adjusted HR of 1.10 (95% CI 1.00–1.21) per 10-unit AHI increase among men aged 70 or younger. But the Results say the age interaction was not statistically significant, and the Discussion calls the subgroup finding only nominally significant and possibly a type 1 error. The claim overstates the evidence by presenting the association as established only in that age group. (LLM, conf=0.88)

  Experimental — accuracy not yet measured

In the same SHHS secondary analysis, severe OSA (AHI ≥30) conferred a stronger increased risk of symptomatic CAD (HR 1.68, 95% CI 1.02-2.76), again only in men aged 70 or younger.

“Severe OSA (as defined by AHI ≥ 30 events/h) conferred a stronger increased risk of developing symptomatic CAD (HR 1.68, 95% CI = 1.02–2.76) again seen only in men aged 70 or younger.641”

• cites [641] sourceSupported

The accompanying increase in sympathetic activity, inflammation, endothelial dysfunction, and elevated blood pressure in OSA is associated with increased risk for cardiovascular morbidity and mortality.

“The accompanying increase in sympathetic activity, inflammation, endothelial dysfunction, and elevated BP is associated with increased risk for CV morbidity and mortality.634,636–638”

• cites [634] sourcePartly supported

  The Discussion (page unavailable) reports increased cardiovascular events and mortality in people with OSA, primarily at ages 30–50. The paper also links obstructive breathing disturbances to hypertension, but its discussion of increased sympathetic activity concerns central sleep apnea and heart failure. It does not substantiate the attributed increases in inflammation or endothelial dysfunction, and its own analysis did not find OSA to be a significant predictor of incident heart failure after adjustment. (LLM, conf=0.86)

  Experimental — accuracy not yet measured
• cites [636] sourceNot checked — no full text of the source
• cites [637] sourceNot checked — no full text of the source
• cites [638] sourceNot checked — no full text of the source

The prevalence of nocturnal non-dipping blood pressure is high in subjects with OSA, in the range of 50%-80%.

“The prevalence of nocturnal non-dipping BP at night, an adverse CV prognostic risk factor, is high in subjects with OSA, in the range of 50%–80%.689,690”

• cites [689] sourceNot checked — no full text of the source
• cites [690] sourceSupported

Moderate or severe OSA (AHI ≥15) is reported in 20%-62% of atrial fibrillation patients.

“Moderate or severe OSA (AHI ≥ 15) is reported in 20%–62%.330,704,706,710–715”

• cites [330] sourcePartly supported

  Table 1 (Baseline Characteristics; page unavailable) reports AHI >30 in 55 of 88 control patients (62%) undergoing AF ablation. This supports an upper-end figure in a selected group, but the paper does not report a 20%–62% range for moderate or severe OSA defined as AHI ≥15 among AF patients generally. (LLM, conf=0.87)

  Experimental — accuracy not yet measured
• cites [704] sourceSupported
• cites [706] sourceNot checked — no full text of the source
• cites [710] sourcePartly supported

  The Results (Prevalence and severity of SA; page unavailable) report AHI ≥15 in 42.1% of patients with paroxysmal AF, with sleep apnea predominantly obstructive. The Discussion reports a 21%–62% range across previous AF studies, but does not specify that the entire range refers to moderate or severe OSA (AHI ≥15). (LLM, conf=0.84)

  Experimental — accuracy not yet measured
• cites [711] sourceNot checked — no full text of the source
• cites [712] sourceNot checked — no full text of the source
• cites [713] sourceNot checked — no full text of the source
• cites [714] sourcePartly supported

  The ‘Sleep apnoea syndrome and risk of atrial fibrillation’ section reports moderate-to-severe sleep apnoea in 31% of 158 AF patients after successful cardioversion. That supports a prevalence within the cited range, but this paper does not establish the 20%–62% range or its endpoints. (LLM, conf=0.86)

  Experimental — accuracy not yet measured
• cites [715] sourceNot checked — no full text of the source

In six studies that systematically assessed OSA before atrial fibrillation treatment, OSA was associated with an approximately two- to three-fold increased risk of recurrent atrial fibrillation following electrical cardioversion or pulmonary vein isolation.

“In six studies that systematically assessed the presence of OSA prior to treatment of AF, OSA was associated with an approximately two- to three-fold increased risk of recurrent AF following electrical cardioversion714 or pulmonary vein isolation (catheter ablation) procedures.708,709,711,713,727”

• cites [714] sourcePartly supported

  Split into 2 atoms — 0 confirmed, 2 partially supported, 0 disputed, 0 misattributed, 0 unverifiable. Expand the claim to see per-atom verdicts.

  Experimental — accuracy not yet measured
• cites [708] sourceNot checked — no full text of the source
• cites [709] sourceSupported
  Experimental — accuracy not yet measured
• cites [711] sourceNot checked — no full text of the source
• cites [713] sourceNot checked — no full text of the source
• cites [727] sourcePartly supported

  Split into 2 atoms — 1 confirmed, 1 partially supported, 0 disputed, 0 misattributed, 0 unverifiable. Expand the claim to see per-atom verdicts.

  Experimental — accuracy not yet measured

Untreated rapid eye movement-related OSA is significantly associated with the development of hypertension.

“More recent evidence suggests that untreated rapid eye movement (REM)-related OSA, a subtype of OSA where SDB events are predominantly confined to REM sleep, is significantly associated with the development of HTN.685”

• cites [685] sourcePartly supported

  The ‘REM-SDB and hypertension’ section reports that two earlier studies found an association with increased incident hypertension. But this paper’s own cross-sectional analysis found no significant association with hypertension, including in a subgroup not taking antihypertensive medication. It does not establish the claimed association specifically for untreated REM-related OSA. (LLM, conf=0.84)

  Experimental — accuracy not yet measured

The temporal distribution, night-to-night variability, and degree of desaturation or autonomic response associated with sleep-disordered breathing events may affect an individual's risk of developing hypertension.

“Finally, the temporal distribution, night-to-night variability, and degree of desaturation and/or autonomic response associated with SDB events may ultimately affect an individual’s risk of developing HTN.681”

• cites [681] sourceSupported

Patients with atrial fibrillation have a high prevalence of OSA.

“Patients with AF have a high prevalence of OSA.”

• cites [704] sourceSupported
• cites [705] sourceNot checked — no full text of the source
• cites [706] sourceNot checked — no full text of the source
• cites [707] sourceNot checked — no full text of the source
• cites [708] sourceNot checked — no full text of the source
• cites [709] sourceSupported
• cites [710] sourceSupported

In the Multi-Ethnic Study of Atherosclerosis, the prevalence of atrial fibrillation increased from 4.0% in those without OSA to 7.5% in those with severe OSA.

“In the Multi-Ethnic Study of Atherosclerosis, the prevalence of AF increased from 4.0% in those without OSA to 7.5% in those with severe OSA.718”